Full text of "Journal of the Society of Motion Picture Engineers"

See other formats

From the collection of the

PreTnger
v JJibrary

San Francisco, California
2007

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR
Volume XVIII JANUARY, 1932 Number 1

CONTENTS

Page

Lighting of Sound Films Louis DUNOYER 3

The Rapid Record Oscillograph in Sound Picture Studies

A. M. CURTIS, T. E. SHEA, AND C. H. RUMPEL 39

Photographic Sensitometry, Part III LOYD A. JONES 54

Thermionic Tube Control of Theater Lighting . . BURT S. BURKE 90

A Portable Non-intermittent Cine Projector 101

Committee Activities:

Report of the Projection Practice Committee 107

Report of the Projection Theory Committee 113

Abstracts 116

Patent Abstracts 122

Officers 130

Committees 131

Contributors to This Issue 133

Society Announcements 134

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1931, by the Society of Motion Picture Engineers, Inc.

Subscription to non -members, $12.00 per annum; to members, $9.00 per annum,
included in their annual membership dues; single copies, $1.50. A discount
on subscriptions or single copies of 15 per cent is allowed to accredited agencies.
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y.

Papers appearing in this Journal may be reprinted, abstracted, or abridged
provided credit is given to the Journal of the Society of Motion Picture Engineers
and to the author, or authors, of the papers in question.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930. at the Post Office at Easton,
Pa., under the Act of March 3, 1879.

LIGHTING OF SOUND FILMS*

LOUIS DUNOYER

Summary. — The author examines different types of illumination apparatus.
He discusses their conditions of operation and describes t}ie apparatus devised by
himself in greater detail. This apparatus obtains an extremely fine exploring zone
simply by projecting the image of a rectilinear incandescent filament on the film
by means of a good objective. In order to correct the aberrations rigorously, the light
coming from this filament goes through the walls of the lamp in a place where they
form parallel faces worked optically.

In conclusion, the author describes some comparative tests made on apparatus with
a slit and on the rectilinear filament apparatus. The flux emitted by the latter is
superior to the flux emitted by the apparatus with the slit, with a consumption ap-
proximately one-tenth as great. At the same time the disadvantages of the slits (dust,
defective uniformity of illumination, delicacy of centering, etc.} are eliminated.

INTRODUCTION

1 . Review of the Principle of Sound Films. — It is well known that
on a sound film the section set aside for sound production is a small
straight band only 3 millimeters wide, in general located between
one of the series of perforations for moving the film and the edges
of the picture images which are to be projected on the screen. On
this small band the sounds first have been recorded by either of the
two following processes which we shall review briefly to make them
clearer.

In the constant density** process the sound band is divided in
two regions, each of which has a uniform photographic density,
one clear and the other dark, and whose common boundary is a line
which is more or less indented or wavy. The bends of this line corre-
spond to the recorded sound vibrations. In most cases this re-
cording is performed by means of an oscillograph which receives
the current from the receiving microphone after amplification by
a triod. In vibrating, the spot of the oscillograph, which consists
of a small luminous line perpendicular to the length of the band,

* Translated from Revue d'Optique, 10 (Jan.-Feb., 1931), Nos. 1-2, pp. 1-21,
57-68.

** I.e., variable width.

LOUIS DUNOYER

[J. S. M. P. E.

produces an image on the film of a width which varies according
to the amplitude and the frequency of the sound vibrations. Fig. 1
shows an example of constant density recording.

In the variable density process the photographic density of the
sound band is the same at all points of its width but this density
varies in the direction of the length of the band. The recording
system which is used most at the moment consists in letting the
film slide by a very fine slit illuminated by a flashing lamp. This
lamp contains a gas under a low pressure illuminated by the dis-
charge; the voltage at which this discharge is produced is modu-
lated by the current from the recording
microphone, conveniently amplified. The
luminescence of the gas follows these modu-
lations which are the more intense the
higher the voltage. These modulations,
therefore, are transferred to the sound
band by corresponding modulations of the
density in the direction of the length of
the band. Figs. 2 and 3 show two ex-
amples of variable density recording. Fig.
2 refers to an invariable musical note (ap-
proximately 440 vibrations per second).

In order to reproduce the sounds the
entire width of the sound band must be
illuminated, but only for a width equal to
that of the spot or the slit which illumi-
nated it during the recording; then having
passed through the film the light is received
by a photoelectric cell. If the photographic

opacities are proportional to the luminous fluxes received during the
recording, and if the photoelectric currents are proportional to the
luminous fluxes received by the cell, these currents finally will be
proportional to the currents of the recording oscillograph or to the
brilliancy of the flashing lamp, according to the recording process
employed. Then they are amplified and sent into a loud speaker.
The distortion of the sound can be due only to the microphone
circuit or the circuit of the loud speaker.

Since the light received by the cell should be only the light which
has passed through the film in a rectangle 3 millimeters long and a
few hundredths of a millimeter high (0.05 mm. at most) much light

FIG. 1. An example of
constant density recording.

Jan., 1932]

LIGHTING OF SOUND FILMS

evidently must be concentrated in this rectangle, and naturally this
entire flux must fall on the cell after having diverged on leaving
the film. The latter condition is readily attained; the former con-
stitutes the problem of illuminating devices for sound films.

We shall divide this paper in two parts. In Part I we shall first
review briefly the various types of devices, and then examine the
theoretical conditions which must be satisfied by the mode of illumi-
nating the film in order to obtain a suitable sound performance.

FIG. 2. Example of variable density recording.

In Part II we shall apply the results of this investigation; we shall
discuss the properties of existing devices, and describe in detail the
apparatus which we have devised and the results of a few tests.

PART I
THEORETICAL STUDY OF THE LIGHTING OF SOUND FILMS

2. On the Width of the Illuminated Exploring Region. — It is neces-
sary in particular to know the length on the film occupied either by
a period of the separating curve in the case of constant density re-
cording, or by a period of the density in the case of variable density

LOUIS DUNOYER

[J. S. M. P. E.

recording, assuming, of course, that the recorded sound itself has a
definite period. Let / be the length on the film occupied by one
period and v the speed of unwinding the film. If the frequency of
the sound vibration is N, we have:

'-i

since the film advances by / in 1/N second.

In general, the film speed is 45.5 centimeters per second.

(1)

One

FIG. 3. Example of variable density recording; a constant frequency note.

period of Ia3 or normal la, corresponding to 440 vibrations per second,
thus would occupy a height of 1.03 millimeters on the film. It is
necessary, however, to record much shriller sounds. First, the
highest note used in music is re-i, corresponding to 4698 vibrations
per second, which on the film gives a period of 0.097 millimeter.
But this is far from sufficient for the correct recording of different
timbres and for the timbre of the human voice. Much higher har-
monics then must be attained and some hold that it will be necessary

Jan., 1932]

LIGHTING OF SOUND FILMS

to record 20,000 vibrations per second, which on the film would corre-
spond to a period of 0.02275 millimeter. We shall see the ratio
which it is possible to allow between the period of the sound and the
height of the illuminated region on the film. It is clear in any case
that this height should be smaller than the period. It can be from
0.2 to 0.05 millimeter.

3. Different Types of Lighting Apparatus. — The majority of ap-

FIG. 4. One type of lighting apparatus in which the slit
is placed as near as possible to the film.

paratus employed heretofore for illuminating the cell through the
film involves the use of a fine slit which limits the height of the
exploring zone. This slit can be used in two different ways.

In one group of devices (Fig. 4), the slit F is placed as near as
possible to the film. The light supplied by an incandescent lamp,
with the filament S as concentrated as possible, is projected on
this slit by means of a condenser. The light which has passed

FIG. 5. Another type of lighting apparatus in which the image of the slit
is projected on the film.

through the film then falls on the photoelectric cell C. Since the
light rays diverge more from the window the wider the angular
aperture of the beam from the condenser, thereby increasing the
illumination of the slit, the illuminated region of the film will be
sufficiently narrow only if the slit is very close.

In another group of devices (Fig. 5) the image of the slit is pro-
jected on the film by means of an objective 0. This objective fre-

LOUIS DUNOYER

[J. S. M. P. E.

quently is a microscope objective which produces a smaller picture
on the film than the slit itself. The slit, therefore, can be com-
paratively wide but still all the light received by the condenser is

FIG. 6. Diagram for studying lighting effects on the film in the
system illustrated by Fig. 4.

far from utilized. In this type of device the film is located at a dis-
tance from the back face of the microscope objective which is equal
to its frontal distance, that is, a few millimeters.

In order to avoid the considerable losses of light which take place

FIG. 7. Similar to Fig. 6, showing a different phase in the passage
of the film past the aperture.

in these devices, the extreme nearness of the film and the slit in the
former, and in a general manner the inconveniences outlined in Part
II, which result from the use of a slit, devices without a slit can be

Jan., 1932] LIGHTING OF SOUND FlLMS 9

used, and of these there also are two types. In one of these types
a cylindrical or cylindrospherical lens is used. In the other type
the reduced image of a rectilinear incandescent filament is produced
on the film by means of a good objective. This solution, which is
our own, then requires the construction of a special lamp.

4. Distribution of the Light on the Film. — Let EE' (Figs. 6 and
7) be the exploring zone, or more exactly, its projections on a plane
through the axis of the sound band and perpendicular to the film,
this plane being the plane of the figure. Whatever lighting appara-
tus is used, this plane also is a plane of symmetry of the illuminat-
ing beam, and the exploring zone receives light from a surface HH'
which is the exit pupil of the illuminating apparatus; the luminous
flux passing through a point of the exploring zone fills a cone whose
peak is this point and the base this pupil. Let FF' be the film;
for perfect illumination its plane should coincide with EE'. If the
exploring zone consists of a slit (apparatus of the first type), the
film should rub on the sides of this slit. Since this involves serious
inconveniences, the film must be separated slightly from the slit
as indicated in the figure. With the other illuminating apparatus
the difference between the film and the exact position of the explor-
ing zone may be due to an error in focusing.

Since this difference in a general manner is different from zero
and equal to d, a point M of the film near the axis receives light
from all points of HH' provided that it is inside the cone HPH'
whose peak P is obtained by joining the edges of the pupil and of
the exploring zone located on the same side of the axis. The fully
illuminated field on the film, therefore, has the height BB ' . But
the film also receives a degraded illumination on the two bands,
BC and B'C', the points C and C' being those where the film is
struck by the rays which join the edges of the pupil and those of the
exploring zone located on both sides of the axis. The band CC'
is the total field which we shall call the explored zone.

In order to calculate the luminous flux received by the point M
of the explored zone (Fig. 6), the exploring zone is projected from
point M on the plane of the pupil HH' ', and the common surface
of the latter and the projection of the exploring zone is used. If
the flux received by a point of the fully illuminated field is used as
a unit, the flux received by the point M of the degraded field will
be equal to the ratio between the area of the circular segment NiN^Hi
and that of the circle

10 LOUIS DUNOYER [J. S. M. P. E.

Let us assume that the film gradually is moved away from the
exploring zone; the fully illuminated field BB' is reduced to zero when
the film passes through point P. For this position the illumi-
nation decreases constantly from the center to the edge of the ex-
plored zone CC' . When the film is above point P (Fig. 7), the parts
taken by HH' and EE1 in limiting the rays are exchanged, the ex-
ploring zone now serving as outlet pupil and the surface HH' as a
window. But the window surface employed decreases constantly
when the point considered on the film moves away from the axis
because, since the flux received by point P is used as a unit, the flux
received by point M will be equal to the ratio of the circular seg-
ment NiNzNi'Ni to that of the circle HyH^ and the area of this
segment is maximum when NiNz and Ni'N* are symmetrical in
regard to the center. The illumination of the explored zone de-
creases constantly from the center to the edge.

Let us calculate the heights h and hf of the field of uniform illumi-
nation and the explored zone. Let D be the height of the outlet
pupil HH', L its distance to the exploring zone, e the height of the
latter, and d its distance to the film; finally d0 is the distance QP
which is important, as we shall see. Triangles which evidently are
similar give (Fig. 6):

or since e always is very small compared with D:

do = § e (2)

Then

h d - d0 .D

- = — 5 — or h = e — d-f
e do L

or:

and:

h' d + QF D = L - QF
e ~~ QF € = QF

from which:

or:

*• - • (' + Q

Jan., 1932] LIGHTING OF SOUND FILMS 11

If we should consider Fig. 7 instead of Fig. 6, we would find the
same expressions for do and for h'; since there is no field of uniform
illumination, h cannot be considered.

It is useful to consider the values which the factor

may assume. We call attention to the fact that since the flux which
flows through the exploring zone should be as intense as possible,

D
the ratio — should be large. If u is the opening half -angle of the

beams which illuminate the exploring slit, we have:

The numerical aperture sin u of the illuminating apparatus in
general should be at least 0.20 (u = 12 degrees) and with a micro-
scope objective or a special photographic objective it may attain

0.4 (u = 24 degrees) which in the former case corresponds to —

L L

= 2.5 and in the latter case to — = 1.12; thus — will be comprised

approximately between 2 and 1. Consequently, according to for-
mula (2) do will be comprised between twice the height of the ex-
ploring zone and this height itself. Since this height may vary from
0.02 millimeter to 0.05 millimeter, it is clear that d0 always will be
very small, of the order of 0.02 to 0.1 millimeter. To assume that
d may reach 10 times dQ, therefore, is not an inadmissible hypothesis.
Moreover, we shall consider the maximum value. In any case an
error in focusing which is entirely possible, amounting to 0.1 milli-

meter, for instance, may correspond to a value of — of several units ;

and formula (4) shows that the explored zone can easily be several
times greater than the exploring zone.

5. Analysis of the Effects Produced by Enlarging the Explored
Zone on the Film and by the Distribution of the Light in This Zone. —
We shall divide this analysis into two parts. We shall first ex-
amine the effect of enlarging the exploring zone, of uniform illumi-
nation, assuming that this zone is formed exactly on the film. We
examine the influence of the width of the slit either when the film
is in contact with the sides of it or when a perfectly focused image
of a uniformly illuminated slit is formed on the film. This investi-

LOUIS DUNOYER

[J. S. M. P. E.

gation also applies to the case when apparatus without slit is used,
provided that the illumination of the exploring zone is uniform.
We also suppose that the exploring zone and the film do not coin-
cide, and we shall examine the effect of the degraded illumination
of the explored zone.

In both cases it is necessary to form a hypothesis of the manner
in which the transparency of the film varies along its length or, which
is the same, the manner in which the total luminous flux would
vary, which would go through the film if it were explored by means
of a zone infinitely narrow in regard to the length occupied by a
period of the transparency. In the case of a variable density film
this transparency is the same as that of the film. For a constant

minimum

film

\trans parency
.. Y

overage

transparency

FIG. 8.

Curve representing sinusoidal variation of the film
transparency along its length.

density film it is defined by the ratio between the width of the part
which is entirely clear and that of the opaque part. The most
natural hypothesis which we can choose for the law of variation
of the transparency of the film is that of a sinusoidal variation which
would correspond to the perfect recording of a musical sound. In
Fig. 8 the curve representing the variations of the transparency of
the film along its length is plotted on the right. The abscissa XQ
of the beginning of a period in regard to the axis of the lighting
apparatus defines a given position of the film. Its transparency
y at a point M, with the abscissa x, of the explored zone then will
be expressed by:

- x)

y = a + b sin

(5)

Jan., 1932]

LIGHTING OF SOUND FILMS

where / as in formula (1) is the length of a period of the transparency
on the film, that is, the length occupied by the recording of a com-
plete sound vibration.

The constant b represents the half-amplitude of the variation
of the opacity. It defines the amplitude or the intensity of the
recorded sound. The constant a is equal to the mean value of the
opacity when the film is unwound; in the photoelectric cell it corre-
sponds to a constant current, hence is of no interest in regard to the
sound.

6. Effect of the Height of the Exploring Zone when Formed on
the Film. Sound Efficiency. — If the luminous flux which falls on the

11 10 9

7 6 54

7 8 9 10 11

3' 2 ^ 0 J 2 3 if 5

'Slit
FIG. 9. Graphs of formulas (9), (10), and (11).

film is used as a unit, the flux which leaves the film through a band of
the height dx centered in M in the exploring zone will be expressed by:

d& = ( a + b sin °. J dx

The total flux which leaves the film thus will be expressed by:

or:

_, .

<b(xo)

. bl . ire . 2-n-Xo
ae H -- sm y sm — j-

(6)

LOUIS DUNOYER

[J. S. M. P. E.

Formula (6) shows that the leaving flux $(XQ) is a sinusoidal
function of the abscissa XQ, of the same period / and the same phase
as the transparency of the film. Whether the exploring zone is
narrow or wide, the height of the sound is not changed and no har-
monics will appear.

In principle, therefore, it is not absolutely necessary to employ
an extremely fine exploring zone to plot a band on which musical
sounds are recorded, even very shrill ones, provided that they are
sinusoidal. But, as we shall see, the intensity of the sound produced
decreases rapidly when the width of the exploring zone increases
and approximates the period of the transparency on the film. In
other words, the sound performance of the latter decreases very
rapidly. If a non-sinusoidal sound is concerned, having high har-

FIG. 10. Diagram for evaluating total flux passing
through the film in a given position, XQ.

monies, the timbre will be deformed, the more so according as the
width of the exploring zone increases, because the harmonics will
be the more reduced the higher they are.

In reality, according to formula (6) the amplitude A of the varia-
tion of the emerging flux will be expressed by:

2-bl . TT6

A = sin -7-

TT I

. 7TC

2bf — sin 7-

TTf I

(7)

If we assume that the ratio, -, of the width of the exploring zone
to the period of the transparency decreases toward zero, it is clear

Jan., 1932] LIGHTING OF SOUND FlLMS 15

that the variations of the flux increase toward a maximum A 0 corre-
sponding to an ideally fine exploration of the band; we have

The sound efficiency m of the exploring apparatus of the film can
be defined by the expression:

A I . TT6

m = -r- = — sm-r (8)

A o TTC I

d
The curve — = 0 of Fig. 1 1 represents the sound efficiency plotted

as ordinates as a function of the ratio - plotted as abscissae. It is

clear that the efficiency still is 90 per cent when the width of the
exploring zone is equal to one-quarter period. It is only 30 per cent
when the width of the exploring zone is three-quarters of a period.
In order to calculate the frequency which is reproduced with a
given efficiency, it is sufficient to eliminate N from formula (1)

in which / is substituted by the value which gives to - the value indi-

cated for the efficiency by the curve in Fig. 11. Thus, with an ex-
ploring zone of 0.02 mm. the frequency obtained with an efficiency

/e 1 \

of 90 per cent will be ( - = - , hence / = 4-0.02 ):

a frequency which is slightly higher than that of re-i of a piccolo.
The efficiency corresponding to this note will be exactly 94 per cent

o -

Another example: the frequency corresponding to an efficiency
of 25 per cent will be:

455- 0.785 =
0.02

e 0.02 \

from which: - = 0.785 and consequently / = ;r^J-
/ 0.785/

The frequency 20,000 ( *- = — : = 0.88 J will be reproduced

only with an efficiency of 13 per cent.

LOUIS DUNOYER

. S. M. P. E.

We note that probably a much higher sound quality could be
obtained in regard to the reproduction of the voice and of timbres
if the width of the exploring zone dropped to 0.01 millimeter. The
frequency 20,000 then would be reproduced with an efficiency of
70 per cent. This is due to the very rapid decline of the efficiency

0,1

FIG. 11. Curves showing the sound efficiency m as a function of K = c/l for

various values of d/d0.

indicated by the form of the curve when the width of the exploring
zone exceeds one-quarter of the period of the transparency on the
film.

7. Effect of an Error in Focusing or of a Reduced Explored Zone. —
We refer to Figs. 6 and 7 for the investigation of this problem, The
calculation of the luminous flux leaving the film, for a given posi-

Jan., 1932] LIGHTING OF SOUND FlLMS 17

tion of the latter, presupposes that the distribution of the incident
flux is known. In order to calculate the flux which falls in a point
M of the abscissa x in regard to the axis of the illuminating appara-
tus, we project, as mentioned above, the exploring zone on the plane
of the outlet pupil HH' of the illuminating apparatus, and we con-
sider the common surface of this projection and this pupil. Let S
be this surface, which has been shaded in Figs. 6 and 7. Since the
flux illuminating the fully illuminated zone is used as a unit, the flux
illuminating the point M will be equal to the ratio between this
common surface S and the surface wR2 of the pupil:

Calculation by integration of the surface S leads to the following
results:

In the case of Fig. 6 (d < do) we have in the fully illuminated
field:

and in the reduced part of the explored zone, that is, for:

1 <_•<<*•<!+<
«o « <*o

5 1 , . d, /, 2x

_ = § + arcsm

In the case of Fig. 7 (d > do) we obtain two different expressions
for 5 according as the projection of the exploring zone goes through
only the outlet pupil or projects beyond it. In the former case
we obtain:

1 . d0 /, 2x\ . 1 . d0 A , 2x , \

- arc sin -^ ( 1 ) + - arc sin -f [ 1 H h I

x d \ e/7T d \ e/

S
— --arc sn

with:

18 LOUIS DUNOYER [J. S. M. P. E.

and in the latter case:

with:

*_!<*<*+,

d € d

The formulas (9), (10), and (11) essentially assume that x is posi-
tive. It is clear that the incident flux $i(x) is the same for two
symmetrical points in regard to the axis. For negative x, therefore,
x should be replaced by — x in the second terms of the formulas
(9), (10), and (11).

These formulas have been translated into curves (Fig. 9) for dif-

ferent values of — .
do

On the other hand, the transparency of the film as above is as-
sumed to be expressed by the second term of formula (5) at point
M of the film (Fig. 10). The luminous flux passing through the
film in M through a band of the height dx is:

$»C

The total flux flowing through the film which is placed in a given
position, that is, for a given value of XQ, will be obtained by inte-
grating this expression from one limit to the other of the total field
or explored zone which has the height h'. (See §4.) We have, there-
fore:

+ -
= f J *<(*) (a

b sin dx

Like the transparency of the film this function of x0 has the period /.
This is evident physically; this also is due to the fact that if x0 in-
creases from /, the function under the sign of integration does not
change and that both limits of integration increase from /. The

emerging flux $,(#0) also goes through a maximum for x0 = - and

31 4

through a minimum for XQ = — ; we have in reality:

Jan., 1932] LIGHTING OF SOUND FlLMS 19

.*'

/ 2" 2-n-b
V T***0

For #o = ~ we have:

. 27T*

m —

and since $j(#) assumes equal values for equal values of x and oppo-
site signs, it is clear that the preceding integral is zero. The same

holds for #o = T« The emerging flux &e(xQ), therefore, has maxima
4

and minima in the moments when the axis of the illuminating ap-
paratus goes through the film in a region of maximum or minimum
transparency. Finally, this flux oscillates around the same mean
value no matter what the extent of the explored zone may be. If,
therefore, we put XQ = 0, we have:

,hr h'

C 2" / 2-*X\ C "

/ ^ <*><(*) (a - b sin -- J dx = a j /
~ ~2 "2"

The last integral represents the total flux illuminating the ex-
plored zone which is assumed to be constant and equal to 1 for any
surface of this zone.

Since the curves represent the variations of the flux which has
passed through the film, as a function of XQ, that is, of the position

of the film and for every value of — , they are undulating curves of

the same period and the same phase as the transparency of the film
(with one exception which will be examined below), all having the
same average ordinate. These curves naturally are a function of
the coefficients a and b, but the value of the efficiency of the illumi-
nating apparatus defined as above is not. In reality the amplitude
A of the variations of the flux, according to what has been said, is

equal to the difference between the values of $e(x0) for XQ = - and

#o = — • Hence:
4

20 LOUIS DUNOYER [J. S. M. P. E.

/?'
*»(x)co8^dx (13)

—

= 2b I 2 4>.-(
J-h-

If the explored zone having the height hf were infinitely narrow
in regard to the period / (perfect exploration of the film), the ampli-
tude of the variations of the flux would be as we have seen in the
preceding paragraph:

A, = 2be (14)

The efficiency w, therefore, will be expressed by:

i c+h* 2™

m = - I fy(x) cos — j- dx (15)

^-2-
With

If we consider the formulas (9), (10), and (11), we find

2x
really is a function of — . If we write, therefore,

€

^ = X K = - (16)

we get:

m = I 3>i(xY cos KirX -dX (17)

a formula which is independent of the coefficients a and Z> which
enter into the law of variation of the transparency.

In order to calculate the efficiency according to formula (17),

a value of K first must be chosen. Choosing a given value of —

we calculate for a sufficient number of values of X the ordinates of

the curve $i(X) corresponding to this value of — ; we use the formulas

(9), (10), and (11), bearing in mind that when there is a fully illumi-
nated field $i(X) = 1 in this field. Thus the curves of Fig. 10 are
plotted. The ordinates of these curves are multiplied by cos KirX
and the curve fy(X) cos KtrX is plotted. Then only the area which
it defines above the axis of X remains to be measured.

Jan., 1932]

LIGHTING OF SOUND FILMS

The curves shown in Fig. 1 1 which represent the efficiency m as a
function of K have been constructed in this way, each one being

plotted for a given value of — . The uppermost curve corresponds

to a perfect focusing (d = 0). We have already examined it in the
preceding paragraph.

0.9

0,8

0,7

0.5

0.3

.0,2

0.1 -

FIG. 12. Curves derived from those of Fig. 11, showing the variation of the
efficiency with the focusing, i. e., with d/d0 the ratio K remaining constant.

The curves shown in Fig. 12, which are derived from those in Fig.
11, indicate how the efficiency varies when the focusing is varied,

that is, — , the ratio K remaining constant as for a given illuminating
d

apparatus and a given film.

22 Louis DUNOYER [J. S. M. P. E.

It appears from the curves in Figs. 11 and 12 that the poorer the

/ d\

focusing is ( large value of — J, the more rapidly the efficiency de-

V do/

creases when the ratio between the width of the exploring zone and
the period of the transparency decreases.

The analysis of these curves clearly demonstrates the great im-
portance of the focusing. Let us assume, for instance, that the
note recorded on the film is 7^7 (4698 vibrations) and that the width

of the exploring zone is a quarter period, K = - (I — 0.097 mm.,

e = 0.024 mm.). We have seen already that for perfect focusing
the sound performance will be 90 per cent. If we assume that the
aperture of the illuminating pencil is 60 degrees (D = L), we have
do = e = 0.024 mm. An error in focusing of only 0.1 millimeter

d
will give — = 4. The curve in Fig. 1 1 corresponding to this value

d d° 1

of — shows for the abscissa K = - that the sound efficiency drops
do

to 17 per cent, that is, less than one-fifth of the value which it had
with perfect focusing.

8. Remarks on the Case when the Explored Zone Covers Several
Periods of the Transparency. — Each one of the curves in Figs. 11
and 12 is limited to an arc comprised between the efficiency limit
1 and the efficiency zero. They could be extended beyond this.

According to formula (17) for a given value of — , m starting from 1

for K = 0 decreases when K increases and reaches the value 0.

For — = 4, for instance, we have m = 0 for - = 0.31. If- increases
do II

still more, m becomes negative. This is not surprising when we
consider the formulas (13), (14), and (15). Formula (13) particu-
larly shows that if - or K exceeds the first value for which the ampli-
tude A is zero, the sign of the latter changes, that is to say, the
emerging luminous flux still has maxima and minima but in phase
opposition with the transparency of the film at the point where it
meets the optical axis. ' The efficiency then is. equal to the absolute

value of m. When - continues to increase, this absolute value goes

Jan., 1932]

LIGHTING OF SOUND FILMS

through a maximum, returns to zero, then again assumes positive
values and so on.

It is clear that the successive maxima always are decreasing.
Let us assume that d = 0 (exploring zone on the film), for example.

The efficiency m decreases from 1 to zero when - increases from 0 to 1.

If the exploring zone is further enlarged, the emerging flux begins
to fluctuate again; their amplitude will be maximum when the
exploring zone covers P/2 period on the film; the emerging flux
will be maximum when the optical axis goes through the film in a
minimum of transparency as shown in Fig. 13 (a) ; it will be minimum

Optical axis

FIG. 13. (a) Showing how the emerging flux is a maximum when the
optical axis goes through the film in a minimum of transparency, and
(6) how it is a minimum when the axis goes through a maximum of trans-
parency.

when the axis goes through a maximum of transparency (Fig.
The fluctuation again will be zero when the exploring zone covers
two periods of transparency; then if the exploring zone is further
enlarged, other fluctuations will result with a maximum when it
covers an odd number of half-periods and again become zero when
it covers a whole number of periods. It is also clear that the dif-
ference between the maximum and the minimum of the flux, that
is, the amplitude of the fluctuations, will decrease when the num-
ber of periods of the transparency simultaneously involved increases.
Consequently an adjustment of the focusing and the width of the
exploring zone, which produces an efficiency equal to zero for a given

24 LOUIS DUNOYER [J. S. M. P. E.

frequency, will produce an efficiency which differs from zero for
higher frequencies. In the case of perfect focusing, for instance,
the efficiency according to formula (8) will have maxima for:

c = 3 5 7 2n + 1

I ~ 2' 2' 2 2 '

and the values of the efficiency will be, respectively:

I - °'21- I - °'13' T, ~ °'09' I - 0'07

The corresponding frequencies will be given by formula (1) when
e has been chosen. An exploring zone of 0.1 millimeter would give,
for instance: for N = 4550, 9100, 13,650, etc., an efficiency = 0;
and for N = 6830, 11,390, 15,900, 20,450, etc., efficiencies equal to
21, 13, 9, 7 per cent, etc.

This example clearly shows the disturbance or unbalance which
a slightly wide exploring zone can produce in a symphonic reproduc-
tion even with perfect focusing. Such disturbances are inadmissible.
For this reason we have in paragraphs 6 and 7 systematically limited
the investigation of the efficiency to the range between its maxi-
mum limit and its first zero minimum. The remark, which we just
have made, should be borne in mind and it no doubt could explain
certain sound distortions actually observed.

9. Effect of a Lack of Uniformity in the Illumination of the Ex-
ploring Zone. — Heretofore we have assumed that the luminous
flux flowing through the exploring zone was the same at every point.
With some methods of illumination this is not so: for example, if
the image of a spiral incandescent filament with the turns spaced
too far apart is formed in the plane of the exploring zone, or if an
error in centering the optical parts causes a lack of symmetry in the
illumination of this zone, or if this zone is the image of a slit whose
edges are not parallel or if this slit is partially closed.

This lack of uniformity presents the greatest inconveniencies
for constant density films. If we assume, to consider the extreme
case, that only a part of the width of the sound band (Fig. 14) is
swept by the exploring zone, it is clear that only the peaks corre-
sponding to the most intense vibrations produce fluctuations in the
flux transmitted through the film and that the resulting sound will
only be remotely related to the recorded sound. Without going to
this extreme, it is clear that any lack of uniformity in the illumi-

Jan., 1932]

LIGHTING OF SOUND FILMS

Sound band

nation of the exploring zone will favor certain parts of the recording
curve and consequently certain sounds at the expense of others.
A more or less strong sound distortion will take place.

On the other hand, this lack of uniformity has no importance for
variable density films since the transparency of the film is the same
on its entire width. Then only a difference in height between the
ends of the exploring zone is detrimental (edges of slit not parallel),
but less, of course, than if this zone had the height of the widest end
on the entire width of the band.

For an equal height of the exploring zone the variable density
films, therefore, are much less sensitive to the im-
perfections of the illuminating apparatus than
the constant density films.

10. Resume of Part I. — The essential points of
the analysis which we have outlined above are
summed up in the formulas (1), (2), (3), (4), (8),
(9), (10), (11), (16), and (17). We have intro-
duced the important idea of the sound efficiency
of an illuminating device connected to a film
which is supposed to be perfect, in the same way
as all the devices which actually transform the
fluctuations of the luminous flux passing through
the film into sound vibrations are supposed to be
perfect. The variations of the sound efficiency
as a function of (1) the ratio between the width
of the exploring luminous zone and the period of
the transparency on the film and (2) the focusing,
are represented by the curves in Figs. 11 and 12
which allow of determining primarily the efficiency
under any given practical circumstance. Not only
do they admit of calculating this efficiency for a pure sound (sinusoidal)
of a given period but in the case of a fundamental sound accompanied
by various harmonics they also admit of calculating the ratios in
which these diverse composing vibrations will be reproduced. Thus
they completely solve the problem of the sound distortion produced
by a given illuminating apparatus.

These curves particularly demonstrate how rapidly the efficiency
decreases as a result of an error in focusing or the widening of the
exploring zone. We shall apply the results of Part I of our paper
to the investigation of different types of lighting apparatus.

Exploring
zone

FIG. 14. Illustrat-
ing the case where
only part of the
sound band is swept
by the exploring
zone, due to un-
uniform illumina-
tion of the zone.

26 Louis DUNOYER [J. S. M. P. E.

PART H
DIFFERENT LIGHTING APPARATUS

In section 3 of Part I we enumerated succinctly the different
types of lighting apparatus in order to .base the theoretical investi-
gation of the lighting of sound films on sufficient concrete data.
In Part II we shall not describe them in detail but study their opera-
tion.

11. Apparatus with Slit near the Film. — The simplest method
of powerfully illuminating a very narrow zone of the sound band
evidently is to place the film on the sides of a very fine slit and illumi-
nate the latter strongly. To produce this illumination a source
could be provided, so extensive, or placed so near the slit, that the
angle under which the center of the latter is seen would be very
great. If the source itself is very brilliant, it is clear that a more
intense illumination is obtained in this manner than with any opti-
cal system. The apparatus at the same time would be extremely
simple.

The available sources of great brilliancy, however, have a high
temperature and cannot be placed sufficiently near the film. A
condenser (Fig. 4) must be used which concentrates the light on the
slit in forming a more or less good image of the source on the slit.
If this image were perfect and the condenser did not absorb light,
each of the surface elements of the image would have a brilliancy
equal to that of the conjugate surface element of the source. In
reality the reflection on the glasses of the condenser, the absorption
of light by the latter, and the aberrations of this apparatus, which
in general are considerable, materially reduce the effective brilliancy
of the image formed on the slit. On the other hand, when the source
is a spiral filament, as usually is the case, the aberrations have the
effect of making the illumination of the slit uniform, a fact which
is valuable for a constant density film. (See §9.)

The very serious disadvantage of these devices is that if the film
rubs even very slightly on the edges of the slit, it is rapidly scratched
and, besides, the slit soon is closed by dust. This dust cannot be
avoided even if the sides of the slit are polished mirror-like and
curved inward so as to touch the film only at points separated
somewhat more than the width of the slit. Irregularly accumulated,
it also can be carried along suddenly; thus it causes sudden varia-
tions of the luminous flux illuminating the film and hence inad-
missible interfering noises.

Jan., 1932] LIGHTING OF SOUND FILMS 27

In order to avoid them, a small space could be left between the
slit and the film. A considerable decrease of the sound efficiency
would result, however. In reality, we noted in §4 that the critical
distance dG is the of same order as the width of the slit if the angular
opening of the pencil which illuminates it is slightly large, as is
necessary (24 degrees at least). If quite high frequencies are to be
explored when the film is illuminated with sufficient intensity, dQ
never will be far from 0.02 to 0.05 millimeter. If the film does not
touch the sides of the slit, the air current which it produces near
its surface and its electrification carry along atmospheric dust parti-
cles which also adhere to the edges of the slit unless the space be-
tween the latter and the film is sufficiently large. We consider a

space d of 0.1 millimeter as a minimum. The ratio — , therefore,

d d°

will be comprised between 5 and 2. If — = 4, for instance, the ef-

ficiency, which is equal to 80 per cent when the height of the slit is
0.1 per cent of the period of the transparency of the film, drops to
17 per cent when the height of the slit reaches one-quarter of a period,

and to 0 when - = 0.32. That is to say, with a slit of 0.025 milli-
meter, a space d = 0.1 millimeter, an angular opening of the illumi-
nating pencil of 45 degrees (dQ = 0.025 mm.), the efficiency is 80
per cent for a frequency of 1820 (approximately la& sharp), drops
to 17 per cent for a frequency of 4550 (approximately sharp do^
and to 0 for a frequency of 5820.

The dust could be avoided entirely by leaving only a space of less
than 0.1 millimeter between the film and the slit but this makes the
slit somewhat complicated. It could be covered with a film. In
order that neither be scratched no other material than glass should
be chosen. Thus a thin glass plate is attached to the sides of the
slit by means of a suitable adhesive (Canada balsam, for instance);
then the external face of this plate is ground by processes ordinarily
used by opticians in order to reduce its thickness as much as possible.
We do not know whether the system has been employed effectively
for illuminating films during reproduction, but an entirely similar
device is used particularly by the Fox Movietone for the sound re-
cording of the film. The distance between the slit and the film
thus can be reduced to a few microns.

12. Apparatus with Projected Slit, — In this type of apparatus

LOUIS DUNOYER

[J. S. M. P. E.

an image of the slit is formed on the film by means of an objective
(in general, a microscope objective). Since this image is smaller
than the slit, the latter may be wider. Being separated from the
film by the objective it can be placed in a closed space. For these
two reasons dust is much less to be feared. The slit is illuminated
by means of a lamp with a spiral filament, of the automobile head-

FIG. 15. A reproduction, enlarged approximately 14 times, of the
image formed on the film in a high-grade apparatus with a slit which
had been used only a short time. Note the breaks in the image caused
by dust in the slit.

light type, with a condenser interposed. Let us analyze more closely
the conditions of construction which obtain for this type of apparatus.
First, the necessary ratio between the length of the image (or
exploring zone) and its width requires a very fine slit. It must
not be forgotten that the width of the sound band is three milli-
meters and that the height (or width) of the exploring
be approximately 0.02 to 0.05 millimeter, the latter dimension,
moreover, being too large and acceptable only as a makeshift. Hence

FIG. 16. One means of lighting the slit consists in
forming an image of the filament on the slit by means of a
condenser.

the slit itself should be from 100 to 150 times longer than it is wide.
A slit of 0.5 millimeter should be from 50 to 75 millimeters long,
which is difficult to employ owing to the space required and the
difficulty of illuminating it sufficiently. In reality we have to use
slits which are 0.1 millimeter wide and consequently 10 to 15 milli-
meters long.

Jan., 1932]

LIGHTING OF SOUND FILMS

Experiments show that dust from the air very easily clings to the
edges of a slit 0.1 millimeter wide. Fig. 15 is the reproduction,
enlarged approximately 14 times, of the image formed on the film
in a high-grade apparatus with slit
which has been in use only a little
as yet. We note, however, that
this image is cut six times by dust
which has fallen on the slit. The
latter was 12.5 millimeters long and
0.1 millimeter wide; a microscope
objective formed an image of it 4.6
times smaller. The exploring zone thus was 0.022 millimeter wide.
In order to obtain the photograph reproduced in Fig. 15 this image is
retaken by means of a photographic objective.

In Fig. 15 we also find that the illumination of the image is little
uniform from one end to the other. The centering had been par-
ticularly careful, however. The slightest lack of adjustment in-
creases this lack of symmetry materially, which depends on the
manner in which the slit is illuminated.

Among the methods of lighting the slit, two should be particu-

FIG. 17. Photograph of a coiled
filament lamp, showing variations of
brightness between turns.

FIG. 18. Another method of illuminating the slit
consists in forming the image of the filament on the inlet
pupil of the microscope objective.

larly considered. One consists in forming an image of the incan-
descent filament on the slit by means of the condenser (Fig. 16).
The other (Fig. 18) consists in forming the image of the filament
on the inlet pupil of the microscope objective; in the very same
manner as when the image of a slide is to be projected on a screen,
the image of the source of light is formed on the projection objec-
tive by placing the slide very close to the condenser.

In the former case the illuminated part of the film is the image
of the part of the conjugate source of light of the slit in regard to
the condenser. The illumination of the exploring zone, therefore,

30 LOUIS DUNOYER [J. S. M. P. E.

is uniform only if the brilliancy of the source is uniform within this
part of the source. This is not so when it consists of a helical fila-
ment even when the turns are so close that the images of the back
half-turns are formed between those of the front half -turns because
the temperature of the internal surfaces of the turns is higher than the
temperature of their external surfaces. This is illustrated in Fig.
17 which reproduces the photograph of the filament of the lamp
(automobile headlight type) used in the apparatus with slit to which
Fig. 15 refers also.

On the other hand, in the case when the image of the source is
formed through the slit on the microscope objective, the illumi-
nation of the exploring zone, image of the slit, is perfectly uniform.
In reality, each point M of the exploring zone is illuminated by the
continuous flux in a cone whose peak is the conjugate point M'
of M on the slit and the base on the objective 0 is the part of the
image of the filament limited by this objective (or more accurately,
by its inlet pupil).

Evidently the former method of illumination is inadequate for
illuminating a constant density film since for a film of this type the
illumination of the exploring zone should be uniform (§9). This
condition is not important for a variable density film. Neverthe-
less, the latter method of adjustment in general is used even for
these films. In order that the total flux which falls on the film
then can be as great as with the former method of adjustment, the
microscope (or more generally projection) objective must be covered
entirely by the image of the filament which the condenser produces.

In order to calculate it we shall neglect the losses of light due to
absorption, reflection, and diffusion through the lenses and indicate
the brilliancy of the source by B. In the first mode of adjustment
(Fig. 16) the flux falling on the film is expressed by:

Gi^B-hS— (18)

where kiS is the surface of the part of the image of the source within
the slit, with surface S, S the surface of the entrance pupil of the
objective 0, and q the distance from the slit to this objective. In
the second mode of adjustment (Fig. 18) the expression of the flux
falling on the film is:

Jan., 1932] LIGHTING OF SOUND FlLMS 31

when kzS is the part of the surface of the objective covered by the
image of the source which is formed on it by the condenser. The
problem then is to know whether k\ is larger or smaller than k2. If
the turns of the filament are sufficiently close so that the image of
the back half-turns is projected between the images of the front
half- turns, the slit is covered completely and we have ki = 1. The
flux $2, therefore, cannot be greater than the flux $1 but it can be
equal to it if kz — 1, that is to say, if the entrance pupil of the pro-
jection objective is completely covered by the image of the filament
formed by the condenser.

Nevertheless, much light is lost with both methods. In the
former method all the light is lost which forms the part of the image
of the filament outside the slit and all the light which, after having
passed through the slit, does not reach the objective 0, the angular
aperture of the condenser in general being larger than that of the
projection objective. In the latter method of adjustment all the
light is lost which does not go through the slit and, besides, all
the light which, having passed through the slit, will pass through the
points of the image of the filament located outside the entrance
pupil of the projection objective.

From the above it results that the devices with projected slit
always utilize only a small portion of the light projected by the
source on the condenser and that their efficiency is low. Numerical
data bearing on this subject will be given later on.

13. Apparatus without Slit, with Cylindrical or Cylindrospherical
Lenses. — In order to avoid the disadvantage of dust on the slit, a
disadvantage which the devices with projected slit do not avoid
completely as we have just observed, and the losses of light involved
in these devices, two solutions have been proposed. The first one
is the solution which we have adopted and with which we shall con-
clude this article. But, first, we shall describe briefly the other solu-
tion which is based on the use of cylindrical or cylindrospherical
lenses.

It is well known that the luminous rays sent from a point located
on the axis of such a lens, after having passed through it, strike two
perpendicular focal lines, one of which is parallel to the generators
of the cylinder. A spiral filament whose axis is parallel to the genera-
tors also concentrates the light on two luminous bands, of which
the one which is nearest the lens, and consequently is the narrowest,
also is parallel to the generators. This small luminous band may

32 Louis DUNOYER [J. S. M. P. E.

replace the slit, and a projection objective (in general, a microscope
objective) forms an image of it on the film.

Since the rays issued from every point of the source outside the
plane of the principal section which is perpendicular to the genera-
tors and the plane of the principal section parallel to the latter pass
through different points of the focal distance, the distribution of the
flux in the small luminous band furnished by the cylindrospherical
condenser is independent of the distribution of the intensity in the
source and very uniform on the entire useful length of this band.
This fact as well as the elimination of dust makes the device very
attractive for constant density films.

Its disadvantage seems to be the difficulty of obtaining a sufficiently
fine exploring zone for reproducing high frequencies with a sufficiently
intense useful flux. If the small luminous band produced by the
cylindrospherical condenser at the slit of the apparatus with the pro-
jected slit examined above is substituted, we note that this band
should be 12.5 millimeters long and only 0.1 millimeter wide on the
entire length. It seems difficult to obtain such an image with a
spiral filament, which always would have a diameter of 2 to 3 milli-
meters at least, and with beams which should have a numerical
aperture of at least 0.2. If it is possible to correct the aberrations
in the plane of the principal section which is perpendicular to the
generators in such a way that the image, furnished by the rays com-
prised in this plane would be only 0.1 millimeter wide, it seems
improbable that the rays which are not comprised in this plane
and are oblique to the generators could be regulated with the same
precision.

14. Apparatus without Slit, with Rectilinear Filament Lamp. —
There is an extremely simple means of avoiding the dust, the losses
of light on the sides of the slits, the aberrations of the cylindrical
lenses, and the lack of uniformity of the brilliancy on the spiral
filaments. It consists in constructing the lighting apparatus with
a lamp which has only one filament set up rectilinearly and forming
a reduced image of this filament on the film. If the filament is very
fine and the aberrations of the optical system which forms its image
are well corrected, it is clear that the exploring zone will be as fine
as the separating power of the objective employed will permit.
There is nothing to prevent it from being reduced to the width of
the finest details which can be detected by a microscope, that is,
less than 1 micron. According to §6 such a fine exploring zone

Jan., 1932] LIGHTING OF SOUND FlLMS

H K

FIG. 19. Lighting apparatus with rectilinear filament for sound
films. A, lamp; B, rectilinear filament; C, closing mirror soldered
into the wall of the lamp; D, elastic plate used for guiding the
lamp when introduced in the apparatus; E, objective; F, ventila-
tion hole; G, window for letting out the luminous pencil, diminating
interfering reflections; H, swivel joint; /, socket of swivel joint;
/, lamp holder tube for orientating the filament; K, covers of
spikes of current supply; L, tube forming the body of the appara-
tus; M, tightening screw making the tube /immovable; N, screw
for centering the filament; 0, focusing ring.

LOUIS DUNOYER

[J. S. M. P. E.

makes it possible to reproduce vibrations of 120,000 cycles with an
efficiency of 90 per cent, that is, well above the audible range. More-
over, since all the light falling on the entrance pupil of the projec-
tion objective is used to form the image, except for losses in this
objective, the optical efficiency of the apparatus will be excellent.
Eliminating the condenser will reduce the losses of light still more.
Figs. 19 and 20 show the first model of the apparatus which we

had made according to this principle
by the Societe S. C. A. D. and the
lamp which it contains.

The filament of this lamp, which is
of tungsten 25 mm. long with a di-
ameter of 0.1 mm., is stretched be-
tween two metallic rods which con-
duct the current. The difficulty of
constructing this lamp lies in the
choice of the metal constituting these
rods and in the tension which the fila-
ment should have. If it is stretched
too much, it breaks at a high tem-
perature ; if it is not stretched enough,
it does not stay rectilinear when it is
brought to incandescence. Methods
of construction have been perfected
so that the filament remains perfectly
rectilinear at its normal operating
temperature (2290°K.) and at the
same time its tension is low enough
not to jeopardize its life, which is
several hundred hours.

In order to obtain a rectilinear
image of the filament and a careful

correction of the aberrations, the rays employed must go through
the walls of the bulb under conditions which are known perfectly.
For this purpose the bulb of the lamp consists of a glass cylinder
closed at the end opposite the base of the lamp by a lens with parallel
faces ground optically and of a known thickness. By means of a
special method this lens is fused into the walls of the cylinder, the
deformations resulting from the junction not extending to the central
part through which the useful rays are passing.

FIG. 20. Photograph of lamp
used in apparatus shown in
Fig. 19.

Jan., 1932] LIGHTING OF SOUND FlLMS 35

These rays fall upon a photographic objective of the anastig-
matic triplet type. An objective of this type has been preferred
to a microscope objective because the enlargement which is to be
obtained is approximately */4 in order that the exploring zone shall
be 3 mm. long and 0.0125 mm. wide, which is considered sufficiently
small at the moment. This width in reality is half as great as the
width of most apparatus with a slit and still gives a sound efficiency
of nearly 70 per cent (see §6) for the frequency 20,000, whereas
twice the width gives an efficiency of only 13 per cent for the same
frequency. The microscope objectives corrected for this magnifica-
tion and an object field exceeding 3 mm. in general have a numerical
aperture which is lower than 0.15, whereas anastigmatic triplet lenses
are found to have an excellent definition in a field exceeding by far
3 mm. and having a numerical aperture of 0.25 at least. Moreover,
with such objectives the distance separating the last lens from the
film is greater than with ordinary microscope objectives, which also
may be considered an advantage.

FIG. 21. The image of the filament.

Fig. 21 shows the image, on the same scale as the image of the
slit in Fig. 15, of the image of the filament as received by the film.
It can easily be proved that this image is perfectly rectilinear. The
brilliancy of the exploring zone then will be perfectly uniform ex-
cept in the immediate vicinity of the ends, whereas it is far from
uniform with the apparatus with the slit, although adjusted with
care, as shown in Fig. 15.

Naturally it would not be difficult to reduce the width of the ex-
ploring zone still more either by using a finer filament of the same
length with the same objective and the same magnification or by re-
ducing the magnification and increasing proportionally the length of
the filament, which is assumed to have a constant diameter.

In the model shown in Fig. 20, the socket of the lamp is provided
with spikes which sink into an insulating piece provided with a
swivel joint H whose case is in one piece with the plug K fixed on
the tube / which can rotate with slight friction against the external
tube L. By turning the plug K the filament can be made hori-

36 Louis DUNOYER [J. s. M. P. E.

zontal (the film being unwound vertically). Four screws N whose
points rest on the walls of the bulb through springs D make it possible
to center the filament in such a manner that the exploring zone
sweeps the sound band exactly. The blades of the spring also
facilitate the introduction of the lamp mounted on the plug K in
the tube L.

The pencil illuminating the exploring zone goes through the small
window G without touching its edges, this window being used only
to eliminate the interfering light reflected on the internal walls of
the tube L.

Focusing on the film is performed by means of the ring 0 which
displaced the objective or an element of the objective by small
amounts.

15. Results of Experiments. — In addition to the satisfactory
results obtained in operation with this apparatus we have tried
to determine the total luminous flux which it sends through the
exploring zone and its energy output, that is, the number of watts
consumed to obtain this luminous flux. For comparison this in-
vestigation was conducted also with the apparatus with slit, already
mentioned.

Either of these two lighting devices was fixed on a rotating sup-
port, the axis being vertical, so that the entire emerging flux was
received by a photoelectric cell (hemipherical S. C. A. D. cell), and
by one rotation of the support this flux could be substituted instan-
taneously by the flux from a standard lamp mounted on a rack
support. The cell as usual was connected to a battery and a galva-
nometer. The lamps, that of the investigated lighting apparatus
as well as the standard lamp, were supplied by a storage battery
with potentiometers to regulate with precision the current in the
lamps, standard ammeters to measure it and standard voltmeters
to measure the voltages at the terminals of the latter. By experi-
menting, the distance of the standard lamp was regulated in such
a manner that the deflection of the galvanometer was the same
when the cell received the flux leaving the lighting apparatus or
that from the standard lamp limited by a diaphragm of a known
surface placed on the cell. The standard lamp was supplied in
such a manner that its potential at the terminals was its standard
potential, 102.9 volts; its luminous intensity then was 20.2 candles.

The results of a measurement made on our apparatus and on the
apparatus with the slit are as follows:

Jan., 1932] LIGHTING OF SOUND FlLMS 37

Apparatus Apparatus

L. D. with Slit

Current in the lamp, amperes 1.5 5.5

Voltage at the terminals, volts 3.44 7.69

Power consumed, watts 5.16 4.24
Diameter of the diaphragm placed on the cell illuminated

by the standard lamp, centimeters 2.535 2.043
Distance of the standard lamp for the equilibration,

centimeters 74.3 67.3

20.2 • x • 2.S352

Flux leaving the lighting apparatus

4-74.S2

20 2 • 7T • 2 0432

(L.D.); - — (with slit); lumen 0.0185 0.0146

4-67.32

Before we conclude this paper we wish to make the following re-
marks in regard to these results:

(1) Variations in the centering of the lamp of our apparatus pro-
duce no effect on the emerging flux, provided, of course, that the
decentering is not so great that the pencil is partly hidden by the
edges of the window G. On the other hand, very slight variations
in the centering of the apparatus with the slit produce very great
variations of the emerging flux. This is readily understood since the
region of the spiral filament is varied, the image of which is formed
by the condenser on the microscope objective. The flux obtained
above is the maximum flux which we have been able to produce;
a very slight irregularity which leaves an excellent centering upon
examining the pencil makes the flux drop to 0.0122 lumen.

(2) If the energy output is expressed by the number of lumens
emitted in the emerging pencil for 1 watt consumed, it is clear that
the apparatus with rectilinear filament, which already is superior
to the apparatus with slit in absolute value, is far superior in regard
to efficiency. Its efficiency is 0.0036 lumen per watt, whereas the
efficiency of the apparatus with slit is one-tenth as great, that is,
0.00034 lumen per watt. Such a result could be expected owing
to the losses of light avoided in the apparatus with rectilinear fila-
ment. This advantage, which perhaps is little appreciated in the
present talking picture installations, evidently may attain great
importance.

(3) The efficiency of the apparatus with rectilinear filament would
be increased still further in regard to that of the apparatus with
slit if our lamp were as powerful as that of this apparatus. With
1.5 amperes the color temperature in the middle of the filament
of our lamp is 2290 °K., the output then being 7.75 lumens per watt

38 Louis DUNOYER

(at a reduced operation of 1.45 amperes, the color temperature drops
to 2250°K. and the output to 6.70 lumens per watt). At an operating
current of 5.5 amperes the color temperature of the spiral filament
of the apparatus with slit is 2570°K. on the internal surfaces of the
turns and 2400°K. on the external surfaces, the average output
being 15.4 lumens per watt. The lamp of the apparatus with slit,
therefore, is much more powerful than ours; its light is richer in
blue rays to which the cell is more sensitive. If our filament were
brought to the same temperature, although at the expense of its
duration, the efficiency of the apparatus measured as above would
be practically doubled since the output of the filament would be
increased from 7.75 to 15.4 lumens per watt.

THE RAPID RECORD OSCILLOGRAPH IN SOUND PICTURE

STUDIES*

A. M. CURTIS, T. E. SHEA, AND C. H. RUMPEL**

Summary. — This paper describes a special oscillograph which was designed for
making rapid records in sound picture studies. The oscillograph is briefly described,
and illustrations are presented of records obtained in making the following studies:
microphonic action of vacuum tubes; noise levels in amplifiers; investigations on
rectifiers; studies on light valve clash; action of the biasing current of light valves
as used in noiseless recording by the variable density method; acoustical studies
showing the rise and decay of transients; loud speaker selection with regard to load
carrying capacity and mechanical flutter investigations of reproducer sets.

The recording oscillograph, although an extremely valuable in-
strument, is not in general very popular with engineers. This is
especially true in sound picture work where time is often limited
and the minutes and sometimes hours which must elapse after the
oscillogram is taken and before it can be readily examined are a
serious drawback. In addition, most types of recording oscillographs
are found to be so insensitive over the frequency band used in sound
pictures that the information which they give is frequently unreliable.

About two years ago the Bell Telephone Laboratories, realizing the
limitations of the available oscillographs of the recording type, under-
took to design an instrument which would as far as possible avoid
these shortcomings. The instrument which was evolved is capable
of recording frequencies accurately up to 6000 cycles per second,
and can furnish a developed record almost immediately after the
oscillogram has been taken. Usually, therefore, oscillograms may be
taken as rapidly as the conditions under investigation can be changed,
and the results of the changes may be known at once.

The oscillograph illustrated in Figs. 1 to 4 may be divided into
two main parts, the galvanometer and the photographic mechanism.
The galvanometer is of the string type, and is not unlike the light
valve familiar to most sound picture engineers. There are, however,

* Presented at the Spring, 1931, Meeting at Hollywood, Calif.
** Bell Telephone Laboratories, New York, N. Y.

CURTIS, SHEA, AND RUMPEL

[J. S. M. P. E.

three independent strings which permit the observation and record-
ing of three simultaneous and separate phenomena with their phase
relations.

FIG. 1. Front view of the oscillograph.

A tungsten filament lamp and a simple lens arrangement magnifies
the motions of the strings and compresses their shadows to black spots
on a line of light which extends across the 35-mm. bromide recording
paper. This provides an oscillogram with white lines on a dark

Jan., 1932]

RAPID RECORD OSCILLOGRAPH

gray background. Means are also provided to photograph amplitude
and timing lines on the oscillogram if desired.

The photographic mechanism takes care of the exposing, develop-
ing, and fixing of the oscillogram. The exposing is done by passing the
paper through the line of light at the desired speed, using a system of
rollers rotated by the exposure motor. The paper, having been

FIG. 2.

Front view of the oscillograph with covers removed to show part
of the exposing and developing mechanism.

exposed, is passed down a small chute into the developer tank through
which it slowly travels by means of conveyor belts. From the
developer the oscillogram is led into the fixing bath and is then
passed out into a large fixing tank where it may be observed. Since
oscillograms are generally taken much more rapidly than they are
developed, a storage tank is provided into which the excess of ex-

42 CURTIS, SHEA, AND RUMPEL [J. S. M. P. E.

posed paper is passed, where it remains until led through the de-
veloper.

The process of taking an oscillogram is briefly as follows: the
sources of current which it is desired to investigate are connected to

FIG. 3. Rear view of the oscillograph.

the galvanometer strings and the controls are adjusted until suitable
deflections are obseryed on the viewing screen of the camera. The
motors are then started and an operating lever is pulled out. After
the deflection of the string images on the screen show that the ex-

Jan., 1932]

RAPID RECORD OSCILLOGRAPH

pected phenomenon has occurred, the operating lever is returned to
the normal position. The developed and fixed oscillogram begins to
pass before the operator's view about ten seconds later. It may then
be examined immediately, measured, and, if desired as a permanent

FIG. 4. Rear view of the oscillograph taken from another angle.

record, returned to the large hypo tank for a few minutes to complete
the processing.

A particular feature of this instrument is the sharp definition of the
string image, permitting accurate observations to be made with

CURTIS, SHEA, AND RUMPEL

[J. S. M. P. E.

deflections much smaller than are common with other types of
recording oscillographs. This allows the oscillograms to be enlarged
for analysis, and tracks having a height as great as four inches still
give sharply defined lines.

In order that the use of such an oscillograph in sound picture
studies may be illustrated, a number of oscillograms have been
prepared showing the application of this oscillograph to the solution
of problems which are continually being investigated so that the
sound picture may attain a greater degree of excellence. These

0, 45

0.50

1,25 1,30 2,45

TIME IN SECONDS FROM rMPACT-

2.50

FIG. 5. Oscillogram of microphonic response of vacuum tubes.

oscillograms are not intended to show complete results of the various
investigations but rather to point out how effectively this instrument
may be used.

1. Microphonic Vacuum Tube Studies. — Microphonic vacuum
tubes have imposed certain limitations both in recording and in
reproducing systems. Those tubes, commonly used at low levels
because of operating limitations, have been more microphonic than
the higher powered tubes. Fig. 5 shows an oscillogram taken to
illustrate the improvement which has been made in the microphonic

Jan., 1932]

RAPID RECORD OSCILLOGRAPH

response of a small vacuum tube as the result of studies which have
been carried on during the past year. In this oscillogram three tubes
were placed at the input to three amplifier channels having the same
gain, each channel terminating at one of the oscillograph strings.
The mounting upon which the three tubes were placed was given a
single rap, causing the microphonic response of the tubes as shown.
The relative freedom from microphonic effects of a recently produced
vacuum tube (Tube B) is easily seen from a comparison with the

A OVERLOADED WITH FILTER

FIG. 6. Oscillograms of single tube amplifier.

response of the earlier type of tubes recorded on the two outer strings
(Tubes A and C).

2. Amplifier Studies. — In recording, it is common practice to
operate a large number of recording amplifiers from a common "B"
battery. Figs. 6 and 7 show the noise level across this battery due to
a single amplifier, which may cause objectionable cross- talk in the
other amplifiers unless precautions are taken to reduce the effect.
Fig. 6 illustrates the effect in a single tube amplifier operated both
within its rating and overloaded; Fig. 7 shows the corresponding

CURTIS, SHEA, AND RUMPEL [J. S. M. P. E.

C WITHOUT FILTER A WITH FILTER

OVERLOADED

FIG. 7. Oscillograms of push-pull amplifier.

B BATT.
FIG. 8. Oscillogram of amplifier blocking.

Jan., 1932]

RAPID RECORD OSCILLOGRAPH

TOP BOTTOM
STRING STRING

FIG. 9. Oscillograms of rectifier characteristics.

FIG. 10. Oscillograms of rectifier flickering.

CURTIS, SHEA, AND RUMPEL

[J. S. M. P. E.

effects in a push-pull amplifier. In each case record A illustrates the
amplifier equipped with a simple filter composed of a single condenser
and a small inductance, and indicates the absence of noise at the
battery terminals. Records B and C show the presence of noise
when the filter is removed. It is particularly interesting to note the
presence of noise in the case of the push-pull amplifier as this is con-
trary to the fairly common supposition that such an amplifier is
totally without this effect. The tubes used in obtaining this record
were carefully selected for equivalence.

FIG. 11. Oscillograms showing light valve clash.

In a-c. operated amplifiers blocking is a familiar phenomenon
due frequently to a common plate impedance between the various
tubes. In certain cases, however, it is very difficult to determine
what is taking place. Fig. 8 shows a simple form of blocking in a
two-stage amplifier. A small change in the plate current in tube 2
causes a change in the plate current of tube 1 through the common
plate impedance RI. This resistance R\ represents the impedance
of the plate power supply which, in the case of an a-c. rectifier or a
run-down "B" battery, might be quite high.

3. Rectifier Investigations. — Investigations of rectifier characteris-

Jan., 1932) RAPID RECORD OSCILLOGRAPH 49

tics may be readily made with this oscillograph as is shown in Figs.
9 and 10. Record A , Fig. 9, shows the effect of working a gas rec-
tifier tube directly into a capacity, as compared with working into an
inductance (record B). This record was taken to determine the
magnitude of the current peaks for the two types of filter, in order to
determine the optimum condition for tube and condenser operation
and thereby assure maximum service from the equipment.

Fig. 10 shows a peculiar phenomenon found in certain full- wave
gas rectifier tubes when operating under light loads. The tubes may
be seen to flicker at various frequencies, as shown in the oscillogram.
In each case the upper trace is the input voltage of the filter, the
middle and lower traces the voltage between each plate and the
filament of the rectifier tube.

4. Light Valve Studies. — Figs. 11, 12, and 13 show how the
oscillograph may be used in studying the light valve used in the
variable density method of recording. In Fig. 11 light valve string
clash is shown. It may be seen how, while the bus voltage (input to
the power amplifier feeding the valve) and valve voltage are un-
affected, the output, as picked up by a monitoring photoelectric cell
placed back of the film, is considerably distorted, but on one side of
the cycle only.

Figs. 12 and 13 illustrate the action of the biasing current of
the light valve as used in noiseless recording by the variable density
method. Fig. 12, record A, shows the action which takes place dur-
ing attack or beginning of a sound wave. During this time the bias
as shown by the center trace is being removed, allowing the light
valve strings to resume their normal average spacing. This par-
ticular condition illustrates the effect of too slow a removal of the
biasing current causing the valve ribbons to clash, as indicated by the
irregularities in the wave shown on the bottom trace. Record B is
the same as record A except that speech is used to modulate the
valve instead of a single frequency. Fig. 13 shows the decay after the
input to the valve has died down. It may be seen that the bias is
placed upon the valve much more slowly than it is removed in order
that the low level portions at the ends of the various sounds as they
die down will not be cut off. The reason for this may be seen from
Figs. 14 and 15.

5. Acoustical Studies. — Figs. 14 and 15 are records of >sound
build-up and decay, and were taken by placing the input to a loud
speaker on the middle string or trace and picking up the sound by

CURTIS, SHEA, AND RUMPEL [j. s. M. p. E.

A SINGLE FREQUENCY 400^

VALVE VOLTAGE
,BUS VOLTAGE

MONITOR VOLTAGE

B SPEECH (SPOKEN NO* 4)

FIG. 12. Oscillograms of noiseless recording light valve bias (attack).

BUS VOLTAGE

CURRENT

MONITOR VOLTAGE

A SINGLE FREQUENCY 400^

fUS VOLTAGE^ /VALVE VOLTAGE

^MONITOR VOLTAGE

B SPEECH (SPOKEN NO. 4:

FIG. 13. Oscillograms of noiseless recording light valve bias (decay).

Jan., 1932]

RAPID RECORD OSCILLOGRAPH

FIG. 14. Oscillogram of sound growth.

B CONTINUATION OF RECORC

FIG. 15. Oscillogram of sound decay.

52 CURTIS, SHEA, AND RUMPEL [J. s. M. P. E.

means of two differently placed microphones whose amplified outputs
are shown on the outer two traces. The sound growth curve of Fig.
14 shows that in the case of a single frequency the sound builds up
to its normal value very rapidly but may then drop or rise slightly
depending upon the position of the microphone and the interference
patterns set up. Fig. 15 illustrates different ways in which sound may

FIG. 16. Oscillogram of loud speaker overload.

decay when the input to the loud speaker is cut off. These differences
are the results of interference, and it will be noted that they occur
during the interval immediately following the cut-off of energy to
the loud speaker. Being transient phenomena, the oscillograph
is well suited to study them and is particularly valuable as an instru-
ment supplemental to the reverberation meter, as by means of it

FIG. 17. Oscillogram of mechanical flutter.

particular portions of a decay curve may be studied in detail. How-
ever, because of the limited amplitude of the oscillograph record,
over-all reverberation times are most accurately measured by means
of a reverberation meter.

In selecting a loud speaker for a particular application it too fre-
quently happens that this selection is made on the basis of frequency

Jan., 1932] RAPID RECORD OSCILLOGRAPH 53

response characteristics, and that the load carrying capacity is en-
tirely neglected. The record of Fig. 16 was taken to illustrate this
point. The upper and lower traces of this record indicate the sound
outputs, as picked up by two similarly located microphones, of two
loud speakers each receiving the same input. Speaker 1 is obviously
overloaded.

6. Mechanical Flutter Investigations. — Fig. 17 shows how the
oscillograph may be used to assist in mechanical design. Three sepa-
rate sound film reproducers were set up, and the output of each
reproducer was put on one string of the oscillograph. The same
sound print of a thousand cycle film record was used on each of the
reproducers with the results as shown. The upper trace is the output
of a normal reproducer. The middle trace shows a reproducer having
too large a driving sprocket at the sound gate. The lower trace
shows the effect of having the driving sprocket slightly eccentric
producing ninety-six and six cycle flutter.

PHOTOGRAPHIC SENSITOMETRY, PART III*
LOYD A. JONES**

Due to its length, Mr. Jones' paper on sensitometry which was presented in part
on three consecutive days at the Spring, 1931, Meeting of the Society at Hollywood,
Calif., will be published in the JOURNAL in four issues. The following is the third
of the four installments. The paper deals in a tutorial manner with the general
subject of sensitometry, its theory and practice. The fourth installment will be
published in the March, 1932, issue of the JOURNAL.

OUTLINE

I. Introduction.

(A) Definition.

(B) Scope of field.

CD) The characteristic D-log E curve.

II. Sensitometers.

(.4) Light sources.

(1) Historical resume.

(a) Natural light (sunlight, skylight, etc.).

(b) Activated phosphorescent plate.

(d) The Hefner lamp.

(e) The Harcourt pentane standard.
(/) The acetylene flame.

(g) Electric incandescent lamps.

(2) Spectral composition of radiation.

(a) The spectral emission curve.
(6) The complete radiator.

(d) Effect of color temperature on sensitivity values

(3) Modern standards of intensity and quality.

(a) Acetylene flame plus dyed gelatin filter.
(6) Acetylene flame plus colored glass filter.

(d) Electric incandescent, plus colored filters

(4) The international unit of photographic intensity.
(B) Exposure modulators.

(1) Intensity scale instruments.

* Presented at the Spring, 1931, Meeting at Hollywood, Calif.
** Kodak Research Laboratories, Eastman Kodak Co., Rochester. N. Y.
54

PHOTOGRAPHIC SENSITOMETRY 55

(a) Step tablets (/ variable by finite increments).
(6) Wedge tablets (/ variable by infinitesimal incre-
ments).

(d) Tube sensitometer.

(e) Optical systems with step diaphragms.

(/) Optical systems with continuously variable dia-
phragms.
(2) Time scale instruments.

(a) Exposure intermittent.

Finite exposure steps (discontinuous gradations).
Infinitesimal exposure steps (continuous grada-
tions).
(6) Exposure non-intermittent.

Finite exposure steps (discontinuous gradations).
Infinitesimal exposure steps (continuous grada-
tions).

III. Development.

04) Developers.

(1) Standards for sensitometry.

(a) Ferrous oxalate.

(b) Pyro-soda.

(2) Standards for control of processing operations.
OB) Temperature control.

(1) For standardized sensitometry.

(2) For control of processing operations.

IV. The measurement of density.

04) Optical characteristics of the image.

(1) Partial scattering of transmitted light.

(2) Diffuse density.

(3) Specular density.

(4) Intermediate density.

(5) Relation between diffuse and specular values.

(6) Effective density for contact printing.

(7) Effective density for projection.

(8) Color index.

(B) Fog and fog correction.

(1) Source of fog.

(a) Inherent fog.
(&) Processing fog.

(2) Fog correction formulas.

(1) Bench photometer,
(a) Rumford.
(&) Bunsen.

56 LOYD A. JONES [J. s. M. p. E.

(2) Martens polarization photometer.

(a) Simple illuminator.

(&) Split beam illuminator.

(3) Integrating sphere.

(a) For diffuse density.

(b) For diffuse and specular density.

(4) Completely diffused illumination.

(a) For diffuse density.

(5) Specialized forms.

(a) Furgeson, Ren wick, and Benson.

(b) Capstaff-Green.

(d) Density comparators.

(6) Physical densitometers.

(a) Thermoelectric.

(b) Photoelectric.

V. Interpretation of Results.

04) Speed or sensitivity.

(1) Threshold speed.

(a) Scheiner speed numbers.

(b) Eder-Hecht.

(2) Inertia speeds.

(a) H & D scale.

(b) Watkins scale.

(3) Luther's crossed wedge method.

(4) Minimum useful gradient.

(B) Gamma infinity, 7^.

(D) Time of development for specified gamma.

(1) Td (y = 1.0).

(E) Latitude, L.

(F) Fog, F.

VI. Spectral Sensitivity.

(A) Dispersed radiation methods.

(1) Monochromatic sensitometers.

(2) Spectrographs.

(a) Ordinary.

(b) Glass wedge.

(B) Selective absorption methods.

(1) Tricolor.

(2) Monochromatic filters.

(3) Progressive cut filters.

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

V. INTERPRETATION OF RESULTS

Having now exposed the photographic material to a definitely
known quantity and quality of radiation, developed the exposed
material under standardized conditions, and measured the densities
resulting from the various exposures, it remains to interpret the
results thus obtained. As previously stated, some sensitometric
testing methods, such, for instance, as the Scheiner, Eder-Hecht,
etc., do not require the measurement of density, the result being
judged directly by inspection of the developed material. Much

1.6

I-Z

0.» I 1.1 f t.Q, 2.0 ZA Z.W

x tA L.OG. EXPOSURE: (r-vcs}

FIG. 38. Typical curve showing the relation between density and log

exposure.

more complete information may be obtained by methods involving
the measurement of density and subsequent analysis of the results.
For this purpose it is customary first to express the results in graphic
form, and then to read off directly or to compute, by means of pre-
viously established analytical relationships, the values of certain
numerical constants useful in specifying the characteristics of a
photographic material. The various graphic forms in which the
sensitometric data may be shown will now be considered, after which
numerical values derived therefrom and their significance for various
theoretical and practical purposes will be considered.

58 LOYD A. JONES [J. S. M. p. E.

By plotting density, D, as a function of the logarithm (to the base
10) of the exposure, logio E, a curve as shown in Fig. 38 is obtained.
This is the graphic form proposed first by Hurter and Driffield (loc.
cit.) for the presentation of sensitometric data and is therefore quite
commonly referred to as the H & D curve although the terms D-log E
curve and characteristic curve are frequently used in reference
thereto. It has been found experimentally that in the case of many
photographic materials a considerable portion of the D-log R curve
is represented satisfactorily, within the limits of experimental errors,
by a straight line. The limits of the straight line region are desig-
nated by the points, A and B. The exposure region covered by the
straight line portion of the characteristic curve is the region of correct
exposure since throughout this exposure range density is directly
proportional to log E. Therefore, for the correct proportional
rendering of the negative of the various object brightnesses, the
camera exposure must be adjusted so that only the straight line
region is used. For the fulfillment of this condition the minimum
density in the negative (corresponding to the deepest shadow in the
object) must not be less than that of point A and the maximum
negative density (corresponding to the highest light in the object)
must not exceed that of point B.

The relation between a given log E interval or increment, A log E,
and the corresponding density interval or increment, AD, is given by
the ratio AD/ A log E which is an expression of the average slope or
gradient, G, for the interval A log E. Since the gradient is not in
general constant, but changes continuously from point to point (for
instance, in the region between C and A), it is necessary in order to
express the gradient of the curve at any point to reduce the finite
increments A log E and AD to the corresponding infinitesimal incre-
ments d log E and dD. The gradient, G, at any point reduces
therefore to the differential form

G = dD/d log E

For the straight line portion, however, G is constant and may be
conveniently expressed in terms of the angle a subtended by the line
AB and the log E axis. The tangent of this angle is called gamma, 7.
For the straight line portion, therefore,

G = dD/d log E = constant = tan a — y
Thus, gamma is the proportionality factor giving the relation between

Jan., 1932] - PHOTOGRAPHIC SENSITOMETRY 59

a given log E difference, A log E, and the corresponding density
difference, AD.

AD/ A log E = y

Thus, if, in an object being photographed, two areas have brightnesses
of 10 and 80 units, the A log £ value becomes log 80 - log 10 = 0.90.
Now, if both are rendered on the straight line portion of the D-log E
curve and if 7 = 0.8, then

AD/0.90 = 0.8
and

AD = 0.72

If a = 45 degrees, tan a or 7 becomes unity and any log E increment is
rendered in the negative by an identical density difference. This is
the condition which must be fulfilled if it is desired to reproduce
exactly in the negative the brightness contrast in the object. If
gamma is less than unity, correct proportional reproduction will be
obtained but with compression of the brightness scale, while, if gamma
is greater than unity, correct proportional reproduction will also be
obtained but with expansion of the object brightness scale.

Since gamma is equal to the ratio of the negative density difference
to the corresponding log exposure difference, it is frequently used as a
means of expressing the contrast of the negative or of the photographic
material. It should be borne in mind constantly that gamma gives
information pertaining only to the straight line portion of the curve
and tells nothing of the contrast characteristics of other portions of
the D-log E curve. This sensitometric constant is of great value and
importance in both the theory and practice of photographic sensi-
tometry.

Projection of the straight line portion of the D-\og E curve on the
log E axis determines the log exposure range over which direct
proportionality between D and log E exists. By dropping perpendicu-
lars from A and B to the log E axis the points M and TV are established.
These fix the limits of this exposure range. The distance between
M and N is called latitude, L, and may be expressed either in log
E units or in exposure units. Thus,

Latitude, L = log En — log Em (Log E units)
or

Latitude, L = En/Em (Exposure units)

The value of latitude for any given D-log E curve determines the
maximum object contrast (ratio of maximum to minimum object

60 LOYD A. JONES [J. s. M. p. E.

brightness) which may be rendered with strict proportionality be-
tween density and log exposure on that photographic material
processed under the specified conditions used in obtaining the charac-
teristic curve. Latitude is not a constant for a given photographic
material, since its value depends profoundly upon the extent to which
development is carried and, to a lesser extent, on other processing
factors. It depends also upon certain exposure conditions, such as
the quality (spectral composition) of the exposing radiation.

The straight line, AB, extended cuts the log E axis at the point x
and the value of E at this point is called the inertia, i. Since a
material of low sensitivity has a high inertia value, and vice versa, it is
necessary to take the reciprocal of the inertia in order to obtain a
value which is directly proportional to sensitivity, hence

Sensitivity oo l/i

The absolute values obtained by taking the reciprocal of the inertia
may be inconvenient for practical purposes since they may be less
than unity, and hence expressible only as decimals or fractions. It is
customary, therefore, in setting up practical sensitivity or speed
scales to multiply this reciprocal by a constant, k, chosen more or less
arbitrarily so as to give a series of convenient numbers. In general,
therefore, speed is defined by the equation

Speed, S = l.k

The values of k commonly used will be discussed later.

Now, from point A (Fig. 38) the ZMog E curve continues to the left
into the region of decreasing exposure with constantly decreasing
gradient, G, until at the point C this gradient becomes zero (G = 0),
that is, the curve becomes parallel to or, if proper correction for fog
has been made, coincident with the log E axis. This region, C to A ,
is called the region of underexposure or sometimes the toe of the
characteristic curve. vSince the gradient, dD/d log E, decreases
progressively from A to C, it follows that the density difference, AD,
corresponding to a given small A log E, decreases continuously as the
exposure is decreased, becoming zero at the exposure value correspond-
ing to the point C. Thus, the power of the photographic material to
show detail due to brightness differences in the object becomes less
and less throughout the underexposure region vanishing entirely at an
exposure value corresponding to the point C.

From the point B, the upper limit of the straight line, the curve

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

continues to the right into the region of increasing exposure with a
constantly decreasing gradient until at the point D the gradient
becomes zero, that is, the curve becomes parallel to the log E axis.
The value of density corresponding to the point D is the maximum
density, Dmax, obtainable with the specified processing conditions,
development time, developer constitution, temperature, etc. Its
value is not fixed entirely by these processing factors, but depends to
some extent upon the quality of the exposing radiation. This region,
B to D, is called the region of over exposure, or sometimes the shoulder
of the characteristic curve. Here, as in the underexposure region, the

z.o

1.2

0.8

0.4

OA o.a

1.7. I. ft

L.OC..

Z.O

X L.OC.. E-XPOSURe. (M.C.ft.^

FIG. 39. D-log E curves obtained with development times of T and 2T.

density difference, AZ2, corresponding to a small log £ difference A log
E, decreases progressively with increasing exposure and becomes zero
at point D. Thus, the detail rendering power decreases progressively
with increasing exposure and vanishes completely at the exposure
value corresponding to point D.

Points C and D, therefore, represent the limits of the exposure range
within which the material is capable of rendering an object brightness
difference by some density difference although near the limits
(points C and D) this may be negligibly small, even for very great
object brightness differences. This exposure range is termed the

LOYD A. JONES

fj. S. M. P. E.

total scale of the material and may be expressed either in log E units or
as the ratio of the limiting exposures. Thus,

Total scale = log EL — log EC (Log E units)
or

Total scale = EL/EC (Exposure units)

The latter form is more commonly used and is perhaps better for most
purposes, since it is more directly interpretable in terms of the ratio of
maximum object brightness to minimum object brightness, which is
the form in which data relative to the brightness of the object are
usually available.

The shape, and frequently the position of the characteristic curve,

!•*

o.e>

z e>

%<b

FIG. 40. Family of Z>-log E curves illustrating the approach to 7 oo for in-
creasing development time.

depends upon the development conditions. A simple case is illus-
trated in Fig. 39 in which curve No. 1 represents the P-log E charac-
teristic obtained for a development time of T, while curve No. 2 is
that obtained for a development time of 2T. In this particular case
the straight lines intersect at a point x lying on the log E axis. There-
fore, the inertia value is the same for both times of development, and
speed expressed in terms of inertia is the same for both curves. This is
by no means true for all materials and all processing conditions, since
in many cases the intersection point of the straight line portions is

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 63

found to lie either above or more frequently below the log E axis.
Since inertia is defined as the value of exposure where the straight line
extended cuts the log E axis, it follows that the value of the speed,
based upon inertia, is not the same when determined from curves
representing different developing times unless their intersection point
lies on the log E axis. Angle a' is appreciably greater than a , hence
gamma increases as development time is lengthened. This is true in
practically all cases except when development is forced to such an
extent that excessive fog is produced which may cause a decrease of
gamma with development time. Such conditions are rarely met in
practice and hence the statement that a increases with time of
development is for all practical purposes a correct generalization.
Projection of the straight line portion, A ' to B ', of curve No. 2 on the
log E axis is appreciably shorter than that of the similar region, A to
B, of curve No. 1, hence in this case latitude has decreased with the
increasing time of development. The curves in Fig. 39 do not extend
sufficiently far into the region of increasing exposure to show the final
values of Dmax, but it is quite evident that the value of Dmax in-
creases with development time.

A somewhat more complete picture of the change in the shape of
the D-\og E curve is shown in Fig. 40 in which the curves numbered
from 1 to 6, inclusive, represent the data obtained from sensitometric
strips developed for 2, 4, 6, 8, 10, and 12 minutes, respectively.
Values of 7, AY, L, and i for these various times of development are

TABLE XI

Data Derived from Fig. 40

0.50

1.94

0.10

0.80

0.30

1.74

0.10

1.02

0.22

1.56

0.10

1.14

0.12

1.40

0.10

1.20

0.06

1.34

0.10

1.24

0.04

1.30

0.10

1.30

...

0.10

shown in Table XI. For the two-minute development time a gamma
of 0.50 is obtained. At 4 minutes gamma is equal to 0.8, an increase
of 0.30. Increasing the time of development by another 2 minutes
gives a gamma of 1.02, an increase of 0.22. For each successive two-
minute addition to the development time the increase in gamma be-

LOYD A. JONES

[J. S. M. P. E.

comes less and less. This change in the rate of growth in gamma is
more clearly shown in Fig. 41 (curve A) which is plotted from the
data in Table XI. This curve is practically parallel to the Td axis at
16 minutes and by extrapolation it is ascertained that it will not ex-
ceed, for any kind of development, a value of 1.30. It is evident that
as Td is prolonged, gamma approaches a limiting value, and this is
called gamma infinity (TOO). This limiting gradient is illustrated in
Fig. 40 by the dotted line designaed at 7. The value of gamma
infinity is of great significance in both theoretical and practical sensi-
tometry and will be discussed more fully a little later.

w
<0.8
I

/ /
/ /

FIG. 41. Time of development-gamma curves, A for high rate of develop-
ment, B for low rate of development. Curve C is the corresponding time
of development-fog curve.

It will be noted by reference to Table XI and Fig. 40 that the value
of latitude, L, tends to decrease as the time of development increases.
Small vertical lines drawn through the various curves mark the limit
of the straight line portions. In many cases when all of the straight
lines (extended) intersect at a point lying on the log E axis, the points
marking the limits of the straight line portions lie approximately on
the circumference of circles drawn with the intersection point as a
center. Under such conditions the actual length of the straight line
is approximately constant, and hence is related in a definite manner to

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 65

gamma. The frequency of occurrence of this state of affairs is
relatively low, and hence it is unsafe to attempt to make any
generalization as to the relation between latitude and gamma except
to say that latitude usually decreases as gamma increases, and, there-
fore, as the time of development increases.

Curves of the type shown in Fig. 41 are frequently of great value in
analyzing the characteristics of a photographic material, particularly
from the standpoint of its behavior during processing. These are
known as time-gamma curves and are obtained by plotting gamma as a
function of development time. As has been mentioned, curve A is
obtained by plotting the data shown in Table XI which were derived
from the family of characteristic curves shown in Fig. 40. Curve B
illustrates the results obtained by processing the same material in a
different developing solution. It is evident from a comparison of the
two curves that gamma increases at a much lower rate in the case of
curve B, although if the development time is sufficiently lengthened
gamma appears to be approaching the same limiting value.

The time-gamma curve is of use where it is desired to determine the
development time which will yield some specified value of gamma.
If such a curve is available for the material and the processing condi-
tions being used, it is only necessary to read from the curve for any
gamma value the corresponding development time. Such curves are
also very useful in obtaining some idea as to the permissible variation
in development time when it is desired to control processing so as to
obtain gamma values lying within certain prescribed limits. For
instance, let it be assumed that it is desirable to obtain a gamma of 0.6
and that the permissible variations from the value are set at ±0.03.
The corresponding permissible variation in development time can be
readily determined for the two conditions represented by curves
A and B. The horizontal dotted lines are drawn through gamma
values of 0.60 + 0.03 and 0.60 - 0.03. Where these horizontal lines
intersect with curves A and B, perpendiculars are dropped onto
the development time, Td, axis. In the case of curve A it is found
that the development time must be held between 2.6 and 3.0 minutes,
thus permitting a total allowable variation of 0.4 minutes which may
be expressed as 2.8 ± 0.2 minutes. In case of the curve B it is found
for the same tolerance in gamma, minimum time is 5.7 and the maxi-
mum 6.5, which may be expressed as 6.1 =*= 0.4 minutes. It is evident,
therefore, that the allowable error in development time for the re-
quired precision in control of gamma is twice as great in the case of

LOYD A. JONES

[J. S. M. P. E.

curve B as for curve A . The relation between a given gamma incre-
ment and the corresponding development-time increment is, of course,
given directly by the gradient of the y — T curve at any particular
point. If it is desired, therefore, to express numerically this relation-
ship, it is only necessary to evaluate the differential dy/dt at any
given point. The value of the differential at any point is inversely
proportional to what may be termed processing latitude. In other
words, the greater the gradient of the time-gamma curve at any point,
the more precise must be the control of processing conditions in order
to maintain a given tolerance in gamma.

FIG. 42. Illustrating the general form of the first derivative, curve OA'B',
of the D-log E curve, AB.

Curve C in Fig. 41 shows the relation between fog and development
time. Fog is determined by measuring the density of an area on the
photographic material which has received no exposure but which has
been developed. In general, for most photographic materials the
value of fog is relatively low for the shorter times of development, but
usually grows at an increasing rate as the development time is
extended. Any value of fog which is given for a photographic
material obviously must be accompanied by some specification of the
development time or the extent of development (in terms of gamma)
in order to have any definite significance. The complete Tj-fog

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

curve is, of course, a complete representation of the relation between
fog and the extent of development and in graphic methods of showing
sensitometric results should be used rather than attempting to express
this factor by a single numerical value.

In Fig. 42 another useful graphic form is shown. Curve No. 1
is the usual .D-log E characteristic curve. Curve No. 2 is the first
derivative of the characteristic curve. It is obtained by plotting
values of gradient, dD/d log E, as a function of log E. This curve

3.0

2.4

3.6

1.8 ZA

LOG E

FIG. 43. Family of ZMog E curves obtained by plotting densities as read,
without fog correction.

shows somewhat more clearly the way in which gradient changes with
log exposure. For the straight line portion of the curve lying between
points A and B gradient is constant and equal to gamma. The first
derivative curve throughout this region is a straight line parallel to
the log E axis and having an ordinate value equivalent to gamma as
shown on the gradient scale at the right of the figure. For values of
exposure less than A and greater than B the first derivative curve takes
the form as shown. This graphic form is useful where it is desired to
determine precisely the exposure value corresponding to some

LOYD A. JONES

[J. S. M. P. E.

particular slope of the ZMog E curve. This form of presenting the
data contains no more information nor can it be drawn with any
greater precision than the D-log E curve itself, but for many purposes
it presents the data in more convenient form and gives a more vivid
mental picture of the relation between gradient and exposure.

All the characteristic curves thus far shown have been plotted from
data which have been corrected for fog. For many purposes for
which sensitometric work is done this procedure is to be preferred, but

1.4

1.0

•®

<s>

FIG. 44. The time of development-gamma curve derived from Fig. 43.

for certain practical purposes it may be preferable to deal with the
sensitometric data without making the correction for fog. This is
true, for instance, in certain problems relating to tone reproduction
where it is desired to obtain information as to actual density differences
in the negative corresponding to known brightness differences in the
object, and also to compute the time of exposure required for the
making of a positive from the negative. In such problems it is
essential to deal with the actual density values on the negative rather
than with the corresponding values which have been corrected for fog.

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

In Fig. 43 is shown a family of D-log E curves drawn from the measure-
ments as read directly from the sensitometric strips without fog
correction. It will be noted here that in the underexposure region
the curves do not come down to the log E axis but become parallel to
it at density values which are equivalent to the fog for the develop-
ment times in question. In this group of curves it will also be noted
that latitude decreases very markedly as the contrast or gamma of the
characteristic curve increases. In Fig. 44 is shown the time-gamma
curve plotted from values read from the curves in Fig. 43. This

FIG. 45. The graphic representation of sensitometric characteristics of a
high-speed negative material, including D-log E curves for various times
of development, time of development-gamma curve, and time of develop-
ment-fog curve.

curve, of course, represents the effective contrast as a function of
development time. It should be remembered that the correction for
fog changes the values of the measured densities by different amounts,
this change being proportionately greater for the lower densities,
thus modifying the magnitude of gamma. When the data are to be
used in tone reproduction problems, careful attention should be given
to this point.

Various ways of presenting sensitometric data in graphic form have
now been considered and it is evident that in order to convey a maxi-
mum of information more than one graphic form is necessary. It has

LOYD A. JONES

[J. S. M. P. E.

been found in practice that a complete family of .D-log E curves
obtained with various development times together with a time-gamma
curve and a time-fog curve serves as a fairly satisfactory graphic

- z. o

- \.o

1.0 20 3.0 40

FIG. 46. The graphic representation of sensitometric characteristics of
motion picture positive film, including D-log E curves for various times
of development, time of development-gamma curve, and time of development-
fog curve.

representation of the sensitometric characteristics. In Fig. 45 is
illustrated one way in which these various functions may be con-
veniently shown together. The characteristic curves themselves are

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 71

drawn in and the value of gamma for each is indicated. In the upper
left-hand portion of the rectangle are shown the time-gamma and time-
fog curves derived from the D-log E curves. The curves shown in
Fig. 45 apply to a high-speed negative material for which gamma
infinity is relatively low, being of the order of 1.4 to 1.5 as indicated
by extrapolation of the time-gamma curve. In Fig. 46 is shown a
similar group of curves for motion picture positive film which of course
is a relatively slow, high contrast material. Having now dealt in
some detail with the various graphic methods of presenting sensito-
metric data, the problem of deriving from these graphic forms certain
significant numerical values, which may be used as convenient
specifications of these sensitometric characteristics, will be considered.
It is quite impossible to express completely by means of a relatively
few numerical values all of the information contained in the various
possible graphic forms which may be used in presenting these data.
Such numerical values, however, are convenient when it is desired to
summarize in tabular form the sensitometric characteristics of various
materials for purposes of record and intercomparison.

SPEED OR SENSITIVITY

In the case of negative materials one of the most important charac-
teristics about which information is desired is that of sensitivity or
speed, and in the earlier stages of the evolution of photographic
sensitometry great emphasis was placed on the determination of this
characteristic. Several different methods of expressing speed have
been evolved and have been used rather widely in this country and
abroad. It may be of interest to consider the significance of these
various methods of speed specification and the inter-relation between
the resultant numerical values.

Threshold Speed. — One of the earliest methods used for the express-
ing of sensitivity was to specify the exposure required to produce a
just perceptible density. In methods of sensitometry not involving
the measurement of the developed densities this is the only feasible
method of speed expression which can be used. This was adopted by
Scheiner who devised a sensitometer which has already been described
in an earlier section of this paper. The sector wheel in the Scheiner
sensitometer was so cut that exposure increased logarithmically from
1 to 100 units. The distance between the points on the photographic
material corresponding to these exposure limits was divided into
twenty equal steps, numbered consecutively from 1 to 20. The

72 LOYD A. JONES [J. S. M. p. E.

Scheiner speed scale, therefore, consists of numbers in arithmetic
progression, 1, 2, 3, 4, etc., from 1 to 20, covering the sensitivity range
of from 1 to 100. Relative sensitivity represented by any given
number in the scale is 1.27 times as great as the relative sensitivity
corresponding to the next lower number in the scale. This relation
is shown in Table XII, in the first column of which the Scheiner
numbers are given, and in the last column will be found the correspond-
ing relative sensitivity values. The consecutive numbers of the

TABLE XII

F/209 Inter comparison of Speed Values as Expressed by Various Weil-Known

Methods

Scheiner

Eder-Hecht

H & D

Watkins

Wynne

Relative

F/21

1.0

F/24

1.27

F/21

1.62

F/30

2.07

F/34

2.64

F/38

3.36

F/43

4.28

F/49

5.45

F/55

6.95

F/63

8.86

122

F/71

11.3

104

153

F/79

14.4

133

196

F/90

18.3

170

250

F/101

23.4

216

317

F/114

29.8

276

405

F/129

37.9

351

515

F/145

48.3

448

660

F/165

61.6

570

840

F/196

78.5

20 90 727 1065 F/209 100.0

Scheiner scale which increase in arithmetical progression correspond
to a geometrical progression in relative sensitivity. The scale inter-
val, therefore, is slightly greater than that given by using consecutive
powers of the cube root of 2 in which the multiplying factor from step
to step is the cube root of 2 or 1.26.

The Eder-Hecht sensitometer, as has already been mentioned, is of
the tablet type consisting of a neutral gray wedge with a continuous
gradient. On this are printed a series of numbers in arithmetical
progression and equally spaced. This speed scale is, therefore, also

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 73

of the logarithmic form, assuming that the neutral gray wedge has a
constant gradient. The numbers actually used on the Eder-Hecht
sensitometer tablet as compared with the Scheiner scale are as shown
in the second column of Table XII.

While the threshold method of expressing sensitivity has certain
features to recommend it, it leaves much to be desired from the stand-
point of precision and significance. The magnitude of the least
perceptible density depends profoundly upon the conditions under
which the inspection is made. In fact, the judgment which is actually
made is not that of least perceptible density, but least perceptible
density difference. Under the most favorable conditions of observa-
tion, the human eye can detect a brightness difference of 1.7 per cent.
This corresponds to a density difference of 0.008. Under other con-
ditions of inspection, however, such as relatively low illumination and
uncomfortable visual conditions, this just perceptible density differ-
ence may be easily as great at 0.04. It is evident, therefore, that,
unless great care is taken to standardize and maintain the visual condi-
tions under which judgment of the just perceptible density is made,
values of threshold speed read from the same actual test strip may fluc-
tuate over a considerable range. Furthermore, if conditions are ad-
justed to give maximum visual sensitivities so that a very slight density
difference may be detected, such, for instance, as the value men-
tioned above, namely, 0.008, the absolute value of speed is rather high,
regarded from the practical standpoint. For instance, the point on
the toe or underexposure region of the characteristic curve where a
density of 0.007 is obtained, is in almost all cases at a point of ex-
tremely low gradient. It is questionable whether the underexposure
region at or near the point where D is equal to 0.008 is of any practical
value. While it may be argued that the effective speed should be con-
sidered to go down into the region of low exposures to the point where
a just perceptible density is produced, this seems somewhat fallacious,
when it is considered that the real function of a photographic material
is to reproduce, as perceptible density differences, the brightness differ-
ences which exist in the object. It seems, therefore, that we should
be more concerned with the definition of speed in terms of the power
of the material to reproduce satisfactorily some minimal contrast.

Inertia Speeds. — Hurter and Drirfield in their work on photographic
sensitometry suggested that the speed of a material could be specified
satisfactorily in terms of the inertia. They proposed, therefore, the
expression of speed as the reciprocal of the inertia multiplied by a

74 LOYD A. JONES [J. S. M. P. E.

constant for which they chose a value of 34. The use of this number
gave a series of speed values of convenient magnitude for practical
use. In the third column of Table XII are shown the H & D speed
numbers in direct comparison with relative values as shown in the last
column.

The Watkins speed scale is also based upon inertia, but instead of
using 34 as suggested by Hurter and Driffield, Watkins adopted 68
as the value of constant k. The actual relation between the Watkins
and H & D numbers, however, indicates that the Watkins constant is
more nearly 50 than 68. In the fourth column of Table XII are
shown the values of the Watkins speed scale in comparison with the
other well-known systems.

The Wynne system of expressing speed is not used to any great
extent but is of some interest. This is also based fundamentally upon
inertia values but uses numbers which are expressed in terms of lens
aperture as indicated by the symbol F which precedes the number.
These numbers are proportional to the product of 6.4 by the square
root of the Watkins number. A Watkins speed of 100 (equivalent to
H & D speed of 68) gives a Wynne number F/64. This scale is shown
in the fifth column of Table XII.

For many purposes and under many conditions, the expression of
speed in terms of inertia is of great value. As long as all of the
straight line portions of a family of D-log E curves pass through a
common intersection point and this point lies on the log E axis,
inertia and hence speed are independent of development time. Under
such conditions the speed becomes a very significant constant for
the photographic material. Unfortunately the existence of a common
intersection point lying upon the log E axis is frequently not found in
practice. In most cases of normal development a common inter-
section point is found, provided that proper corrections have been
made for fog. This intersection point, however, very 'frequently
lies below the log E axis and in relatively rare cases is located above
that axis. This subject has been dealt with at great length by
Nietz.100 It has been found that in the presence of free bromide,
whether it be in the developing solution or present in the photo-
graphic material itself, the intersection point is, in general, depressed
to a position below the log E axis. Such a condition is shown in Fig.
47 which represents the straight line portions of a family of D-log E
curves. Assuming for the moment that a common intersection
point does exist, its coordinates may be represented by a and b as

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

shown in Fig. 47, and it has been proposed to define the speed of the
material in terms of the coordinates of this point. Under such
conditions it is evident that the inertia is a function of gamma, and
hence speed based upon inertia value will become a function of
gamma, and a speed value of this nature can only be significant
provided the corresponding gamma value is specified. For the
purpose of certain theoretical investigations into the nature of
exposure and development, a knowledge of the coordinates of the

FIG. 47. The straight line portions of a family of ZMog E curves illus-
trating the existence of a common intersection point below the log exposure
axis, thus causing inertia to depend upon time of development.

intersection point, as shown in Fig. 47, may be of great value, but it
does not appear to be very significant for the purposes of determining
the practical speeds.

The extent of the dependence of the inertia speed upon gamma is
illustrated in Table XIII. These data are derived from measure-
ments made on a high-speed negative material processed in a develop-
ing solution containing some bromide. The straight line portions of
the P-log E curves intersect at a point well below the log E axis. The

76 LOYD A. JONES [J. S. M. p. E.

development times, Td, extend from 3 to 20 minutes, the correspond-
ing gamma range being from 0.27 to 1.12. For this range in develop-
ment, the reciprocal inertia changes from 22 to 140, a six-fold increase
in speed as derived from inertia values.

The complication involved in expressing speed by means of the
inertia does not end here. Many materials are found for which there
is no single point of intersection for the straight line portions of the
D-log E characteristics. In fact, some materials do not give very
satisfactory straight line relations between density and log
exposure. There is wide divergence in the relative shape of the
underexposure regions and, in fact, an almost endless variety of
conditions are found which makes it extremely difficult to generalize
satisfactorily the expression of practical or effective speed of photo-
graphic materials for all purposes. Anomalous behavior in both
shape and position of the D-log E curves resulting from development

TABLE XIII

Data Illustrating the Dependence of Inertia upon Time of Development

Td y i l/i

0.27

0.045

0.43

0.020

0.68

0.012

0.83

0.010

100

1.12

0.007

140

for different times seems to be particularly common in materials of
high sensitivity. This subject has been discussed at considerable
length by Sheppard.101 He classifies emulsions generally into
orthophotic and anorthophotic categories. Orthophotic materials
show a definite convergence point of the straight line portion of the
characteristic curves, while anorthophotic materials depart widely
from this condition showing no tendency to give a common point of
convergence. Sheppard concludes from his study of the subject that
emulsions of the anorthophotic type have characteristics which are
much less reproducible from batch to batch than those of the ortho-
photic class. In many fields of work reproducibility is highly im-
portant; for instance, when these materials are used as a means of
making quantitative measurements in science and technology, when
it is desired to employ automatic processing methods, and in those
cases where it is desirable to make application of the laws of tone re-
production. It would appear that this demand on the part of the

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 77

users of photographic material for reproducibility may tend auto-
matically toward the rejection of materials of the anorthophotic type.
Hence photographic materials may respond to evolutionary laws,
their characteristics tending to become predominantly orthophotic as
a result of the survival of the fittest. While it is impossible to ignore
the existence of certain materials which, during the course of develop-
ment, do not even approach to the classical behavior required by H &
D theory, it would seem that in the development of sensitometry
greatest attention should be paid to the evolution of sensitometric
systems applicable particularly to the materials which do approach
to normal types. The case, therefore, is not quite as hopeless as it
may appear and it does not seem unreasonable to assume the existence
of a normal type behavior from which the great majority of materials
used in large volume depart but little and in a known and specifiable
manner. In any case it seems more profitable to take the position
that normality and orderliness are the rule rather than to assume the
attitude of destructive criticism and maintain that "all photographic
materials are exceptions," thus abandoning the entire field of syste-
matic sensitometry to chaos.

Luther's Crossed Wedge Method. — In the section dealing with sensi-
tometers the crossed wedge method proposed by Luther23 for
obtaining directly the D-log E curve is mentioned. A few words rela-
tive to the interpretation of the results obtained in this manner seem to
be in order. The envelope of the darkened area gives directly the
D-log E characteristic. If the density and density gradient of each
wedge are known, it is possible to establish the correct density and log
exposure scales. The linearity of these scales, that is, the representa-
tion of a specified density or log E difference by a constant linear
interval throughout the respective scales, requires that the density
gradient of the wedges be constant, and also that the wedge density
be uniform along any line perpendicular to the gradient direction.
Considering the difficulties involved in the manufacture of wedges of
satisfactory uniformity, both of density and density gradient, having
sufficient freedom from selective absorption, and in the exact
calibration of these wedges, it does not seem likely that the precision
obtainable in the final result can be as great as that resulting from
the exposure of the photographic material in a well-designed and
^carefully operated sensitometer of the time-scale type followed by the
measurement of density with suitable densitometers. However, as a
rapid and convenient means of testing, the method has much to

78 LOYD A. JONES [J. S. M. P. E.

commend it. Yielding as it does the typical D-log E curve, the
interpretative methods are identical to those applicable to similar
curves obtained by other methods, and values of the usual factors,
such as gamma, inertia, fog, etc., may be derived.

Minimum Useful Gradient. — Thus far two methods for the expres-
sion of sensitivity or speed have been discussed, one based upon
exposure corresponding to a just perceptible density (Schwellenwert) ,
and the other based upon the value of inertia read at the point where
the straight line portion of the D-log E curve cuts the log E axis.
While both of these methods have certain points to commend them,
both also have very serious deficiencies. Threshold speeds are not
independent of the time of development and therefore cannot be said
to be strictly a constant of the photographic material. Moreover,
the absolute value obtained under inspection conditions yielding
maximum visual sensitivity gives a speed which is too high, when it is
desired to compute the exposure required for the satisfactory render-
ing of detail in the shadow regions of the object. It has been seen,
further, that speed values based upon inertia are dependent in many
cases upon time of development and hence require an accompanying
expression of gamma in order to be significant. Even under these
conditions, the absolute value of speed is not very useful in computing
the exposure time required in order to give satisfactory rendering of
shadow detail. A study of a large number of characteristic curves
shows that the gradient corresponding to the inertia value varies
between wide limits. When it is considered that the chief function of
a photographic negative material as used in practice is to reproduce as
density differences the brightness differences existing in the object
photographed, it seems logical to demand that the minimum useful
exposure be determined by some specified gradient of the D-log E
characteristic. This idea has been discussed by Luther102 and its use
advocated by him, especially in cases where the fog of the emulsion is
relatively high. The subject has also been discussed at some length
by Jones and Russell.103

There seems to be little doubt that this idea is based on a sound
theoretical foundation. The difficulty met, however, is that of
deciding upon the value which is to be taken as representing the
minimum useful gradient. Luther suggested that this value should be
0.5. Judging from data available in publications by Goldberg and
found in a paper by Jones104 dealing with the contrast of photographic
printing papers, it appears that this value is too high and that satis-

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

factory reproduction of object detail can be obtained by utilizing
portions of the characteristic curve of lower gradient. The subject
has also been discussed by Sheppard101 who states that the minimum
useful gradient "will in general depend not only upon the negative
but also upon the positive aspect of tone reproduction so that its
fixation is not expressible by a unique function of the negative
material itself." This conclusion is undoubtedly correct and its
validity is supported by the data and the discussion given by Jones
(loc. tit.).
It seems quite possible, however, for certain definite classes of work

RADIENT

& 5

<S> Ad

I
M

LOG EXPOSURE

FIG. 48. D-log E curve, O4J3, and its first derivative, FED. The con-
struction illustrates the method of finding the relation between gradient
and log exposure.

to establish a value of minimum negative gradient in terms of which
sensitivity or speed may be expressed in a manner of considerable
practical utility. For instance, from a knowledge of common
practice which results in acceptable tone reproduction in the field of
motion picture photography, it is possible to draw fairly definite
conclusions as to the minimum useful gradient of negative materials
used in this work. This knowledge is based upon careful densi to-
metric analyses of a large number of motion picture negatives and
positives. The value indicated by the available information lies

LOYD A. JONES

[J. S. M. P. E.

between 0.2 and 0.3. A similar value is known to represent fairly well
the conditions existing in the field of amateur photography. It is not
possible with the information available at the present time to say
definitely whether or not a fixed value of minimum limiting gradient
can be chosen which would be satisfactory in all fields and, if so, just
what the absolute value of this quantity should be. It is definitely
known, however, that in the motion picture field and in the amateur
field practically all negatives utilize the greater portion of the under-
exposure region of the D-log E characteristic. Furthermore, it is
logical to conclude that the toe of the characteristic curve below

'3.4 3.0 l.<* 12. T.8 0.4

LOG EXPOSURE

FIG. 49. Typical P-log E characteristic curves illustrating the disagree-
ment between speed values based upon inertia and those based upon minimal
useful gradient.

some definite value of gradient is too flat for satisfactory reproduction
of object brightness differences. It seems desirable, therefore, to give
this suggested method for the specification of speed very careful
consideration and some data based upon an arbitrary assumption of
the value of minimum limiting gradient may be of interest. For
this purpose a value of G equal to 0.2 will be assumed.

In Fig. 48 the most precise method of determining the exposure
value corresponding to this specified gradient is illustrated. Curve
OAB represents the underexposure and part of the correct exposure
region of the .D-log E characteristic, curve FED being its first deriva-
tive. Through the gradient value of 0.2 a horizontal line is drawn

Jan., 1932]

PHOTOGRAPHIC SENSITOMETRY

which establishes point D. A perpendicular dropped from this point
cuts the characteristic curve at point O, which is the point having the
gradient of 0.2. The exposure value, Em, corresponding to the point
0, is that desired in order to express speed in terms of a minimum
useful gradient equal to 02. Speed and sensitivity are of course
inversely proportional to Em.

A typical case which illustrates the merits of this method of express-
ing speed is shown in Fig. 49. The two materials illustrated have
been developed to the same gamma, and on the basis of the inertia
method of expressing speed; the material of which the straight line

Z.3

27 T.%

LOG EXPOSURE

FIG. 50. Family of D-\og E curves illustrating the dependence of minimum
useful gradient speed upon time of development.

portion cuts the log E axis farthest to the left has the higher speed of
the two. The small arrows indicate the points on the underexposure
region where G is equal to 02. It will be seen that the minimum
limiting gradient method reverses the speeds of the two materials.
If we express inertia speed as 10 times the reciprocal of the inertia
and the gradient speed as 10 times the reciprocal of Em, the numerical
results derived from Fig. 49 are as follows:

220

1360
720

82 LOYD A. JONES- [J. S. M. P. E.

Further illustration of this suggested method is given in Fig. 50
in which are shown the underexposure regions of four D-log E charac-
teristic curves obtained by using different times of development.
Again, the points of gradient equal to 0.2 are indicated by the small
arrows attached to each curve. In the case of the shortest time of
development it will be noted that the gradient for 0.2 is obtained for

TABLE XIV

Data Illustrating the Relation between the Values of Speed Based upon Inertia,
i, and Those Based upon the Exposure, Em, Corresponding to the Minimum Useful

Gradient

Fog

1.5

0.68

0.15

0.045

0.0280

1.6

2.0

0.80

0.22

0.025

0.0058

4.3

4.5

1.7

0.32

0.026

0.0053

4.9

6.0

2.0

0.47

0.024

0.0044

5.5

9.0

2.6

0.51

0.033

0.0050

6.6

an exposure value practically equal to the inertia value, while for
longer times of development the speed values based upon minimum
gradient are very much higher than those based upon inertia. The
data derived from the curves in Fig. 50 are shown in Table XIV.
The ratio of i to Em is given in the last column.

GAMMA INFINITY, y „ , AND CONSTANT OF DEVELOPMENT, K

The importance of gamma infinity, both for theoretical and practical
sensitometry, already has been emphasized and in Fig. 41 an experi-
mental method of determining the value of gamma infinity has been
illustrated. In certain cases, similar to that illustrated in Fig. 40
where development takes place in the manner assumed by the classical
H & D theory, it is quite possible to compute from the data contained
in two or more J9-log E curves a theoretical value for gamma infinity.
Along with this may be derived the value of K , the velocity constant of
development. According to the method introduced by Sheppard and
Mees, l05 this computation is based upon the gamma values obtained
from two D-log E curves, for one of which the development time is
twice that of the other. The necessary data are therefore derived
from a pair of curves such as those shown in Fig. 39 where this ratio of
development times was used. From theoretical considerations it can
be shown that the relation between development time and growth of
gamma can be expressed by the equation

7 - 7.(1 - «-*') (1)

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 83

By substituting in this equation values of gamma and time of develop-
ment obtained from curve No. 1, equation (2) is obtained, and
similarly by using values of gamma and the time of development read
from curve No. 2, equation (3) is obtained.

71 = 7^(1 - e-*«0 (2) .

72 = 7 00(1 - e-™*} (3)
Combining (2) and (3) we obtain

Ti(l - «-««) = T2(l - «-*") (4)

Since fe = 2fc

I? = i + «-», (5)

(6)

7i _ . 72 ,Rx

n-Kt^ 1 ,,-R-fo V°/

From the known values of 71, 72, /i, and /2 it is therefore possible to
compute values of 7 «, and X. From any other pair of characteristic
curves for which the times of development are related by the expres-
sion /2 = 2/i, additional values of these constants may be computed
which should, of course, check those based on any other similar pair
of sensitometric curves. It should be emphasized that these
theoretical relationships do not hold in all cases, their validity de-
pending upon the normality of the family of D-log E characteristics
as judged by the requirements of the H & D theory.
By differentiation of equation (1) the relationship
dy/dt = K(ym - 7)

is obtained. Now dy/dt is the slope of a time-gamma curve such as
that shown in Fig. 41. The values of this gradient can be determined
graphically frpm an experimental time-gamma curve for any value of
gamma, and if gamma infinity is known, having been determined, let
us say, experimentally as illustrated in Fig. 41, it is possible to compute
the corresponding value of K. All values of K computed in this way
for various values of gamma should, of course, be the same provided
the curve is of the exponential form. It is found in practice that
when such a time-gamma curve is derived from a normal family of
characteristic curves, and when the experimental determination of
7 OP is valid, this condition, K equal to a constant, is fulfilled. In many

84 LOYD A. JONES [J. S. M. P. E.

cases, however, the value of K, computed from different assumed
values of gamma and corresponding graphically determined values of
dy/dt, is not constant. This is evidence of abnormality in the time
of development-gamma relation and can usually be traced back
and found to be due to improper correction for fog, lack of a common
convergence point, or other departures from what may be termed the
normal behavior.

Certain development characteristics of any particular photographic
material may be deduced from the values of 7 and K.

For instance, if K is high and 7^ is high:

Development will start quickly, proceed at a high rate, and gamma
will continue to build up to a high value. Process plates and motion
picture positive film are typical examples of the materials having
these characteristics.

If K is high and ym is low:

The image will flash up quickly and 7 will build up rapidly at first
but soon cease to increase, reaching a limit at a relatively low value.
In the case of these materials the image appears very quickly but
fails to carry through and build up high densities.

If K is low and 7ro is high:

Development starts slowly and 7 increases at a relatively low
rate, but by extending the time of development the value of 7 may be
built up to a high value.

If K is low and 7 „ is low:

Development starts slowly, 7 increasing at a relatively low rate
which very soon decreases and flattens out at a final low value.
Further development will not serve to increase contrast.

TIME OF DEVELOPMENT FOR A SPECIFIED GAMMA, Ty = x

In practical specification of sensitometric characteristics it is
sometimes desirable to measure directly, in terms of a single constant,
the dependence of gamma or contrast on time of development. Such
a figure combines to a certain extent the information contained in the
values of 700 and K and usually is more easily obtained. This is
accomplished by stating the time of development required to give a
specified gamma. In choosing the gamma value for which this time
of development is stated, it is, of course, desirable to use the contrast
to which the material is in practice usually developed. It is, of
course, impossible to find any single value of gamma which fits the
requirements of all possible classes of photographic work. In some

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 85

cases as, for instance, motion picture photography, the negative is
usually developed to a relatively low gamma such as 0.5 or 0.6. In
the amateur field somewhat higher gammas are usual. In portrait
work it seems probable that a gamma of 0.8 represents a fair practice.
In commercial work this value is unity or even somewhat above,
while in process work gamma is pushed as near as possible to 7^,
practical values lying between 1.8 and 3.0. In the case of motion
picture positive film it is probable that a gamma of 1.8 represents a
fair average. For the purposes of preparing tables showing relative
sensitometric characteristics, a gamma value of unity is usually
chosen, the time of development required by various materials to
obtain this value under standardized processing conditions being
determined. This factor is usually expressed symbolically as
Ty » i.o.

LATITUDE, L

Latitude has already been defined in the earlier discussion and the
method of determination explained. As stated previously, latitude
varies with gamma and therefore only has significance when ac-
companied by a statement in terms of gamma of the extent to which
development has been carried. Various attempts have been made to
find a theoretical or analytical expression relating latitude and gamma.
While some equations have been proposed, none of these seem to be of
sufficient general validity to warrant consideration. It is customary,
therefore, to determine latitude graphically directly from the plotted
characteristic curves. It is usual to express latitude in the form of
the ratio of the maximum to the minimum exposure lying on the
straight line portion of the curve. A knowledge of latitude is in
certain classes of work of considerable importance. It defines the
ratio of object brightnesses which may be rendered by the material,
when developed to the gamma specified, without non-linear distortion
of the object contrasts. For most high-speed negative materials, the
magnitude of gamma is considerably greater than the ratio of maxi-
mum to minimum brightness in average photographic subjects
under normal illuminations. It is extremely unusual in out-of-door
work to encounter subjects in which the contrasts, that is, ratio of
maximum to minimum brightness, is greater than 100. Extreme cases
show values as high as 250, but these are rare. In studio work it is,
of course, possible to obtain artificial lighting in which the contrast is
greater than that mentioned above. However, the measurement in a
great many portrait and motion picture studios indicates that in this

86 LOYD A. JONES [J. S. M. P. E.

class of work contrast seldom exceeds 100 or at the most 200. For the
low values of gamma usually used in motion picture studio and
portrait work, it is not infrequent to find that the photographic
materials have latitudes well above 500 and in some cases^reater
than 1000. For the low-speed, high-contrast materials; gmSmfirbf
course is much lower. Here again the value will depend upon the
extent to which the material is developed. For the high contrast to
which motion picture positive film is usually developed, a latitude of
32 to 64 is usual.

FOG, F

The definition of fog and the method of measurement has already
been defined in the previous discussion. Its value is dependent upon
the extent to which development has been carried and of course is
profoundly influenced by the composition of the developing solution.
In giving fog as a sensitometric value it is necesaary, therefore, to
specify both the composition of the developer used and the extent to
which development has been carried usually in terms of gamma.
As stated previously complete information as to fog giving propensi-
ties of photographic material is shown best in graphic form by the
time of development-fog curve as illustrated in Fig. 41. In cases
where the time of development required to give gamma of unity and
inertia for gamma of unity are given, it is customary to express fog
also for gamma of unity.

TABLE XV

Sensitometric Constants of Typical Photographic Materials

Td-
Material F K ym (7 = 1.0) L * 10/t

Motion picture film super-
speed 0.15 0.24 1.6 4.0 400 0.010 1000
Motion picture film normal 0.10 0.20 1.8 4.0 300 0.017 600
Motion picture film positive 0.03 0.30 2.8 1.5 50 0.330 30
Portrait film normal 0.08 0.20 1.6 5.0 200 0.020 500
Portrait pan film super-speed 0.15 0.24 1.6 4.0 300 0.010 1000
Amateur film, fast 0.10 0.16 1.8 5.0 150 0.017 600
Amateur film, normal 0.07 0.18 1.6 5.5 80 0.025 400
"Press" plate 0.15 0.16 1.8 5.0 100 0.010 1000
Commercial ordinary 0.05 0.18 2.5 3.0 75 0.040 250
Commercial ortho 0.08 0.19 2.2 3.2 75 0.028 350
Commercial pan 0.10 ' 0.20 2.0 3.5 100 0.020 500
Process plate ordinary 0 . 04 0 . 28 3.0 1.5 25 0 . 250 40
Process plate pan 0.08 0.28 3.0 1.5 25 0.067 150
Lantern plate 0.03 0.32 3.2 1.2 25 0.650 15

Jan., 1932] PHOTOGRAPHIC $ENSITOMETRY 87

In Table XV are shown some typical numerical sensitometric
values for a variety of photographic materials. The values given do
not refer to any particular material but represent a fair average of
materials which fall within the classifications as indicated in the
first column. They neither represent the best now available nor the
"ideal" material in each class, but specify the characteristics that may
reasonably be expected of these materials. The various sensito-
metric constants tabulated and the specification of conditions under
which the determinations were made are as follows:

All of the sensitometric strips were developed in a solution made up
according to the two-solution pyro formula given in the section on
development* used at a temperature of 20° C.

F. The particular value of fog given is that obtained for the
development time giving 7 = 1.0

Td(y = 1-0). This is the development time in minutes required
to give a 7 = 1.0.

700. The value of gamma infinity shown in the table is that
determined experimentally by extrapolation of the time-gamma curve.
In the construction of this curve, development time was sufficiently
prolonged so that the curve showed a definite tendency to become
parallel to the D-log E axis, thus decreasing the amount of extrapola-
tion required to obtain a fair estimate of the value of 7,,, .

K. The values for this term as shown in the table were computed
from the equation dy/dt = K(ym— y). By determining graphi-
cally the slope of the time-gamma curve at the point where 7 = 1.0,
the value of dy/dt at that point was obtained and when substituted in
the equation above, together with the already experimentally de-
termined value of 7 a, , permits the computation of K. The values
shown in the table, therefore, may be termed the instantaneous value
of K for the condition 7 = 1.0, and consequently for the time of
development as shown in the fourth column. Since many photo-
graphic materials do not conform to the classical H & D theory
with respect to growth of gamma with increasing development times,
it follows that K cannot be regarded strictly as an invariant constant
for all materials. It seems, therefore, that from the practical stand-
point a value of K, as determined above, may be of somewhat greater
value than one computed as previously described by using two ZMog
E curves obtained for development times of T and 2T, respectively.

* J. Soc. Mot. Pict. Eng., XVH (November, 1931), No. 5, p. 700.

88 LOYD A. JONES [J. $. M. p. E.

Obviously it is, of course, possible to put the value of K and 7^
shown in the table back into the above equation and compute dy/dt.
The value of this term, as has already been pointed out, is useful in
obtaining some idea as to the processing latitude of the material, that
is, the variation in the time of development which is permissible for a
specified gamma tolerance.

L. The values of latitude shown are those given by the material
when developed to a contrast at which the material in question
is customarily used. This conveys definite information as to the log
exposure scale which can be rendered with non-linear distortion.

i. Values of inertia are those of exposure at the point where the
straight portion of the characteristic curve having a gamma value of
1.0 cuts the log E axis. They are expressed, of course, in terms of
visual candle meter seconds, m.c.s., of radiation of daylight quality.

10/i. Sensitivity values given in the last column of the table are
obtained by using 10 as a constant in the expression for speed

The constants as shown in Table XV are probably as useful as any
for the purpose of specifying numerically the characteristics of a
photographic material. Emphasis should again be given to the
statement already made to the effect that it is hopeless to attempt to
convey in any set of numerical values as much information as can be
deduced from a complete set of graphical representations of the sensi-
tometric data. While numerical constants are very convenient and
useful for many purposes, they should not be expected to serve as a
substitute for the more comprehensive graphic representation.

From a consideration of what has already been said relative to
the interpretation of sensitometric data, it should be evident that it is
quite impossible to formulate any single interpretative method which
will meet the requirements of all of the purposes for which sensito-
metric data may be required. Interpretation must depend to a great
extent upon the use for which the information is intended. For
purposes of standardization sensitometry, it may be desirable to adopt
a single developer in which all materials to be compared are developed,
and to express a group of numerical constants derived in such a
manner as to facilitate intercomparison between the various materials.
For the control of uniformity of product an entirely different procedure
may be necessary. For instance, it may be necessary to adopt a
particular developing solution and technic for each different material,

Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 89

and to maintain this with high precision over long periods of time.
Emphasis may be laid upon the determination with utmost precision
of some particular characteristic, such as speed or contrast for fixed
time of development. From the standpoint of the user of a photo-
graphic material it may be necessary to establish a sensitometric
procedure which duplicates precisely the processing conditions which
exist in practice, and lay particular stress on the maintenance of this
equality at the expense of other factors. If, for instance, sensito-
metric data are to be used for the control of the uniformity of the
product turned out by continuous developing machines in the motion
picture laboratory, great care must be taken to insure that the sensito-
metric strips are developed under conditions identical to those occur-
ring in the developing machine. If such is not feasible, it is necessary
to establish, by an extended series of experiments, a correlatoin
between the results obtained by some adopted sensitometric condi-
tions and those existing in practice. It is quite possible that it may be
necessary to develop particular methods for analyzing the data. For
instance, in the sensitometric work done in connection with the
photographic reproduction of sound it is frequently more convenient
to plot transmission of the silver deposit as a function of log exposure.
In such cases, of course, it is only necessary to transform density to
transmission and plot this as a function of log exposure. The analyses
of these curves require special treatment. It may be useful for some
purposes to express transmission as a function of exposure rather than
of log exposure. There are almost numberless variations in inter-
pretative methods. It is quite impossible to treat all of these
completely at this time, but it is hoped that the subject matter which
has been presented may form a foundation upon which further
elaboration of analysis and interpretation may be built.
( Concluded in the March issue of the JOURNAL)
REFEBENCES

100 NIETZ, A. H.: "Theory of Development," Monograph No. 2 from the
Kodak Research Laboratories, Eastman Kodak Co. (1922).

101 SHEPPARD, S. E.: Phot. /., 50 (n. s. 66) (1926), p. 190.

102 LUTHER, R.: Trans. Faraday Soc., 19 (1923), p. 340.

103 JONES, L. A., AND RUSSELL, M. E.: Proc. Seventh Intern. Congress Phot.,
(1928), p. 130.

104 JONES, L. A.: /. -Frank. Inst., 202 (1926), pp. 177, 469, 589; 203 (1927),
p. Ill; 204 (1927), p. 41.

106 MEES, C. E. K., AND SHEPPARD, S. E.: "Investigations on the Theory of
the Photographic Process," Longmans, Green & Co., London (1907).

THERMIONIC TUBE CONTROL OF THEATER LIGHTING'

BURT S. BURKE**

Summary. — The development of thermionic tubes has opened an entirely new
field in control of theater lighting. This development has made possible the ob-
taining of preset dimming proportional dimming, and a small compact switch-
board such as has been heretofore impossible.

The preset dimming feature allows an 'operation whereby a board may be set up
for any desired number of effects in advance, so that these effects may be called for
at the will of the operator by operating a single control. This feature might be termed
an ability of the switchboard to learn effects and bring them out when called upon
by its master, the operator.

Proportional dimming, a new feature, allows the lights to be controlled in such a
manner that they may be dimmed out in combinations while retaining the same color
tone throughout the dimming process.

The third desirable feature is that a small compact control board may be so arranged
that it can be placed as desired in the orchestra pit, or some similar location so that
the operator becomes a light artist, taking his place in the performance along with
the organist or other artists.

In the past 15 years the equipment for controlling illumination in
theaters has developed from the simple knife-switch type of switch-
boards to the complex arrangement of circuits and dimmers which are
required by the elaborate stage productions of the present day. Two
general qualities are essential in a modern theater switchboard:
flexibility in the selection of circuits, and flexibility in controlling the
light intensity of these circuits. The first has been well provided for
in the various types of multi-preset switchboards which have been
built for the past several years. The second requirement, as provided
for in the dimmer systems which are built into the usual multi-preset
switchboard, leaves much to be desired.

In the past three years a new means for controlling the intensity of
the light circuit has become available. The development of thermi-
onic devices for industrial uses has made possible new systems for
accomplishing the complex lighting effects required in modern

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Westinghouse Electric & Manufacturing Co., East Pittsburgh, Pa.
90

CONTROL OF THEATER LIGHTING 91

theaters. Architects, illuminating engineers, and theatrical producers
are now able to use lighting in ways that have never before been
possible.

The thermionic tubes used in theater lighting equipment have been
of the hot cathode grid-glow type, whose output may be regulated by
properly controlling the grid circuit of the tube. The whole control
system consists briefly of a reactance type dimmer, similar to the
reactance dimmers which have commonly been in use for many years,
a thermionic tube unit for supplying direct current to saturate the
reactance dimmer and a control means consisting of a network of
potentiometers for properly controlling the output of the tube unit.

A system of this type was used for the control of the stage and
auditorium lighting of the new Los Angeles theater which was opened
January 30, 1931. This paper will describe this equipment, which is
typical of the thermoionic tube type of control for theater lighting.

The Los Angeles theater, similar to many theaters of its size, is
equipped for both motion pictures and stage presentations. On this
basis, the control switchboard was split into two parts, one of which
was located in the projection room for controlling the auditorium
lights, and the second of which was located at the stage floor for
controlling the stage lights. On the stage floor is also located a
remote panel which allows the operator to obtain color master control
of the auditorium circuits as well as control of the five-scene preset
arrangement which will be desired later. Dual control for the foot-
lights, the first border, the orchestra floods, and the stage floods is
provided so that these circuits may be controlled either from the stage
floor or from the projection room. A transfer scheme is used to
transfer these controls from one place to the other so that no inter-
ference of control is possible.

Each of these two switchboards consists essentially of a reactance
dimmer bank, a set of tube units for use in connection with the d-c.
coils of these reactors, and a control board used to control the
output of the tube unit, thus indirectly controlling the intensity
of the lighting circuit.

The reactance dimmers are similar to the standard theater duty
reactance dimmer in that they consist of a-c. coils and a d-c. coil
mounted on an iron core. The reactance of the dimmer is varied by
changing the amount of direct current in the d-c. leg of the reactor,
thus varying the saturation of the iron. With no direct current in
this coil, the iron is unsaturated and of very high reactance, thus

92 BURT S. BURKE [j. s. M. p. E.

dimming out the light circuit with which it is associated. By increas-
ing the direct current the iron becomes saturated, so that the reac-
tance of the dimmer becomes lower, thus increasing the voltage across
the lamps and bringing them up to full brilliancy.

The direct current supplied to the coil of the reactance dimmer
comes from the thermionic tube unit which consists of two grid-glow
tubes, a control tube, and the requisite transformers for supplying
the proper filament and plate voltages for these tubes. The tube
unit receives its power from the 115-volt, 60-cycle, single phase mains,
which is applied to the two grid-glow tubes through a suitable trans-
former in order to obtain the proper voltage on the plates of the tubes.
The a-c. voltage is rectified by these tubes and the resulting rectified
alternating current is impressed on the coil of the reactor. The
magnitude of the d-c. output of the tube unit is varied by changing
the relation of the grid voltage to the plate voltage of the grid-glow
tubes. This is done by means of the control tube, which is a vacuum
tube similar to the standard UX-226 tube used in radio circuits. The
output of this control tube is varied by a system of potentiometers
located on the control board which will be described in the following
paragraphs. One set of these tube units was mounted on the reactor
bank located in the projection room for the auditorium lights, and a
second set of the tube units was mounted on the reactor rack located
in the basement which was for the stage lights.

The two control switchboards, one located in the projection room
for the auditorium circuits, and the second located on the stage floor
for the stage circuit, are similar except for the circuits controlled, and
may be briefly described as follows:

For the control of each individual circuit there is provided a pilot
switch, a selector switch, an indicating lamp, 5 preset potentiometers,
and an individual control potentiometer. In addition to this there
is provided for each circuit a scene-fader by means of which it is
possible to fade from one effect into the next effect giving a gradual
transition from one to the other. These faders are ganged together
and have a common drive, operated either by a handwheel or a
motor.

The pilot switch serves a purpose similar to that of the pilot switch
on the standard switchboard. That is, one throw of the pilot switch
connects the circuit directly to a hot-bus connection so that the circuit
may be controlled independently of any of the master set-ups. A
second position of this switch connects the circuit through a color

Jan., 1932] CONTROL OF THEATER LIGHTING 93

master control so that it is possible to control an entire color by
means of a single color master. The middle position of the switch is
the off position. The pilot lamp is wired in connection with the
pilot switch so as to indicate when the circuit is hot; that is, when
the switch is thrown directly to the hot bus position or when the
pilot switch is thrown to the color master position and the color
master is energized.

The selector switch is used for transferring the control from the
individual control potentiometer to the preset potentiometer. In
one position of this switch the grid lead from the tube circuit is
connected to the moving arm of the individual potentiometer. In
this position, the output of the tube circuit may be controlled by
manipulating the individual potentiometer, and it is not affected
by any changes made in the preset potentiometer. In the second
position of the selector switch the grid lead is connected through the
fader to the preset potentiometer. In this position the circuit is
controlled through the preset control by means of which the intensi-
ties of the circuit may be set for five scenes in advance.

Regarding the operation of the preset potentiometer, let us first go
back to the operation of the tube unit. As has been previously
stated, the output of the tube unit, and consequently the intensity of
the lighting circuit, is changed by altering the grid potential on the
control tube. The grid potential is obtained through a system of
potentiometers from the d-c. control source as indicated on the
attached diagram. (Fig. 1.)

This preset operation is accomplished by two methods, one of
which allows a gradual fading from one effect to the next and a second
of which permits the operator to flash immediately from the effect
in progress to any other effect that has been previously set up. These
two operations are accomplished by means of the dimming fader and
the multi-contact flashing relays as described in the following para-
graphs.

One dimming fader is provided for each circuit. In addition to
this there are five multi-contact flashing relays, 1 to 5, and a fader
disconnecting relay (/). (Fig. 1.) One contact of each relay is con-
nected in each control circuit as shown on the diagram. The normal
position of the relays for operating through the dimming fader is to
have relays 1 to 5 open and relay (/) closed. Now to set up circuit
No. 1 for full intensity the slider of preset potentiometer No. 1 is
moved to the positive end of the potentiometer. To set up circuit

BURT S. BURKE

[J. S. M. P. E.

No. 2 for black-out the slider is moved to the negative end. In
order to obtain intermediate intensities on the other scenes, the
sliders are moved to intermediate positions corresponding to the
intensities desired.

Now assuming that the switchboard is operating on scene No. 1,
the pointer of the dimming fader will be connected to point 1 on
this piece of apparatus. In order to transfer to scene No. 2, the
dimming fader, which consists of a unit for each circuit to be con-
trolled on a common drive is moved either by a handwheel or by
means of a motor drive to position No. 2. It can be seen, therefore,

!»>• LW

H h h h h

MULTI CONTACT #£LArS

FIG. 1.

Schematic diagram of scene flashing control thermionic type
theater switchboard.

that since point No. 1 of the dimming fader is connected to the slider
of preset potentiometer No. 1 and is therefore at the same potential
as its preset for this scene, and since point No. 2 is connected to the
slider of preset potentiometer No. 2, there is a gradual transition from
the potential set for preset No. 1 to that for preset No. 2. This gives
a gradual change on the potential impressed on the grid circuit of the
tube unit, and, therefore, a proportional change in the lighting
circuit.

A similar operation is also performed to transfer from scene No. 2
to scene No. 3. It is, furthermore, possible with this type of equip-

Jan., 1832] CONTROL OF THEATER LIGHTING 95

ment to set up the scenes not in use for additional effects without
affecting the scene in progress.

The purpose of the scene flashing equipment is to allow the operator
to transfer immediately from the effect that may be set up from one
scene to the effect set up for any other scene and have all the lighting
circuits come to the desired preset intensity. This is accomplished
by means of the relays, 1, 2, 3, 4, 5 and / which operate to disconnect
the grid lead from the scene fader and connect it to the preset
potentiometer associated with the relay selected. Thus the potential
which has already been set up on this potentiometer is applied to the
grid lead of the tube unit and a corresponding intensity of the lighting
circuit results. When any other relay is operated, the relay previously
closed is automatically disconnected by means of the switch keys,
which are interlocked.

In order to obtain color master operation of any of the circuits, as
has been previously described, the pilot switch is thrown to the color
master position. This transfers the lead to the positive end of the
control potentiometer from the positive bus and connects it to the
sliding arm of the color master potentiometer. Thus it may be seen
that the voltage on all the potentiometers connected to this particular
color master is varied by moving its sliding arm. Consequently, a
proportional change in the voltage impressed on the sliding arm of the
individual potentiometers is obtained, which results in a proportional
change in the lighting intensity of the circuits connected to these
controls. Thus, if one of the circuits connected to this color master is
at full brilliancy, a second at 3/4 brilliancy, a third at l/s brilliancy,
etc., should these circuits be dimmed out by the color master they
would start dimming at the same time, and proportionally change, so
that they would reach the black-out position at the same time. This
is in contrast to the operation of the standard interlock type of color
master in which a similar operation would result in the circuit at full
brilliancy being dimmed until it corresponded to the one at */4
brilliancy, at which point the second circuit would interlock with the
color master and both these circuits would travel until they reached
half brilliancy, where the third circuit would interlock and finally
all would black-out together. This often results in a spotty effect and
is undesirable. By using an electrical color master rather than a
mechanical color master, a proportional dimming effect is accom-
plished as previously described.

In order to obtain grand master control, the pilot switch on the

BURT S. BURKE

[J. S. M. P. E.

color master section is transferred from the hot-bus position to the
grand master position which connects the circuits on the color master
to a master generator. By varying the voltage on this master

FIG. 2. Thermionic lighting control console, Severance Hall,
Cleveland, Ohio.

generator, the potential impressed on the color master is varied, thus
causing a proportional change similar to that previously described for
the circuits connected to the grand master.

Jan., 1932] CONTROL OF THEATER LIGHTING 97

Motor operation is provided for the dimming fader in the Los
Angeles theater so that it is possible for the operator to change from one
effect to another simply by pushing a small telephone switch starting
the motor drive. This motor drive is so provided with limit switches
that it will travel to the succeeding scene at which point it will stop,
and will not start again until the operator pushes the "start" button.

FIG. 3. 5-Scene thermionic control board for stage
lighting control, Los Angeles Theater, Los Angeles, Cali-
fornia.

This allows an easy means of obtaining remote control of the intensity
of all the lighting circuits in the theater. In the old type of multi-
preset board it was possible to obtain remote control scene changes,
but it was impossible to preset the intensities, as it was necessary for
the operator to set the dimmers beforehand or to change their setting
when transferring from one effect to the other. By using the thermi-

98 BURT S. BURKE [J. S. M. P. E.

onic tube control, it is possible to preset the dimming as well as to
preset the circuits which are to be used and, furthermore, to obtain
remote control of these circuits if it is desired. This is of special
advantage in motion picture houses where it is desirable to control
the light from the projection room and at the same time have a switch-
board on the stage floor which can be used in case of stage presentation
work.

In the Los Angeles theater, the remote control board at the stage
floor allows the stage switchboard operator to have full control of the
color masters for the auditorium circuit, and in addition to this allows
him to change the lighting effects on the auditorium for five presets
which had previously been determined by means of the switchboard
located in the projection room. Furthermore, by means of the color
masters, it is possible for the operator to dim out any particular color
from any scene that had previously been preset, thus giving a very
flexible control.

The previous description is typical of this type of control. Due
to the rapid development in the art, at least one other scheme has
already been conceived. Instead of a grid-glow type tube, a vacuum
tube of rather large plate capacity is used and a motor generator set
supplies 500 volts d-c. to the plates. The output is regulated by grid
control of these tubes. That is, the 500-volt supply from the genera-
tor is connected in series with the vacuum tube and the d-c. coil of
the reactor. Thus, by varying the impedance of the vacuum tube by
change of grid potential, the amount of direct current that is allowed
to pass through the d-c. coil of the reactor is changed. The control
equipment, that is, the potentiometer set-up is practically a duplicate
of that used in the Los Angeles theater, the new developments
affecting the tube units rather than the control.

Another advance has been in the method of control by means of
which it is now possible for a switchboard to be built wherein the
operator may change from one scene to any other scene and get a
gradual fading effect from the one to the other, or to obtain a flashing
effect as previously described. The Los Angeles theater was built
prior to this development, so that it was necessary to fade from one
scene to the next succeeding scene. The new development allows a
much more flexible control due to the fact that in stage presentation
work it is often desirable to repeat an effect ; with this type of control
it is possible to do this as often as is desired, and to fade into this
effect from any other that may be in progress.

Jan., 1932] CONTROL OF THEATER LIGHTING 99

Another interesting development that has recently been brought
forth, the first application of which is for the control lighting of the
Buckingham Fountain in Chicago, provides a continuous preset pro-

FIG. 4. Reactor and thermionic unit rack, Los Angeles Theater,
Los Angeles, California.

gram which is laid out in advance for an evening's performance.
This is accomplished by means of an insulating track on which there
has been drawn a conducting path. This moves and is continuously in

100 BURT S. BURKE

contact with a potentiometer similar to the control potentiometers
for the tube units. By varying the location of the conducting strip
on the moving track, a variable preset potential is obtained which
correspondingly changes the output of the tube unit, and a change
in the intensity of the lighting circuit results. While this is developed
for floodlighting control it may be used for such an application as
varying the lighting of a theater according to a definite program for
the overture. It could also be used for providing a light change
program for the patrons at the time between the opening of the
house and the beginning of the performance. This program is motor
driven, and can be started and allowed to run for a definite period of
time after which, by means of a transfer relay, the control can be trans-
ferred back to the regular stage switchboard for use in connection with
the picture or the stage presentation.

In summarizing it may be said that the application of thermionic
tubes to theater dimming has made possible a stage board giving the
following very desirable features:

(1) Presetting of intensity for any desired number of scenes.

(2) Proportional dimming.

(3) A light compact board using telephone switches, thus insuring ease of

operation with added assurance of proved reliability of this type of equip-
ment.

(4) Low control voltage (less than 50 volts, d-c.) allows use of telephone

cable for control wiring.

(5) Remote control easily added.

In fact, the field of application of tube control to lighting is in its
very infancy, and due to the rapid development in tubes, as we have
witnessed in the past in the radio field, a great deal may be expected of
this type of equipment.

A PORTABLE NON-INTERMITTENT CINE PROJECTOR*

Summary.— A portable projector made by the £tablissement Gaumont Franco-
Film Aubert is described. The projector is of very small weight and is arranged
for carrying in a case. The film moves with a constant motion past the axis of the
light source and the projection lens, the image being maintained stationary upon
the screen by a combination of the movement with an optical "compensator." It
is claimed that due to these features, wearing of the film has been very much reduced
and the motion is extremely silent in operation. The article describes briefly the
optical principle of the motion, how the principle is applied, and the construction
and assembly of the apparatus.

The "Simplicine"" is a cin6 projector for standard film, self-con-
tained and complete, yet small enough in bulk and weight to be port-
able. The whole projector is enclosed in a metal casing and can be
carried easily on a sling strap. Its erection is almost instantaneous
and its manipulation so simple that no special experience is required
for its use.

The chief importance of this machine, particularly so far as the
non-professional user is concerned, is that it employs the principle of
constant movement projection. The film moves with a uniform mo-
tion across the axis of the light source and the projection lens. This
is a vital difference from the usual intermittent projector, in which a
Maltese cross or other mechanism is used to drag the film into position
and then bring it momentarily to a standstill in the gate of the
machine. In the new projector the image is kept stationary on the
screen by means of a special combination of the movement with an
optical device termed a "compensator." One greater advantage of
such a system is the very much reduced wear on the film perforations
owing to the elimination of the violent and repeated tugs to which
films are subjected in ordinary types of projectors. Film is said to
last five times as long when it is run in this continuous manner. To
this advantage may be added the not less important one that abso-
lutely silent mechanism can be obtained when all the moving parts are
given nothing but continuous rotary movement, as is the case in the
"SimplicineV'

* Translated from Revue d'Optique, 10 (April, 1931). No. 4, p. 178.

101

102

NON-INTERMITTENT PROJECTOR

[J. S. M. P. E.

The Optical Principle. — Fig. 1 represents a film moving in a down-
ward direction and carrying a series of images 1, 2, 3, etc. Imagine
that in front of these images is a series of similar lenses Oi, 02, 03, etc.,
each having its focal point in the plane of one of the images, and
suppose this chain of lenses to move in a direction parallel with the
film and at the same speed. If the beams of parallel light so formed
meet a fixed lens C the images of the different elements of the film
will be superimposed in the focal plane of this lens. Indeed, if we
consider an element formed by an image on the film and the corre-
sponding lens, the image of a point of this element given by C will
have its position, in the focal plane of C, determined solely by a
straight line passing through the optical center of C and parallel to
the straight line joining the given point in the element to the optical
center of the corresponding lens. Now this straight line as it moves
remains parallel, consequently the final image is fixed. This is true

o,t

•*&•

' C

FIG. 1. Diagram illustrating the optical principle.

for all images of the points of the element of the film. The straight
lines joining corresponding points of the element to the optical centers
of the corresponding lens being parallel, the images of successive
elements are superimposed in the focal plane of C. Hence the pro-
jection screen E is made to take the position of the focal plane of C
and focusing for various distances is obtained by providing a set of
lenses C of different focal lengths.

How the Principle Is Applied. — The realization of this principle in
actual fact has been achieved in the following manner: The lenses are
set round the periphery of a cylindrical drum T (Fig. 2), which is
free to turn on its axis. T is made to rotate by the fact that the film
catches a tooth D formed on the drum and carries T round with its
own movement. The lenses therefore move at the same speed as the
film. Light passing through the illuminated film reaches the lenses
O after traversing a prism P (Fig. 2), which is formed integrally

Jan., 1932]

NON-INTERMITTENT PROJECTOR

103

with the fixed axis of the drum. This prism has two reflecting sur-
faces MI and Mz set perpendicular to one another. The system thus
produced is that indicated in Fig. 1, with the difference that the film
and the lenses do not travel in a straight path but follow curves
of the same radius. This difference has a practically negligible effect
on the quality of the images, assuming that the film is illuminated
only over the length of two images. This means that two images
and two only of the chain of lenses are actually utilized, film elements
and lenses which have any appreciable inclination to the normal
being kept out of action. Furthermore, the projected image shows

FIG. 2. The construction of the drum
carrying the lenses on its periphery.

its maximum illumination at the moment at which the corresponding
lens has its axis coincident with the axis of projection, and the effect
of this is to reduce greatly the aberrations of the images thrown by
adjacent lenses which are slightly inclined. This mechanism is
exceedingly simple: it consists of a single component moving with a
continuous rotary movement at low speed— 80 revolutions a minute
for a projection speed of 16 pictures a second. Wear is therefore
reduced to a minimum and the running is quite noiseless.

The Projector Described— The "Simplicine*" has been given the
form of a rectangular case, the top and side of which are formed with

104 NON-INTERMITTENT PROJECTOR [J. S. M. P. E.

hinged swinging sections which are raised vertically for use. All the
mechanism is then made visible.

The feed-reel, 3 (Figs. 4 and 5), with its pulley, 4, is then fixed on
this raised portion of the casing. The film passes under the feed
sprockets, 5, and on to the drum T carrying the compensating lenses ;
then on to the toothed sprocket 6, the take-up reel 7, rotated by its
driving pulley 8.

Masking the film on the screen is effected thus: when the lever,
9 is pressed downward, the roller, 10 is pushed up between the two
pressure rollers, thus raising the film and raising the roller, 11. If
the pressure on this lever is released, the roller 11 returns into contact

FIG. 3. External view of projector.

FIG. 4. View of projector opened
for use, showing internal arrange-
ment.

with the drum, and the loop formed by the film is taken up by the
movement of the drum. The importance of the formation of this
loop is that in this operation the film advances by one perforation.
Masking is thus effected by the displacement of the image by an
amount equal to a quarter of its height.

Motor Drive and Lighting. — The driving parts are carried on a
fixed aluminum platform and can be removed as one unit from the
box. This block consists of an electric motor, 12; a pulley drive, 13,
reversible for rewinding; and a transformer, 14, to feed the lamp 15.
In front of this assembly, against one of the panels of the box, are
mounted side by side two rheostats, for the lamp and for the motor,

Jan., 1932]

NON-INTERMITTENT PROJECTOR

105

with finger controls, 16 and 17, for their adjustment; and an am-
meter, 19, for the control of the lighting system. A plug let into the
casing provides for the connection of the apparatus to a source of
electric power.

The lamp used is a 225-watt Phillips, taking 30 amperes, at 71A
volts. The beam is of about 525 cp. in a horizontal direction per-

FIG. 5. Diagram of internal arrangement.

pendicular to the tungsten filament. The lamp is conical in shape
and works upside down in order to avoid the blackening of the bulb
around the filament. The lamp is suitably supported and adjusted.
The optical system includes a two-lens condenser 25, and spherical
mirror 26 in line with the axis of the filament.

As a safety device there is a wire gauze, 27, which is arranged to

106

NON-INTERMITTENT PROJECTOR

come into place automatically between the light source and the
condenser when the film is stationary. This takes place by a centrif-
ugal action. Its effect is to protect the film from any dangerous
degree of heating without restricting the light unduly when the
machine is used for still projection from selected pictures.

As regards focusing, the apparatus possesses five collimating
lenses C (Fig. 1), arranged on a rotatable disk, and by means of these
the image can be focused on a screen at any distance from 6 to 32
feet. The milled edge of this disk projects through the casing at the
side so that it can be rotated by the finger, and a spring detent sets
it in accurately centered position whichever lens is in action. A

FIG. 6. Showing the method of
rewinding.

rectangular window in the front wall of the apparatus, made to allow
the beam to pass, is fitted with two sliding covers adjustable verti-
cally to cut off parasitic images which would otherwise be thrown on
the screen.

Re-winding at the end of the projection is very simply carried out.
The film is released from the drum and from the guiding sprockets, so
that it runs as shown in Fig. 6. An adjustment is then made, to
allow the take-up reel to turn freely on its axis and to fix the feed-
reel to its axis. The full take-up reel thus becomes the feed-reel,
and vice versa. The motor, running just as in projection, then
rapidly re-winds the film, leaving it ready to project again.

COMMITTEE ACTIVITIES

REPORT OF THE PROJECTION PRACTICE COMMITTEE*

The Projection Practice Committee wishes to direct attention to
what it considers one of the foremost causes of waste and monetary
loss suffered by the motion picture industry, namely, the mutilation
of positive prints. This mutilation not only results in a consider-
ably shortened life of the individual print, which is serious enough
in itself, but in addition to this, it is impossible to obtain the optimum
screen results, which are so highly important in creating the proper
illusion so necessary to the motion picture play. Both picture and
sound are affected by mutilation of film.

It is generally understood that the mutilation of film is frequently
due to the maladjustment of projector parts, wearing of projector
parts, accumulation of emulsion during projection, excessive oiling
of projector or leakage of oil, and careless handling of film. The
Projection Practice Committee is of the opinion that there is urgent
need for the establishment of standards dealing with the various
tensions to which the film should be subjected during projection,
the clearances of adjacent projector parts and sound apparatus,
allowable tolerances, and the amount of wear projector parts may suffer
without impairing the quality of the picture or causing mutilation of
film.

The committee, therefore, plans to conduct a thorough investigation
which will be nation wide in scope, with the view of obtaining all
necessary data for submittance to the Society for the purpose of
adopting such standards. In order to accomplish this, the committee
requests the earnest cooperation and support of the Society as a
whole, as well as of associated individuals and organizations. Their
assistance will be needed as this work will be of considerable mag-
nitude and should, when completed, prove invaluable to the industry.

The Committee wishes also to call attention at this time to the lack
of uniformity in the processing of prints, which constitutes another

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.

107

108 PROJECTION PRACTICE COMMITTEE [J. S. M. p. E.

serious loss. In regard to the processing of film, there seems to be no
standard for this work at the present time. One producer uses a
certain method of processing film; another producer simply waxes
the margins of the print; and a third producer does not process
the print at all. This condition works a hardship on all concerned,
inasmuch as it frequently happens that the producer who has
processed his product suffers by reason of the fact that the theater
uses unprocessed film at the same time. This evil adversely affects
both the sound quality and the quality of the picture.

It is well known that with unprocessed film there is a tendency to
accumulate emulsion at the tension points in the projector. Forma-
tion of emulsion greatly increases the tension applied to the film
and imposes a strain on the sprocket holes. Occasionally a positive
print is irreparably damaged during its first projection. The Pro-
jection Practice Committee recommends that a thorough investiga-
tion to find the best method or methods of processing film be con-
ducted by a designated committee of the Society so that such methods
may be recommended as a standard for the industry.

Unless such a standard is adopted, generally accepted, and put into
use by the producers of film, the industry will continue to suffer
the great loss now occasioned through faulty (or the lack of) process-
ing methods, and such benefits which should accrue through the
adoption of the standards relating to projector tensions, adjust-
ments, etc., would be largely nullified. In the opinion of the Pro-
jection Practice Committee, such a work is one of the most impor-
tant contributions the Society could make to the industry.

RESOLUTION

The Projection Practice Committee wishes to include in the records of the
Society a statement of its appreciation of the splendid work and cooperation
which President Crabtree extended to this Committee and, also, for his realiza-
tion of the important role which practical projection plays hi the motion picture
industry.

Through President Crabtree's foresight, initiative, and efforts, a committee
to deal with the practical problems of projection was formed for the first time
in the history of the industry, and specific problems greatly in need of attention
and correction were brought to the light of day and taken under consideration.

Therefore, we, the Projection Practice Committee, gratefully acknowledge
what President Crabtree has done for the craft, for the Society, and for the
industry at large, and extend to him our thanks and a vote of confidence in his
conduct of the affairs of the Society.

Jan., 1932] PROJECTION PRACTICE COMMITTEE 109

HARRY RUBIN, Chairman

THAD. C. BARROWS R. H. McCuixouGH

G. C. EDWARDS P. A. McGuiRE

SAM GLAUBER RUDOLPH MIEHLING

J. H. GOLDBERG F. H. RICHARDSON

CHAUNCEY GREENE MAX RUBEN

HERBERT GRIFFIN H. B. SANTEE

JESSE J. HOPKINS L. M. TOWNSEND

DISCUSSION

MR. McGuiRE: For quite a few years I was one of those who vigorously
protested against the neglect of projection by this Society, but no longer have
I any cause for complaint as we have our own Projection Practice Committee
and it is up to ourselves to make good. In discussing the Report of the Pro-
jection Practice Committee it is not my intention to complain or criticize, but
to offer some suggestions which I hope will be helpful to the motion picture
industry and a benefit to the Society of Motion Picture Engineers. I ask you
to be patient because much of what I will say is to be of a somewhat general
nature and, perhaps out of place in the proceedings of this Society. But most
of the papers and much of the discussion of the Society are more or less incompre-
hensible or relatively unimportant to some part of the membership of this organi-
zation. In order to deal only with subjects which would be of interest to every-
one it might be necessary to hold a hundred conventions or divide the meetings
into an equally large number of groups. In its proceedings, the Society of
Motion Picture Engineers must give some attention to invention, development,
manufacture, maintenance and operation, the electrical, chemical, and mechanical
divisions of the industry, visual and sound recording and reproducing, and
always theory and practice. These are rough classifications, but give a general
idea of the vast field the Society must cover.

The Society of Motion Picture Engineers is not a scientific body seeking ab-
stract truth, but a technical organization with a very definite commercial back-
ground. When we lose sight of the fact that we are part of the motion picture
industry we fail to realize the true purpose of the Society. It, therefore, seems
to me that anything the Society can do to render a practical service to the in-
dustry should result in the organization receiving increased support. The
benefits that the industry has derived from the Society of Motion Picture Engi-
neers have not always been recognized because they were often of an extremely
indirect and intangible nature. The fact that the Society has for many years
focused attention upon the technical side of the motion picture industry and to
some extent has won the interest of non-technically minded executives is in
itself a great achievement. The executives of this industry have never given
the Society adequate support, and I believe that the producers and exhibitors
have contributed more to a single activity of another organization in this
field than they have to the Society of Motion Picture Engineers in its entire
history.

The Society is facing new conditions and it is desirable that the service which
it renders the industry should be more direct and more obvious. If this can be

110 PROJECTION PRACTICE COMMITTEE [J. S. M. P. E.

done, the Society will receive increased support and be in a better position to
carry on its important work. This organization is in a particularly strong
position to secure technical data regarding the cause and prevention of film
mutilation. Various attempts have been made to get this information, but there
is good reason to believe that the results have not been entirely satisfactory.

Someone has said that "science is common sense made exact." The Projection
Practice Committee will conduct a scientific survey, collecting the facts sys-
tematically and thoroughly, and present them in an authoritative report.
When this is done definite action should result and the Society will have rendered
a service comprehensible in terms of dollars and cents.

The work we are undertaking, however, will involve considerable time and
expense, and should receive adequate support from the Society as well as the
industry. It is an unfortunate fact that the industry does not take proper
interest in the collective thought developed by such an organization as the
Society of Motion Picture Engineers. Progressive projectionists in this organi-
zation, and their own projection societies, are constantly giving their own time
to do valuable technical work without receiving the least recognition from the
executives of their own firms. Conceding that this is a period in which executives
are very properly insisting upon economies, it nevertheless seems unwise to
ignore totally all the collective effort for the betterment of the industry.

Back of the artistic side of the motion picture industry is a vast technical
field whose work offers infinite opportunity for flaws and failures. Motion pic-
tures provide entertainment and education through chemical, mechanical, and
electrical processes. What the public pays for is not the product of a single
commercial organization, and it is important that the Society of Motion Picture
Engineers should bring this to the attention of the industry — emphasize the
interdependence of the various departments and point out the need for coordi-
nation. In all work which is not of a competitive nature the industry benefits
tremendously from the collective thought developed in such organizations as
the Society of Motion Picture Engineers. I sincerely hope that a way will be
found to encourage and finance adequately the efforts of the Projection Practice
Committee to find the cause and prevention of film mutilation. The men on
this Committee have the technical and practical experience to do the work.
Their report should result in a tremendous saving through prevention of waste
and the improvement in screen presentation.

PRESIDENT CRABTREE: I indorse Mr. McGu ire's remarks one hundred per cent.
Of course the world was not made in a day. But it is encouraging that the pro-
ducers have shown a much greater willingness to do things for us at this Con-
vention than at any time previously.

Will Mr. Griffin give us a few details as to how the stoppages occurred in the
projection room?

MR. GRIFFIN: There was only one cause, Mr. President, and that was the
processing. Emulsion, or whatever was on the film, supposedly to prevent its
"seizing up" during transit through the mechanism, did not prevent it. In
some cases it was wax, and in some cases something else.

PRESIDENT CRABTREE • Where did it seize?

MR. GRIFFIN: In the gates. It can seize anywhere in transit, wherever
there is tension — at the picture gate or the sound gate. This time it happened

Jan., 1932] PROJECTION PRACTICE COMMITTEE 111

to be at the sound gate. It is not the fault of the manufacturer of the equipment,
because one can run a film through that has been run two or three times, and
properly processed, and it will not cause any trouble at all. We could have run
the film through without stopping; it was a new print, however, and we wanted
to save it. And in as much as it was not a very serious matter to stop the picture
here, as would have been the case in a theater, we stopped it. But the sound
was terrible in some cases, caused by a piling up of the wax, behind the film,
thus changing the thickness of the scanning beam.

PRESIDENT CRABTREE: Was this a new machine?

MR. GRIFFIN: It was not new in the sense that it had never been run before.
Films had been run through it on different occasions, but the equipment to all
intents and purposes is new.

PRESIDENT CRABTREE: In the case of two metal surfaces, one of which is
polished to an extremely high degree with rouge, and one which is not polished,
the polished surface will not pick up as much gelatin or emulsion as the rougher one.
I was wondering, therefore, that if this machine had been a little older, would
the trouble have occurred?

MR. GRIFFIN: The finishes on all parts that come into contact with film
are finished with rouge, and I believe that RCA uses crocus-cloth for polishing.
I don't know of anything better. The surfaces are highly polished and burnished,

MR. SUMNER: I happen to be an exhibitor, and this report of the Committee
was very interesting to me. We happen to run a theater that is called a "first
subsequent run;" that is, we run after the key point in this district, which is
Boston. I have attended a number of the conventions, and have heard the
reports from the various specialists in the studios; and I realize the great amount
of thought and work that is put into the pictures, the great mass of work that
has been done to accomplish perfect sound, and so forth. And yet, when these
prints get to the theaters, the greater part of that work has been ruined by
improper handling of film. As an exhibitor, I wish to state that I believe that
the work that has been begun by this Committee is most important. I want
to urge them not to stop with the problem of processing film. They must go.
much further than that.

In spite of the noiseless recording system, the prints reach the theaters so
dirty and scratched that the work of noiseless recording has almost gone for naught.

I think this Committee is one of the most important factors in the organization
and I want to urge that it be given all possible support in its work.

PRESIDENT CRABTREE: I should like to ask Mr. Griffin: Was the accumulation
of emulsion due to friction along the perforations, or at some portion of the
picture area? In other words, is it necessary to process the entire surface of
the film, or merely the edges of the perforations?

MR. GRIFFIN: For projection purposes it is necessary only to process— or
lubricate, as it may be called — the edges of the film in the sprocket hole area.

PRESIDENT CRABTREE: Was the film in question lubricated or "processed?"

MR. GRIFFIN: I cannot answer that. I do not know either the processes
or who does the processing. I only know what occurs during projection.

MR. FAULKNER: Four different prints caused the trouble, and each one
of the four had four different applications and four different kinds of chemicals
on them. The gathering of emulsion on three different prints that I looked at

112 PROJECTION PRACTICE COMMITTEE [j. s. M. P. E.

was identically in the same spot, showing that no matter what caused it to gather,
it did so in exactly the same place on the film. I did not see the fourth print
but the sound quality and the way in which it behaved were similar.

As Mr. Griffin says, as far as passing the film through a projector is concerned,
it is only necessary to lubricate the margin of the film. The emulsion that is
on film, unless the metal parts with which it comes into contact are lubricated,
is quite likely to stick. Therefore, the film is lubricated for the purpose of
keeping the tension shoes lubricated.

Mr. Rubin asked me to present to you his idea that "processing" is an in-
correct term to use for this process. He wants to find a name for waxing, treat-
ing, processing or ' 'whatnot, ' ' and to standardize that name. I went to a dictionary
and ran down every name I could think of. I have a great number of them,
none of which I think would be appropriate, except perhaps "treatment" or
"finishing" or the like. "Processing" is used to indicate anything that may
happen to film from the time it is printed to the time it is developed for screening.

PRESIDENT CRABTREE: Why not use the word "conditioning?"

MR. FAULKNER: Some of the names I accumulated are: hardening, com-
pleting, seasoning, curing, impregnating, finishing, duratizing, dura-proofing,
inuring, toughening, preserving, protecting, treating, perfecting treatment.
None of these I think would be satisfactory except perhaps "conditioning" or
"treating." I do not like "processing," nor does Mr. Rubin.

MR. McGuiRE: I ask you not to exaggerate the importance of "processing"
merely because it has received special attention in this discussion. It is a serious
problem, but we shall have other important subjects to consider in our efforts
to find the cause and prevention of film mutilation. There has been much talk
in the past about film mutilation and various organizations have dealt with it
rather unsuccessfully.

The Projection Practice Committee is starting out with the idea that there
seems to be an evil which is called film mutilation, but that it knows absolutely
nothing about its cause and prevention. We hope to be able to gather some
data in the next six or twelve months, which will save the motion picture industry
a tremendous sum every year and greatly improve the quality of screen pres-
entation.

MR. J. CRABTREE: I think a little more attention to the projector is what is
necessary. I often project green film, and find that as long as the projector is
kept in shape, little trouble is experienced. Mr. Faulkner pointed out that
last night the accumulation occurred in the same spot in each case, which goes
to show that there is a high spot somewhere. One cannot expect lubrication
to take care of all high spots. Eliminate the high spots, and the lubrication
won't be so necessary.

MR. GRIFFIN: I must take exception to that. I don't know under what
conditions Mr. Crabtree projects his prints, but I defy anybody to take a piece
of green film off the drying rack and project it under conditions existing in the
theater today and not have it seize up, no matter how well the projector is de-
signed.

PRESIDENT CRABTREE: Are you speaking now of a film to the edges of which
wax has been applied?

MR. GRIFFIN; Mr. Crabtree said he would use it without treatment — right

Jan., 1932] PROJECTION THEORY COMMITTEE 113

out of the laboratory. It is not waxed there. Now, waxing is not the solution,
apparently, because the wax peels off and rolls up. With the old silent machines,
waxing was all right. Today we have sound. The wax rolls off, gets in the
sprocket holes, and is carried to the sound gate, where it either leaves the film
or raises it off the sound gate.

PRESIDENT CRABTREE: You are speaking of the old method of waxing with
solid wax?

MR. GRIFFIN: Yes.

PRESIDENT CRABTREE: You should use a solution of wax in a solvent. It
is only necessary to put on a layer of wax a millionth or so of an inch thick, to
provide the necessary lubrication.

MR. GRIFFIN: I have seen, in cases where the film is put on a rewinding device,
two pieces of tallow right at the sprocket holes, over which this film is drawn.
The projectionist should be taught not to do a thing like that. We must find
a proper means of treating the filtri so that during projection under high amperages
it does not seize in the tension parts of the projector.

PRESIDENT CRABTREE: Of course, Mr. Crabtree is not projecting under the
high amperages that you speak of.

MR. FAULKNER: When the film comes off the drying cabinets and is pro-
jected for inspection, felt runners are used in some places, and I know one labora-
tory that does not use them. They never scratch film, but it is due to the fact
that there is no heat on them.

MR. GRIFFIN: We supply thousands of different types of runners to the
laboratories of studios, and I know how they work. They use a Mazda lamp,
and very little light.

PRESIDENT CRABTREE: I happen to have done a considerable amount of
research on the lubrication of film. Our researches have shown that if you
have even the merest trace of wax or oil or grease or any lubricant, on the film,
it makes a tremendous difference in the ease with which it passes through the
projector. To date we have not found that any special processing treatment is
any better with regard to lubrication.

REPORT OF THE PROJECTION THEORY COMMITTEE
SUBCOMMITTEE ON LITERATURE

At the Spring Convention at Hollywood a report was made of
the activities of the Projection Theory Committee. A subcom-
mittee to examine the literature of the subject was formed, consist-
ing of C. Tuttle, F. K. Moss, and H. P. Gage, Chairman. The
present plan of this Committee is to prepare a tutorial paper on the
progress of the optics of motion picture projection, based principally
on the papers published by of the Society of Motion Picture
Engineers, but also referring to significant papers in other publica-
tions.

114 PROJECTION THEORY COMMITTEE [J. S. M. P. E.

A letter from F. K. Moss states, "I have devoted some time in sur-
veying the literature on the effect of motion pictures upon the eyes.
As I progressed in my survey it became increasingly apparent to me
that available data on the effect of motion pictures upon the eyes
were largely negative in character. In other words, pictures made
according to the best of modern practice had little if any observable
deleterious effect upon the eyes. In the past, when pictures pre-
sented excessive brightness contrasts, unsteadiness and flicker, -there
was no doubt that they were the cause of ocular strain and fatigue.
These objectionable characteristics seem to have been reduced to a
point where they cease to be important in the better grade of pictures.

"I also approached the problem more or less directly from the view-
point of general physiological optics. Such an analysis indicates,
for example, that the brightness contrasts on the screen and with the
general surroundings, are not of such an order as to induce unusual
degrees of ocular fatigue. Hence the conclusions reached by scientific
considerations and those resulting from actual experience are in agree-
ment. Since the subject is largely one of eye-strain or 'eye-fatigue'
which has never been satisfactorily measured, a quantitative discus-
sion is impossible. In brief, these effects upon the eyes become
important in cases where projection is faulty."

Mr. Tuttle under date of April 28 sent a list of forty titles containing
significant information on the subjects of Illumination, Optics, Pro-
jection Angle, Mechanics of Projectors, Aberration of Lenses, Pro-
jection under Special Conditions, and Visual Angle.

Subcommittee on Literature
H. P. GAGE, Chairman
F. K. Moss
C. TUTTLE

DISCUSSION

MR. MATTHEWS: In connection with the work of the Progress Committee, it is
worthy of mention that a considerable amount of information has been published
on the subject of visual fatigue in motion picture theaters, in the International
Review of Educational Cinematography. A series of measurements were made of a
great many school children in theaters in Italy, giving data that might be con-
sidered by the Committee. There is a series of four or five articles in this publi-
cation.

MR. MURRAY: Does the work of the Committee include a search of the
literature in regard to the psychological effects involved in the projection of
motion pictures in color?

MR. GAGE : Mr. Moss is studying the literature dealing with the effects on the

Jan., 1932] PROJECTION THEORY COMMITTEE 115

eye. The Committee is considering more generally the possible deleterious effects
rather than the whole group of psychological effects, which constitutes such an
unlimited field that little would be accomplished if it were to be considered in its
broadest aspect.

MR. MURRAY: I have heard complaints that colored pictures produce eye fa-
tigue that I have not heard in connection with black and white. Have these been
considered by the Committee?

PRESIDENT CRABTREE : In the case of the old two-color additive Kinemacolor
pictures, I would say that they caused fatigue. With modern two-color subtrac-
tive pictures fatigue may have been caused by the fact that some of the pictures
were out of focus. They lacked definition, and the person viewing them did not
know whether it was his eye that was at fault or the picture. He assumed that
his eyes were at fault, and strained them in trying to focus the picture.

I asked for a vote the other day as to whether the colored pictures we saw one
evening this week caused any annoyance or eye-strain, and no one said that they
had any effect — they did not seem to notice any difference between the effects
produced by the colored and by the black-and-white pictures. Perhaps at this
time, if any one has thought it over and has the courage to say it gave him annoy-
ance, he might care to say something about it.

MR. J. CRABTREE: During the showing of the picture I thought I should be
able to view it to the end, but I had to close my eyes. When the next black-and-
white picture was projected, the annoyance entirely disappeared. Checking
with other people, no one else to whom I spoke seemed to have had the same ex-
perience. Apparently it was merely an idiosyncrasy. But the irritation was
undoubted hi my case.

MR. FALGE : Is it not true that with color in general, it is harder to focus on
some colors than on others, and that one experiences certain visual effects with
pictures in the blue and red ?

Another factor which is a function of eye-strain is the size of the picture. The
magnetoscope pictures, if viewed throughout an entire show, would be very hard
for those seated in the front rows, as their eyes have to chase back and forth across
the picture as in a three-ringed circus. Does not the addition of color to the
picture in general cause a reduction of its brightness? And from the standpoint
of lighting, it usually f ollows that a decrease hi brightness is less harmful to the
eyes than an increase. And aren't we back to the same subject we were on a min-
ute ago, that we have not enough brightness hi our pictures today, and that
that is harmful to the eyes?

PRESIDENT CRABTREE: We are talking of annoyance of a much higher order
of magnitude than what you have in mind.

MR. BURNETT: In color photography, it has been known for a long time
that the red colors are very harmful to the eye, while the green colors are not. In
most of the colored pictures that I have seen the reds have been predominant, caus-
ing a great deal of strain on eyes which have not been strong enough to stand it.
I do not think that this effect was as conspicuous the other night as heretofore.

Eye-strain can be also considered from the standpoint of brightness; but it is
the reds, I think, that cause the greatest trouble in color photography, as far
as eye-strain is concerned.

ABSTRACTS

Supply and Cost of 16-Mm. Film for the Home. F. S. IRBY. Electronics,
August, 1931, p. 48. An analysis of the various factors that contribute to the
cost of 16-mm. films for sound pictures in the home. The author considers that
if such films are to reach more than a very limited class market, the rental cost to
the consumer should not exceed $2 for a four or five reel feature picture. The
library must anticipate liquidation of the cost of the film in from twenty to
twenty-five rentals, which means that the cost of the film to the library must not
exceed $10 to $12 per reel. A. C. H.

Light-Valve Sound Recording. J. P. LIVADARY. Electronics, August, 1931,
p. 54. The third and final installment in a series of articles dealing with the fre-
quency distortion introduced by the finite width of the slit in recording. This
article is concerned chiefly with a mathematical analysis of the distortion intro-
duced in the light valve method of recording. In the conclusion, the author
summarizes the results of this and the preceding articles by comparing the various
methods of recording that have been studied; namely, the glow lamp method,
the single ribbon light valve method, the double ribbon light valve, and the
variable width method. He concludes that "from a practical standpoint, all
three systems are capable of high-grade recording, and any difference such as we
have shown will not become very apparent or objectionable until such time when
the film grain noise is suppressed and sound recording systems are capable of
commercially reproducing frequencies up to 10,000 cycles or over. Until then all
three systems will be competing on practically equal terms." A. C. H.

Dynamic Loud Speaker Design. J. E. GOETH. Electronics, August, 1931,
p. 66. A very elementary account of the magnetic circuit of dynamic loud speakers.
A second installment of this article will appear in a later issue. A. C. H.

A Rapid-Record Oscillograph. A. M. CURTIS AND I. E. COLE. Electronics,
August, 1931, p. 70. An oscillograph of the string galvanometer type that is
especially designed for the study of transient phenomena. A. C. H.

Noiseless Sound-on-Film Recording. GEORGE LEWIN. Electronics, Septem-
ber, 1931, p. 102. The author discusses the theory of noiseless sound-on-film
recording by the light valve. The subject will be treated from a practical stand-
point in a subsequent issue. A. C. H.

Dynamic Loud Speaker Design — II. J. E. GOETH. Electronics, September,
1931, p. 112. The second and final installment of an article concerned primarily
with the magnetic circuit of dynamic loud speakers. A. C. H.

Stage Equipment: An Outline of Modern Practice. W. L. TANN. Theater
Management, 27, December, 1931, p. 6. Essential stage equipment in an average
sized theater presenting straight pictures or pictures and stage performances is
described and illustrated. Modern advances in fire protection by asbestos cur-
tains and steel smoke pockets are pointed out. Various mechanical contrivances
for enlarging the screen to permit the showing of wide films are discussed. A
116

ABSTRACTS 117

notable advance in design of stage equipment is the silence with which the intri-
cate mechanism operates. E. P. J.

Room Noise Reduction for Improved Sound Reception. V. A. SCHLBNKBR.
Theater Management, 26, November, 1931, p. 3. Describes tests conducted to
determine the effect of extraneous noises on sound reproduction. Illustrates out-
side noises in typical theater before and after acoustical treatment of vestibule,
lobby, foyer, and exit doors. An oscillograph trace of three bands of noises re-
corded simultaneously in the street, lobby, and foyer of theater under discussion
reveals that while high and middle frequency bands are effectively silenced by
entrance doors, bands of low frequency enter the theater practically undiminished.

A chart showing the effect of various sensation levels expressed hi decibels
above minimum audibility of the human ear is discussed. The painful effect
produced by fader manipulation to produce audibility of picture sound above
room noise is indicated. E. P. J.

The Use of Rochelle Salt Crystals for Electrical Reproducers and Micro-
phones. C. B. SAWYER. Proc. IRE, 19, No. 11, November, 1931, p. 2020. A
brief history of the use of piezo-activity for acoustic uses is followed by a descrip-
tion of a cheap method of production of Rochelle salt crystals, used in the author's
experiments. The principle of opposition was used. Two Rochelle salt sec-
tions are cemented together so that upon application of an electrical field, one
section tends to expand and the other section tends to contract, thus amplifying
the resultant motion. The method of cutting Rochelle salt crystals for this work
is explained. Brief descriptions of Rochelle salt microphones, loud speakers, and
phonograph pick-ups are given. The Rochelle salt development has the follow-
ing outstanding advantages.

(1) Cheapness and simplicity.

(2) Long life.

(3) Flexibility of design.

(4) Generation of high voltages in input circuits.

(5) Directly matched with output tubes in output circuits.

(6) No necessity for an exciting field. A. H. H.

Trans-Lux Rear Stage Projection. W. MAYER. Theater Management and
Theater Engineering, 26, No. 22, October, 1931, p. 3. A non-technical discussion
of the Trans-Lux system of rear stage projection as installed in theaters. The
history of the system, various problems encountered and their solutions, and a de-
scription of the present installations give a concise outline of Trans-Lux. By
means of special lens and optical systems, no changes in the projector and sound
head mechanisms are necessary. Standard film is used and is threaded in the
projector in the standard way. The average distance between screen and pro-
jector is 13V2 feet. A- H- H-

Moving Coil Telephone Receivers and Microphones. E. C. WENTE AND A. L.
THURAS. Bell Telephone Tech. J., X, No. 4, October, 1931, p. 565. A descrip-
tion of a moving coil head receiver and a microphone. The mechanical construc-
tion is based on using light-weight materials for moving parts, thus giving greater
response over the frequency range. Theoretical and actual response are com-
pared. The sensitivity of the moving coil microphone was found to be about ten
db. higher than that of the condenser microphone. A. H. H.

1 18 ABSTRACTS [J. s. M. P. E.

Playing Light on a Thermionic Organ. W. C. FULTON Motion Picture
Herald, 104, No. 13, September 26, 1931, Section 2, p. 12. A description of a
unique lighting switchboard, built for the Severance Memorial Hall in Cleveland.
The major innovation in the lighting system is the switchboard, built along the
lines of a console of a modern organ. Controls for 4000 lighting combinations of
110 load circuits are at the finger tips of the operator. Included are a four scene
preset control, proportional control, remote control of intensity, and inter-
connection of circuits. The system is based on the thermionic type of lighting
control. The control apparatus for each circuit requires a dimming reactor, a
conventional vacuum tube, two grid glow rectifiers, and a system of control po-
tentiometers. The lamp load current flowing in the a-c. coils of the reactor, is
directly dependent on the d-c. saturation current flowing in the d-c. coil of the
same unit. As the direct current increases, the iron core of the reactor becomes
saturated alternating current increases. The direct current is supplied by a pair
of grid glow tubes whose output is controlled by the plate current of the vacuum
tube. The plate current of the vacuum tube is in turn controlled by varying
the bias on its grid. All the above apparatus is placed at a remote point from
the control console. The control circuit of the vacuum tube grid is brought to the
console. By means of selector switches, potentiometers, etc., any or all circuits
in the hall may be controlled at will. Circuit diagrams and pictures clearly show
the operation of this installation. A. H. H.

Audible Frequency Ranges of Music, Speech, and Noise. W. B. SNOW. Bell
Telephone Tech. J., X, October, 1931, No. 4, p. 616. A description of tests to
determine the maximum frequency range necessary for perfect or nearly perfect
reproduction. With the aid of experienced listeners, and using a series of filters,
varying degrees of cut-off were tried. It was found that frequencies between 80
and 8000 cycles were necessary to give good quality. Although rather indefinite
as to the advantages of using frequencies outside this range, it is believed that the
most nearly perfect quality is obtained by reproducing the full audible frequency
range. A. H. H.

The Development of the Microphone. H. A. FREDERICK. Bell Telephone
Quarterly, July, 1931, p. 164. An interesting history of the early experiments
leading up to the present design of microphones. Dr. Page in 1837, Sullivan in
1845, Bourseil in 1854, Reis in 1861, Helmholtz in 1863, and Varley in 1870, made
contributions to the development of the microphone. The experiments of
Dr. Alexander Graham Bell, begun in 1874, are described in more detail. In
1877, Edison patented a transmitter of the varying resistance type, using a button
of solid carbon or plumbago. The granular carbon design was first used in 1885.
The condenser type and the piezoelectric crystal type are of more recent design.
The difficulties of developing the carbon microphone are described in detail. It
is interesting to note that minute granules of carbonized anthracite coal were first
used by Edison in 1886. This source of carbon is still used to a great extent at
the present time. A. H. H.

The Effect of Humidity upon the Absorption of Sound hi a Room, and a De-
termination of the Coefficients of Absorption of Sound in Air. V. O. KNUDSEN,
JR. /. Acoustical Soc. of America, HI, No. 1, Part 1, July, 1931, p. 126. It is
shown that the absorption of sound hi air for frequencies above 2000 cycles is ap-
preciable. This effect is great enough to affect very appreciably the calculation

Jan., 1932] ABSTRACTS

of the reverberation time and absorption in a room for frequencies of 4000 cycles
and above. The absorption of air becomes less as the humidity increases.

An idea of the magnitude of the effect may be obtained from the following
statement. "Thus, if a tone of 4096 d.v., in the form of a plane parallel beam,
were used for long range signaling there would be, at a temperature of 21 ° C. and
a relative humidity of 44 per cent, an attenuation of 9.8 db. per second, or about
46 db. per mile. On the other hand, the attenuation would be less than 1 db. per
mile for a frequency of 512 d.v." Furthermore, it appears from this data, that a
reverberation chamber with perfectly reflecting walls would have a reverberation
time of no more than about six seconds for a tone of 4096 d.v. if the humidity of the
air in it is 44 per cent or less.

Theoretical formulas are deduced. The method used in separating the effect of
the absorption in air and that at the surface of the rooms was to take comparable
data in two rooms of different sizes but with the same boundary material, namely,
painted and varnished concrete. This yields sufficient data to separate the effects.
Even at 4096 d.v. the absorption of the painted concrete was about 0.02 and prac-
tically independent of humidity as long as condensation did not occur.

W. A. M.

A Critical Study of the Precision of Measurement of Absorption Coefficients by
Reverberation Methods. P. E. SABINE, JR. /. Acoustical Soc. of America, HI,
No. 1, Part 1, July 1931, p. 139. The data presented include a comparison of
absorption coefficients obtained at the Bureau of Standards and by two methods
at the Riverbank Laboratories on identical samples of each of four materials.
It is concluded that normal experimental errors in measuring absorption coef-
ficients may easily be 3 or 4 per cent, that probably an error of 10 per cent in the
coefficients would not appreciably affect the acoustic properties of an audience
room; and the actual computation of the reverberation time in a room is a matter
of approximate estimate rather than precise determination. W. A. M.

The High Intensity Arc for Motion Picture Projection. F. PATZELT. Kino-
technik, 13, September 20, 1931. p. 344. Measurements and graphs were made of
the light distribution of an "Artisol 75" projection lamp with high intensity car-
bons and with ordinary carbons. The average brightness of the entire crater of
ordinary carbons 14 mm. in diameter at 35 amperes and 45 volts was found to be
140 Hefner candles per sq. mm. Copper-coated high intensity carbons 11 mm.
in diameter were found to have a brightness of 357 Hefner candles per sq. mm.
at 75 amperes and 45 volts. The variation in the brightness of high intensity
carbons with different amounts of current was also measured. It was found that
carbons of small diameter require higher current densities than larger carbons to
attain the same brightness. The effect of changing the relative positions of the
carbons was studied, and it was found that greater brightness was attained with
the axis of the negative carbon in line with the center of the positive carbon than
with the axis of the negative carbon opposite the lower edge of the positive carbon.
The variation of the brightness at constant current with varying length of arc was
found to be small. It is stated that a 25-degree inclination of the axis of the nega-
tive carbon to the axis of the horizontal positive carbon is the most favorable. It
is concluded that the difficulties in the use of high intensity carbons are compen-
sated for by the increased illumination. M. W. S.

Safety Film. K. BRATRING. Kinotechnik, 13, July 20, 1931, p. 237. In its

120 ABSTRACTS []. S. M. P. E.

mechanical properties, such as resistance to wear and damage, cellulose acetate
motion picture film base is considered inferior to cellulose nitrate base. In view
of the universal precautions against fire in the projection of professional motion
picture films, it is considered that the low inflammability of cellulose acetate film
is sufficient cause to justify the increased expense attendant upon its use in thea-
ters. For schools, homes, and other places where proper safety precautions for
nitrate film are not taken, cellulose acetate film should undoubtedly be used. It
is thought that nitrate support constitutes no great hazard when used for amateur
roll films and film packs, or for professional portrait films. For x-ray films, the
introduction of cellulose acetate support is viewed with favor. M. W. S.

The Phillips Reproducing Set. Kinemat. Weekly, 172, June 4, 1931, p. 61.
The sound equipment in the Phillips set is a pedestal mounted at the left-hand side
of the projector; and a flexible shaft coupling driven by the motor is connected with
the projector flywheel. An integral gear shift permits the use of either sound-on-
film, sound-on-disk, or silent operation. The sound head of the projector em-
ploys a curved gate which is said to prevent film buckle. A high emission photo-
electric cell (18 microamperes per lumen) is used at present but a gas-filled
caesium cell is being investigated for future use. The speed control is ingenious,
the electric control being effected by rotating make-and-break cams, one driven
by the projector motor and the other by a constant-speed motor. When the
contact is made on both cam switches, a resistance is short circuited. The
period during which this resistance is short circuited, therefore, depends upon the
relative positions of the two cams. The cams revolve at approximately 80 rpm.
The fader used in the set gives a logarithmic change. The projection room ampli-
fier consists of a single stage which supplies current to the main amplifier which
may range in capacity from 20 to 200 watts with speech levels of 10 to 45 watts,
respectively. L. E. M.

A Continuous Motion Picture Projector. M. Hue. Bull. soc. frang. phot.,
73, June 1931, p. 128. A newly designed single oscillating mirror type of con-
tinuous projector is described. The principle involved is one in which the film
passes over a cylindrical drum having an aperture through which the single frame
is projected upon an oscillating mirror, which in turn reflects it into the objective
of the machine. During the movement of the film over the aperture, the adjacent
frame is isolated by a moving window behind the aperture, which moves with the
same angular velocity as the film. When the projection phase is terminated, a
shutter in front of the objective masks it during the return of the mirror and win-
dow. The light from the illuminating sources does not fall directly on the film
but is interrupted and reflected by a 45-inch mirror which is fabricated of a metal
capable of absorbing a large percentage of the heat rays, thus protecting the film.
All gears and cams are encased in oil, where possible, thereby minimizing noise.
It is claimed that a projector as described is capable of projecting a film 3000
times without injury to the film. Drawings are included. C. H. S.

Faith in the Title. F. SLIP. Filmtechnik, 7, May 2, 1931, p. 6. Although
titles have been replaced temporarily by the use of sound, they have a place in
certain classes of films, such as teaching films. Correctly composed titles may
also be of value in the presentation of certain sound films. During a study of
correct methods of title composition the maximum title width of 19 mm. has been
selected as desirable with the height accordingly proportional. The background

Jan., 1932] ABSTRACTS 121

should preferably be dark and the letters light. The type must be simple, clear,
and attractive. The optimum length for the title has been investigated from a
consideration of (1) length of the lines, and (2) number of letters. A useful table
is given showing length of the lines, number of letters, length of the title, and
length of the film per line of title, assuming projection at the rate of 24 frames per
second. L. E. M.

Motion Picture of the Eclipse of the Moon. F. Albrecht. Filmtechnik, 7, May
2, 1931, p. 1. On April 2, 1931, the first motion picture of a total eclipse of the
moon was photographed at the Trepton observatory hi Germany. With the
usual motion picture camera the image of the moon is far too small and even with
a teleobjective of 30 cm. The image is only 3 mm. in diameter. In the successful
motion picture an //10 objective of 65 cm. focal length was used, mounted on an
Ernemann E camera. The camera and lens were secured in place on the 21-meter
Trepton telescope. Positive film was employed and exposures of l/4 to Vz second
were made, using a blue filter with the teleobjective operated with a 35-mm. open-
ing. The camera shutter opening was increased to 160 degrees at the beginning of
the eclipse and decreased to 90 degrees as the eclipse passed. Single frame ex-
posures were made at intervals of 5 seconds, thus giving for the 3!A hour time a
length of film which, when projected at the rate of 24 frames per second, occupied
P/2 minutes. L. E. M.

Television Demonstration at Broadway Theater. Film Daily, 57, October 23,
1931, p. 1. A television demonstration was given at the Broadway Theater,
New York, for two weeks beginning on Oct. 22, 1931, a 10 by 10 foot screen being
used. The receiving disk revolved 900 times per minute and a projection system
projected the images on the screen. The sending station was located a short dis-
tance away in the Theater Guild Studio. G. E. M.

BOARD OF ABSTRACTORS

CARRIGAN, J. B. MACFARLANE, J. W.

COOK, A. A. MACNAIR, W. A.

CRABTREE, J. I. MATTHEWS, G. E.

FOWELL, F. McNicoL, D.

HAAKE, A. H. MEULENDYKE, C. E.

HARDY, A. C. MUEHLER, L. E.

HERRIOT, W. PARKER, H.

IRBY, F. S. SANDVICK, O.

IVES, C. E. SCHWINGEL, C. H.

KURLANDER, J. H. SEYMOUR, M. W.

LOVELAND, R. P. WEYERTS, W.

ABSTRACTS OF RECENT U.S. PATENTS

The views of the readers of the JOURNAL relative to the usefulness to them of the
patent abstracts regularly published in the JOURNAL will be appreciated. Favorable
views are of particular interest. In the absence of a substantial body of opinion
to the effect that these patent abstracts are desired by the membership, their early
discontinuance may be considered.

1,821,930. Film Feeding Mechanism. M. COUADE. Sept. 8, 1931. A film
feeding mechanism for projectors in which a claw engages the perforations in the
film and intermittently moves the film in accordance with the operation of a
cam mechanism which imparts angular movement to the claw. Adjustments
may be made for determining the length of stroke of the claw by adjusting the
eccentricity of the driving cam mechanism which engages the claw.

1,821,946. Film Feeding Mechanism. F. H. OWENS. Sept. 8, 1931. A
sound and motion picture apparatus including mechanism for intermittently
moving the picture films in front of the projection lens system while continually
moving the sound record portion. The shutter for the light beam has a peri-
pheral groove thereon for defining a belt wheel which is engaged by the drive belt.
A manual adjusting means is provided for properly positioning the shutter. There
is a lost motion connection between the film moving mechanism and the parts of
shutter by which the shutter may be selectively adjusted under manual control
before being operated under automatic control.

1,822,057. Composite Photographic Sound Records. F. H. OWENS. Sept.
8, 1931. The method of making a composite photograph sound record from dif-
ferent sources of sound such as a song with orchestra accompaniment and with
the addition of some special instrumental features such as bells and the like where-
in a plurality of photographic sound records are synchronously converted into
electric impulses. These impulses are received for modulating the intensity of a
single recording lamp. The modulated light rays from the lamp are photographed
upon the sensitized film. By this process it is possible to produce a sound record
by selecting desirable portions of previous sound records and thereby construct a
program of highly entertaining qualities.

1,822,183. Light Slit for Recording and Reproducing. D. A. Whitson. As-
signed to Whitson Photophone Corp. Sept. 8, 1931. A light slot for a sound
recording and reproducing system in which a guide block is disposed adjacent the
film. The guide block has a wide slot and a communicating narrow slot. The
sound record is passed over the wide slot. There is a lens in the bottom of the
wide slot and in contact with the sides and bottom of the slot for focusing radia-
tions to pass through the slots on the strip. The purpose of the lens slot is to con-
centrate the light at maximum intensity upon the film at the same tune that pro-
tection of the slot against the accumulation of dust or foreign matter is effected.

1,822,350. Arrangement of Perforations in Cinematographic Films. J. H.
JARNIER. Sept. 8, 1931. A motion picture film which is perforated laterally of
the picture frames instead of in two rows on opposite sides of the picture frames.
122

PATENT ABSTRACTS 123

Claws are used to shift the picture frames intermittently before the projector.
The structure of the film is such as to increase the resistance of the film against
tearing at the lateral lines of perforations. Rectangular perforations are provided
in the transverse spaces between the images wherein the ratio of the number ( JV)
of transverse perforations to the width (Z,) of the film having a specific resistance
to rupture by traction X is determined by the formula:

p— a\

in which a is the width of the perforations, p the resistance to rupture for the width
a in such a way that the resistance to tearing of the line of perforations engaged
is the same as the resistance to rupture by traction of the spaces separating them,
this resistance being the maximum.

1,822,528. Moving Lens Cinematograph Machine. W. E. JOHN. Sept. 8,
1931. A motion picture camera or projector having a continuously moving film
and a series of loose lens carriers moving with the film. The loose lens carriers
move through a closed circuit including a straight guide in which they are exposed,
and curved guides, one at each end of the straight guide; the circuit between the
curved guides being completed by a driving and conveying member in the form of
an internally toothed and pocketed wheel. The separate lens members are
brought into alignment with the optical path through the camera by driving
means connected with the lens carrier. The lens carriers slide longitudinally
around the guide which defines the path of movement for each of the lens members.

1,822,551. Lens Shifting Mechanism for Projecting Machines. A. TON-
DREAU. Assigned to Warner Bros. Pictures, Inc. Sept. 8, 1931. A system of
lenses which may be shifted in a projection machine to enable an instant change of
magnification on the projection screen without loss of focus. An attachment is
provided carrying lenses which may be first set in focus and which may be operated
to bring either one lens of a certain magnification or another lens of a different
magnification into the optical path. The lens carrier is provided with individual
supports for the different lens members, allowing independent longitudinal ad-
justment of the different lens carriers.

1,822,865. Glow Discharge Tube for Recording. T. W. CASE. Sept. 8, 1931.
A glow discharge tube for recording variations in light intensity upon film. A
bulb is provided for enclosing a non- thermionic anode and a cathode. An atmos-
phere of helium is provided within the bulb at such a pressure that a concentrated
glow is provided about the negative electrode with a voltage not substantially
greater than 400 volts d-c. across the electrodes. The cathode has a photoelec-
trically activated coating comprising barium actuated for electron emission by the
said glow concentrated about the cathode. The device is designed to produce
response of the glow in terms of light emission according to variations of electrical
impulses produced in a sound control circuit.

1,822,932. Combination Recording and Reproducing Stylus Head. M. H.
LOUGHRIDGE. Sept. 15, 1931. A stylus head is arranged to support both a
recording and a reproducing stylus with respect to a wax record of a phonograph.
The stylus head may be shifted to bring either the recording or reproducing stylus
into engagement with the phonograph record. A switching mechanism is pro-
vided for controlling the connection of the styluses to an amplifying system.

124 PATENT ABSTRACTS [J. S. M. P. E.

When the reproducing stylus engages the sound record, the input circuit of the
amplifier is connected with the stylus. When the recording stylus is placed in
engagement with the sound record, the magnetic windings thereof are connected
with the output circuit of the amplifier for cutting a groove in the record in ac-
cordance with the sound vibrations impressed upon the input circuit of the ampli-
fier.

1,823,243. Method and Apparatus for Lapping Color Film Embossing Rollers.
O. WHITTEL. Assigned to Eastman Kodak Co. Sept. 15, 1931. A method of
lapping lenticular film embossing rollers which comprises providing a cylinder with
a plurality of fine guide lines, turning the cylinder, and lapping the cylinder
with a plurality of wires, a fine lapping compound being used on the cylinder. The
embossing roller is used for operation upon color motion picture films. The len-
ticular areas or elements formed in the film are extremely minute as the distance
across these elements may be from 0.0015 to 0.002 of an inch.

1.823.245. Film Winding Device. O. WITTEL. Assigned to Eastman Kodak
Co. Sept. 15, 1931. Winding device for motion picture film in which a reel
is provided with a pair of concentric hub members. One hub member is slidably
carried by a flange disposed in one side thereof. The two hub members are sepa-
rated by sliding the flange on one hub. The structure of the film winding device
is such that the film may be drawn from an inner convolution of a supply reel and
wound on an outer convolution of a take-up reel. The construction of the reel is
such that the film is properly aligned on the reel without rewinding.

1.823.246. Method of Tinting Film for Use in Sound Reproduction. A. A.
YOUNG. Sept. 15, 1931. A method of tinting the picture areas of a photograph
film in which the sound record portion is preserved untinted while preventing
shrinkage of the film by applying to the picture areas of the film a dye dissolved
in a solution comprising a solvent for the film and the dye and a non-solvent for
the film which has the property of reducing the rate of evaporation of the solvent
whereby the tendency of the film to buckle is eliminated. The dye, which is
applied to the picture areas of the film, is dissolved in a solution containing from 5
to 10 per cent of acetone, from 70 to 75 per cent methyl alcohol, and the remainder
triacetin.

1,823,349. Producing Fade-in and Fade-out of Photographic Sound Record.
S. C. CHAPMAN. Assigned to Electrical Research Products, Inc. Sept. 15, 1931.
The sound record is chemically treated for reducing the end portions of the sound
record progressively varying lengthwise of the film. The reproduced sound will
thus gradually increase in volume from silence to the normal volume of the
record, vary normally with the record till near the end when the volume of
the sound will gradually diminish to silence.

1,823,355. Telescope Framing Device. L. S. FRAPPIER AND E. BOECKING.
Assigned to International Projector Corp. Sept. 15, 1931. Projecting machine
for photographic sound records wherein a microscope is supported in the path of
a scanning ray in such position that the ray can be observed while adjustments
are being made to secure the proper characteristics thereof. A prism is positioned
in the path of the light rays to deflect a portion of the light at right angles into
the microscope in order that the sound record may be analyzed.

1,823,400. Photographic Film Copying Machine. L. HORST. Assigned to
Sinus Kleuren-Film Maatschappil, of Bosch en Duin, Netherlands. Sept. 15,

Jan., 1932] PATENT ABSTRACTS 125

1931. A machine for copying two color films and more particularly a machine
of this kind in which the pictures are transferred from one film to the other by
means of mirrors and objectives provided in duplicate. Two sources of light are
provided, each of which is separately regulable, for timing the degree of copying of
the individual part pictures.

1,823,462. Photographic Camera. K. MORSBACH. Assigned to Siemens &
Halske, Aktiengesellschaft. Sept. 15, 1931. The film refill which is supplied
for the camera is carried by an interchangeable cassette which cooperatively en-
gages a film guide channel located in the interior of the camera behind the objec-
tive lens. There is a guide plate carrying the window for the image, permanently
located behind the objective and in its focus. There is a pressure plate indepen-
dently mounted on the cassette. When the camera is refilled, any differences in
the focal lengths of the objectives of different cameras are compensated by the
pressure plate and the guide plate so that equal operation of cameras which are
not uniform is obtainable.

1,823,737. Sound-on-Disk Motion Picture Projector. CHARLES L. HEISLER.
Assigned to General Electric Co. Sept. 15, 1931. A motion picture projector
which includes a projector housing mounted adjacent a phonograph turntable.
The driving motor which operates the projector also drives the phonograph turn-
table so that the film and the record may be operated in synchronism. The arm
which carries the phonograph pick-up is pivoted adjacent one side of the record
table and permits the phonograph pick-up to be moved over the area of the
revolving record.

1,824,294. Sound and Picture Film Matching Means. FREEMAN H. OWENS.
Assigned to Owens Development Corp. Sept. 22, 1931. A method which permits
the accurate repair or splicing of separate film strips, one of which carries the pic-
ture record and the other of which carries the sound record to maintain synchro-
nism between the picture and the sound wherein an insertable film section is pro-
vided attachable to the broken ends of the film. The insert is provided with a
sound record and images adjacent the sound record, the images being partial
duplicates of the images on the picture film. The splicer finds it very easy and
convenient to judge accurately the length of the insert by simply matching the
two films by observing the partial duplicates of the images on the insert and fitting
the sound strip in to match the sound on the film. That is to say, a guide is pro-
vided on the insertable sound strip so that the splicer is advised accurately as to
where this sound should occur on the sound film in order to match accurately the
images on the picture film.

1,824,417. Treating Sound Records Produced by Splicing. A. T. TAYLOR.
Assigned to Metro-Goldwyn-Mayer Corp. Sept. 22, 1931. The method of
splicing a film carrying a sound record to prevent audible clicks and foreign noises
at the splice marks as the film passes the light path. The ends of the broken film
are cemented and then a patch in the form of a half -cycle sine wave cemented over
the adjoining ends of the sound record. This sine wave patch has a frequency
below normal audibility and an amplitude equivalent to the width of the sound
record so that there is no extraneous sound created as the splice passes the sound
reproducing aperture.

1,824,446. Producing Motion Pictures in Color. E. L. PEARSON. Sept. 22,
1931. A projection screen is arranged for rotative movement in timed relation tQ

126 PATENT ABSTRACTS [J. S. M. P. E.

the rotation of a color filter at the projection machine. The projector is arranged
to project successively images through the different colored filters upon the moving
projection screen from which the picture may be viewed and through which the
images are produced. By shifting the relative positions of the projector and the
projection screen to project successively the images upon the portions of the pro-
jection screen corresponding to the particular filters upon which the images are
produced, an effect upon the eye of colored motion pictures closely portraying in
color and motion real animated objects is produced.

1,824,709. Camera for Taking Cinematographic Pictures. A. L. V. C. DE-
BRIE. Sept. 22, 1931. View taking apparatus comprising two parts, namely, a
front part containing the film driving device, the shutter, and the optical arrange-
ment and a rear removable part which can be secured instantaneously to the front
part and which contains a feeding storing box wherein the unimpressed film is
disposed together with the film guiding devices, the transmission gear, and a second
storing box into which the impressed film is wound up. The latter box can be
the same as the feeding box or else both boxes can be made separate. The opera-
tor can thus be provided with several rear parts ready for use which he may se-
cure to the front part of the apparatus according to the requirements. The result
thereof is, besides the advantage already mentioned, a saving of time which is of
great interest in the case where the taking of the complete scene which is to be
cinematographed requires a length of film greater than what can be contained in
one single storing box.

1,824,731. Picture Transmitting System. D. M. MOORE. Assigned to
General Electric Co. Sept. 22, 1931. A picture receiving system in which the
light is modulated in accordance with the shading of the successive elemental
areas of the picture transmitted. A screen is provided and there are a plurality
of rotatably mounted mirrors arranged to reflect successively the modulated light
to produce a trace on the screen. The mirrors are rotated continuously in one
direction at different speeds with a lens system arranged between the mirrors.
The mirrors are each mounted on the shaft of the associated driving means in
such manner that the mirrors are inclined at an angle to the axis of the driving
shaft so that rotation of each of the mirrors produces a scanning operation over
the area of the receiving screen.

1,825,078. Incandescent Electric Lamp for Projection Apparatus. J. MA-
RETTE. Assigned to Pathe Cinema Anciens Etablissements Pathe Freres.
Sept. 29, 1931. A glow lamp is directly centered in the optical path of a projection
machine by means of a ring member which is secured over the base of the lamp and
serves to center the lamp accurately in its support for accurately directing the
maximum amount of light through the projection path.

1.825.121. Lamp Holder. F. H. OWENS.. Assigned to Owens Development
Corp. Sept. 29, 1931. A plurality of separate lamps are mounted on a carrier
which may be laterally shifted to move any one of the lamps successively into a
predetermined operative position. There are stops provided on the lamp sup-
porting base to limit the movement of the lamps to selected positions. The lamp
holder may be moved through a shaft member to the outside of a lamp housing.

1.825.122. Objective for Color Photography. A. OSWALD. Assigned to Kel-
ler Dorian Colorfilm Corp. Sept. 29, 1931. An objective lens system for color
photography employing films having a goffered base wherein the lens system is

Jan., 1932] PATENT ABSTRACTS 127

made up of a plurality of different elements; a diaphragm and a collimator film.
The several elements of the optical system are so arranged that the pupil of emer-
gence of the objective is in the anterior focal plane of the collimator lens, the
aberrations introduced by the collimator lens being corrected by compensating
aberrations introduced into said objective.

The objective is anastigmatic and is constituted by the three spaced elements
and by a collimating lens located in the vicinity of the focal plane of the objective.
The objective of this invention follows Petzval's law —

and in calculating these objectives in view of increasing the sharpness of the mar-
ginal images, P is left with a negative value suited to the extent of the field to be
represented; when calculating an objective of this sort intended to be provided
with a collimating lens, the residual value ascribed to P will therefore have to be
increased by varying the quantity <f>rj.

1,825,142. Motion Picture Film Magazine. W. A. BRUNO. Assigned to
Clarence W. Fuller. Sept. 29, 1931. A protective housing for films wherein the
film is supported for avoiding breakage or other injuries. A plurality of film
carrying reels of considerable diameter are provided so that the film may be
stored in the magazine, without sharp bends. The reels are constructed to engage
the film near its marginal edges only, the cylindrical surfaces of the reels being
concave or otherwise centrally disposed to prevent contact thereof with the
central portions of the film. Power means are provided for driving the reels for
storing the film in the magazine while preventing scratching or other abrasion to
the picture frames on the film.

1.825.253. Synchronous Camera Mechanism. A. F. VICTOR. Sept. 29, 1931.
A camera having means for controlling and synchronizing the motion and arresting
the movement of the feeding devices with respect to the shutter. A cam co-
operating with an abutting arm is provided in association with the rotatable
shutter by which the shutter may be brought to rest by moving the arm. By
withdrawing the arm from the path of the cam the shutter may be rotated under
control of the drive mechanism. The movement of the shutter with the film
feeding devices is synchronized. The shutter is provided with additional devices
that cooperate with control mechanism so that when the latter is released to return
to normal, the stoppage of the film is momentarily postponed until the shutter
is in position in front of the exposure aperture, whereupon the movement of all
mechanism is arrested. This is accomplished in such manner that it positively
insures the proper positioning of the shutter in front of the aperture at the moment
the movement of the film ceases and the stoppage is made without jar to the
camera.

1.825.254. Intermittent Feed for Motion Picture Apparatus. A. F. VICTOR.
Sept. 29, 1931. A mechanism for intermittently feeding a film through a camera
or projection machine which includes a shuttle that is reciprocated by a continu-
ously rotatable cam. The shuttle is hinged upon the end portions of lever arms
that are pivotally mounted upon the housing of the camera or projector. Means
are provided for adjusting the pivoted ends of the arms toward each other in
such manner that any noticeable wear between the cam and the parts engaged

128 PATENT ABSTRACTS [J. S. M. P. E.

thereby may be taken up by means of a simple adjusting structure. The ful-
crums of the lever arms are supported in a "floating" pivot because the pivotal
members are not actually secured to the camera or projector but are carried upon
suitable rocker-arms which themselves are pivoted on the support or housing.
The operation of these rocker arms is similar to the action of a cam or cams en-
gaged with the lever arms.

1,825,340. Electrooptical Cell. N. DEISCH. Sept. 29, 1931. A Kerr cell is
used for electrically modulating a beam of light. One electrode of the Kerr cell
comprises a frame having an opening comprising the active space thereof and a
plurality of flexible ribbon-like division members dividing the opening into a
plurality of light passages, the flexible ribbon-like members being secured to said
frame and held taut across said opening. Electrostatic stresses impressed on the
cell operate to modify the light passing through the divisions of the cell.

1,825,529. Sound Pipe Reproduction from Photographic Films. R. KOLLER.
Sept. 29, 1931. A motion picture film is combined with an air control band which
moves synchronously with the motion picture film. The ah" control band moves
over a tracker board for controlling the supply of air to various sound pipes for
the reproduction of sound appropriate to the pictures. Synchronization of the
sound with the pictures is obtained by virtue of the interconnection of the moving
band with the picture film. Various forms of pipe organ valves may be oper-
ated by allowing the air to pass through predetermined apertures hi the moving
band.

1,825,486. Scanning Disk. A. O. TATE. Sept. 29, 1931. The apertures in
a scanning disk are arranged in reverse spirals, one of the spirals beginning at the
outer edge of an image and ending at the inner edge thereof and the other of the
spirals beginning at the inner edge and ending at the outer edge. The adjacent
apertures of the spirals are disposed the same radial distance from the center
of the disk so that the images are scanned twice in succession. Each of the aper-
tures is bounded by arcs of concentric circles and by radii of the disk. The ob-
jects of the arrangement of the scanning disk apertures are to eliminate the incon-
venient restrictions with respect to the area available for use as scanning space as
defined by the distances between the open ends of the spirals, to provide means
whereby an object may be scanned laterally by intermittent light beams or pen-
cils which maintain perpendicularly a continuous, rhythmic, undulatory movement
through the period of revolution of the disk; to provide means whereby the total
area of the scanning space may be varied with respect to its dimensions; and to
provide means whereby an object may be scanned with one revolution of the disk
a plurality of times.

The scanning disk is divided circumferentially by a plurality of radial lines to
form circumferential divisions and is divided radially by a plurality of concentric
circles to form radial divisions and may be conveniently plotted by the following
formula, in which:

A represents the number of circumferential divisions of the entire disk;

B represents the number of radial divisions included within the scanning area;

C represents the number of circumferential divisions between successive aper-
tures; and

D represents the number of times the scanning area is scanned hi one revolution
of the disk and also the number of spirals in the system.

Jan., 1932] PATENT ABSTRACTS 129

The following equation represents the relationship of the above quantities:

A = BCD
This equation may be solved for C or B as follows :

C- ±- B- A
~ BD' ~ CD

By assuming the various constants of the disk, the apertures may be con-
veniently laid out in accordance with any desired scheme by following the above
formula so as completely to scan the image any desired number of times for each
revolution of the disk.

1,825,487. Scanning Device. A. O. TATE. Sept. 29, 1931. An endless belt
is provided with a staggered series of apertures. The belt is looped around a
multiplicity of guide drums and is driven by rollers at opposite ends of a frame
structure, so that the apertures are moved successively across the field of a lens
system for scanning an object within the field of the lens. The object is scanned in
lines from bottom to top or top to bottom. The band is approximately 160 inches
in length and has approximately 80 apertures therein, each spaced from the ad-
jacent aperture at a distance of 4 inches.

1,825,497. Light Projection Display Apparatus. T. WILFRED. Sept. 29,
1931. A poly sided screen consisting of a plurality of upright differently faced
concave sides meeting in thin edges is provided for a display surface. There are
light projecting means spaced outwardly in front of each of the concave sides,
the several projecting means being adapted to project cooperatively upon the
respective adjacent concave sides to produce an ornamental light display for at-
tracting the attention of a spectator. The projection apparatus is used hi various
forms of floodlighting architectural displays.

1,825,598. Process for Producing Combined Sound and Picture Films.
H. VOGT, J. MASSOLLE, AND J. ENGL. Assignors by mesne assignments to
American Tri-Ergon Corp. Sept. 29, 1931. The sound and picture records are
photographed on separate film strips to form separate negatives. The negative
picture record is photographed upon a portion of a sensitized film not exposed to
the sound record. The negative sound record is photographed on the same face
of the sensitized film but on a portion thereof not exposed to the picture record.
By the separation of the sound record from the picture record, a film record com-
bining both of these records can be produced without subjecting either record to
conditions of overexposure or underexposure.

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.)

SOCIETY OF MOTION PICTURE
ENGINEERS

OFFICERS
1931-1932

President

A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE, Fox Film Corp., New York, N. Y.

Secretary
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y.

Board of Governors

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada
H. T. COWLING, Eastman Teaching Films, Inc., 343 State St., Rochester, N. Y.
J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y.
P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y.
O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y.
A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New

York, N. Y.

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
R. F. MITCHELL, Bell & Howell Co., 1801 Larchmont Ave., Chicago, 111.
J. H. KURLANDER, Westinghouse Lamp Co. Bloomfield, N. J.
W. C. KUNZMANN, National Carbon Co., Cleveland, Ohio

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLB, 277 Park Ave., New York, N. Y.
130

COMMITTEES

131

COMMITTEES
1931-1932

( The completed list of committees -will be published in a later issue}

Color

R. M. EVANS, Vice-Chairman

P. D. BREWSTER W. T. CRESPINEL N. M. LA PORTE

J. L. CASS H. B. TUTTLE

W. C. HUBBARD

Convention
W. C. KUNZMANN, Chairman

M. W. PALMER

H. T. COWLING
W. B. COOK

Finance

L. A. JONES, Chairman
J.I. CRABTREE
W. C. HUBBARD

J. H. KURLANDER

L. C. PORTER

J. R. CAMERON

B. W. DEPUE

C. D. ELMS

Historical
C. L. GREGORY, Chairman

Membership and Subscription

H. T. COWLING, Chairman
W. H. CARSON, Vice-Chairman

R. EVANS

J. G. T. GILMOUR

J. KLENKE
E. E. LAMB
E. C. SCHMTTZ

J. A. BALL
C. DREHER
P. H. EVANS
A. C. HARDY

Papers

O. M. GLUNT, Chairman
N. M. LA PORTE
G. E. MATTHEWS
P. A. McGuiRE*
D. McNicoL

P. MOLE
K. F. MORGAN
C. N. REIFSTECK
T. E. SHEA

J. O. BAKER
T. BARROWS
W. H. BELTZ
G. C. EDWARDS
S. GLAUBER

Projection Practice
H. RUBIN, Chairman
J. H. GOLDBERG
C. GREENE
H. GRIFFIN
J. HOPKINS
R. H. MCCULLOUGH
P. A. McGuiRE

R. MlEHLING

F. H. RICHARDSON
M. RUBEN
P. T. SHERIDAN
L. M. TOWNSEND

132

COMMITTEES

J. L. CASS
H. GRIFFIN

Projection Screens
S. K. WOLF, Chairman

J. H. KURLANDER

W. F. LITTLE

A. L. RAVEN
H. RUBIN

F. C. BADGLEY
B. W. DEPUE

Publicity

W. WHITMORE, Chairman
D. E. HYNDMAN
F. S. IRBY

G. E. MATTHEWS
D. McNicoL

M. C. BATSEL
P. H. EVANS
N. M. LA PORTE

Sound

H. B. SANTEE, Chairman
E. W. KELLOGG
C. L. LOOTENS
W. C. MILLER

H. C. SILENT
R. V. TERRY
S. K. WOLF

L. E. CLARK
L. DE FOREST
J. A. DUBRAY
P. H. EVANS
R. E. FARNHAM
H. GRIFFIN

Standards and Nomenclature
M. C. BATSEL, Chairman
A. C. HARDY

R. C. HUBBARD

L. A. JONES
N. M. LA PORTE
D. MACKENZIE
G. A. MITCHELL
G. F. RACKETT

W. B. RAYTON
C. N. REIFSTECK
V. B. SEASE
T. E. SHEA
J. L. SPENCE
E. I. SPONABLE

R. S. BURNAP
W. H. CARSON

Ways and Means
D. McNicoL, Chairman
H. GRIFFIN
F. S. IRBY

J. H. KURLANDER
J. A. NORLING

Chicago Section

R. F. MITCHELL, Chairman R. P. BURNS, Manager

B. W. DEPUE, Sec.-Treas. • O. B. DEPUE, Manager

New York Section

P. H. EVANS, Chairman M. C. BATSEL, Manager

D. E. HYNDMAN, Sec.-Treas. J. L. SPENCE, Manager

Pacific Coast Section

D. MACKENZIE, Chairman G. A. MITCHELL, Manager

E. HUSE, Secretary H. C. SILENT, Manager

L. E. CLARK, Treasurer

CONTRIBUTORS TO THIS ISSUE

Burke, B. S.: E.E., University of Michigan, 1923; engineer, Westinghouse
Electric & Manufacturing Company, 1923-27; theater switchboard engineer,
Westinghouse Electric & Manufacturing Company, 1927-29; developed ther-
mionic tube control for theater lighting, Westinghouse Electric & Manufacturing
Company, 1929 to date.

Curtis, A. M.: Born June 4, 1890, at Brooklyn, New York. United Wireless
Company, 1907-10; radio engineer, Lloyd Brasiliero S. S. Company and
Brazilian Department of Agriculture, 1910-13; engineer, Western Electric
Company, 1917-19; engineer, research laboratory, Western Electric Company
and Bell Telephone Laboratories, 1919 to date.

Jones, L. A.: See May, 1931, issue of JOURNAL.

Rumpel, C. H.: Born 1903 at Kitchener, Ontario, Canada. Graduate,
Massachusetts Institute of Technology; technical staff, Bell Telephone Labora-
tories, 1929 to date.

Shea, T. E.: See March, 1931, issue of JOURNAL.

133

SOCIETY ANNOUNCEMENTS

BOARD OF GOVERNORS

At a meeting of the Board of Governors held at the Waldorf-
Astoria Hotel, New York, N. Y., on December 10th, consideration was
given to the establishment of new committees which would expand to
a great extent the scope of activities of the Society in directions which
have so far not been adequately investigated. Among the new com-
mittees considered was one to deal with non-theatrical home equip-
ment, a committee on the development and care of film, a Museum
Committee whose duty it will be to gather historical pieces of motion
picture apparatus for purposes of exhibition in suitable depositories,
and a committee on the preservation of film.

It was decided that the S. M. P. E. Fellowship, created through the
generosity of Mr. George Eastman, is to be established at the Uni-
versity of Rochester. Its administration is to be left to the Projection
Theory Committee, the object of the Fellowship being to conduct
investigations on problems particularly concerned with or cognate to
the motion picture art.

Much discussion was held concerning the financial operations of the
Society for the fiscal year and on the general matters of entrance fees,
dues, and subscription rates.

SPRING, 1932, CONVENTION

The Board of Governors decided that the names of the cities, New
York, N. Y., and Washington, D. C., be placed upon the ballot which
is to be mailed to the entire membership for voting upo tnhe location
of the Spring, 1932, Meeting.

NEW YORK SECTION

A meeting of this section was held on Wednesday, December 9th,
in the auditorium of the Engineering Societies Building, 33 West
39th Street, New York, N. Y. Mr. H. A. Frederick of the Bell
Telephone Laboratories repeated his paper entitled, "Vertical Sound
Records; Recent Technical Advances in Mechanical Records on
134

SOCIETY ANNOUNCEMENTS 135

Wax," which was presented at the Swampscott Convention on Octo-
ber 7th. The demonstration which accompanied the paper included
considerably more elaborate apparatus than that which was used at
the Swampscott Meeting. Following Mr. Frederick's presentation
Mr. Leopold Stokowski, director of the Philadelphia Orchestra, ad-
dressed the meeting, presenting from a musician's standpoint some
of the problems of recording. The meeting created considerable
interest, more than seven hundred and fifty people attending in spite
of the inclement weather.

The next meeting of the Section is scheduled to be held about the
second week in January. Announcements will be mailed to all mem-
bers enrolled in the Section. Those whose names are not on the
mailing list of the Section, but who wish to receive information con-
cerning the meetings, should communicate with the general office of
the Society.

PROJECTION PRACTICE COMMITTEE

At a meeting of the Projection Practice Committee held on Novem-
ber 24th, a program outlining the work to be conducted by the Com-
mittee during the current year was formulated, being particularly
directed toward the recommendation of standards of tolerances and
clearances of projector and sound parts, and the determination of the
degree of wear of projector and sound equipment which can be al-
lowed without impairing the quality of the projected picture or dam-
aging the film. A study of the methods of so-called processing, or the
treating of finished positive prints to prevent damage during the first
showing of the film is to be included in the work of the year. It is
felt that there is a need for more perfect methods which will com-
pletely eliminate the shedding of emulsion or the accumulation of oil
and wax in the projector, due to the film.

At a second meeting of the Committee on Tuesday, December 15th,
further discussion of the problems of tolerances and clearances in
projector and sound parts was held, particular attention being paid
to the points at which tension of the film and wearing of the parts
occur. The problem of the specifications desirable for projector
apertures was also discussed at some length, and the work on this
problem, although not completed, is recommended for the study of the
Standards and Nomenclature Committee of the Society.

136 SOCIETY ANNOUNCEMENTS [J. S. M. P. E.

MEMBERSHIP CERTIFICATE

Associate members of the Society may obtain the membership
certificate illustrated below by forwarding a request for the same to
the General Office of the Society at 33 W. 42nd St., New York, N. Y.,
accompanied by a remittance of one dollar.

Sonets/Motion Picture Engineers

INCOSPOOATEO

Society of Motion Picture Engineers

LAPEL BUTTONS

There is mailed to each newly elected member, upon his first
payment of dues, a gold membership button which only members
of the Society are entitled to wear. This button is shown twice
actual diameter in the illustration. The letters are of gold on a
white background. Replacements of this button may be obtained
from the General Office of the Society at a charge of one dollar.

Jan., HK32J

SOCIETY ANNOUNCEMENTS
JOURNAL BINDERS

137

The binder shown in the accompanying illustration serves as a
temporary transfer binder or as a permanent cover for a complete
year's supply of JOURNALS. It is made of black crush fabrikoid,
with lettering in gold. The binder is so constructed that each in-
dividual copy of the JOURNAL will lie flat as its pages are turned.
The separate copies are held rigidly in place but may be removed or
replaced at will in a few seconds.

These binders may be obtained by sending your order to the
General Office of the Society, 33 West 42nd Street, New York, N. Y.,
accompanied by a remittance of two dollars. Your name and the
volume number of the JOURNAL may be lettered in gold on embossed
bars provided for the purpose at a charge of fifty cents each.

SUSTAINING MEMBERS

Bell & Howell Co.
Carrier Engineering Corp.
Case Research Laboratory

Eastman Kodak Co.

Electrical Research Products, Inc.

National Carbon Co.

RCA Photophone, Inc.

BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS

Prior to January, 1930, the Transactions of the Society were published quar-
terly. A limited number of these Transactions are still available and will be
sold at the prices listed below. Those who wish to avail themselves of the op-
portunity of acquiring these back numbers should do so quickly, as the supply
will soon be exhausted, especially of the earlier numbers. It will be impossible
to secure them later on as they will not be reprinted. The cost of all the available
Transactions totals $46.25.

No.

Price

No.

Price

1917 { I

$0.25
0.25

1924-

r!8
19

$2.00
1.25

1918 7

0.25

[20

1.25

1920 ft
1921 11

1.00
1.00
1.00
1.00

1925

21
22
23
24

1.25
1.25
1.25
.25

1922 \\
1923 ^

1.00
1.00
2.00
2.00

1926-

25
26
27
28

.25
.25
.25
.25

1927

1928

1929

No.

30
31
32
33
34
35
30
37
38

Price
$1.25
.25

.25
,25

.50

2.50

3.00

Beginning with the January, 1930, issue, the JOURNAL of the Society has been
issued monthly, in two volumes per year, of six issues each. Back numbers of all
issues are available at the price of $1.50 each, a complete yearly issue totalling
$18.00. Single copies of the current issue may be obtained for $1.50 each.
Orders for back numlu-rs of Transactions and JOURNALS should be placed through
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and
should be accompanied by check or money-order.
138

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR
Volume XVIII FEBRUARY, 1932 Number 2

CONTENTS

Vertical Sound Records: Recent Fundamental Advances in

Mechanical Records on "Wax" H. A. FREDERICK 141

Sound Recording — From the Musician's Point of View

LEOPOLD STOKOWSKI 164
On the Assignment of Printing Exposure by Measurement of

Negative Characteristics CLIFTON TUTTLE 172

Utilization of Desirable Seating Areas in Relation to Screen

Shapes and Sizes and Theater Floor Inclinations

BEN SCHLANGER 189

A Method of Measuring Directly the Distortion in Audio

Frequency Amplifier Systems W. N. TUTTLE 199

Directional Effects in Continuous Film Processing

J. CRABTREE 207

Resume of the Proceedings of the Dresden International

Photographic Congress S. E. SHEPPARD 232

Committee Activities:

Report of the Projection Screens Committee 242

Abstracts , 255

Patent Abstracts 258

Book Reviews 266

Officers 267

Committees 268

Contributors to This Issue 271

Society Announcements 272

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers.

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc.

Subscription to non-members, $12.00 per annum; to members, $9.00 per annum,
included in their annual membership dues; single copies, $1.50. A discount
on subscriptions or single copies of 15 per cent is allowed to accredited agencies.
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y.

Papers appearing in this Journal. may be reprinted, abstracted, or abridged
provided credit is given to the Journal of the Society of Motion Picture Engineers
and to the author, or authors, of the papers in question.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930, at the Post Office at Easton,
Pa., under the Act of March 3, 1879.

VERTICAL SOUND RECORDS

RECENT FUNDAMENTAL ADVANCES IN MECHANICAL RECORDS ON

"WAX"*

H. A. FREDERICK**

Summary. — This paper describes recent progresswhich has been made in laboratory
studies of mechanical records of sound cut on a wax disk. Both theoretical and
experimental investigations indicate that a phonograph record, cut with vertical undula-
tions instead of the more usual lateral undulations possesses fundamental advantages.
The principal improvement comes from a marked increase in the volume and frequency
range over which faithful reproduction may be obtained. A higher volume level can be
recorded for the same groove spacing and speed. More playing time can be provided
with a given size of record and volume level since, for these conditions, both the groove
spacing and speed may be reduced. Improvements in methods of processing the
stampers and in the record material give a large reduction in surface noise and hence a
corresponding increase in the volume range. With these improvements the frequency
range which can be reproduced satisfactorily can be extended nearly an octave to
8000 to 10,000 cycles. Other improvements incidental to the improvements noted
above are great improvement in the quality of reproduction obtainable directly from
a soft "wax" record and a great extension in the life of the hard record.

At the convention of this Society held at Lake Placid in the fall of
1928, data were presented showing that a very good frequency charac-
teristic could be obtained in recording and reproducing by means of
the "lateral" disk recording system.1 The data presented at that
time had to do chiefly with the response-frequency characteristics of
the elements which entered into that system. The information then
available, however, about non-linear distortion was somewhat
limited. That discussion, in addition, did not attempt to cover the
limitations imposed by background noise commonly called "surface"
or "needle scratch."

In most commercial uses of lateral records, surface noise has
imposed very serious limitations. In many cases this noise has been

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
(Repeated at a meeting of the New York Section, December 9, 1931, at New
York, N. Y.)

** Bell Telephone Laboratories, New York, N. Y.

141

142

H. A. FREDERICK

[J. S. M. P. E.

suppressed by the use of so-called "scratch" filters. These have
effectively quieted the reproduction but only by the sacrifice of an
important portion of the recorded band of frequencies which are
above 3000 to 4000 cycles. Investigations have been carried on to
determine the fundamental causes and the characteristics of the
surface noise in order that, with a better understanding, it might be
more effectively reduced and without such a sacrifice.

In addition to the limitations imposed by surface noise, other
studies have indicated that, with the available reproducers for lateral
cut records, the needle point may fail to follow the center of the groove
accurately when the curvature becomes too sharp, and may skid from
side to side by varying amounts, depending on the record and the
characteristics of the reproducer being used. Studies have proceeded

RECORDER
STYLUS

A = POINT OF CONTACT OF REPRODUCER
STYLUS WITH GROOVE

FIG. 1. Distortion in a lateral groove. .

relating to the physical characteristics necessary in a reproducer in
order that it may faithfully follow a groove. These studies have led
us to expect superior performance from a groove cut with vertical un-
dulations than from one with lateral undulations. These records are
similar in principle to those used by Mr. Edison. With the lateral
groove there is distortion due to the fact that the sound is recorded
with a chisel-shaped stylus and reproduced with a round stylus;
also that in reproduction the bearing point of the stylus against the
groove shifts forward and backward as the needle rounds a curve.
These effects are illustrated in Fig. 1. With vertical records the first
of these effects, sometimes called the "pinch" effect, is absent, but a
shifting of the bearing point of the reproducing stylus forward and
backward occurs if a round stylus is used. It is doubtful if a chisel-
shaped reproducing stylus or a stylus with an elliptical point can be
justified due to the increased cost and complication, and in considera-

Feb., 1932]

VERTICAL SOUND RECORDS

143

tion of the rather small amount of distortion which this would
eliminate. Some qualitative idea of what takes place with vertical
undulations may be gained from Fig. 2, in which a sine wave is shown
together with the resulting positions of the stylus point. For a given
stylus tip radius and for a given recording level this effect increases
with frequency.

This failure of a stylus point to follow a vertical record with great
accuracy is, of course, due to the finite length of the stylus point
along the groove. A fact which relieves this situation is that speech
and music and most other sounds which we are interested in re-
cording contain much less energy in the high than in the low fre-
quency range.2

Frequency analyses of surface noise have been made using a variety

RECORDER
STYLUS

REPRODUCER STYLUS POSITIONS

FIG. 2. Distortion in a vertical groove.

of reproducers and record materials. In general, these frequency
characteristics have been found to be very largely influenced by the
characteristics of the reproducers, but do not show any marked differ-
ences as between lateral and vertical recordings. Frequency charts of
surface noise taken with a vertical reproducer having a very flat
frequency characteristic over the audible range have shown the
surface noise to be relatively richer in high frequencies. The distribu-
tion of surface noise energy below 10,000 cycles from a cellulose
acetate pressing is shown in Fig. 3. The amount of recorded sound
energy in the low frequency range, i. e., below about 2000 or 3000
cycles, however, is large relative to that in the higher frequency
ranges. Moreover, the characteristics of many lateral reproducers
have been such as to accentuate surface noise between 3000 and
5000 cycles. Hence the use of "scratch" filters for the elimination of

144

H. A. FREDERICK

[J. S. M. P. E.

the high frequency components of the surface have made a large effec-
tive reduction in noise without any material loss in loudness of the
sounds of interest. The loss in loudness at the higher frequencies
has also reduced the audible distortion due to poor traction and,
although the loss of the higher frequencies is serious, it has been
held by many that the end has justified the means. Surface noise
is probably caused by a more or less random distribution of im-
pulsive shocks on the needle due to minute irregularities in the
record. It has been common practice in lateral recording to use
record material containing a certain amount of abrasive in order
to grind the needle to fit the groove. The irregularities due to
the abrasive would logically be expected to produce a scratchy
noise of much the character with which we are all familiar. A
5000-cycle note of the same loudness as a 10,000-cycle band of
surface noise using a reproducer with a flat characteristic would have

40 50

5OO 1000

FREQUENCY IN CYCLES PER SECOND

5000

O,OOO

FIG. 3. Energy distribution of surface noise from a cellulose acetate record.

an amplitude of only about 0.000001 inch. In order to reduce the
surface noise to the point where it is no longer troublesome, it appears
necessary to eliminate irregularities at least down to this order of
magnitude. It has been found that, if the usual abrasive record were
replaced by an unabrasive record pressed of a very clean homogene-
ous material such as cellulose acetate, the surface noise caused by the
record material itself would be greatly reduced. Such a change,
however, by itself, has been found to give a comparatively minor
improvement ; for, when this cause is moved well into the background,
other causes of surface noise of practically the same order of magni-
tude as that due to the abrasive of a shellac record become controlling.
The next process which it has been found necessary to improve has
been that of rendering the surface of the original wax electrically
conducting. The usual methods of graphiting or brushing with fine
electrically conducting powders have been found unsatisfactory.

Feb., 1932] VERTICAL SOUND RECORDS 145

Recourse has therefore been had to one of the earlier methods used in
phonograph practice, namely, cathode sputtering3 of the wax. This
method was not devoid of difficulty, however. With the best sputter-
ing technic the usual thick "waxes" are heated to such an extent as
to injure or destroy their finely engraved surfaces. By using a very
thin layer of wax flowed on a metal surface it is possible to keep it cool
during the sputtering operation. It is thus possible to apply an
extremely uniform, smooth, and tenacious surface of metal of adequate
thickness in a very few minutes. This can be electroplated by the
ordinary methods, and the electroplate used for pressing the final
record. By using this thinly flowed wax, it is possible to obtain
a surface texture which is extremely smooth and homogeneous and
which is also free from the mechanical strains incident to shaving
the waxes by the methods previously commonly used. In addition,
waxes of this type possess obvious advantages in ease of transporta-
tion, ruggedness, etc. When the noise due to the two causes dis-
cussed above has been removed or largely reduced, a third source
of noise is apt to become prominent. This involves the reaction
of the wax shaving on the recording stylus, which appears on the
final record as "clicks" when the shaving breaks or is removed
in a non-uniform manner. It has, however, been found possible
by suitable design to provide a recorder, stylus, and suction arrange-
ment such that the shaving is removed in a very smooth stream, thus
eliminating this type of noise to a large extent.

It has been common practice in the past to provide duplicate
stampers by electroplating the first stamper or "master" to obtain a
negative metal record. This in turn has been plated to provide the
duplicate stamper. A convenient and quick alternative method is
provided by sputtering and plating a suitable pressing made directly
from the "master."

These improvements in the methods of engraving and processing,
and in the final record material are more or less applicable to either
type of recording, lateral or vertical. Their full value, however, can
only be realized provided full advantage may be taken of the increased
frequency range which greater quietness permits. It is possible to
take advantage of this improvement to effect other improvements or
economies rather than to use it all in the one direction of decreased
noise. In amount, the reduction in surface noise from that of present
commercial records will differ depending on the frequency range
reproduced.

146 H. A. FREDERICK [J. S. M. P. E.

If a blank groove record, made with the improvements noted above,
is reproduced by a reproducer which is uniformly responsive up to
10,000 cycles, the surface noise is 20 db. below that of an old type
record reproduced in the same manner. If, however, all frequencies
above 5000 cycles are eliminated in each case, the difference is 15 db.
If now the noise of the new record reproduced to 10,000 cycles is
compared with the old record reproduced to 5000 cycles only, which
is the comparison of greatest practical interest, the difference in noise
is about 15 db. In addition, it is possible to take advantage of the
fact that most sounds to be recorded contain less energy in the high
frequency range than in the medium or low frequency range, and to
record the higher frequencies at a level somewhat higher than normal.
In reproduction these higher frequencies are then correspondingly
reduced by the reproducing amplifier or circuit. It is thus found
that a further reduction of about 10 db. in surface noise can be ob-
tained, the amount depending somewhat on the high frequency cut-off
of the reproducer or circuit. This effect occurs chiefly between 5000
and 10,000 cycles.

The ''volume range" for any particular frequency band is usually
considered to be the difference in decibels between the loudness of the
surface noise and the loudness of the maximum recorded sound which
the record can accommodate when reproduced faithfully over this
frequency range. With the lateral records of the past, reproduced to
5000 cycles, this volume range may be stated as about 25 to 30 db.
This figure obviously will differ somewhat for different cases, depend-
ing on the character of the sounds recorded and on the degree of
excellence obtained with the recording and processing methods
throughout. With vertical recording the reductions in surface noise
described above increase the volume range for a 5000-cycle band of
frequencies to from 50 to 55 db. For 10,000-cycle reproduction the
volume range is 45 to 50 db. Obviously, these new facilities open
the door for very great improvements in fidelity of reproduction and
for the reproduction of many effects not possible in the past. In
many cases it means that the surface noise may be reduced to in-
audibility.

Lateral records have usually been cut with a stylus having a tip
radius between 0.002 inch and 0.003 inch. The angle between the
two sides has, in this country, commonly been about 90 degrees. The
groove has been 0.002 inch to 0.003 inch deep and about 0.006 inch to
0.007 inch wide. The groove spacing has been 0.010 inch to 0.011

Feb., 1932] VERTICAL SOUND RECORDS 147

inch so that the uncut space between blank grooves has been 0.003
inch to 0.004 inch. If one groove is not cut over into the next, the
maximum amplitude which can be used is limited to about 0.002 inch.
If the usual loudness of the record is to be maintained it is necessary
to maintain this spacing between grooves.

With vertical records it has been found desirable, particularly
where a very loud record is to be made, to use a recording stylus with
approximately the same tip radius as previously used with lateral
records, but to reduce the divergence between the sides of the stylus
above the tip. In addition, it has not been found necessary to
provide any clearance space between grooves. In fact, it has been
found entirely satisfactory to have the side of one groove cut con-
sistently into the next. It is therefore entirely feasible to increase
the number of grooves per inch from the usual 98 to between 125 and
150, at the same time that the recording level is increased. When
using this recording stylus with the lesser divergence for cutting a
record with 125 to 150 threads per inch, it has been found desirable to
make the groove about 0.007 inch wide and about 0.003 inch deep.
The maximum amplitude may, under these conditions, be increased
about 4 db. It has been found possible, however, to obtain satis-
factory results with most waxes even though the normal depth of the
groove is increased to as much as 0.004 inch to 0.006 inch. In this
case, the recorded level may be increased 6 db. This increase in the
recording level obviously increases the volume range by a like amount.
If occasionally, due to a loud crash of sound, the recording stylus
completely leaves the wax, the reproducer will still "track" satis-
factorily; that is, continue in the correct groove. The corresponding
situation with a lateral record where one groove cuts into another is,
of course, fatal since in such a case the reproducer will usually cross
into the next groove. It has been found desirable with vertically cut
records to use a permanent reproducing stylus in order to reduce the
vibrating mass of the reproducer to a satisfactory value. This
stylus point remains sharp in contrast with the old steel needles used
with lateral records, and therefore will reproduce satisfactorily
undulations of sharper curvature. In other words, for the same
amplitudes the linear speed of the record may be reduced. Practically,
it may be undesirable to reduce or change the rate of rotation of a
record from the commercially used value. It is, however, feasible,
to decrease the internal groove diameter recorded on the 33 rpm.
record to about 6 inches for a 10,000-cycle frequency range. By

148

H. A. FREDERICK

[J. S. M. p. E.

the combination of the various elements mentioned above, it is
feasible to record for 15 to 20 minutes on a 12-inch record and for
10 to 12 minutes on a 10-inch record. This involves the use of
about 200 grooves per inch and a decrease in the recorded level
to about the level of laterally recorded records using 98 grooves
per inch. Of course, longer recordings can be made in the same space
if the recorded level is decreased (more grooves per inch), or if the
upper frequency cut-off is decreased (decreased rpm. and inner
diameter). However, these changes may introduce tracking difficul-
ties if carried too far and must be well justified by other considerations.
Laterally and vertically cut records drive the reproducer point
quite differently. Laterally cut records drive the point from both
sides but the point rarely follows the center of the groove with great
exactitude. It deviates from the center by amounts chiefly depen-

-50
-60
-70

-80
4

0 db= 1 VOLT ACROSS LOAD IMPEDANCE OF 1 OHM FOR
VIBRATORY VELOCITY = 1 CENTIMETER PER SECOND

0 50 100 500 1000 5000 IO,C

FREQUENCY IN CYCLES PER SECOND

FIG. 4. Response frequency characteristic of an experimental vertical
reproducer driven by cellulose acetate records.

dent upon the mechanical impedance of the reproducer. A vertically
cut record, on the other hand, drives in only one direction. The
restoring force is due . chiefly to the elasticity of the supporting
structure of the reproducer, the normal restoring force being equal
to the total weight on the needle minus the weight of the moving or
vibrating part. The stylus point will always remain in contact with
the record unless the forces set up by the undulations exceed this
normal restoring force. Operation should always be below this
limiting condition. This sets definite requirements on the mechanical
impedance of the vibrating parts and, unless this condition can be
met, reproduction of extreme frequencies by vertical records is im-
possible. With the vertical reproducers which we have used, the
stylus can follow sudden downward motions of the record groove even
to accelerations about a thousand times that due to gravity. With
laterally cut records, there is no definite limiting condition analogous

Feb., 1932]

VERTICAL SOUND RECORDS

149

to the above. However, it appears easier in practical design to
reduce greatly the mechanical impedance of vertical than of lateral
reproducers. Practical experience has shown that the mass can be
so reduced as to reproduce up to well above 10,000 cycles and the
stiffness reduced so as to reproduce down to the order of 20 cycles.
In fact, there appears to be considerable margin on this score. This
makes it possible to reduce the weight with which the reproducer
point bears on the record to between 2 and 20 per cent of what has been
used with most commercial lateral reproducers. This reduction in
stylus or needle point pressure has been found to decrease the wear
on the record very greatly, with the result that its life has been
considerably increased. Tests have shown that the first few thousand
playings cause negligible deterioration, and even several hundred
thousand playings do not show excessive wear if the record is properly
protected from dust and dirt.

RESPONSE IN DECIBELS

fc. 0 0 0 0 .C

Odb = l CENTIMETER PER SECOND PER VOLT PER OHM

x-^

\
\

„**•

,^--

0 50 100 500 1000 5000 10,0

FREQUENCY IN CYCLES PER SECOND

FIG. 5. Response frequency characteristic of an experimental vertical recorder.

A highly satisfactory method of providing a reproducer for
vertically cut records has been to use the type of structure with which
we are all familiar in loud speaker design ; namely, that in which a coil
moves in a radial magnetic field. Such a reproducer is simple and
sturdy. Its performance is linear over a wide amplitude range; it
may be made extremely light and, at the same time, is quite efficient.
The coils used have had a diameter of between 0.1 and 0.2 inch, and
the total mass of the vibrating system, including the diamond or
sapphire stylus, has varied with different models from 5 to 35 milli-
grams. The total force on the record has been reduced from about
150 grams to between 5 and 25 grams, the lighter structure being
used when playing from a soft wax. With the larger of these designs
it has been found possible to obtain efficiencies which are comparable
with the efficiency of the Western Electric oil-damped reproducer used
with lateral records. No difficulty has been experienced due to

150

H. A. FREDERICK

[J. S. M. P. E.

failure to follow the groove if the reproducer is mounted on a simple
pivoted arm, as in the case of lateral reproducers. Due to their very
small mass they operate quite satisfactorily even though the record
turntable fails to operate in a true plane, and even though the record
be considerably warped.

The response of the moving coil vertical reproducer is practically
constant over a very broad frequency range. It is shown in Fig. 4,
which is the characteristic of an experimental model taken with
cellulose acetate pressings.

The design of a recorder for use with vertically cut records involves
no fundamentally new problems over those used with laterally cut
records which have been described previously.4 It is still desirable
to design the recorder to approximate a constant amplitude character-
istic for the lower frequency range and a constant velocity character-

RELATIVE RESPONSE IN d
P ri, 1 _ rv

^ ,uO 0 0 0 C

. —

—~

•=.

— ^

7— =

-tf^w.

-*— -

^^1

—«_•«-

•i ^»

•- —

^=*

0 50 100 500 1000 5000 10,000
FREQUENCY IN CYCLES PER SECOND

. Over-all response frequency characteristic (recorder + reproducer +

network + amplifier).

istic for the higher range. This frequency characteristic has been
often shown and is familiar to all. The same recorders which have
been used for lateral recording can usually be converted for vertical
recording by the addition of a comparatively simple link system and
are quite satisfactory if a high frequency cut-off of 6500 to 7000
cycles is acceptable. It is, however, desirable to have a higher high-
frequency cut-off. Such a recorder has been used in making many of
the records which we have studied. Its frequency characteristic is
shown in Fig. 5.

The response of the oil-damped lateral reproducer is greatest at
the very low frequencies. Its response decreases with frequency,
this decrease in the lower frequency range compensating more or less
for the increase of response of the recorder with frequency. Because
of the flat characteristic of the vertical reproducer, it has been found
desirable to compensate in the reproducing amplifier or circuit for

Feb., 1932]

VERTICAL SOUND RECORDS

151

the low response of the vertical recorder at the lower end of the
frequency scale. A frequency characteristic for the combination of
recorder, reproducer, amplifier, and network is shown in Fig. 6.

It has been found with vertical records that speech is reproduced
with considerably improved naturalness and that the word endings,
sibilant sounds, etc., are much more distinct. The sounds of the

RESPONSE IN DECIBELS
£ .5 o 6 8 je

w_^

j — .

~x_

•^_>

Odb= SOUND PRESSURE OF 1 BAR AT MEASURING
POSITION WITH 0.25 WATTS ELECTRICAL INPUT

D 50 100 500 1000 5000 10,0
FREQUENCY IN CYCLES PER SECOND

i. 7. Response frequency characteristic of combined high and low
frequency loud speakers.

different instruments in an orchestra, particularly when playing a
loud passage, are reproduced with very great individuality and
clarity. Results of this kind are difficult to describe and should be
heard to be appreciated fully. If records such as those described are
reproduced using various low pass filters, the loss of distinctness due
to the elimination of frequencies above even 7000 cycles is easily

5000 10,000 20,000

FIG. 8. Field calibration of a moving coil microphone.

500 1000

FREQUENCY IN CYCLES PER SECOND

noticeable, whereas little or no difference in needle scratch or surface
noise may be observed, this being almost wholly absent in all cases.
The latter statement holds whether the records contain speech or
music or if blank grooves be reproduced. In listening to such records
a loud speaker has been used which is essentially flat over a large
portion of the range of audibility, its characteristic being as shown.5

152 H. A. FREDERICK [J. S. M. P. E.

(Fig. 7.) The reproducer frequency characteristic, as shown in Fig, 4,
is essentially flat to 10,000 cycles. A corrective network has been used
which compensates for the low frequency droop in the recorder
which, at the high frequency end, is, as shown in Fig. 5, essentially flat
to 9000 cycles. Thin metal-backed waxes have been used which,
after recording, have been rendered electrically conducting by metal
sputtering. The moving coil microphone has been used,6 the charac-
teristic being as shown in Fig. 8. The records have been pressed of
cellulose acetate.

REFERENCES

1 FREDERICK, H. A.: "Recent Advances in Wax Recording," Trans. Soc.
Mot. Pict. Eng., 12 (1928), No. 35, p. 709.

2 FLETCHER, H.: "Speech and Hearing," D. Van Nostrand Co., Inc., New York,
N. Y. (1929).

3 GUNTHERSCHULZE, A.: "Cathode Sputtering," Zeit. f. Phys., 36 (Jan. -May,
1926), p. 563.

4 MAXFIELD, J. P., AND HARRISON, H. C.: "High Quality Recording and
Reproducing of Music and Speech," Trans. A. I.E. E., 45 (Feb., 1926), p. 334.

6 BOSTWICK, L. G.: "An Efficient Loud Speaker at the Higher Audible Fre-
quencies," /. Acoustical Soc. of Amer., 2 (Oct., 1930), No. 2, p. 242.

6 JONES, W. C., AND GILES, L. W.: "A Moving Coil Microphone for High
Quality Sound Reproduction," /. Soc. Mot. Pict. Eng., 17 (Dec., 1931), No. 6,
p. 977.

DISCUSSION

MR. RICHARDSON: What is an L. P. filter?

MR. FREDERICK: An L. P. or low pass filter is one that cuts out everything
above the particular frequency noted, and transmits everything below this
"cut-off" frequency.

PRESIDENT CRABTREE: I think I have pointed out on several occasions that
the public has been satisfied to date with the reproduction of speech, but not
with the reproduction of music, because of its lack of range, both frequency and
volume. This demonstration has shown that the extent to which the frequency
range can be covered is excellent.

I am afraid that the range of volume is still inadequate for providing a facsimile
of orchestral music. But I think this demonstration shows an epoch-making
advance in sound reproduction. I don't believe that I have ever heard a re-
production of a film record that is as satisfying as some of the passages we have
just heard.

While I do not predict that the producers will hasten to adopt wax records
immediately, this performance will make them sit up and get busy, and either
match this quality on film or turn over to the disk.

MR. RICKER: We do not get quite the full benefit of these excellent records
in this room. This room lacks in acoustical qualities for a proper appreciation
of the magnificent work done.

Feb., 1932] VERTICAL SOUND RECORDS 153

MR. PALMER: We are always being told that the radio can produce better
sound in the home than talking pictures can in the theater. Can Mr. Frederick
give us any information as to whether the quality 'of the music reproduced here,
the fidelity of reproduction, is as good as or better than what the best radio
receiver can furnish?

MR. FREDERICK: I hesitate to hazard an answer to that, as I am not familiar
with the characteristics of all radio receivers. The reproduction of many re-
ceivers that I have heard was greatly inferior in quality to the reproduction to
which we have just listened, but I prefer to let someone who knows more about
that particular field attempt to answer your question.

I think you have been given a very definite picture of what those frequency
characteristics mean by listening with the different filter settings. That is why
we played the records with filters in and out so much of the time. One has to
hear these things again and again, and even then he would have to check up his
ears every once in a while, in order to have an accurate appreciation of what
they are hearing.

MR. MAXFIELD: Was not the reproduction level of the orchestra record as
reproduced here louder than would be heard in the center of the orchestra seats
in the theater where it was made? I frequently make tests in that theater, and
usually sit in the center of the orchestra. My impression here at the back of
the room is that the reproduction of the loud parts is louder than they appear
to be in the theater, in fact, a little uncomfortably loud.

MR. FREDERICK: I believe it was.

MR. CARLTON: What type of acetate was used for the new record? What
method was used for the production of the cellulose acetate from which the
record was made?

MR. FREDERICK: I cannot tell you in great detail. It was a very pure acetate.
We obtained it from various sources. Du Pont, for example, has supplied it.

MR. CARLTON: Is it molded?

MR. FREDERICK: Yes, with slightly higher temperatures than are used for
most other record materials.

MR. HICKMAN: I believe if the Bell Telephone Company were to present this
entire outfit to an average person living in an apartment, and provide an easy
means of adjusting the quality of reproduction to suit his taste, you would find
that in general he would eliminate all components having a frequency greater
than about thirty-five hundred. If the same person were to listen at a hole in
the wall leading to an auditorium holding a good orchestra, the hole being dis-
guised by a loud speaker design, he would tell you that the reproduction was
rather good but was deficient in low frequency response. As gramophones are
getting bigger and bigger, if you produced one big enough for a man to get inside,
and let him speak through a speaker aperture, the observer would tell you that
it was pretty good but not quite like the human voice. There has grown up,
since the reproduction of canned music, a sort of new standard of what is desirable.

Why is it that instinctively we object to what should be the more correct
form of recording? I am speaking as a layman — as an enthusiastic amateur
musician.

Is it possible that, when the most perfect reproduction has been made, in
picking up the sound from the record a high frequency chattering is created

154 H. A. FREDERICK [J. S. M. P. E.

which cannot be expressed as harmonics but as a slight disagreeable individuality
imparted to the record after, say, a frequency of five thousand, which we would
rather have cut out?

MR. FREDERICK: I do not think that your question about reproducing these
high frequencies was directed particularly to me or that you expect me to answer
about the tastes of people. And I am not sure that I understand your last question.

PRESIDENT CRABTREE: I think Mr. Hickman wants to know why the fre-
quencies above five thousand seem to annoy one in the home.

Recently a friend recommended for my radio a new speaker which had a
straight curve up to I don't know how many thousand cycles. I obtained this
speaker and compared it with my old speaker, which certainly does not reproduce
above five thousand cycles. After repeated tests my observation was*exactly
the same as Mr. Hickman's — that in spite of the high frequencies, I preferred
the low ones. The frequencies higher than about five thousand apparently
were annoying, and seemed to irritate the ear. From this observation, it seems
that straight line reproduction is not always necessary, but depends on the condi-
tions of the room in which the reproduction occurs and on the tastes of the indi-
vidual.

MR. HICKMAN: I am not questioning whether the reproduction shall be linear.
I am asking why, when apparently the reproduction is most faithful, we do not
find the reproduction of the high frequencies pleasing. I want to know whether
some particular form of high frequency distortion is introduced in the pick-up,
or later on, which is not in the musical record.

MR. FREDERICK: I do not think that anybody knows enough to really answer
your question. And I doubt very seriously if a simple answer could be made if
anyone did know enough. However, this point is extremely important. When
you extend the high frequency range, if there is any distortion anywhere in the
system it may be made audible and distinctly annoying, whereas it was previously
inaudible. If the frequency range is to be extended upward a distinctly higher
grade of performance must be obtained of all parts of any reproducing system
than may seem perfectly tolerable with a lower high frequency cut-off.

The loud speakers so far designed are certainly not perfect. The curves
shown in the paper indicate that by far the most jagged and roughest curve of
all was that of the loud speaker. It seems reasonable to think that these irregu-
larities, which certainly must mean a certain amount of resonance and "hanging
on" of the sounds, must have an effect on the ear.

Now, if we could get a perfect loud speaker — and remember, that a perfect
loud speaker means that it must be considered in conjunction with the particular
room in which it is used — the results would undoubtedly be greatly modified.
The characteristics of a loud speaker will be different in one room from another,
and may be quite different in different parts of the same room. But placing our-
selves at a particular place in a particular room, and having an ideal loud speaker
to project the sound, I personally am convinced that we would, as soon as we were
used to it, all vote for as broad a range as we could possibly get, and the most
perfect or straight-line reproduction.

The trouble is, as we advance in our halting manner, we often make an im-
provement which shows up defects which were previously inaudible.

MR. EVANS: Is it not possible that the curves may not tell the complete story?

Feb., 1932] VERTICAL SOUND RECORDS 155

The curves that we have seen are for continuous tones — single frequencies — and
do not take into account transient effects that may exist. If we knew more
about transients, might it not be possible to answer the questions that have
been asked here?

MR. FREDERICK: I think so. At least, it would take us further.

MR. MILLS: I speak not only as a layman, as some of the others have spoken,
but also as one peculiarly inept in music. I have at times attended symphony
concerts and orchestral renderings, and suffered from the higher and rasping
violin overtones, and from irritating high frequency sounds of brass instruments
and cymbals, and it may be that I prefer to listen to a little thirty-five hundred
cycle cut-off loud speaker, and interpret its output as music.

But it may be that those who have a wider appreciation of music than I, and
a greater discernment, would prefer the more nearly complete reproduction
which includes the higher overtones.

I should like to ask Mr. Frederick whether he would briefly summarize four
or five points: What is the increased range of loudness which this new record
is capable of providing, over and above the previous loudness range? What is
the increase in frequency range? What is the increase in time recorded under
normal conditions? And what is the increase or decrease between the cellulose
acetate record and the normal shellac, in ground noise, at the various frequencies
which are important?

MR. FREDERICK: The volume range was stated in the paper as being about
twenty-five to thirty decibels for most records. I have called them shellac
records. They are not simply "shellac" records, but shellac plus a lot of other
technic which accompanies it. The volume range with the type of record demon-
strated here lies between fifty-five and sixty decibels, according to the best data
we have. The improvement is not due merely to cellulose acetate. It is due
to a combination of changes. If one of three or four causes, all approximately
equal, is eliminated from consideration, the improvement which will have been
made in the total effect is of course fairly small. Our observations have led us to
believe that with the old records the noise due to the shellac was somewhat
greater than the other noises, but only a few decibels greater. As soon as it was
reduced a little the other components of the noise came into evidence.

Regarding the time of recording, I tried to summarize this matter in the
paper, but it is difficult to give any simple and definite figure to cover the entire
question of playing time. . On the older, lateral record, a greater number of
grooves per inch was sometimes used. Something has to be sacrificed to do this,
but it may be worth while. Edison put out hill and dale records which played
thirty or forty minutes. They were not successful because they did not have
certain other characteristics which were needed. But as far as playing time is con-
cerned, that is something on which I don't believe you can make any simple
statement.

PRESIDENT; CRABTREE: For a ten-inch record, how long will a lateral play
and how long a vertical?

MR. FREDERICK: The usual ten-inch lateral record will play three minutes,
I believe, and on this type we have thought it good practice to make it play ten
or twelve minutes.. But if the "game is worth the candle" the time can be made
longer. You have to sacrifice something else, however, to do so.

156 H. A. FREDERICK [J. S. M. P. E.

MR. VICTOR: Is there a relation between frequency and volume?

Perhaps the high frequencies carry farther, and perhaps it may be possible
to introduce an automatic modulator of some kind for the home, that might tone
down the high frequencies to a level that would be more pleasing to the ear.

MR. FREDERICK: Of course it is the easiest thing in the world to get rid of
them. The trouble is to get them. High frequencies, I think, are generally
found not to carry in distance as well as the low frequencies. A good example
of this is furnished by a man making a speech in an open space. As you walk
toward him, from a distant point, you first hear the sound of his voice but cannot
understand a word. As you get closer, you perceive more of the high frequencies,
until, when you get close enough to get the frequencies on which his articulation
depends, you can understand what he is saying. This is not simply a matter of
how the various components are transmitted, but is also concerned with the fact
that the lower frequencies in speech, as in music, are usually very much stronger.

PRESIDENT CRABTREE: What Mr. Victor had in mind was a means of con-
trolling the various frequency components, not to cut them out entirely, but
selectively to diminish their volume.

MR. VICTOR: That is right. When a soprano voice comes over my radio I
usually reduce the volume.

MR. FREDERICK: If you should go to a concert to hear a first-class soprano,
you would not think of doing such a thing. If you should go to hear a first-class
orchestra you would not expect to do it. When we do things of that kind I
believe we only try to compensate for the faults of the equipment.

MR. RICHARDSON: When I said that the sound was best, according to my
judgment, with the seven thousand cycle components included, I did not mean
that it was most pleasing, but most natural. A railway whistle, of high fre-
quency, is annoying to everyone. But everyone rather likes to hear a steamboat
whistle, which is of low frequency.

Some people enjoy a soprano, but they are the exception. But I do not believe
that there is anyone in this room who would not enjoy the sound of a contralto
singing "Silver Threads among the Gold" or something of that kind. The
sounds are pleasing to everyone. The high frequencies — the soprano, the loco-
motive whistle — are annoying to the nerves, not to the ear.

MR. SHEA: I think there is a great deal in what Mr. Richardson has said.
It is probably true that most sounds which are startling or grating, as some
people call them, have a large high frequency content. It seems to be the general
experience that for the high frequencies you must have clearer reproduction.

MR. KELLOGG: Quite a little comment seems to have been inspired by the
idea that during some of the numbers there was a kind of unnaturalness, par-
ticularly where there was fairly complex music — for example, the orchestra.
It sounded to me as though it might possibly have been due to some imper-
fection in the correspondence between input and output — non-linear distortion
as we sometimes call it, or it might have been an effect such as one gets when
in a room with an orchestra and the reverberation is rather high, particularly
in the high frequency range. I should be interested to know in the case of the
orchestra recording whether the acoustics of the recording room were such as I
have described.

I have another question: The piano was very steady and firm, a condition

Feb., 1932] VERTICAL SOUND RECORDS 157

which obtains only when the turntable is rotating very steadily. A year and
a half ago a turntable mechanism was described wherein a great deal of refinement
was gone into, to avoid speed irregularities. I should be interested to know
whether that type of turntable is used both in recording and reproduction in
this case.

MR. FREDERICK: In answer to the first question, as to whether there was not
some non-linear distortion somewhere in the system, I do not think there is any
question but that there is, but I believe that there was perhaps less of it than we
sometimes hear.

The conditions of the pick-up of the sound were not what we should have
chosen if we had had a place where we could move the microphone about. A
single microphone had to be placed close to the conductor; those conductors
to whom I have spoken, and in whose opinions I have great confidence, insist
that they cannot judge from the conductor's position what the orchestra should
sound like. They have to permit another conductor to rehearse the orchestra,
so that they may go out into the body of the hall to get the correct effect. As
to turntable speed control, the recording machine was the usual Western Electric
Company recording machine. The reproducing turntable was driven by a
synchronous motor with multiple belt speed reduction.

MR. MAURER: How abrupt was the cut-off of the low pass filter used in this
demonstration ?

MR. FREDERICK: The cut-off is quite abrupt. A matter of a few hundred
cycles means a great many "db.'s."

MR. THOMPSON: Is this type of reproducer more responsive to vibrations of
the turntable, or irregularities of that kind, than is the present lateral pick-up?

MR. FREDERICK: Of course these records would not play on a seventy-eight
rpm. turntable, because they were recorded at thirty-three rpm. They are
far less sensitive to certain types of irregularity, due to the small mass of the
pick-up and the small pressure on the record. We had a record one time that
was pressed with a stamper, that had been bent at least an eighth of an inch out
of plane, out on one edge, so that there was a big bulge in the record. The re-
producer tracked over this without difficulty, and no trouble was experienced in
the reproduction in doing so — until the pick-up reached the point where the
frame of the reproducer hit the bulge in the record.

As to vibration, I should hesitate to say, because it would seem off-hand that
a vertical reproducer would tend to be somewhat sensitive to it. So far as my
experience has shown, it has seemed to be certainly no more sensitive to that
type of trouble, and my impression is that perhaps it is a bit less sensitive. But
I should hesitate to make a definite answer on that.

MR. OLNEY: The question of the frequency range of radio receivers has been
raised in the discussion several times, and someone inquired as to how it compared
with the reproduction we heard today. There is no comparison between them.
In order to reproduce the low frequencies you have been hearing today, a receiver
cabinet would have to be the size of the panel you see there, which is out of the
question. As far as high frequencies are concerned, the requirements of selectivity
prohibit reproduction of anything higher than five thousand cycles. This is a
theoretical limit. Practically, the response drops off in the best radio receivers
between four and five thousand cycles. In the poorer receivers it may drop off

158 H. A. FREDERICK [J. S. M. P. E.

at three thousand cycles. In radio receivers equipped with the so-called tone
controls, it may be possible for the user to reduce the cut-off frequency to fifteen
hundred cycles. Some persons prefer that.

I do not think it is because they object to a normal amount of high frequencies.
Some have claimed that what annoyed them were the frequencies above five
thousand cycles. Those frequencies, I believe, are not reproduced by any radio
receiving set on the open market today with which adjacent stations can be
separated. I believe that one of the difficulties is that most loud speakers used
in radio receivers have a very exaggerated response to frequencies in the neighbor-
hood of twenty-five hundred or three thousand cycles; and unless these peaks
are suppressed in some manner, the reproduction is bound to be unpleasant.

In commercial receivers, correctly designed, an attempt is made to equalize
these defects in the loud speakers. When they are not equalized you will get
this impression of harshness in the upper register; but this is not due to the
frequencies above five thousand.

MR. VICTOR: I am afraid these discussions will read as if our Society members
criticized this performance. I should not like to see the records so appear after
such a splendid performance. It is the best I have ever heard.

PRESIDENT CRABTREE: Exactly. But we must not close our eyes to the
imperfections. We will never progress if we do not criticize our own work. The
man who is satisfied with his own work never gets anywhere.

MR. FREDERICK: I should like to add a comment suggested by something
Mr. Victor said, that I may not have properly answered. As the volume level
of any reproduced sound is raised or lowered, the quality appears to change,
due to the physiological characteristics of our hearing. If it is played too loud
the balance is off in one direction, and if it is played softer than in the first place
the reproduction is off in another direction, even though the frequency character-
istics of the reproducing system, taken either in part or in whole, are entirely flat.

Now, one interpretation of his comment might be that if a person simply
must play the reproduced sounds too loudly, he needs to have a certain amount
of distortion to make up more or less for what his ear is doing. Of course,
I do not think that is the proper thing to do. He ought to endeavor to play
the reproduced sounds at the level at which they were originally produced and
picked up. If that were done he would not need that kind of a compensating
system.

MR. HICKMAN: Mr. Victor put a point for me, that I had in mind. I think
that anybody in reading this discussion might imagine that our criticism had
been of this demonstration. This demonstration has been the most perfect
music I have heard. My own criticism referred to reproduced music in general.

DR. GOLDSMITH: There were two points about this extremely interesting and
significant demonstration, which I believe merit consideration: In the first
place, comparisons were made to radio. There is no possibility of comparison
to radio, because transmitting stations today send out approximately forty-five
hundred cycles. The networks of the country carry very little above that.
A few stations are going up to approximately eight thousand cycles in trans-
mission. But taking five thousand cycles as the present transmission of radio
stations, it is obvious that a wide open receiver, receiving from zero to ten thou-
sand cycles, would get only the upper five thousand cycles of extra noises super-

Feb., 1932] VERTICAL SOUND RECORDS 159

imposed on the signal; and a five thousand cycle low pass one would be per-
fectly justifiable.

If, below that, there are cut out frequencies from five thousand down to three
thousand, then reception suffers considerably. A great many people prefer
that, however, because they have been working in noisy locations all day, and
want to be soothed rather than given an esthetically true reproduction.

Then, a corollary of that is that home conditions are not the same as conditions
in an auditorium. If you listen to a great conductor, you discover that the
faintest whisper of sound can barely be heard, whereas, the climax following
immediately, nearly brings the plaster off the ceiling, if the theater has not been
acoustically treated. That condition must not exist in the home for the reason
that a radio receiver in the home would be an intolerable nuisance and would
start a neighborhood feud. We have to limit the volume range. We have
something here that is adequate, indeed, for home purposes, and has the maximum
of volume range which is permissible and consistent where people live near each
other.

Another feature that we have to take into account is our own reaction when
we hear things for the first time, in that we have a habit of reverting to old stand-
ards. I remember the first time a certain man, who was quite a capable man in
his field, listened to a modern record, and remarked, "That is not good at all.
It does not sound like a phonograph." And so we have to be careful. We
must remember that we have all made a mental adjustment, a charitable ad-
justment if you wish, to reproduced music. We have accustomed ourselves
to making mental allowances, applying the necessary automatic corrections.
And we apply them to something where they are no longer relative.

I regard this as a most impressive demonstration. And throwing a bright
light on nature, and holding up a brightly polished mirror to look at nature,
is crude, but it is the only way to progress.

One other point, and that is the matter of the capabilities of film records. It
is to be hoped and believed that results like this will properly stimulate the
production of film records for theater use, which will be of equivalent quality.
There is nothing theoretical or impossible about that. The trouble now is not
with what is on the film, but the acoustic qualities of the theater.

MR. FREDERICK: I think it would be hopeless to try to summarize all the
discussion which has taken place.

I regret that I am not fully familiar with all the facts regarding the transmission
circuits of the networks connecting studios to radio stations. I know a great
many of them are good to eight thousand cycles, and a great deal of effort has
been made to make them good to eight thousand cycles. I think it would be
unfortunate and quite incorrect if we should take away with us the impression
that the transmission circuits were limiting or will limit radio transmission to
five thousand cycles.

(The following discussion was held on the occasion of the re-presentation of Mr.
Frederick's paper, and the address of Mr. Stokowski published on page 164 of this
issue of the JOURNAL, at the meeting of the New York Section on December 9, 1931.)

MR. RICHARDSON: One of the worst things we have to contend with in re-
production and projection of sound is dust, both in records and in the film itself.

160 H. A. FREDERICK [J. S. M. p. E.

It will adhere to the film, particularly when a little oil has gotten on the film
and will set up heavy ground noise.

It would seem to me that a record of hills and dales would be much more easily
injured; it would be more difficult for the projectionist to keep it clear of dust
and abrasive materials than a lateral record.

It also seems to me that a sapphire or a diamond point needle running over
dust, which unquestionably will collect in the grooves, would have a more in-
jurious effect and set up a greater amount of ground noise than would be the
case with the lateral record groove.

Finally, I should like to know what are the limits of range of frequency in con-
versation and in music.

MR. FREDERICK: As to the effect of dust, our experience with these records
has not indicated any particular difficulty. We have taken no particular
precautions to avoid trouble. Except where we wanted to play a record thou-
sands and thousands of times, continuously, we have found it necessary to take
no special precautions whatever.

As to the limits of speech or music, opinions may differ as to that. If anyone
will provide an adequate or an accurate audiogram showing what the upper and
lower frequency limits are for his own ear, that will provide the answer to the
question. If he is very young and can hear from twenty cycles to seventeen
thousand cycles, twenty and seventeen thousand are the limits of his speech and
music. If his hearing isn't quite so good and he can hear only from twenty to
three thousand, why, that is the limit for him.

(At Mr. Crabtree's request, Mr. Frederick repeated a record, using only the speakers
which reproduced the high frequencies.}

MR. CRABTREE: Mr. Chairman, I hasten to congratulate Mr. Frederick and
his collaborators at the Bell Laboratories on this epoch-making development.
The demonstration this evening, especially of the organ, shows how inadequate
the present apparatus is and the present theory. I don't know whether Mr.
Zukor, Mr. Lasky, or Mr. Harley Clarke were here tonight but if they weren't
they should have been, and perhaps it might offer them some hope of getting the
people into the motion picture theaters today if they would put on music of this
high quality.

MR. SCHENCK: In your opinion, Mr. Frederick, when these high frequency
speakers are playing — we haven't been accustomed to hearing frequencies of
that order — were those actual reproductions on that frequency band or would
you say there was distortion present?

MR. FREDERICK: There is no question that the day of reproducing sound
without distortion is not yet here. Surely there is some distortion there. All I
should say is that I think there is a little less than I have often heard, and I
hope that five years hence there will be still less.

I waver between two feelings on this whole matter. Some days I quite enjoy
listening to some of this music, and most of the other days I feel greatly impressed
with the fact that we yet have a long, long way to go. This is not perfect and
the day of perfection is a long way off.

MR. SCHENCK: We are not accustomed to hearing the high frequencies re-
produced, and I am merely wondering whether we jump to conclusions about the
distortion at that high frequency, particularly in connection with the orchestral

Feb., 1932] VERTICAL SOUND RECORDS 161

record, wherein it sounded to me as if the cymbals were playing. At one point
in the record it seemed almost certain that there was distortion. Following
that, it started to clear up somewhat, and I could hear the high frequency instru-
ments such as the cymbals or the bells that you mention.

MR. STOKOWSKI: Those loud crashes are cymbals. But they are cut off at
nine thousand cycles. You need at least thirteen thousand, according to our
experiments, and perhaps more. That is why they sounded strange.

MR. CRABTREE: What is the thickness of the records, and how are they made?

MR. FREDERICK: These particular records are about a quarter inch thick.
They can be made two-hundredths of an inch thick. They are thermoplastic,
not like bakelite. Under the application of heat they soften. They were pressed
in the usual manner.

MR. WILSON: I would like to ask Mr. Frederick if the sound level, as it
appears to the average person sitting here tonight listening to the orchestral
record, is approximately what would be expected in an equivalent position in an
auditorium listening to the actual orchestra. It is difficult, looking at a bank of
loud speakers, to judge whether one is hearing the true level or something con-
siderably above what he would get from the real orchestra.

MR. FREDERICK: I fully appreciate the difficulty you express of judging
whether the level is right. Remember, you always hear an orchestra with two
ears. The binaural effect changes your impressions always. You also use your
eyes when you hear an orchestra and I think that what you see also changes the
general impression quite a lot.

The loudness can, of course, be definitely measured and can be compared
under different cases. We haven't actually made such measurements in this
hall, but it is my impression that the loudness, both in the case of the organ
record and in the case of the orchestral record, was fairly close to the original
loudness.

MR. EDWARDS: I should like to ask Mr. Frederick about tracking. That
is the great handicap of lateral recording, the thing that has given the most
trouble in projection.

This is the first time that I have seen a reproducer that hasn't depended for
its trackage on a thread and screw.

In the illustration showing the difference between the hill and dale and the
lateral recording as placed on the record, I noticed that in the case of the hill
and dale recording the wall of the record is cut very much lower than the surface
level. Would not a little wear cause a great deal of difficulty in tracking, es-
pecially with the free reproducer?

MR. FREDERICK: I don't think I have ever seen one of these fail to track. We
have had practically no trouble at all from this. I don't doubt there may be cases
where they haven't tracked but I don't remember ever seeing one. That hasn't
been one of our difficulties.

MR. EDWARDS: I think possibly the most notable example of detracking in a
lateral record was in the picture, Lilac Time. In that picture there was a shot
in the center of one record, and I think that shot must have cost the producing
company a matter of twenty-five thousand dollars for that record because, once
played, the next time it went through the wall. It brought disastrous results
to everybody concerned.

162 H. A. FREDERICK [J. S. M. P. E.

MR. FREDERICK: Of course, with a lateral record, if an extra broad deviation
of the groove occurs there is danger of cutting over into the next groove. With
the vertical cut, even when the cutting stylus leaves the wax entirely, we have
never experienced any difficulty in tracking. There is some distortion, of course,
due to the fact that the top of the wave has been cut off. But it tracks perfectly
well. And I think that is a rather important practical advantage of the vertical
as opposed to the lateral type of record.

MR. CRABTREE: Might I ask Mr. Stokowski to tell us what is lacking in the
music from a musician's standpoint? First of all, is the volume adequate?
Do you get the thrill from the reproduction that you do from the actual orchestra?
Is it lacking in depth or static effect? Do you notice the lack of perspective in it?

It is only by criticism of this kind, of course, that we can really advance;
find out what is lacking and then try to improve it.

MR. STOKOWSKI: As to volume range, it is approximately, in my opinion,
the same as in the original orchestra but in frequency there is a departure. The
cymbals don't sound like cymbals because, as I said before, they are cut off
at about nine thousand cycles and they need thirteen or fourteen thousand.
The range between nine thousand and thirteen or fourteen thousand is necessary
for several other instruments to give the proper tone color. It is a pity we
do not have a word in the English language for timbre. We ought to invent
one, because we need technical terms which have an exact significance and are
invariable in their meaning.

We have in Philadelphia, in the monitor room (a room, I suppose, about one
hundred and twenty feet long, so that there is plenty of space in which the tone
can develop), the Bell Telephone Laboratories' loud speaker, different from
this one. This is a double speaker; we have a triple loud speaker there, wired
from the microphones in the hall. We have usually three, four, or five different
microphones in different positions, so we can switch from one to the other.

When we sit in that room, which is soundproof, we don't hear the original.
We hear the music only from the loud speaker. And we have there a most
wonderful and faithful reproduction of the orchestra. I go in and conduct the
orchestra for a time, to get the direct sound of the orchestra. Then I go down
the hall about two-thirds of the way and listen to the orchestra from that point,
which is the average listening point for the public. Then I go into the monitor
room and compare what I hear there with what I heard outside, and it is a very
faithful reproduction. From that comparison I notice that if we cut off from
about 15,000, as we have done there, down to 9000, we not only cut off those
higher frequencies, but there is some relation between those high frequencies
and the ones which exist from 9000 downward, and they, too, are changed. The
frequencies ranging from one to five thousand should remain the same when the
frequency range is cut down to 9000 but, to the ear, they don't remain the same.
They are changed in some way. You get a totally different sound. And that
is, I think, one thing that will be gained when we have still higher frequencies
than this, which we undoubtedly will have, because we already have them in
Philadelphia.

MR. RICHARDSON: I believe that this style of recording will meet with trouble
in the projection room due to the abrasive effect of dust in the bottom of the
groove where the pressure must come from the needle. It must be borne in

Feb., 1932] VERTICAL SOUND RECORDS 163

mind that the conditions in the laboratory and those in a projection room are
quite different, particularly in the smaller theaters.

MR. HORNBLOWER: I should like to know whether, in checking the original
against the reproduced sound of the symphony, consideration was given to the
fact that the symphony orchestra would occupy a stage as large as the one you
are standing on, that your base drum would be, say, thirty feet from the first
viol, and so on, while in reproducing you get everything within an area six feet
square.

MR. FREDERICK: I tried to bring out that point before, that one of the limita-
tions of this type of reproduction is that we are effectively listening with only
one ear, picking up with a single microphone, whereas under normal conditions,
in a hall, we hear with both ears, and the orchestra is spread out.

We have taken that fact into account in some recording work by placing the
microphone at an adequate distance from the orchestra.

In the particular orchestral record which we played here, we were obliged to
have the microphone close to the conductor's stand, which we know is an atrocious
place for it ; but it was impossible to place it anywhere else, and I am quite sure
that the record was very much injured as a result of it.

MR. SIMMIONS: Mr. Chairman, I should like to ask Mr. Frederick why the
research which has been done at the Bell Telephone Laboratories has been
confined to the speaking voice and has not included the singing voice.

MR. FREDERICK: I didn't think it had been.

MR. SIMMIONS: Then very little has been accomplished in regard to the
singing voice.

MR. FREDERICK: Of course, in the telephone business our greatest interest
is in speech, although we have done some work in other directions. We have
played records here of singing voices.

MR. SIMMIONS: I should like to know if the physicist alone can solve the
problem of the singing voice, from the physical point of view. In order to carry
on this research work I have suggested that there should be three physiologists,
three musicians, and three teachers or psychologists — of course, the singing teacher
is a psychologist — and then these nine men together could accomplish something
in regard to finding out the exact amount of pressure which is necessary in order
to produce a beautiful sound.

I would suggest Mr. Stokowski, Mr. Damrosch, and Mr. Bodanski as the
three musicians on that research committee. The Academy of Singing Teachers,
whom I have approached, suggested that they should select three men from
among their ranks. Regarding the physiologists, I have spoken to Dr. Williams of
Columbia University and he is interested. I have spoken to Professor Wisluki,
who is professor of anatomy at Harvard University, and Dr. Frank E. Miller;
and to Dr. Fletcher, Dr. Watson, and Dr. Knudsen, of the University of California.

These men, with the help of the singing teachers, should get together to solve
the problem. If they did so I am quite sure the problem of singing in relation
to the films would be solved. But as I say it is not a one-man job.

The same standards which have been used in checking the singing could be
applied in teaching control of the human voice, so that I should not have to
depend on the monitor man when I go on, as Mr. John McCormick does. You
know, he said the monitor man changes his voice.

SOUND RECORDING— FROM THE MUSICIAN'S POINT OF

VIEW

LEOPOLD STOKOWSKI*

An address delivered before the New York Section, December 9, 1931, following
the re-presentation of the paper "Vertical Sound Records: Recent Fundamental
Advances in Recording on Wax," by H. A. Frederick, published in this issue of the
JOURNAL on page 141.

As we listen to music, if I may speak purely from the musical
standpoint, we have two kinds of reaction. First, there is the
physiological reaction.

When you heard the great volume of tone coming from the organ it
thrilled you. It uplifted you. It excited you. I am sure you would
find upon analysis, that your heart was beating more quickly, that
your blood was flowing more quickly, that your nervous system was
tremendously stimulated. That is the physiological reaction.

If you hear a good military band playing in the street, with a
really good rhythm, you want to march. That again is physiological.

If you hear very good dance music, you want to dance. Again,
the physiological reaction.

The other kind of reaction is the psychological, the emotional. If
you hear music of a certain type it arouses in you intense feeling. If
you hear music which has very powerful contrasts — very loud, then
very soft; very quick, then very slow; and so on — that has a
psychological effect on you.

If you hear music which has very rich colors in it, and differences of
tone colors, that again has a psychological effect. Melodic form, the
flowing up and down of melodies, tunes, motifs — that also has a
psychological effect.

There is also a very mysterious thing about music. It is psychic
suggestion. I work all day long and every day in music. I experi-
ence every day, the whole day, the next day, that week, that month,
ten years past, this psychic impression and suggestion that comes
from music. So it is with all musicians; we talk about it; we think

* Director, Philadelphia Orchestra, Philadelphia, Pa.
164

SOUND RECORDING 165

about it; but we don't know what it really is. We feel it vividly, but
we don't understand it at all.

That is, to my mind, the most important part of the reaction of the
music lover or of any one listening to music. That suggestive power
which can carry us into the most remote spheres and realms of feeling
and thought, and things that are higher than thought and higher
than feeling — that is the important part of music. And in order for
this to be done we must have this greater range which we have had
demonstrated here tonight; greater range of frequency, of volume,
and the elimination of foreign noises, needle scratch, static, and all
the noises that we hear in radio. You hear on your radio the dial
telephone in the next room; you hear the refrigerator; you can
hear all the vegetables in the refrigerator talking to each other; and
when the cook takes them out of the refrigerator and puts them on
the electric stove and switches it on, you hear that. And so it goes.
We must find methods of eliminating all foreign sounds.

When our orchestra plays in Philadelphia, or as we played last
night at Carnegie Hall, here in New York, we give out a volume range
of about 75 db. ; but when we are recording we do so at about 35 db.
And I think it is important for everyone connected with music, and
the public at large, to know definitely and quite clearly — it is no
secret, but quite plain — that when they listen to a record, or when
they listen to the radio, they are listening to a sound level of approxi-
mately 35 db., sometimes less, sometimes a little more. But when
they listen to a symphony orchestra, which is, I think you will agree
with me, the most difficult thing to record or to transmit, they are
listening to a level of about 75. We must find a way of increasing
that 35 to 75 before we really can give the public what it ought to
have in the way of expression in music.

That is one dimension, so to speak. Then there is the other
dimension, the up and down dimension, the frequency range. When
we play as we did last night at Carnegie Hall, in the overtones, or in
the fundamentals, we are producing frequencies certainly up to 13,000,
probably more. But we know certainly that it is up to 13,000.
When you hear a record or when you hear music over the radio, you
are hearing frequencies of about 4500, often less, sometimes a little
more. The average, however, is about that. You can easily measure
it and find out for yourselves whether I am telling the truth or not.

Last Friday night we had a concert in Philadelphia, and after the
concert we made a number of tests, in connection with the Bell

166 LEOPOLD STOKOWSKI [J. S. M. P. E.

Telephone Laboratories, and these are the exact figures we got from
those tests:

We asked the first oboe player to play. We were in a room a long
way from the room in which he was playing. We had previously
arranged everything so that what we heard was an exact reproduction
of what was happening on the stage. The oboe player was sitting in
the same seat he always occupies during a concert, so that it was an
exact reproduction. And we found that he needed frequencies up to
13,000 to express his tone color.

Then we took the trumpet and we found that up to 8000 cycles it
gave a satisfactory effect.

The piccolo took up to 6000, and that was a very astonishing thing:
that the piccolo, which is a very high pitched instrument, should
require up to 6000, whereas the trumpet, which is a lower instru-
ment, requires up to 8000 and the oboe, a moderately low pitched
instrument, requires up to 13,000. That is something you couldn't
determine without exact experiments like these.

Then we took the violin. It needed up to 8000. The cymbals
needed up to at least 13,000, probably more; the tympani, 6000;
triangle, 13,000; xylophone, 6000; snare-drum, 13,000.

I was doing these experiments, but the Bell Laboratory scientists
were all watching very closely so that there was no chance of exag-
geration or mistake. Those are the exact results.

In order to express all this, in my opinion, we must find out what
the average living room, with the average curtains, rugs, paintings,
and all the things that our wives like to have in our living rooms, which
affect the tone, its absorption, and so forth — we must find out what
the average living room will take in the way of volume range. We
really don't know that exactly, yet. At least, I have never found
anybody who did. In my opinion we must know that and we must
experiment along that line.

The same thing applies to the average theater in which sound
pictures are shown. They vary greatly, and when we record sound,
music, or speech, no matter what kind of sound it is, we must have
those conditions as nearly as possible invariable. They must be the
same, because we record in a certain way, to project the sound in a
certain way, and then if the projecting instrument and the hall or
room in which it is sounding is different in each case, a different effect
will be produced in each case.

This is the place, in my opinion, where standardization is very

Feb., 1932] SOUND RECORDING 167

desirable. In many other things in life, such as thought, emotion,
etc., it is very undesirable to have standardization. But it is impor-
tant for us to see clearly where standardization is necessary and I
think it is necessary here.

It is the same with frequency characteristics. This hall has
certain frequency characteristics; your living room where you play
your radio, where you play your gramophone, has frequency character-
istics. In producing our music for the gramophone or for the radio
we should know roughly what is going to be the frequency character-
istic of the place in which it is going to be played or the whole thing
will be distorted.

In my opinion the gramophone and radio are twin brothers. There
is often a certain antagonism between those who follow one god and
those who follow the other god. But they are fundamentally the
same, and they help each other very much.

For instance, we broadcast our concert last Saturday night.
Forgive me if I speak about what we are doing. I do it with a
definite purpose, not to be personal or to boast in any way — far from
that — but I want to tell you tonight about my own direct experience,
not what I have read in books or what someone has told me, but what
I have tried by experiment. I think that is the only thing that has
any value.

When we were recording last Saturday night, for example, we sent
this music out. We asked the public to send us criticisms. That is
what we want. We want them to tell us what is wrong about the
broadcast, because we honestly want to make our broadcasting better
and better all the time. Those criticisms came in, hundreds of letters
and telegrams, telling exactly what those people felt was not good
about this thing.

That is one method we have of checking. Another one is that
during the performance someone is recording the concert in the
concert hall where we are playing, and of course the connection
between the microphone and the recording instrument is close and
can be well taken care of, so that it is in good order and we have good
reproduction.

Then the selection is sent over the air, and in the laboratory in
New York someone is again recording, over the ether.

So that we are using those three methods of checking our perform-
ance and comparing them one with the other. First is the criticism
from the public— what we want is the reaction of the average man in

168 LEOPOLD STOKOWSKI [J. S. M. P. E.

the average living room who is listening to our broadcast; we want to
know how it impresses him, and we are receiving that information
through the letters. Then we have the recording in the academy, and
the recording over the ether in New York. By comparing those three
things we get a fairly clear idea of what is wrong and how perhaps
we can improve.

People often say when they listen to music, especially modern
music, "That isn't music." For example we recently produced an
opera called Wozzek and one of the music critics wrote this in his
newspaper (as I say, this was a very modern work, different from other
works, extremely original): "This department is organized to criticize
music. Wozzek is not music. Therefore we shall say nothing."

But what is music? What are the limitations of music? There
are people who think that the Last Rose of Summer is the summit, the
highest peak of music. Well, it is a very beautiful melody. I enjoy
it very much when it is well sung or well played. But there are other
kinds of music, too. A little bird singing in the forest is producing
very marvelous music, and a different kind.

What are the limitations of music in sound? Personally, I think
one sees, as music progresses and has wider and wider horizons, that
its limitations are becoming less. We are seeing it in a bigger and
bigger way all the time. And ultimately it may be that we will think
that all sound is music. All sound has something to which we can
respond.

The sound that comes from that little machine* down there I should
call music because it has a definite frequency. It has definite dura-
tion, and it has a very interesting rhythm if you will listen to it.
The narrow-minded musician would say, "No, that is merely a noise."

But, I think, for the sake of the motion picture with sound, with tone,
which is going to be an ever and ever more important type of art, that
we have to think about what is sound, and what is music, and what
are the limitations of music; and we have to take in more and more of
sound, the sounds of nature, like the wind going through the trees
The sound of the sea has a most interesting rhythm if you will take the
trouble to listen to it. It has very deep, strange sounds, which are
quite extraordinary. The sounds of the birds are marvelously
beautiful as Wagner has shown in Siegfried.

There are all kinds of sounds in nature which are interesting and

* The stenotype machine.

Feb., 1932] SOUND RECORDING 169

which we wish to reproduce in tone films. The machine has very
interesting rhythms if you listen to it with an open mind, not with the
nineteenth century narrow-minded view, but with the view of today,
which takes in more of life. All this will come, in my opinion, in the
tone film, and then music will not be a narrow thing. It will extend
itself until it takes in all sounds.

I said a little while ago that we must standardize certain things. I
think a great battle is coming in the world between standardization
and non-standardization, individuality. It is coming in all planes.
It certainly is coming in the field of economics. I believe it is coming
in the things in which we are interested, in sound, in science, in
photography, in light, and I am watching it with great interest.

For example, my orchestra, I notice, plays differently every day.
We play the same music; we rehearse it, say, for five days in succes-
sion. That is what we do every week. We rehearse every morning —
Monday, Tuesday, Wednesday, Thursday, and Friday — the same
music. Friday afternoon we play it in the concert. But every day
the men play differently. Each day one can draw from them a
different quality of tone and a different volume of tone.

This is going to be the great question: whether we shall standardize
that or whether we shall allow it to be free and individual. Certainly,
when we record it we standardize it. We must. We fix that day's
impression on the disk, and send it out to the receiving apparatus as a
standardized thing. But when we play it in the concert it is un-
standardized. It is different every day. Emotionally, it is also
different. That is something for us to think about, and I believe it
will be years before we get any results on that.

I believe that this tremendous development that has been going
on hi sound in the last six or seven years through the radio, the
gramophone, etc., will lead to something that is very desirable.

At the present time, when a composer hears in his inner being
some music, he desires to make it permanent, that impression that is
going on within him, so he takes a paper and pencil, and writes down
marks on the paper to preserve that melody, those harmonies, those
rhythms. Then the singer or the player comes and reproduces those
sounds. The composer listens and he says, "That is not at all what I
intended."

We have that all the time. He composes something for the
orchestra. We play the notes that he has written and he says,
"That is totally different from what I intended."

170 LEOPOLD STOKOWSKI [j. s. M. P. E.

Why is that ? It is because the method of writing sounds on paper
is tremendously imperfect.

If a painter wishes to paint a picture he takes his canvas and his
colors; he puts his colors on the canvas where he wishes them. He
makes his design, his relativity of color to color or form to form,
and when he finishes it and he is satisfied; that ends the matter.
It is complete.

But when Beethoven or any composer composes a symphony and
writes it on paper, he has only half completed the process. It must
then be given to the orchestra. They play it and he is dissatisfied,
because it doesn't reproduce his idea, because our method of notation
is so imperfect

I see in all this development something new coming. I believe it
will be only a few years before the composer will paint directly in
tone. He won't write down his impressions on paper. He will
express them through frequency, through volume, and through
duration. In that way he will express his ideas exactly, and not with
the imperfections we now have. That is almost possible today, and
through electrical production of tone, such as we get through the
Theremin instruments and others which are being developed now,
that will soon be possible and will be a very desirable thing.

What is the ideal for us who are scientists, or engineers, or musicians
or photographers, or producers of tone films? What can we do in
the future which is greater than what we are doing now? A great
deal, in my opinion.

We may communicate with someone by telephone. We can talk to
someone over the telephone. We can communicate ideas. We can
come to understandings about ideas. We can talk for a long time on
the most intricate, complex subjects, and make decisions and have a
discussion. But when we combine sight and sound, through the
tone picture, we can communicate much more, not only ideas but
emotions and suggestions of things which are not completely said but
which are conveyed in a more subtle way. We can suggest on levels of
consciousness higher than thought, and feeling, and imagination, and
all those strange things that go on in our nervous system which make
our inner life so complex and so rich. Above all, these things for which
we have words we all know perfectly well there are other things. We
have no names for them, no words for them, but they exist. They
are part of our daily experience. Especially do we feel those things
through the finest type of music. Music of the higher type expresses

Feb., 1932] SOUND RECORDING 171

just those things. And it is through the tone film that we can very
richly and completely express that, and it is through radio, and
eventually television that we can project those things through space
all over the world.

That is the magnificent ideal, something quite supreme, toward
which we must all work. We must not be satisfied to stand where
we are at present, which is about a half-way point toward that thing.

The development of the radio, the gramophone, of photography
and reproduction of sound has been perfectly miraculous during the
last six, eight, or ten years, but there is far more yet to be done.
Let's admit that frankly, and let's work for that immense ideal which
is possible.

ON THE ASSIGNMENT OF PRINTING EXPOSURE BY
MEASUREMENT OF NEGATIVE CHARACTERISTICS*

CLIFTON TUTTLE**

Summary. — The theory of photographic tone reproduction, though specific for
ideal cases, cannot always be applied in the determination of printing exposure for
motion picture negatives. A statistical study of the correlation of various optical
characteristics — maximum transmission, minimum transmission, and total frame
transmission — with the required exposure has been made. Of the possible measure-
ments to be made, the value of total frame transmission seems to be the best criterion
of printing exposure. The apparatus used in making the measurements is described
and the data obtained are presented graphically.

PRESENT PRACTICE IN PRINTING EXPOSURE ASSIGNMENT

In motion picture finishing laboratories, one of the problems
which must be considered is the assignment of printing exposure
to each scene of negative of which a print is desired. The usual
type of printer operates at constant speed, thus fixing the time of
exposure. Compensation for differences in negative density is
made by varying the intensity of the light incident upon the negative.
In practice, a series of intensity steps is provided either by control
of resistance in series with the printing lamp or by the setting of an
opening in an optical diaphragm. Before a negative is printed its
correct printing intensity must be selected and the light source
must be regulated to give this intensity.

Methods for selecting the best printing exposure vary somewhat
in different laboratories. In some instances, a tablet sensitometer
is used as described by Jones and Crabtree.1 In this method, a
print of the negative scene is made through a density step tablet.
The steps, each the size of a single frame, have been calibrated to
correspond with those of the printer light-change board. The
resultant positive after processing is inspected visually and a se-

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Communica-
tion No. 470 from the Kodak Research Laboratories.

** Research Laboratories, Eastman Kodak Co., Rochester, N. Y.
172

ASSIGNMENT OF PRINTING EXPOSURE 173

lection of the best exposure is made by an expert judge of print
quality.

More frequently the assignment of printing exposure is made
directly from the negative. A negative "timer," who by virtue of
long experience and particular aptitude has become adept at judging
negatives, is able to assign the proper printing exposure to a series
of negative scenes merely by visual inspection. In the timing of
negatives used for release prints, the initial results obtained by the
timer are, of course, subject to correction after the projection of a
trial print.

The author has never had the opportunity to gather any data
concerning the waste of time and material occasioned by errors
in the initial timing of a negative. It is probable, however, that
such waste amounts to a negligible per cent of the total processing
cost on release pictures, and it is probably true that the present
methods of exposure assignment are entirely satisfactory where a
large number of prints are made from a single negative.

If a single print is to be made from a negative, if speed is desired
in the production of a first print, or if the services of an expert judge
of photographic quality are not available, assignment of printing
exposure on the basis of a measurement of the optical character-
istics of the negative may be desirable. These practical considera-
tions and the obvious interest of the question in the theory of tone
reproduction have suggested the value of a study of the relation of
the optical characteristics of motion picture negatives and their
required printing exposures.

NEGATIVE CHARACTERISTICS AND PRINTING EXPOSURE

The rigid theory of tone reproduction is specific on the subject
of required printing exposure. To reproduce with perfect accuracy
the brightness relationships existing in the object by an equivalent
series of tone relationships in the picture requires, first, a negative
in which the total range is included on the straight line portion of the
H & D characteristic curve. Given such a negative, a perfect
print must translate the negative density range into positive density
values lying in inverse order on the straight line portion of the
positive characteristic curve. For the thinnest perfect positive,
which for efficient projection would seem to be the thing desired,
the printing intensity, according to the tone reproduction theory,
is given by the following:

174 CLIFTON TUTTLE [j. s. M. P. E.

lOg / = log EP mia. - log t -h DN max.

where / = intensity incident upon the negative.

/ = time of exposure.
DN nu«. = maximum negative density.

EP mm. = minimum exposure for positive — the exposure for the lowest
density on the straight line portion of the characteristic.

In practice, many negatives, probably most of them, are not
perfect in the sense just described. The printing operation also
is usually a compromise, and throughout the literature we find
numerous suggestions as to practical criteria for printing exposure.

Hurter and Driffield, the pioneers of quantitative photography,
make the following statement in one of their early papers:2 "We
first of all measure the highest density of the negative . . . and
knowing the inertia of the plate (positive) we take care that the
exposure shall be such that behind the highest density of the negative
the plate shall receive an exposure at least equal to the inertia."
Since the inertia is defined as the intersection of the extended straight
line portion of the characteristic curve with the log R axis, it is evi-
dent that an exposure equal to the inertia will not give a density
lying on the straight line portion of the positive characteristic.

Driffield, in a later paper,3 modified this criterion. He suggested
that the printing exposure be computed as the antilog of the average
of maximum and minimum negative densities multiplied by the
geometrical mean of the exposure range of the positive. This pro-
cedure bases the printing exposure on the transmission of the middle
tones of the negative.

According to this criterion, a negative with a range of density
greater than can be accommodated by the linear portion of the
positive characteristic would give a print in which shadow and high-
light would overlap shoulder and toe of the positive characteristic
to the same extent. Since that time a number of others have recom-
mended the use of a similar criterion of printing exposure.

One authority quoted by Renwick4 considers the total range of
the print and not the relationship between tones to be the important
thing, which is equivalent to saying that printing exposure is in no
way critical so long as a given maximum contrast is obtained.

F. C. Tilney,5 speaking for the artist, remarks, "There seems to
be one thing only in matters of tone that is absolute, and that is
the correct relation of one tone to another in the same picture what-
ever the key adopted." Translated into photographic parlance
this statement may be taken to mean that the straight line portion

Feb., 1932]

ASSIGNMENT OF PRINTING EXPOSURE

175

of the characteristic curves only should be used and that the locating
of negative density values with respect to the log E axis of the
positive is unimportant so long as this condition is fulfilled.

If we may be permitted to apply the practical photographer's
axiom for negative making — "Expose for the shadows and let the
highlights take care of themselves" — to the making of the positive,
we should do well to base our judgment of exposure on the trans-
mission of the thinnest (or shadow) portion of the negative.

L. A. Jones'6 discussion of tone reproduction, which is based upon
a knowledge of the limitations of photographic materials, fixes the
printing exposure at the value which will give a "just perceptible"
density (0.008) for the highlight portion of the negative. This
procedure insures gradation throughout the highlights of the picture.

FIG. 1. A projection densitometer for the measurement of motion picture
image characteristics.

The foregoing comments indicate that there are some differences
of opinion regarding the assignment of "best printing exposure"
from a consideration of negative characteristics. It is hoped that
this fact will supply an excuse for the statistical treatment of a
problem which does not appear amenable to purely theoretical
solution.

MEASUREMENT OF OPTICAL CHARACTERISTICS OF A NEGATIVE

Apparatus. — Three characteristics of a negative — minimum trans-
mission, maximum transmission, and total transmission — are readily
measurable. Any one of these, or a combination of two, might be
expected to give some correlation with the required printing exposure.
To facilitate the measurement of these three values for a large
number of motion picture negatives the instrument shown dia-
grammatically in Fig. 1 was constructed.

176 CLIFTON TUTTLE [j. s. M. P. E.

Referring to this figure, a monoplane filament lamp, A, ma. suitable
housing is imaged by lens, B, in the plane, C. In the plane, C, a
sliding carrier containing a lens, D, and thermopile, E, may be so
positioned that either the lens or the thermopile receives the filament
image.

Over the condenser lens, B, is placed a rectangular mask with an
opening the size of a single motion picture frame. The aperture
is supplied with a spring gate so that a motion picture film may be
readily inserted and framed in the opening.

When the lens, D, is in position an image of the motion picture
frame may be formed either at F or at F', depending upon the position
of the totally reflecting prism, G. This prism may be rotated about
a vertical axis through its hypotenuse face to either of the two
positions shown in Fig. 1. The plane at F' is provided either with
a ground glass screen for viewing the image or with a sheet of bromide
paper for making a permanent record of it. A stylus back of the
plane F' may move either in the vertical or horizontal direction so
that it is possible to indicate any area of the image. The movement
of the stylus in the plane F' is mechanically linked to the movement
of the Moll thermopile in the plane F. Thus, if the stylus is posi-
tioned at an area of the image corresponding, say, to the most dense
portion of the negative highlight, the thermopile is brought auto-
matically to an identical position with respect to the image which is
projected on F by the rotation of the prism.

The thermopile is connected to a Leeds and Northrup high sensi-
tivity galvanometer (17 mm./MV.) which is provided with an Ayrton
shunt. Two readings are required in the making of a transmission
measurement: A value for zero density, and a second value of the
amount of light which has passed through the area of the negative
to be measured. For the first value, the motion picture frame is
removed from the beam by sliding the whole aperture plate and gate
assembly horizontally in a pair of gibs. It is possible with this
instrument to read transmission values as low as 0.1 per cent with
an error less than 5 per cent, while higher transmission can be read
to a much higher degree of accuracy.

Procedure Followed to Obtain Data. — Through the courtesy of a
number of studios on both the west and east coasts, about 1000
clippings from release picture negatives were obtained. A wide
variation in subject-matter and composition was represented by
these samples. From each of the negatives a sensitometric tablet1

Feb., 1932]

ASSIGNMENT OF PRINTING EXPOSURE

177

print was made to be used subsequently in the determination of
required printing exposure.

A single frame of each scene sample was registered in the gate of
the motion picture densitometer. With the thermopile, E, centered
with respect to the frame, a measurement was made of the per-
centage of light transmitted by the whole frame of the negative.
It should be noted at this point that the measured value of trans-

&'5h

(S>

LU
Z

o
d 10 -

fes^

K
Z

0 10 80 30 40 50 60 70 6O 90

DIFFUSE TRANSMISSION OF THE WHOLE FRAME IN PER CENT

FIG. 2. Statistical summary of distribution of inspected studio negatives
according to the whole frame transmission. The areas represent the com-
parative numbers of negatives to be found within each region of transmission.

mission so obtained is a specular value, and therefore is not identical
with the diffuse transmission value. This matter must be con-
sidered in an application of the data to the contact printing problem.
In order to make measurements of the transmission of the densest
and thinnest portions of each frame, the lens, A was used to project
an image of the negative magnified 10 times. With the prism, G,
positioned to throw this image on the ground glass at F't the areas

178 CLIFTON TUTTLE [J. s. M. P. E.

selected for measurement were designated by the indicator stylus,
the movement of which automatically positioned the thermopile
to receive the identical area when the prism was rotated through
90 degrees. The blackened receiver of the thermopile covered a
circular area 1.0 centimeter in diameter which corresponded to a
circle on the negative film of 1.0 millimeter diameter.

In the measurement of the total transmission of the large number
of negatives, it soon became apparent that the great majority of
professionally photographed and processed scenes occupied a rela-
tively small portion of the possible transmission range for printable
negatives. The data presented graphically in Fig. 2 is of interest
in that it indicates the remarkable uniformity of the product of a
number of studios so far as average density is concerned.

In the plotting of Fig. 2, the specular transmissions obtained
directly from the galvanometer readings have been transformed to
diffuse transmissions* to make the results more, directly applicable
to the contact printing problem. The figure shows the distribution
of the per cent of the total number of scenes measured among various
regions of transmission. It is seen from this figure, for instance,
that 50 per cent of professional negatives have a total transmission
of from 20 to 30 per cent and that about 95 per cent have a trans-
mission between 10 and 60 per cent — a range of but 6 to 1.

An expert judge of print quality working from the sensitometric
tablet prints assigned the printing exposure for the 1000 negatives.
An analysis of his results showed that the required printing intensity
range for 95 per cent of the negatives also covered a range of only
6 to 1.

Because of this great preponderance of the available samples
within these narrow limits of printing exposure and transmission,
it was decided to select a limited number of negatives distributed

* The author has shown in a previous paper7 that the relation between specular
and diffuse density is of the form D \\ (specular density) = KD-\\- M (diffuse
density). For motion picture negative film K — 1.37 and M = 1.088, approxi-
mately. In computing the value of diffuse density of picture negatives from
the specular density in this manner there are three possible sources of error:
(1) The constants given apply to a truly specular optical system; (2) the values
of the constants vary somewhat for different emulsions; (3) in substituting a
value of D \\ in the exponential relation one must assume that the density is
uniform which is, of course, not the case with a motion picture image. It is
believed that none of these errors is of any great importance for the type of data
to be presented.

Feb., 1932]

ASSIGNMENT OF PRINTING EXPOSURE

179

more uniformly throughout a greater printing range rather than to
encumber the graphical presentation with data for the entire group.
The table summarizes the data for the selected series.

TABLE I

Table Showing Characteristics of Group of Motion Picture Negatives

Scene
Number

Total
Specular
Trans-
mission

Total
Diffuse
Trans-
mission

Maximum
Specular
Trans-
mission

Maximum
Diffuse
Trans-
mission

Minimum
Specular
Trans-
mission

Minimum
Diffuse
Trans-
mission

Ratio
Maximum
to
Minimum
Trans-
mission

Printing
Exposure
M.C.S.

57.0

69.0

80.0

17.0

28.0

2.9

1.03

56.0

68.0

69.0

80.0

12.0

22.0

3.6

1.03

41.0

55.0

62.0

74.0

7.8

15.0

4.9

1.67

34.0

47.0

62.0

74.0

5.9

12.0

6.2

1.67

30.0

44.0

48.0

62.0

6.9

14.0

4.5

2.28

29.0

42.0

57.0

71.0

6.2

12.0

5.9

1.67

25.0

38.0

49.0

62.0

3.8

8.3

7.5

2.28

20.0

32.0

46.0

60.0

15.0

25.0

2.4

1.03

17.0

28.0

40.0

54.0

5.1

11.0

4.9

2.65

12.0

21.0

26.0

39.0

4.8

10.0

3.9

3.64

11.0

20.0

18.0

29.0

4.8

10.0

2.9

2.65

11.0

20.0

19.0

31.0

1.0

2.8

11.0

3.64

8.0

15.0

17.0

28.0

1.1

3.1

9.0

5.05

7.8

15.0

26.0

39.0

0.9

2.5

15.6

3.64

6.9

14.0

13.0

22.0

1.6

4.2

5.2

5.05

6.5

13.0

11.0

20.0

0.8

2.7

7.4

5.05

6.5

13.0

15.0

25.0

0.5

1.6

15.6

5.05

6.5

13.0

12.0

22.0

0.3

1.1

20.0

5.05

4.6

9.8

. . .

9.55

3.2

7.2

5.5

11.0

0.2

0.9

12.2

13.60

3.0

6.7

2.4

5.7

0.4

1.2

4.7

7.70

2.5

6.0

2.7

6.5

0.17

0.8

8.1

13.60

2.4

5.9

2.1

5.1

0.4

1.2

4.2

13.60

2.4

5.8

3.7

7.8

0.3

1.1

7.1

13.60

25
26

1.8
1.6

4.6
4.1

1.6

4.2

0.16

0.8

5.2

25.00
13.60

1.5

4.0

1.9

4.8

0.4

1.2

4.0

31.60

1.5

4.0

1.1

3.1

0.07

0.3

10.0

25.00

1.3

3.4

0.7

2.0

0.1

0.4

8.0

19.00

This table is probably self-explanatory with the following brief
enumeration of the methods of obtaining each column of figures.
Column 2 gives the ratios of galvanometer deflections with and
without each negative scene in place. In this case a lens in the
plane of the frame forms an image of the densitometer lamp on the
thermopile and the reading is, therefore, a specular measure of the
transmission of each whole frame. In column 3, the values of

180

CLIFTON TUTTLE

[J. S. M. p. E.

specular transmissions have been converted to diffuse transmissions
as explained in the preceding footnote. Column 4 gives the specular
transmission of the least dense area in each scene. Column 5 shows
these same values converted to diffuse transmissions. Columns
6 and 7 give similar data for the specular and diffuse transmissions
of the densest negative areas. Column 8 lists the ratios of the values

1.5

1.2

'-1.7

.3 .6 .9 \.Z

LOG PRINTING EXPOSURE (M.C.5.)

1,5

FIG. 3. Relation between login of required printing
exposure and logio of minimum negative transmission,
diffusely measured.

of column 5 to those of column 7. These ratio values are of interest
in a consideration of the exposure scale of the positive material
which is to be used. The final column is the result of the expert's
judgment concerning the printer step required to print each negative
scene, the positive being developed to a gamma of about 1.6. These
data are given as the exposure in meter candle seconds which would

Feb., 1932]

ASSIGNMENT OF PRINTING EXPOSURE

181

be required to print each scene. A calibration of the printer to
which the "required step" data applied was made by methods of
photographic photometry, the procedure for which has been pre-
viously described by the author.8 The intensity factor of the
exposure was measured by its photographic effect on positive film
compared to the effect of a source operated at 5000 degrees K.

.o .3 .6 .9 \.Z

LOG PRINTING EXPOSURE (M,C. 5.)

FIG. 4. Relation between login of required printing ex-
posure and the mean of the logsu, of maximum and minimum
negative transmissions.

In Figs. 3, 4, 5, and 6, logic of required printing exposure as se-
lected by the expert is shown as the abscissa axes. Figs. 4, 5, and
6 test the various printing exposure criteria which have been enum-
erated in the early part of this paper. In Fig. 3, logio of minimum
negative transmission is plotted. This tests the exposure criterion
suggested by Hurter and Driffield2 and by Jones.6 If some definite

182

CLIFTON TUTTLE

[J. S. M. P. E.

highlight density is to be produced in the print to make the best
positive we must assign any departure of the points from a straight
line of unit slope to the uncertainty of the judgment of the expert.

The criterion suggested by Driffield3 is tested in Fig. 4. If we
suppose that an average of highlight and shadow negative densities
is to be rendered by a definite positive density this data should define
a straight line of unit slope.

a 1,5
g

0 .3 ,<o .9 1-2 1-5

LOG PRINTING EXPOSURE (M.C.5.)

FIG. 5. Relation between logic of required printing ex-
posure and logio of maximum negative transmission.

Fig. 5 shows logio of maximum transmission, suggested by the
photographer's rule of exposure and Fig. 6 indicates the correlation
to be expected from a measurement of total frame transmission.

A value of total transmission can, of course, be determined much
more readily than can the value either of minimum or maximum
transmission. From the point of view of convenience and speed

Feb., 1932]

ASSIGNMENT OF PRINTING EXPOSURE

183

it would be the best measurement to make for the assignment of
printing exposure. It seems possible that some total transmission
measurement other than for the whole frame might give even better
results and still have the advantages of speed in making the deter-
mination. It is possible, for instance, that the accuracy might be
increased if only the foreground of the picture were measured, thus

1.6

1.5

S-:

,3 .€> .9 , 1.2

LOG PRINTING EXPOSURE (M.C.5.)

1.5

FIG. 6. Relation between logio of required printing ex-
posure, and logio of total negative transmission.

leaving the area usually occupied by sky in exteriors and ceiling
in interiors out of consideration. This alternative was tried but the
results obtained were disappointing. The correlation of this measure-
ment with required printing exposure was not nearly as high as was
the whole frame transmission value.

A second alternative was tried using a circular mask of 0.75
inch diameter centrally located with respect to the film frame.

184

CLIFTON TUTTLE

[J. S. M. p. E.

This idea was followed out on the supposition that the center of
interest, and therefore the area most desirable to measure, usually
occupies an area toward the frame center. In the case of these
measurements, the correlation was somewhat better than that shown
by any of the other measurements. It should be pointed out,

u
or

|.«

a
a

1.2

LOG SELECTED PRINTING EXPOSURE

FIG. 7. Relation between logsio of printing exposure as selected by two expert
judges of print quality.

however, that, while statistically the central area measurement may
give the best correlation, the occasional errors due to grouping of
the subject interest at the edges of the frame may be of greater
magnitude than would ever occur in case the whole frame were
measured.

Feb., 1932] ASSIGNMENT OF PRINTING- EXPOSURE 185

Certainty of Exposure Assignment by Visual Judgment. — In con-
sidering the data of Figs. 3 to 6, it should not be assumed that the
value of required printing exposure chosen by the expert for each
negative scene is the "correct" value. Undoubtedly, at least in
the case of some scenes, this value may vary somewhat and still
result in passable prints. It has already been suggested that the
judgment of printing exposure is regarded as somewhat of an art by
the motion picture profession. If this is the case, it may be that
such factors as personal taste of the observer, and conditions of the
observation play some part in the selection.

While it is difficult to arrive at any decision on this question of
how accurate is the work of the negative timer, the following set
of data may throw some light on this matter. In our own processing
laboratory two individuals have had considerable experience in
assigning printer exposure from inspection of sensitometric tablet
prints. These two persons work interchangeably and it is generally
agreed that both are expert judges of print quality.

After the first timer had completed his work with the tablet prints
the second was given the same set of prints and asked to assign
the printing exposures. The diversity of opinion which is indicated
in Fig. 7 is rather surprising. The two axes of the graph are used
for the log of printing exposure assigned to the series of negatives
by the two experts. If the agreement in all cases had been perfect
all of the points would lie on a straight line of unit slope.

The figure represents data for 180 scenes. The numbers in the
circles show the number of scenes which determine the location of
each point. For only 64 scenes is the agreement of the two observers
perfect. The remainder of the observations is distributed through-
out an area which is enveloped by the two dotted lines. To include
all the scattered points these lines are drawn at positions ±0.3 in
logio E removed from the mean straight line. This means that, at
least in the case of some of the scenes estimated, there is a printing
exposure tolerance equal to a factor of two or one-half. The sta-
tistical method does not reveal the presence of some scenes in which
conceivably there is very little tolerance in printing exposure.

DISCUSSION OF RESULTS

Accuracy Demanded in the Selection of Printing Exposure. — The
wide tolerance in the choice of printing exposure which is suggested
by the data of Fig. 7 is surprising in view of the known facts con-

186 CLIFTON TUTTLE [j. s. M. P. E.

cerning existing practice in the commercial laboratories. Consider
for a moment the usual light change scale of the production labora-
tory printer. Few of them are calibrated to accommodate an
intensity range of more than ten or twelve to one. This range is
split up into twenty-odd steps and the average magnitude of a
step is between 10 and 15 per cent. A printing exposure tolerance,
such as that indicated in Fig. 7, would correspond to plus or minus
perhaps half a dozen such steps. The experts in the laboratories
presumably work to a tolerance of plus or minus one printer step.
The author is in no position to express an opinion concerning the
desirable accuracy of printing exposure assignment but merely
wishes to present the following facts which may have some bearing
on the question.

In column 8 of the table are given the transmission ratios, maxi-
mum to minimum, of the studio negatives which were examined.
These ratios vary from 2.9 to 20.0. It is probable that 20 is an
extreme case. Special precautions9 must be observed to obtain
a lens image brightness ratio for highlight to shadow of more than
25 to 1.0. With negative developed to a gamma of 0.5 or 0.6, the
transmission ratio will seldom exceed 15.0. Since the average
positive material at a gamma of 2.0 has an exposure scale of ap-
proximately 60 to 1.0 there would appear to be a latitude in printing
exposure of two or one-half from the mean value without making
use of the toe or shoulder of the positive characteristic. In other
words, positives which would render the negative tones perfectly
could be made from most negatives throughout a four to one range
of printing intensity. Such positives would differ from each other
only in average transmission.

The amount of light reflected to the audience from the screen is
known to differ in various theaters. The public at present sees
motion pictures under so many different conditions in different
theaters, that it seems quite possible that within wide limits the
average transmission of the positive is a matter of small consequence.

Choosing the Negative Characteristic to Measure. — Whether or not
measured values can be as satisfactory as expert judgment in as-
signing printing exposure, it is conceivable that there may be appli-
cations in the processing laboratory for a quick approximation of
exposure such as would be afforded by a densitometric method.
With this end in view we can consider the relative merits of the
criteria tested in Figs. 3 to 6.

Feb., 1932] ASSIGNMENT OF PRINTING EXPOSURE 187

The relations between printing exposure, E, in meter candle
seconds and negative transmissions in per cent obtained from Figs.
3, 4, 5, and 6 follow:
12.9

(1) E = 7,0.79 m which Tmin is the diffuse transmission of the

•*• min.

negative highlights.
33

(2) E = ,w).89 in which raverage is the geometrical average

•* average

of the diffuse highlight and shadow transmission.
107

(3) E — — — in which jTmax< is the diffuse transmission of the

•*• max.

negative shadow.
67

(4) E = ~ — in which rtotal is the diffusely measured total

* total

transmission.

The measurement of minimum negative transmission has little
to recommend it. The possible error which would follow its use
(roughly indicated by the distance separating the dotted lines) is
greater than that for the other suggested values. The dotted lines
which designate the area required to include all points are 0.8 in
log E apart. This separation corresponds to an exposure factor
of 6.3 which means that a departure of 3.1 times hi exposure from
the value picked by the expert might be made. The relation in-
volved is exponential, which means that a linear calibration curve
between opacity (1/7") and required exposure could not be used.
In addition to these objections, there is the fact that the lower the
transmission value the more difficult is the measurement to make
and the greater is the probability for error in the measurement.

The other suggested criteria appear to be almost equal in that
the maximum departure from the visually selected printing exposure
would be by a factor of about two or one-half. The geometrical
mean of maximum and minimum transmission, which gives slightly
better correlation than the other criteria give, is probably ruled out
as a practical measure of required printing exposure because two
selected areas would have to be measured and a computation made
before this value could be applied.

There is no question but that the value of total frame transmission
is the most readily applicable to the speedy determination of printing
intensity. In many instances, no doubt, exposure assignment on
the basis of a total transmission measurement would be considerably

188 CLIFTON TUTTLE

in error. In any number of conceivable cases where the object of
principal interest occupies a relatively small portion of the frame
against a background of a markedly different transmission, the total
transmission will not give an indication of the best printing exposure.
In scenes where special effects are to be obtained by over- or under-
printing no generally applicable method of printing intensity evalua-
tion by measurement is conceivable.

The value of a measuring method used either alone or to supple-
ment the judgment of an expert is a matter which must be decided
upon evidence gathered under practical conditions of operation.

It is certain that there is waste of some time and material with
the present methods of printing exposure assignment in the making
of first prints. Only extensive trials of the possibilities of exposure
assignment by measurement can decide as to its relative merits.

REFERENCES

1 JONES, L. A., AND CRABTREE, J. I.: "A New Sensitometer for the Deter-
mination of Exposure in Positive Printing," Trans. Soc. Mot. Pict. Eng. (1922),
No. 15, p. 89.

2 HURTER, F. H., AND DRiFFiELD, V. C. : "Relation between Negatives and
Their Positives," J. Soc. Chem. Ind., 10 (Feb. 28, 1891), p. 98.

3 DRIFFIELD, V. C.: "The Principles Involved in the Calculation of Ex-
posures for Contact Prints on Bromide Paper," Brit. J. Phot., 40 (1893), p. 606.

4 RENWICK, F. F. : "Tone Reproduction and Its Limitations," Phot. J., 56
(n. s. 40) (1916), p. 222.

5 TILNEY, F. C.: "The Appeal of the Picture," p. 49.

6 JONES, L. A. : "On the Theory of Tone Reproduction with a Graphic Method
for the Solution of Problems," /. Franklin Inst., 190 (1920), p. 39.

7 TUTTLE, CLIFTON: "The Relation between Diffuse and Specular Density,"
J. Opt. Soc. Amer., 12 (1926), p. 559.

8 TUTTLE, CLIFTON: "Illumination in Motion Picture Printing," Trans. Soc.
Mot. Pict. Eng., 12 (1928), No. 36, p. 1040.

9 TUTTLE, CLIFTON, AND WHITE, H. E.: "Factors Which Affect the Contrast
of a Lens Image in the Motion Picture Camera," Trans. Soc. Mot. Pict. Eng.,
11 (1927), No. 31, p. 591.

UTILIZATION OF DESIRABLE SEATING AREAS IN RE-
LATION TO SCREEN SHAPES AND SIZES AND
THEATER FLOOR INCLINATIONS *

BEN SCHLANGER **

Summary. — The aim of this paper is to establish a relation between the bodily
posture of the viewer, the size and shape of the picture, and the architectural form of
the theater in all its details. The present type of theater floor is compared with the
reversed type described in a previous paper in order to show how the latter type of floor
permits placing a greater number of seats within the desirable seating areas than the
present type. An analysis is made also of the effect of reversing the floor on the
ability of the viewer to assume a comfortable bodily posture. Definite angles of sight
specified by the various tilts of chair backs found necessary for comfortable posture
are shown. Several forms of theaters of various seating capacities and screen sizes
are described in order to show the broad application of the theories involved in reversing
the pitch of the orchestra floor.

The principle of reversing the slope of the orchestra floor in theater
structures, as presented in a previous paper, suggested the possibility
of correcting many of the faults of present-day theaters. Bodily
posture in seating, vision, projection angles, accessibility of various
levels, and construction costs are all affected. Further study of this
new principle in planning theaters has resulted in the development of
definite relations between the various functions that contribute to the
practicability of the whole. Study has also brought out the fact that
this new principle is not only applicable for improving the present
form of the theater, but also for deriving from it many new forms more
adaptable to motion picture exhibition. (Fig. 1.)

A complete analysis of bodily posture has been made in connection
with this new principle. Certain maximum and minimum pitches
of chair backs and floor slopes have been arrived at, and measure-
ments have been made of the vertical range of vision which can be
obtained while sitting against differently pitched chair backs.

Practical projectionists have verified the need of lessening the angle
of projection. This need has been recognized, and has been answered

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Architect, New York, N. Y.

189

190

BEN SCHLANGER

[J. S. M. P E.

in these studies by establishing a maximum angle of ten degrees to
the center of the screen from the lens center. In most cases the angle
will be less, varying from ten degrees to a perfectly horizontal line
of projection. In existing theaters the projection angle is often as
great as thirty degrees or more. Regardless of the size or seating
capacity of a theater, the reversed floor principle of planning requires
no angle of projection greater than ten degrees.

It has been found that the enlarged screen can be more easily ac-
commodated in a theater structure if the reversed floor principle of
planning is applied; and that it would be impossible to install an
enlarged screen in the present type of theater without incurring a

•LONGITUDINAL- SECTION-

• MOTION- PICTURE THEATRE -

FIG 1. Longitudinal section of the present type of motion picture theater as
affected by the use of reversed floor.

great waste of structure area and inefficiency in seating arrangement,
resulting from the failure to utilize the areas most valuable for com-
fortable vision. The difficulty of using an enlarged screen in present
theaters has already evidenced itself, and is partly delaying its popular
adoption. Balcony obstructions, and the difficulty of obtaining a
complete and comfortable view of the higher screen are serious im-
pediments, which may be overcome by the use of the reversed floor.

The practicability of applying the reversed floor principle to vari-
ously sized and proportioned plots of ground has been given special
attention, the object always being to obtain a maximum number of
"good" seats within a minimum area. Many different adaptations of
the reversed floor principle have been devised to fit the peculiar con-

Feb., 1932]

UTILIZATION OF DESIRABLE SEATING AREAS

191

ditions of various theater projects. The feasibility presents itself of
placing a part or the whole of a theater auditorium above or below
the portions of a structure which may be used for other purposes. It
therefore becomes important to design a theater auditorium so that
it will not require too much valuable area in the vertical sense. Thus,
the remaining portions of a structure above or below the theater may
provide an additional income which, in turn, results in a reduced
rental for the theater itself. The use of the reversed floor permits
constructing a theater within a limited height, where it would be im-
possible to include a theater planned according to present practices.
A revised building code for the city of New York, affecting theater
structures, is about to be put into effect. The committee revising
this code has taken into consideration the possibilities of the re-
versed floor by providing for its development in the wording of the

FIG. 2. Chart for determining, in the vertical sense, the desirable seating areas.

code. The revised code will permit the construction of a theater
auditorium, having a capacity of more than 600 seats, directly be-
neath the portions of a building used for other purposes. This is
already permitted in many other building codes.

For the purpose of making possible better vision, smaller projec-
tion angles, and reducing the cubage of the theater structure, a very
intensive search has been made to ascertain which of the physical
areas are most valuable as seat locations in relation to the screen,
the object being to use these areas only for seating arrangements.
The present system of theater planning necessitates the utilization
of portions beyond these valuable areas, thus causing large pro-
jection angles, distorted vision from high balconies, unnecessarily
large construction costs, and the payment of excessive rentals for
space which not only has no value to the exhibitor, but which also
creates conditions highly unsuitable for motion picture exhibition.

192 BEN SCHLANGER [j. s. M. P. E.

For these reasons the present method of theater planning results
in unscientific and uneconomically built structures. If original
construction and maintenance costs can be reduced, at the same
time giving to the theater patron the comforts and surroundings due
him, there is little doubt as to what the effect on the box-office will
be.

A chart showing the location of the desirable areas has been de-
veloped. (Fig. 3.) These areas have been found by determining how
much above and below, and how far from and how near a fixed screen,
a spectator may sit, maintaining a comfortable bodily posture, and
obtain a view of the entire screen. The spectator should be seated
in such a position that the picture on the screen will appear at a
level which is most imitative of the level from which natural surround-
ings are viewed in real life. Still another determining factor in
locating these valuable areas is that it is more natural to sit low and
lean slightly backward against the chair to obtain a higher view,
than it is to sit high and lean forward to look down at a screen below
the eye level. For these reasons, therefore, the desirable areas of
seating are limited to levels below the top of the picture. While
both the reversed orchestra floor slope and the present orchestra floor
slope come within the desirable areas, the present slope of floor, which
rises up away from the screen, causes all upper levels of seating to
come within the areas undesirable for natural and comfortable view-
ing. The reversed orchestra floor slope has a much smaller pitch
downward than the present type of floor has upward. The present
type of floor eats into the valuable areas unnecessarily, while the
reversed floor hugs the lower region of the desirable areas, leaving the
remaining valuable areas for additional seating levels.

A method of adjusting the level of the screen, the levels of the vari-
ous eye lines, the distance between the eyes and the screen, the slope
and inclination of the various levels of seating, and the pitches of the
backs of the seats has been developed, keeping a definite relation
between all the elements involved. (Fig. 3.) Formulas have been
evolved to define the position of the screen in relation to the slopes
required for the orchestra and balcony levels. The shape and size
of the screen are also a definite part of the calculations. Given the
shape and size of the screen, and the distance between the screen and
the nearest seat (which should equal the width of the screen, for per-
fect horizontal vision), the vertical distance between the level of the
nearest seat and the level of the bottom of the screen is determined.

Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS

193

194

BEN SCHLANGER

[J. S. M. P. E.

This distance determines the slope of the orchestra floor and the
pitches of the chair backs. The resultant slope is a parabolic curve
which starts near the screen with a downward pitch, decreasing uni-
formly until the floor is practically flat at about the twenty -fourth
row of orchestra seats.

The slope of the orchestra floor in existing theaters is not sufficient
to permit seeing the bottom of the screen over the head of the person
immediately ahead. If the pitch were sufficient for this purpose, it
would be too great for comfortable walking, and would make it
difficult to adjust the standards and leg supports of chairs to the slope
of the floor. The mild pitch of the reversed floor necessary to allow
full view of the bottom of the screen eliminates these difficulties.

FIG. 4. Adaptation of reversed floor to a small theater.

A full, comfortable view of the entire screen can not be obtained until
the ninth row of the present orchestra floor slope is reached, causing
severe neck- and eye-strain in about 300 seats in the average theater.
The reversed floor corrects this condition entirely, allowing a comfort-
able view from every seat in the orchestra and higher levels.

The slope of the reversed floor automatically establishes the proper
pitch for the backs of the seats in every row. This eliminates the
need for specially adjusted backs, and changes in the standards and
leg supports of the chairs. All chairs can then be exactly alike in
every detail of construction. Instead of designing differently con-
structed seats to fit a floor slope, as is now necessary, the floor is de-
signed to suit uniform seating. It is just as though the seats were
placed in an ideal position for viewing the screen, the floor being built
afterward to support the seats in the proper manner.

The matter of determining the maximum tilt for chair backs has
been discussed with the engineering department of a leading seating

Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS 195

company. A tilt of twenty-seven degrees for the row of seats nearest
the screen has been suggested. This tilt will equally distribute the
weight of the body to the seat and back of the chair, keeping the head
of the spectator in a comfortable position. The tilt diminishes to
sixteen and two-thirds degrees at about the twenty-fourth row, re-
maining constant thereafter. The fact that the head assumes a pitch
of two or three degrees less than the pitch of the body has been taken
into consideration in the calculations for determining the floor slope.
The feet of the occupant of a seat are properly supported, and the
sensation of sliding forward out of the seat, as experienced in the
present type of orchestra seating, is eliminated, because the angle and
distance between the floor and the seat remain constant. Due to the
fact that the same chair can be used unchanged throughout the house,
a considerable economy can be effected. The "spring edge" seats,
required where the present type of floor inclines steeply from the
front of the chair, are costly and unsatisfactory, and can be elimi-
nated. The reversed floor permits the use of the "box spring" seat,
which is less costly and more durable. The uneven wearing of the
seats and backs of chairs is also corrected by equally distributing the
weight of the body by properly tilting the seats on the reversed floor.
The results of all these studies and tests have been utilized in pre-
paring a series of new forms for motion picture theater structures.
Many variations of the seating arrangement are made possible that
could not have been arrived at with the present type of floor. The
chart locating the valuable areas has been used as a basis for designing
these forms.

DISCUSSION

MR. KELLOGG: From a novice's standpoint, I have sometimes thought of how
I might try to figure out the best arrangement of theater floor and seats. Imagine
yourself looking toward the audience from the center of the screen; if the solid
angle which your eyes encompass were filled as compactly as possible with eyes
and ears, that would be the way you could get the most people in. Thinking
of it from this standpoint, there are several parts of the possible solid angle,
within which people might hear and see satisfactorily, that are poorly utilized.
One is the region below, which the arrangement proposed by Mr. Schlanger is
primarily aimed to utilize to better advantage; and the other is the part of the
solid angle occupied by the fronts of the balconies. There is only one way
to cut out waste space due to the balconies, besides making them as thin as
possible, and that is, not to have balconies. But that does not furnish good
space utilization, either.

From the standpoint of the utilizing angle below, one is confronted with the

196 BEN SCHLANGER [j. s. M. P. E.

fact that the part of the audience nearest the stage subtends a disproportionately
large angle compared with the number of people. I am wondering whether,
following the idea described in the paper, we might not actually, within the proper
angle, pack more people into a theater of given dimensions by not trying to be-
gin too close to the screen. If you drop a little further back you can begin a
little lower, leaving more height available for balconies.

There is one question I do not believe was answered in the paper, and that is,
what is considered the desirable rise per row of seats. To be more specific, if
a line were drawn from the back of one row of seats to the bottom of the screen,
how much would that line miss the back of the next row? As a general rule,
one can probably figure on looking between the heads of the people immediately
in front, but he will have to look over the heads of those seated two rows ahead
of him. I should be interested to learn how much allowance is made for this
factor.

MR. SCHLANGER : The point about having the first eye line farther away from
the screen is quite possible, but it depends on how much area in front of the
screen the theater exhibitor is willing to devote to it. That area is costly. It
is better practice to place the first seat a little farther away from the screen
than it is usually placed.

Referring to Fig. 3, it is seen that a given person sees above and not between
the heads of the persons in front of him. His sight line passes over the head
in front, passing to the very bottom of the screen. The orchestra floor in the
present type of theater is not pitched sufficiently to allow this complete view of
the screen, therefore making it necessary for people to keep shifting their positions
in order to see between the heads of the people ahead.

When constructing the diagram, a line is drawn from the bottom of the screen
to the top of the head of the first spectator. This gives the eye level for the
next row. The distance from the eye to the top of the head is four inches. So,
connect another line from the bottom of the screen to a point four inches higher,
and we have the sight line of the next row, and so on.

The pitch of the floor varies from row to row. The first row requires a greater
pitch than the succeeding one, and so on. On reaching the twenty-fourth row,
in the reversed floor system, the floor becomes practically flat. If the theater
were big enough the curve would rise again somewhere about the fiftieth row.

In the present type of house, the low placing of the screen makes it necessary
for each succeeding head to be higher, in order to see over the head of the person
in front. The pitch increases so rapidly that in existing theaters a compromise
has been effected, and one looks between the heads of those in front, and not
over their heads, as on the reversed floor.

MR. PORTER: I should like to know how much tolerance, if any, is allowed
for variations in height among individuals. Is any allowance provided for such
variations, or are the angles worked out for an average height?

MR. SCHLANGER: The average height measured from the floor to the eye level,
the person sitting in a normally pitched chair on a flat floor, varies from three
feet seven inches, to four feet; four feet is the dimension for a person six feet
two inches tall. To allow for a person five feet eleven inches, or six feet, would
be about right. Therefore, about three feet eleven inches should be assumed.
That would take care of almost all such conditions.

Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS 197

MR. RICHARDSON: On what do you base your choice of what are termed
desirable areas?

MR. SCHLANGER: The desirable areas are determined by the distance above
or below the level of the screen at which it is comfortable to view the screen.
To establish areas of desirability, a line is drawn from a test point to the impor-
tant focal point on the screen; a line is then drawn, representing the back of the
spectator, ninety degrees to this line. This will show how much the spectator
must lean forward or backward. The areas where the spectator does not have
to lean forward or raise his head are most desirable. The degree of deviation
from comfortable posture determines the relative value of a given area.

MR. SPENCE: The suggestion was made that this floor plan would be adapt-
able to a Trans Lux theater, but I do not think so. The Trans Lux theater was
designed with a ten-foot head room, in order to be able to move in and out of
established office buildings. To use this type of floor it would be necessary to
break through to the cellar. If rear end projection were used, it is possible that
people walking in the aisle would get into the beam, and the seats would have to
be placed farther back from the screen. The angle of view is so much wider
with a rear projection screen, that what would be gained by putting a projection
room near the box-office, so to speak, would be lost at the front.

MR. SCHLANGER: The distance from the first row to the screen is not deter-
mined by the method of projection. It is determined by the size of the picture.
Because the Trans Lux theater presents a small picture frame on the wall and
we can get very close to it, does not mean that we can use Trans Lux projection
and have effective motion picture exhibition.

As to the ten-foot height mentioned, the head room of the first floor of the
average building is nearer twelve feet. It is possible to use standard projection
in such limited heights by employing a reversed floor, thereby requiring only
half the projection booth area necessary for Trans Lux projection. The saving
in annual rental for this space on the street level could be used to rent a foot or
two below the street grade, which would allow the use of a much larger screen
than is now used in Trans Lux theaters.

MR. PORTER: What would be the total drop in your floor?

MR. SCHLANGER: The slope of the reversed floor is considerably less than
the slope of the present type of floor.

MR. Fox: Mr. Schlanger showed that the weight was equally distributed be-
tween the back and seat of the chair. He meant that a certain amount of weight
was taken from the seat and applied to the back. About fifteen per cent of the
excessive tilt is transferred to the back from the seat. We have done what the
airplane man has done, so that the passenger would not fall out of his seat when
the plane was tilted for landing.

MR. HICKMAN: I should like to ask if any theaters have been constructed in
this manner?

MR. SCHLANGER: Mr. Kinsila, in his book on theaters, stated that a theater
having a reversed floor was built in Moscow many years ago. Better Theatres
published an account of the Pathe Theater in Paris, which also has a reversed
floor. In both cases, there is no evidence of a curved reversed floor. The
Pathe Theater was constructed in such a manner because of a steep street grade
condition, and a limited height usable in an existing building. The posture

198 BEN SCHLANGER

and sight line problems in this theater were not given any thought, as the uniform
floor pitch and chair tilts show.

At present, I am working on a small theater, which is now under construction,
in which the parabolic reversed floor is used. To my knowledge, this will be the
first theater built wherein the principles of applied optics and good sitting postures
are recognized.

A METHOD OF MEASURING DIRECTLY THE DISTORTION
IN AUDIO FREQUENCY AMPLIFIER SYSTEMS*

W. N. TUTTLE**

Summary. — The question of a suitable measure of harmonic distortion is dis-
cussed. The distortion factor employed is defined as the ratio of the effective value
of the combined harmonic voltages to the fundamental voltage. A simple rapid
method of measuring this ratio is described which has several advantages over earlier
methods. Results are given showing the performance of apparatus for the applica-
tion of this method to the testing of audio frequency amplifier systems.

It is customary to rate an amplifier used in the reproduction of
speech or music on the basis of the amount of undistorted power which
it is capable of delivering to the loud speakers.

The distortion produced by the amplifier is the deviation in the
shape of the electrical wave of the amplifier output from that applied
at the input terminals. Several types of distortion are evidently
possible. Frequency distortion takes place when the system does
not amplify equally all the component frequencies of the input wave.
Phase distortion occurs when these several component voltage waves
are shifted in time with respect to one another. Harmonic distortion
or amplitude distortion is observed when the crests of the voltage
wave tend to be partly cut off. The first two of these effects, fre-
quency distortion and phase distortion, depend largely on the funda-
mental design of the amplifier and are not appreciably affected by
variations in the magnitude of the input voltage. Harmonic distor-
tion has the opposite characteristic. It is important when the
amplifier is operated at low energy levels, but increases rapidly when
the voltage applied is increased beyond a certain value. Harmonic
distortion is the factor which limits the useful output of an amplifier.
It is the measurement of this effect which is to be considered.

The measure taken for harmonic distortion should, if possible,
be a measure of the objectionableness to the listener of the distortion
present. Different types of amplifiers cause different types of distor-

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** General Radio Co., Cambridge, Mass.

109

200

W. N. TUTTLE

[J. S. M. P. E.

tion, so that the quantity measured should be such as to supply a basis
for the comparison of amplifiers of different fundamental design.

If a single pure tone is applied to the input terminals of an amplifier
harmonic distortion will result in the appearance in the output
voltage wave of various harmonic frequencies of the applied tone.
Let us call EI the fundamental voltage, and £2, £3, £4 ... the
voltages of the several harmonic frequencies. Let us consider the
ratio

This is seen to be the ratio of the effective value of the combined
harmonic voltages to the fundamental voltage. This expression
gives equal weight to all harmonics having the same value, but as each

TEST
VOLTAGE <>_

FIG. 1. Simplified functional diagram of the distortion-factor meter.

harmonic enters as its square, prominent components are considerably
emphasized. A single component of two volts, for example, is given
twice the importance of two harmonics, each of one volt. In view of
the masking effect of one tone by another, the single two- volt compo-
nent would be expected to be correspondingly more objectionable.
In view of these considerations and of the fact that it lends itself
readily to direct measurement, the ratio D has been generally
adopted1'2 as the best measure of harmonic distortion.

The methods of evaluating the distortion factor which have been
available have been suited primarily to laboratory use. One method
is that of measuring the separate harmonics in the amplifier output by
means of a harmonic analyzer, and of computing from these values the
distortion factor. Similarly, oscillographic records can be analyzed
approximately by any one of several methods. A less laborious

Feb., 1932]

MEASURING DISTORTION

201

procedure is that of eliminating the fundamental in a suitably designed
bridge circuit and measuring the combined harmonics which remain.3
This method has one serious disadvantage in that the bridge must be

TYPE 536 A

DISTORTION FACTOR METER
FILTER CHARACTERISTIC

100

1000 5000 10000

FREQUENCY - CYCLES PER SECOND

500OO

FIG. 2. Over-all attenuation characteristic of the filter and input resistances.

carefully balanced, and that slight fluctuations in the test frequency
may consequently make it difficult to obtain satisfactory results.
Another disadvantage is that frequencies below the test frequency,
including power-supply hum, are included with the harmonics.

INDICATOR Z
> 500000/1

FIG. 3. Detailed circuit diagram of the distortion-factor meter.

Ballantine and Cobb4 have developed an ingenious null method of ob-
taining the distortion factor which avoids these difficulties. But
this method, also, is suited principally to laboratory use.

202 W. N. TUTTLE [J. S. M. P. E.

It seemed desirable to develop a rapid method of measuring the
distortion factor which would require neither bulky apparatus nor the
services of a skilled operator. The essentials of the method finally
adopted are shown in Fig. 1.

The test voltage is applied to a high-pass filter and an attenuator in
parallel. The output of the filter is proportional to the combined
harmonic content of the test voltage. The output of the fixed
attenuator is proportional to the test voltage itself, which is approxi-

FIG. 4. Panel arrangement of the distortion-factor meter.

mately equal to the fundamental voltage in cases encountered in
practice. We can therefore determine the distortion factor by
comparing the output of the filter with the output of the attenuator.
In the apparatus constructed, the fixed attenuator is so proportioned
that when the voltage across the entire voltage divider is equal to the
filter output voltage, the distortion factor is 30 per cent. A dial
reading directly distortion factors from zero to 30 per cent is at-
tached to the voltage divider control. All that is necessary to make a

Feb., 1932] MEASURING DISTORTION 203

measurement of the total harmonic content of the test voltage is to
observe the deflection of the indicator when connected to the output
of the filter. The indicator is then switched to the output of the
voltage divider, the setting of which is varied until the same deflection
is obtained. The voltage divider scale reading then gives the
distortion factor directly.

The success of this method evidently depends on the care with
which the apparatus is designed rather than on the skill of the operator.
The filter must reduce the amplitude of the fundamentals so that it is
negligible compared with the harmonics at the lowest distortion
factor to be measured. It must transmit the harmonics equally and

FIG. 5. Distortion-factor meter with associated amplifier and square-law

galvanometer.

must not act as a generator of harmonics even when large input
voltages are applied. The measured transmission curve of the filter
developed for this purpose is given in Fig. 2.

It is seen that the fundamental (400) cycles are reduced relative to
the harmonics more than 70 db. All harmonics up to the fifteenth
are transmitted equally within 0.3 db. The attenuation of the
fundamental is sufficiently great so that the test frequency can vary
over a range of more than 50 cycles without affecting the result. It
is seen that enough attenuation is provided at the lower frequencies to
eliminate power supply hum.

The circuit details are shown in Fig. 3. It will be observed that a
series resistance is placed in the input branch. This keeps the

204

W. N. TUTTLE

[J. S. M. P. E.

impedance out of which the filter works practically constant and
makes the calibration independent of the impedance of the voltage
source. It also results in the impedance at the input terminals being
high enough (about 175,000 ohms) so that the instrument may be
connected across practically any element in amplifier circuits without
causing appreciable disturbance. Due to the input impedance
characteristic of the filter alone, the series resistance reduces the
harmonic content of the voltage across the voltage divider to such
an extent relative to the fundamental that no correction need be
applied. The "L" network RiR2 may be switched into the circuit
to magnify the scale of the dial by ten when measuring small distor-
tion factors.

0 I 2 3 4

POWER OUTPUT - WATTS

FIG. 6. Curves of distortion factor vs. power output for a laboratory
amplifier employing a single 245-type tube, for two values of load resistance.

The apparatus is simple enough so that a compact mechanical
arrangement is possible. Fig. 4 shows the panel arrangement.

The indicator must have an input impedance high compared with
the filter impedance. As high sensitivity is also required in many
measurements, it is convenient to employ an amplifier in conjunction
with an a-c. voltmeter. To keep the impedance high, no input
transformer should be used preceding the first amplifier tube. The
voltmeter should be of a type that will indicate the effective value of a
composite voltage. Thermocouple instruments have this property
but are sluggish in action and are easily burned out. Vacuum

Feb., 1932]

MEASURING DISTORTION

205

tube voltmeters of the square-law type are satisfactory. A rectifier
type 2-C galvanometer has been developed for use with the distortion
factor meter which has a characteristic closely approximating a
square law and which combines ruggedness with high sensitivity.
The distortion-factor meter, together with the associated amplifier
and the a-c. galvanometer, is shown in Fig. 5.

1000

2000 3000 4000 5000
LOAD RESISTANCE - OHMS

6000 7000

FIG. 7. Curves of distortion factor vs. load resistance for three values of
output power for the amplifier of Fig. 6. (The ratio RL/RP of load resistance
to plate resistance is indicated for reference.)

Curves obtained with this apparatus in testing a laboratory ampli-
fier unit are shown in Figs. 6 and 7. Jn obtaining the data of Fig. 6,
the load resistance was held constant as the input was varied. Ob-
servations were made of the load voltage and distortion factor as the
amplifier input voltage was increased. The distortion factor is

206 W. N. TUTTLE

plotted against the computed power output for two values of load
resistance.

Fig. 7 gives the data obtained by simultaneously varying the load
resistance and the input voltage to keep the output power constant.
It is interesting to note the manner in which the optimum load
resistance varies with the allowable distortion factor.

These results support the conclusion that the distortion factor
which has been defined is the logical index of the performance of an
amplifier. The apparatus developed for directly measuring this
quantity may be conveniently used in making the simultaneous
measurements of power output and total harmonic content which
are necessary in obtaining a definite output rating for an amplifier.
The apparatus is also suited to checking the operating condition of an
amplifier installation. A single measurement of the distortion factor
at the rated power output indicates definitely whether or not the
system is functioning properly.

In view of its adaptability to the testing of amplifier systems, it is
hoped that the distort ion -factor meter will prove useful to the motion
picture industry in maintaining definite standards of amplifier
performance.

RFFERENCES

1 BALLANTINE, S.: "Detection at High Signal Voltages," Proc. I. R. E., 17
(July, 1929), p. 1153.

2 WOLFF, I.: "The Alternating-Current Bridge as a Harmonic Analyser,"
/. Opt. Soc. Amer. and Rev. Sci. Instr., 15 (September, 1927), p. 163.

3 BELFILS, G.: "Measuring the Residue of Voltage Curves with a Distortion
Factor Meter," Rev. gen. de I'electricite, 19 (April 3, 1926), p. 523.

4 BALLANTINE, S., AND COBB, H. L.: "Power Output Characteristics of the
Pentade," Proc., I. R. E., 18 (March, 1930), p. 450.

DIRECTIONAL EFFECTS IN CONTINUOUS FILM
PROCESSING*

J. CRABTREE**

Summary. — Continuous motion picture film developing machines set up a uni-
directional current of developer as a consequence of the motion of the film unless
vigorous measures are adapted to counteract it. This current is shown to cause dis-
tortions of the H & D characteristics of the photographic material, when the usual
type of sensitometer strip is used for control of the development. The effect of the
geometry of the sensitometer strip on the extent of "direction effect" is discussed.

The difficulty of obtaining equal degrees of development of photo-
graphic images over even relatively small areas of any given emulsion
layer is well known to workers with these materials. The necessity
for constant and vigorous agitation of the developing solution across
the light-exposed area to be developed becomes evident even to the
amateur.

The reason for the desirability of vigorous agitation during the
development of a photographic image is the necessity for removal from
the neighborhood of that image, of the products of chemical reaction,
chiefly bromides and developer oxidation products, which diffuse
out from the emulsion layer, and which, apart from locally reducing
the developing power of the solution by exhaustion, have in them-
selves an actual depressing effect on density. These reaction prod-
ucts, being of greater specific gravity than the original developing
solution, probably set up slight local eddy currents during static
development which give rise to unevenness in the degree of develop-
ment from point to point, even though the exposure to light is
uniform. The unevenness is relatively the more pronounced, the
less the degree of development, tending to disappear as gamma
infinity is approached. The well-known "Eberhardt effect" or
"Mackie Line" is occasioned in this manner. These designations
refer to the light halo appearing around an area of high density, in a

* Presented in the Symposium on Laboratory Practices at the Spring, 1931,
Meeting at Hollywood, Calif.

** Bell Telephone Laboratories, New York, N. Y.

207

208 J. CRABTREE [J. S. M. P. E.

field of lower density. The halo results from the reduced rate of
development of the area of lower density adjacent to the area of
higher density, consequent on the diffusion outward from this high
density area of concentrated reaction products of development. A
study of this effect was recently made by Walenkov1 who gives a
bibliography of the literature.

Before the introduction of machine development for motion picture
film, this material was processed almost entirely by the rack and tank
system, in which the film was wound on large square or rectangular
racks which were immersed in deep tanks of suitable shape. The
unevenness of development obtained along the length of film handled
in this manner was well recognized, and has been thoroughly discussed
by J. I. Crab tree and C. E. Ives.2 The method gave locally increased
development in the region of the rack-ends and cross-bars, due to

FIG. 1. Strip of dyed film showing effect of directional currents.

production of eddy currents from temperature differences as well as
from differences in specific gravity of reaction products, unless agita-
tion of the developer was obtained by motion of the rack, which was
not commercially practicable. Pictures developed in this manner
showed periodic light and dark bands when the resulting print was
projected upon the screen, as well as Eberhardt effects when de-
velopment was restricted to allow for control in contrast of the
picture.

In recent years an already growing tendency to change from rack
and tank development to so-called continuous development by
machine was stimulated by the addition of sound to the picture,
so that, as a result, almost all such film is now processed by machines
in which the film is mechanically propelled at a uniform rate through
horizontal trays or deep vertical tanks containing the developing solu-
tion in circulation. This method of processing has resulted in such an

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

209

apparent uniformity of product that certain local effects have been
lost sight of.

However, if the developing bath of a continuous film processing
machine is observed during operation, it will be seen that the onward
movement of the strands of film sets up a current of developer in the
direction of the film movement, but which, relative to the film itself,
is in a direction opposite to that of the film travel.

The result of this is shown clearly by Fig. 1, a photograph of a
section of film which was first dyed red over a small area, then passed

HIGH DENSITY LEADING
LOW DENSITY LEADING

FIG. 2. Typical "direction pair" H& D curves; machine development.

through the developing and fixing trays of an Erbograph machine at
sixty feet per minute in the direction shown by the arrow. It will be
seen that as the color diffused out from the dyed area it was carried
backward along the film surface and partly absorbed by it.

During the development of a photographic image, the products of
the chemical reaction taking place diffuse out from the gelatin layer
much as the solution of dye did in the manner shown in Fig. 1.
Therefore, when developing motion picture film in such a continuous
processing machine as the above, the products of reaction of the

210

J. CRABTREE

[J. S. M. p. E.

development of any image must flow across the images immediately
following it. Since, as was previously mentioned, these products of
reaction have a restraining effect on development, processing by this
means will result in a variation in degree of development from
point to point on the film depending upon the concentration of the
reaction products at those points, which in turn depends upon the
magnitude of the density of the image area just ahead.

Let us now consider what happens in the development of a sensi-

3.2

2.8

2.4

2.0

1.6

0.8

HIGH EXPOSURE LEADING
LOW EXPOSURE LEADING
BRUSH DEVELOPMENT (THIS
CURVE IS SPACED FROM THE
MACHINE PAIR FOR EASY

COMPARISON) r

LOG RELATIVE EXPOSURE

FIG. 3. "Direction pairs" H & D curves; machine develop-
ment, showing comparison with brush development.

tometer exposure. This strip of exposed film bears a series of latent
images increasing progressively in magnitude from step to step and
which on development will result in a series of density areas, ab,
in Fig. 2, represents diagrammatically such an exposed film, the
shaded area being the image portion, and the clear area representing
unchanged silver halide. If now, this strip is moved steadily through
the developer in direction AB (that is, B meeting the developer first),
the reaction products of development from the first step will, by

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

211

virtue of the motion of the film, flow over and successively affect
the other steps. The initial reaction products will be reenforced by
the reaction products from succeeding steps, causing a progressive
weakening of the developing effect as the end A is approached.

With the strip traveling in the opposite direction (A leading)
the developer around end A will be little affected by reaction products
since these will be small in amount and hence nearly full development
will be obtained. As B is approached, however, the products ac-
cumulate from the gradually increasing densities so that there is a

2.8
2.4
2.0

1.6
>
H
</)

a.i.

0.8
0.4

GH DENSITY UP
DW DENSITY UP

-B' !

*^~

^' '

S?-

<S'

O 0.4 0.8 1.2 1.6 2.0 2.4 2.8 32
LOG RELATIVE EXPOSURE

FIG. 4. "Direction pairs" H & D curves; rack tank and development.

decided loss of density in the region of B. The characteristic H & D
curves plotted from the readings of density obtained in the two cases
cited will be found to have the general shapes shown in Fig. 2 where
A'B' is that resulting from progression through the developer with
the high exposure leading; while AB is obtained when development is
carried on with the low exposure leading.

The general effect is to straighten out the shoulder in A'B' and to
depress it somewhat in AB. The effect on gamma is slight but there
is an appreciable difference in the estimate of latitude (straight line

212

J. CRABTREE

[T. S. M. P. E.

portion) and inertia (intercept on log R axis) in the two cases. That
neither curve is true in form is shown in Fig. 3, in which similar
pairs of curves are shown at two gammas with the corresponding
curve obtained by "brush" development. This "brush" method is
one of those generally used in precise sensitometry and consists in
securing a very thorough removal of reaction products from the film
surface by passing a soft brush rapidly backward and forward across
it during development.

A similar effect was found to result from rack and tank develop-
ment where the manipulation of the rack was such as to result in a

FIG. 5. Perspective of Erbograph type developing tray.

moderate amount of agitation of the developer. (See Fig. 4.) With
this method, curves of type A'B' resulted from those cases where
the sensitometer strip was developed with the high exposure end of
the test strip uppermost, while AB resulted from development with
the lowest exposure uppermost. Since in this case the reaction
products which are of higher specific gravity than the original de-
veloper tend to generate a downward current, the same explanation
applies to the occurrence of the two types of curve as in the case of
machine development.

Since the development of variable density sound records in con-

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

213

tinuous machines is usually controlled by using a sensitometer strip
in some form, an inquiry was made to determine what sensitometric
troubles might be encountered in practice from this "directional
effect," as it will be called in this paper, and to determine the best
type and manner of use of the sensitometer strip in machine develop-
ment.

In the processing of variable density records the usual procedure is
to develop the sound negative to an approximate gamma of 0.6 in' a
developer of the D-76 borax type, and to develop the print in a D-16

2.8

2.4

0.4

0 2 4 6 8 10

DEVELOPMENT TIME IN MINUTES

FIG. 6. Time-gamma curves for the negative and positive de-
veloper used.

type of bath to a gamma of 1.80, or higher. Attention was mainly
confined to a study of these two types of developer at gammas in the
region of those just mentioned. Also, the directional effect in the
particular machine used has been found to be present for a variety of
commercial types of film, although in this study we used only standard
positive film.

The machines used in the experiments to be described were of the
Erbograph type, in which the film is stranded horizontally around
drive rolls in 15 loops of 13 feet each in a horizontal tray of 50 gallons

214

J. CRABTREE

capacity, as shown diagrammatically in Fig. 5. Circulation was by
gravitational feed, and overflow to the return pumps was at the
rate of 10 gallons per minute. The film speed may be adjusted over a
range of from 10 to 100 feet per minute.

A'
TIME

I IMC

FIG. 7. Typical time-gamma curve of a photographic developer
The formulas of the developers employed in this work are:

Negative

Positive

Elon

2 grams

Elon

0.3 gram

Hydroquinone

5 grams

Hydroquinone

6.0 grams

Sodium sulphite

100 grams

Sodium sulphite

37.0 grams

(anhydrous)

Borax

8 grams

Sodium carbonate

12 . 5 grams

(anhydrous)

Boric acid

8 grams

Potassium bromide

0.9 gram

Water to

1 liter

Water to

1 liter

Their time-gamma curves for machine development at 67 °F.
are given in Fig. 6.

The sensitometer exposures were made in a variable intensity
sensitometer using photographic step tablets. Different dimensions
of tablet were used, as will be explained later. Where not otherwise

0.4

0 0.4 0.8 1.2 1.6 2.0 2.4 2.6 3.2

LOG RELATIVE EXPOSURE

FIG 8. Effect of gamma on "directional effect;" negative
developer.

3.0
3.2
2.8

2.4

>.2.0
2

«...

1.2

0.4
0

HIGH DENSITY LEADING
LOW DENSITY LEADING

^*-

•^ '•

"]_

*°

/A,

r>^J

''y//

^ /

'/''

*'/'

--^-

0.4 0.« \2. 1.6 2.0 2.4 ZA
LOG RELATIVE EXPOSURE

FIG. 9. Effect of gamma on "directional effect;" positive
developer.

216

J. CRABTREE

[J. S. M. p. E.

specified, the length of each step was 5/i6 inch exposed across the full
width of the 35-mm. motion picture film.

Pairs of sensitometer strips exposed in exactly the same manner
were passed through the machine, one with the lightest exposure,
the other with the heaviest exposure leading and will be referred to
as "directional pairs." The differences between the two H & D

3.2

2.8

2.4

2.0

1.6

5 ,.2

0.8

0.4

POSITIVE

NEGATIVE

LOG RELATIVE EXPOSURE

FIG. 1Q. H & D curves showing effect of circulation of developer in machine

development.

curves resulting are considered as an approximate measure of the
"directional effect."

EFFECT OF GAMMA ON DIRECTIONAL EFFECT

As is well known, the effect of an increase of circulation during
development is to decrease the time required to attain a certain degree
of development. This is a result of the more rapid removal of reac-
tion products from the emulsion surface of the film. We may there-
fore, for our present purpose, consider the effect of the accumulation of

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

217

reaction products at any point to be equivalent to a loss of effective
time of development. Reference to a typical time-gamma curve in
Fig. 7 indicates that this will be more important at low gammas,
since a given interval of time has, in that region of the curve, more
effect on gamma or density.

Since the "directional effect" is considered to be a result of local

CROSS BARS CARRYING FILM WIPERS

TRAY

PLAN

CROSS BARS

WIPERS

r luivi

£*

j] ITI m [J] DO DD [J] [B_

FIG. 11. Plan and elevation of device for eliminating "directional effect"
in machine development.

accumulations of reaction products, its degree might well be expected
to vary with gamma.

Figs. 8 and 9 show three "direction pairs" at low, intermediate,
and high gammas in the negative and positive developers, respec-
tively. From these curves it will be seen that directional effect is
present at all gammas likely to be used in negative development but
that it tends to disappear at higher gammas with the positive bath.
This is consistent with the remarks on the slope of the respective
time-gamma curves (Fig. 8) at the particular gammas used.

218

J. CRABTREE

[J. S. M. p. E.

It should be mentioned here that, as the effect has been found
to be less pronounced in the positive than in the negative developer,
the inquiry was, in many cases, restricted to a consideration of
negative development only.

INFLUENCE OF DEVELOPER CIRCULATION ON "DIRECTIONAL EFFECT"

The fact that high gammas (i. e., low film speed) in the negative
developer showed as much "directional effect" as low gammas (high

LOG RELATIVE EXPOSURE

FIG. 12. Typical curves obtained without squeegee device.

film speed) in the same developer indicates that the general circula-
tion provided by the pumps to the developer had but little effect in
breaking up the "direction current." This was further confirmed by
tests in which for one case no gravity, and pump-return, circulation of
developer was used; while, for the other, all the circulation which the
system was capable of giving (ten gallons per minute) was used.

Fig. 10 shows the curves applicable to the two cases for each
developer at a film speed of 60 feet per minute. No difference is

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

219

apparent that could be considered to be in favor of the circulating
developer.

From these results it is evident that additional means of developer
agitation must be provided if "directional effect" is to be avoided.
This could be achieved by agitation by such means as paddle, pro-
peller, jets, or by injection of air.

It was found, however, that the directional current could be broken

FIG. 13.

LOG RELATIVE EXPOSURE

Typical curves obtained with squeegee device.

up by diverting it from the surface of the film at frequent intervals
by the use of a squeegee device. The contrivance used is shown in
Fig. 11 and consists of a series of stationary rubber squeegees in-
stalled in the developing tray. Each squeegee is about 6 inches apart
and set at an angle of 45 degrees to the longitudinal axis of the film.
As the film passes each squeegee the developer in contact with it is
diverted sideward into the surrounding mass of developer and is
so replaced by fresh solution. This is found effectively to prevent

220

J. CRABTREE

[J. S. M. P. E.

any setting up of continuous currents in the direction of the movement
of the film.

Fig. 12 shows curves for negative development at two different

1.6

1.2

-——HIGH DENSITY LEADING
LOW DENSITY LEADING

"**

/''

FRESH .s

'''

•'EXHAUSTED

S^ s
s

^"'

__-r-e

LOG RELATIVE EXPOSURE

FIG. 14. Relation between condition of developer and
"directional effect."

gammas, obtained without, and Fig. 13 similar exposures developed
with such an appliance. It will be observed that the "directional
effect" has been almost entirely eliminated by the use of this device.

2.4

2.0

1.2

0.8

0.4

HIGH DENSITY LEADING
LOW DENSITY LEADING

LOG RELATIVE EXPOSURE

FIG. 15. Relation between bromide content of developer and
"directional effect."

CONDITION OF THE DEVELOPER

Since the distortion of the H & D curves produced by directional
currents is caused by the presence of reaction products, it was thought

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

221

that if such reaction products were already present in the developer,
as in the case of a partially exhausted bath, the effect should perhaps
be relatively less than with a fresh solution. Fig. 14 shows the results
of such a test and gives "direction pairs" of curves from strips de-
veloped in fresh negative developer and in the same developer after
exhaustion to a degree beyond that normally used in practice. The
improvement shown by the use of the exhausted developer is com-
paratively slight.

A similar test was conducted in which the effect of the addition
bromides to the negative developer was studied. Fig. 15 shows
curves of "direction pairs" developed in a fresh bath (^4) and also in
the same bath to which potassium bromide had been added (B).

LOG RELATIVE EXPOSURE

FIG. 16. Directional distortion effects in two types of developing machines;
negative development.

Little improvement was manifested even though the bromide con-
centration was for some tests considerably in excess of any prac-
ticable figure.

MACHINE DESIGN

The cause of the "directional effect" is such that its presence and
amount must depend upon the design of the particular processing
machine employed. A series of "direction pairs" of strips from step
tablets of different dimensions were, therefore, processed in another
laboratory where a vertical-tank type of machine was available.
A comparison of results obtained with the horizontal tray type used
in our laboratory is shown in Figs. 16 and 17. The indications are

222

J. CRABTREE

[J. S. M. P. E.

that the directional effect is somewhat less in the vertical-tank type.
No information was available as to developers used. It would
seem reasonable, however, that in the vertical type gravity will assist
in removal of the reaction products since they are of higher specific
gravity than the fresh developer.

SENSITOMETER STRIP DESIGN

The trailing of reaction products from any given area will influence
following images only for a certain linear distance for a given density
magnitude. It would therefore follow that the longitudinal dimen-

LOG RELATIVE EXPOSURE

FIG. 17.

Directional distortion effects in two types of developing machines;
positive developer.

sions of a sensitometer exposure strip will have a bearing on the
differences in the H & D characteristics shown by a "directional pair."
Also, since a narrow trail would have a better chance of diffusion
into the surrounding mass than a wide one, the width of the exposure
should also have an influence. Other factors of importance are the
density interval between steps and the length of the toe and shoulder
portions of the curve, respectively. These various factors have been
examined with the following results.

(a) Effect of Length of Step. — Step tablets were obtained having
step widths in the longitudinal direction of film travel of 3/4, Vie,
Vie, 1/s, and Vie inch. The 3/4 and Vie inch tablets were of different

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

223

origin and had different density intervals from the remainder but the
1/s and Vie inch were identical in origin with the 3/ie inch tablet,
having been constructed from a portion of it by dissection and re-

1.2
0.8
0.4

1.2

0.8
0.4
0

HIGH DENSITY LEADING

***

•**

^*-

— LOW DENSITY LEADING

^ "

#^J

*?'

{?'

^j^l

^r^

«-«

**«

x_^

X**

^. —

1»

•

L><1

r^5

f '

-^-

«^

— «r

«^t

LOG RELATIVE EXPOSURE

FIG. 18. Influence of step width on "directional effect;
negative development.

3.2
2.6

2.4
£2.0

S'-6
0

1.2
0.8
0.4
0

\ \ \

—

HIGH DENSITY LEADING
LOW DENSITY LEADING

*•

—

*—

^~~

/x-

^»—

/f*

/,'

//'

„••—

— *

>/«

/£

J/S

i/f

[

/,'

</'

<**

^f^

LOG RELATIVE EXPOSURE

FIG. 19. Influence of step width on "directional effect;" positive
development.

assembly. "Direction pairs" of curves are shown in Fig. 18 for
negative and in Fig. 19 for positive development. They are spaced
along the log relative E axis for convenience of comparison. Gamma

224

J. CRABTREE

[J. a M. P. E.

relations do not hold, since development was performed on differ-
ent occasions, except for the 3/ie, l/s, and Vie inch curves which were
developed together. It will be seen that the directional effect is
more evident in negative development than in positive, and that
in the negative the effect increases as the width of the step decreases ;
and that in the 3/ie, 1/s, and Vie inch curves in both negative and
positive the gamma tends to rise as the step width decreases. The
results obtained in the negative developer appear to be reasonable
in that the farther away the center of one step is from the center of the
preceding step, the less the density at the center of the former will be
affected by the latter, and vice versa. Also, as the longitudinal dimen-

— HIGH DENSITY LEADING

— LOW DENSITY LEADING

FIG. 20.

LOG RELATIVE EXPOSURE

Effect of track width on "directional effect.'

sion of the sensitometer exposure is reduced, the less will be the general
dilution of the supernatant developer by the total mass of reaction
products, and so the gamma reached will be higher.

(b) Width of Track. — Limiting the transverse dimension of the
sensitometer exposure was found not to show any noticeable diminu-
tion of "directional effect" so long as the exposure was confined to
the center of the film. However, limiting the width of the exposure
to sound track dimensions and its position to that of the sound track
resulted in a diminished effect compared with the full width exposure.
The directional effect was very considerably reduced with a 3/4
inch step at positive gammas (Fig. 20) made under these conditions.

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

225

(c) The Density Interval between the Steps of the " Sensitometer
Tablet." — The magnitude of the difference in exposure from step to
step of the sensitometer tablet may be expected to have a bearing
on the degree of the distortion of the characteristic curve from the
exposed strips, since the depression of density, by the "directional
effect," of any given step will depend on the magnitude of the den-
sity of the step preceding it. A comparison was therefore made be-
tween tablets having the same exposure range but in which one tablet
had twice as many steps as the other, and hence but half the density
interval.

H 0.8

0.4

—

RANGE = 2.93
AVERAGE INTERVAL=O.I3

^ •—

*£$

*?*

**'

1.2

H0.8
|

0.4
0

s
s

RANGE =2.97
AVERAGE INTERVAL^ 0.25

. — •

*>^

x<-

s^z

^~*

0.4

O.8 L2 1.6 2.0 2.4

LOG RELATIVE EXPOSURE

2.8

3.2

FIG. 21. Effect of sensitometer exposure interval on
"directional effect."

The curves are shown in Fig. 21 and indicate that decreasing the
number of steps in the tablet, while not affecting the degree of separa-
tion between curves of a "pair," shows greater distortion at the
shoulder and so is less desirable.

(d) Effect of the Exposure Range of the Sensitometer Tablet. — The
range of exposure in the step tablet type of sensitometer depends on
the density range of the tablet, and upon this and upon the time and
intensity of the exposure applied will depend how many of the
resulting readings of density will fall on the toe and shoulder portions
of the curve.

226

J. CRABTREE

[J. S. M. P. E.

1.2
0.8
0.4

LOW DENSITY LEADING

^ — "

^^^

^- — '

^— -

^^^

P^*

,**-

•^

p. — '

(^-*

K — J

X-*

^^-

_
b>

1.2
>-

^:-

50.8

0.4

HIGH DENSITY LEADING

LOG RELATIVE EXPOSURE

FIG. 22. Effect of restricted exposure range in sensitometer; absence of toe.

1.6

1.2

Z0.8

0.4

1 1

HIGH DENSITY LEADING

,^-

x
x

s*^

^^-

^-*^

?*"

1.2

t-
<n

Z0.8

0.4

1 1 1

LOW DENSITY LEADING

^-—

_ >—

x
X

»— — '

•^j

LOG RELATIVE EXPOSURE

FIG. 23. Effect of restricted exposure range in sensitometer; absence

of shoulder.

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

227

It was found that varying these factors resulted in extraordinary
distortions of the characteristic. Fig. 22 shows curves derived from a
series of exposures made through a sensitometer tablet in which the
density steps were progressively masked off from the toe end of the
curve while in Fig. 23 is shown a similar series in which the steps were
removed from the shoulder end. The time and intensity of the
exposure incident on the tablet were constant throughout to ensure
that no gamma differences could accrue from differences in the
reciprocity relation. In this case the "direction pairs" are separated
into two groups, (a) being the group having the high exposure leading
and (b) the group having the low exposure leading. The curves

HIGH EXPOSURE LEADING
LOW EXPOSURE LEADING

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

LOG RELATIVE EXPOSURE

FIG. 24. Example of errors resulting from "directional effect."

are spaced along the log relative exposure axis for ease of comparison
of shapes. They show that unless a full toe and several points on the
shoulder are obtained in the sensitometer exposure an erroneous
impression may be gained from the curve resulting from machine
development, not only of the characteristic, but also of the gamma
obtained. This is by reason of the fact that whenever the first ex-
posure to meet the developer happens to be on or near the straight
line portion, the abnormal increase in density of the first few steps
will alter the angle of the straight line drawn through the points.
Under these conditions the effect is to raise the apparent gamma when
the shoulder end of the straight line meets the developer first and
to depress it when the toe end leads through the bath.

228

J. CRABTREE

[J. S. M. p. E.

The conclusion to be drawn is that the sensitometer exposure
should be arranged to give a full toe and shoulder, and that where
the range of the sensitometer does not permit this, a full toe should
be obtained and the strip developed with the toe end leading.

A typical example of the errors into which one may be led by lack of
consideration of "directional effect" is that of a particular study of
an alleged change of gamma with printer point. When a sensitome-
ter strip is exposed in the printer a full curve will usually be obtained
at the highest printer point, but as the printer point is decreased, the
shoulder part, then the upper portion of the straight line, disap-

1.2

H0.8

8
•>^

^X^l

DIRECTION THROUGH
DEVELOPER

EFFECTIVE y=0.84

•Jo STEP

Ns^

0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2

LOG RELATIVE EXPOSURE

FIG. 25. "Directional effect" in a simulated sound record.

pears. The resulting strips are now analogous to the case of the
shoulder cut-off shown in Fig. 23. If the strips should happen to be
processed toe end leading, no change in gamma will be noticeable in
the curves; but should the high exposure end lead, there will be
gamma distortion at the lower printer points.

A further case in point relating to field practice is illustrated in
Fig. 24 which shows a directional pair of H & D curves derived from
exposures to positive film on a particular time-scale sensitometer. In
this case an entirely false determination of both gamma and character-

Feb., 1932]

DIRECTIONAL EFFECTS IN PROCESSING

229

istic would be obtained by this sensitometer from a strip processed
with its high exposure end leading through the developing bath.

INFLUENCE OF DIRECTIONAL EFFECT ON THE SOUND RECORD

A frequency cycle of a variable density sound record consists of a
series of gradations of density arranged much like pairs of minute
sensitometer strips with their high densities abutting. It is reason-
able, therefore, to conclude that "directional effect " in a processing
machine will distort the recorded sound wave. In a properly ex-
posed frequency record there should, however, be no shoulder densi-
ties such as are met with in a sensitometer strip. Exposures from a

1.2
1.0
0.8
0.6
0.4
0.2
0
0.6
OA

0.2

HIGH DENSITY LEADING
LOW DENSITY LEADING

[

)IRECTION THROUG
DEVELOPER

.^.

«**'

x^'

*>•

r-'1

•^

S'~

sf'

r^.

>x^

S'-

^^x.

<*"

0.4
0.2
0

,'''

.*"'

FIG. 26. Effect of peak density on "directional effect."

very small step tablet having steps each 0.025 inch long and arranged
so as to give a series of densities simulating a sound wave of 45 cycles
showed a definite difference of slope for the two sides of the wave as
shown in Fig. 25. Other tablets arranged to give a similar condition
with progressively decreasing "peak densities" show the effect to
persist to density values as low as 0.5 peak density. (Fig. 26.)
Microdensitometric measurements of single frequency records reveal
a distortion of the wave shape in the manner which would be antici-
pated from a knowledge of the direction in which they were processed.

CONCLUSIONS

Evidence has been presented to show that in a continuous film
processing machine of the type used in these experiments, the uni-

230 J. CRABTREE [j. s. M. p. E.

directional motion of the film through the developer tends to set up a
current within the developer solution which is parallel to the film's
longitudinal axis, and which is sufficiently strong to dominate the
effect of developer circulation provided by the gravity feed.

This current results in a preferential development of any graduated
series of light exposures such as is presented by the conventional
sensitometer strip used in the control of sound picture development.

This preferential development, referred to herein as a "directional
effect," causes a distortion of the characteristic H & D curves ob-
tained from such sensitometer exposures, unless gamma infinity is
approached or unless means are used to obtain the necessary circula-
tion of developer.

The reason for this preferential development is that the reaction
products from any image area are carried across the succeeding im-
ages by the afore-mentioned dominant current. This trail of reaction
products causes a retardation of the development of the images
over which it flows. The "directional effect" is the more pronounced,
as would be expected from a consideration of the respective time-
gamma curves, when developing motion picture positive film in the
customary "borax" negative developer to a relatively low gamma
than when developing to a higher gamma in a positive developer.
When processing in a machine in which this "directional effect" is
known to exist, more than usual care in the exposure and processing
of the customary type of sensitometer strip is necessary if consistent
and representative results are to be obtained. It is recommended
that constant exposures be given to sensitometer strips such that
a full toe and several points on the shoulder are obtained, and that the
strip be processed in such a way that its toe end is leading through
the developer bath.

Care is particularly necessary in drawing conclusions relative to the
characteristics of the H & D curves as to the shapes of the toe or
shoulder or the limits of the straight line portion, especially when work
of a fundamental nature is involved.

Where the "directional effect" is present in a film processing
machine, it may result in the production of an asymmetric negative
sound wave when the wave is recorded on a high gamma material
such as positive film, and developed to a low gamma in a developer
of the "borax" type customarily used. As a corollary to this last
conclusion, the combination of any high gamma material and low
gamma development will always be susceptible to irregularities in

Feb., 1932] DIRECTIONAL EFFECTS IN PROCESSING 231

development, and for sound recording may for the present be perhaps
regarded only as a temporary but necessary evil.

"Directional effect" may be eliminated by the use of any device
that will maintain a degree of circulation which will overcome the
current set up by the forward motion of the film itself.

For use in machine processing, the arrangement of density areas in
the conventional type of sensitometer strip merits consideration so as
to provide that no density be subject to the influence of the reaction
products of an immediately preceding density. With the present
arrangement, distortion of the H & D curves by "directional effect"
will be less in the case of those sensitometer strips having the larger
physical dimensions and where the steps are as numerous as practic-
able, thus ensuring a smaller density interval between steps.

REFERENCES

1 Zeit.ftir Wiss. Phot. Bd., 27, p. 236.

2 CRABTREE, J. I., AND IVES, C. E.: "Rack Marks and Airbell Markings on
Motion Picture Film," Trans. Soc. Mot. Pict. Eng. (1925), No. 24, p. 95.

RESUME OF THE PROCEEDINGS OF THE DRESDEN
INTERNATIONAL PHOTOGRAPHIC CONGRESS*

S. E. SHEPPARD**

The 8th International Congress of Photography was held at Dres-
den, Germany, from August 3 to 8, 1931, inclusive. Occurring at the
time of a financial crisis in Germany, there was a question at one time
as to the possibility of holding the Congress at this date; but for-
tunately, it was found possible to carry it through, and in spite of
this unfavorable circumstance there was a very large attendance.

The preliminary arrangements for the Congress, and the carrying
out of these by the German committee under Professor R. Luther of
the Technical High School of Dresden, were in the last degree praise-
worthy and successful.

The last day's session of the Congress was held in Berlin. After a
visit to the magnificently equipped and sumptuously decorated new
printing house for periodicals of the world-renowned Ullstein-haus the
members of the Congress were taken in motorbuses to the studios of
the Universum Film A. G. (Ufa) in Neubabelsberg. A very inter-
esting survey was made of both the silent and sound film studios and
laboratories.

The work of the Congress was covered by the following sections:

I. (a) Theoretical bases of photography.

(6) Practice of photography.
II. Cinematography (including the sitting of the Cine-Standards Commission).

III. Applications of photography and cinematography in science and tech-

nology.

IV. History, bibliography, legal, and medical applications.

The members of the Society of Motion Picture Engineers will be
chiefly interested in the proceedings taking place in Sections II and
III. However, under Section I was included the discussion of sen-

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Dr. Sheppard
was the official representative of the S. M. P. E. at the Congress.

** Eastman Kodak Co., Rochester. N. Y.
232

INTERNATIONAL CONGRESS 233

sitometry, which contained several papers and reports of immediate
interest to our Society.

I propose to review' the proceedings of the Congress under the
following headings:

A. Cine-Standards Commission.

B. Sensitometry.

C. Miscellaneous papers of cinematographic interest.

A. ACTIVITIES OF THE CINE-STANDARDS COMMISSION

It should be mentioned that some time prior to the holding of the
Congress certain changes proposed in cinematographic standards by
the Committee of Standards of the Deutsche Kinotechnische Gesell-
schaft had been brought to the attention of the Standards Committee
of the Society of Motion Picture Engineers in order that it might
express opinions on these proposals prior to the meeting of the
Congress. I received letters and some criticisms of the German
proposals from Mr. T. E. Shea of the Bell Telephone Laboratories,
and a joint criticism by Messrs L. A. Jones, J. G. Jones, and E. K.
Carver, and as far as was possible these criticisms were brought to
the attention of the Cine-Standards Commission of the International
Photographic Congress during their sittings and incorporated with
the final proposals.

The final conclusions of the Cine-Standards Commission, copy of
which was sent to the chairman of the Standards Committee of
this Society, and which are included in the report of the Standards
Committee, were provisionally translated by the writer and Mr. W.
Webb of the Eastman Kodak Company. This translation was
based on the German text submitted to us by Dr. Erich Lehmann,
chairman of the Committee. This statement is necessary in view of
any possible discrepancy with the authoritative version which will be
reproduced in the proceedings of the Congress when published, or
with any version produced in the German technical journals. It was
evident from the original proposals and the dimensional drawings,
submitted as Deutsche Industrie Normen (DIN) by the Deutsche
Kinotechnische GeseVschaft that the German proposals represented
an attempt to bring their standard dimensions as closely as possible
in conformity with the dimensional standards of the S. M. P. E.
This, I think, is confirmed by the conclusions of the Committee which
now follow:

234 S. E. SHEPPARD [J. S. M. P. E.

CONCLUSIONS OF THE CINE-STANDARDS COMMITTEE

(I) Perforation Pitch. — It is recommended that the length of film
equivalent to 100 perforations be equal to 475 JI ?;o mm. This is
the same as the standard for normal negative film.

(II) Width of Take-up (Also Feed) Sprocket between Centers of
Sprocket Teeth. — It is recommended that the width of the take-up
(also feed) sprocket measured from tooth center to tooth center be

(III) Over-all Width of Take-up (and Feed) Sprocket. — It is recom-
mended that the over-all width of the take-up (and feed) sprocket be
35.00 ± g;o5 mm.

(IV) Gate Opening (Frame) for Silent Film Projectors. — Recommen-
dation is postponed. The Society of Motion Picture Engineers is
requested to express an opinion on this question since the pamphlet
Dimensional Standards ASA — Z22 — 1930 contains nothing on this
point.

(V) Gate Opening (Frame) for Sound Film Projectors. — Conclusion
is the same as for gate-opening for silent film projectors (See No. IV
above) .

(VI) Tolerance in Standard Specifications. — The S. M. P. E. is
requested to amplify their standard specifications by the inclusion
of definite tolerances in the dimensional specifications. These toler-
ances should be expressed in metric as well as in English units and as
far as possible should conform with the tolerances published in
the German Industrial Standard Specifications (DIN = Deutsche
Industrie Normen).

(VII) Shrinkage. — It is recommended that the shrinkage for
nitrocellulose film should not exceed 1 per cent after being dried by
suspending loosely for 240 hours at a temperature of 40 =*= 1°C. in
air having a relative humidity of 50 to 55 per cent, the air to be
changed once or twice per hour.

(VIII) Discrepancies in S.M.P.E. Standards.— The S. M. P. E.
is asked to express an opinion on the disagreement in the dimensions for
the distance between teeth on sprockets as given in charts No. 6 and
No. 7 of the pamphlet Dimensional Standards ASA — Z22 — 1930.

(IX) Sound Gate for Projector. — The question of the length of the
sound gate (sound slit) in projection machines must be investigated
in all countries. There are differences between the German and the
American standards. Special attention is called to the fact that the
dimensions of the sound slit in projectors must be such that the slit

Feb., 1932] INTERNATIONAL CONGRESS 235

does not extend over the film perforations. Recommendations are
postponed until existing differences are cleared up.

(X) Diameter of Projection Lenses. — It is recommended that a
study of the dimensions of projection lenses be undertaken from the
viewpoint of establishing international standard diameters.

(XI) Definition of Safety Film. — The following definition of Safety
Film is recommended as standardized basis for all codes and regula-
tions in all countries of the world:

(A) Safety Film is a film which is "slow burning" and" difficultly
inflammable."

(B) A film can be considered to be "slow burning" if the burning
time of a piece 30 cm. long is more than 45 seconds. In the case of
film less than 0.08 mm. in thickness the burning time for a piece 30
cm. long must be more than 30 seconds.

(1) A film strip is to be used from which the emulsion has been removed by
washing in warm water, after which the emulsion-free film base is to be dried by
suspension in air at a temperature of 18 ° to 22 °C. and a relative humidity of
40 to 50 per cent for a period of 12 hours.

(2) The test-strip has a total length of 35 cm. At a distance of 5 cm. from
one end a mark is placed.

(3) The test piece is suspended edge upward, if possible, in a horizontal
position between two thin wires which are threaded through the perforations
at intervals of not more than 32 mm. and in such a manner that the perforation
holes utilized on the edge for this purpose do not lie opposite those so utilized
on the other edge, i. e., the threaded holes are staggered. The wire for threading
must have a diameter not greater than 0.5 mm.

(4) The burning time is calculated from the time the flame reaches the mark
5 cm. from the lighted end to the time that the whole strip has been completely
burned. The burning test is to be carried out immediately after the film is dry
and in a room free of draughts. The mean of at least three tests is taken as the
final results.

(D) A film can be considered to be "difficultly inflammable" if,
when tested according to the method given below, it does not kindle
(flash) at a temperature of 300°C. in less than 10 minutes.

(E) Method of making inflammability test:

(1) The test is made in a small electrically heated furnace, the interior of
which has the form of a vertical cylinder, with hemispherical bottom, having a
diameter of 70 mm. and an average height of 70 mm. The opening at the top
of the furnace is provided with a sheet iron cover in which are two symmetrically
placed openings, one having a diameter of about 7 mm. and the other about
15 mm. The distance between the openings is about 15 mm. from center to

236 S. E. SHEPPARD [j. s. M. P. E.

center. The small opening is for the introduction of an iron-constantan thermo-
couple with a porcelain sleeve which just fits through the opening. The measure-
ment can also be made by means of a thermometer for which stem correction
has been made and of which the projecting stem is protected by means of a
cork disk placed around the thermometer a short distance above the furnace
cover.

(2) The piece of film to be tested is hung on a U-shaped wire hook and in-
troduced through the larger opening in the furnace cover. The solder joint of
the thermocouple, or the bulb of the thermometer, and the center of the film
test piece must be at the same height in the furnace which should be about 35
mm. from the top of the furnace.

(3) The piece of film to be tested should be 35 mm. long and 9 mm. wide and
should have the emulsion removed by washing in warm water and drying exactly
in the same manner as that used for preparing a piece for the burning test else-
where described.

(4) Before the introduction of the sample the furnace is brought to a tem-
perature of 300 °C. which must remain comparatively constant, i. e., the variation
should not be more than =tl°C. per minute. At 300 °C. the sample is quickly
introduced.

(5) Before repeating each test the cover of the furnace must be removed and
the products of combustion completely removed by means of an air blast.

(XII) Edge Marking of Safety Film. — It is recommended that,
upon safety film, having a width of more than 34 mm., there be
placed a special characterizing mark which will be visible and recog-
nizable when the film is spooled in the form of a roll. As a means of
accomplishing this it is recommended that the edge of the film be
provided with a thin protective coating which hinders alteration in
the emulsion layer during the subsequent processing operations.

On those matters for which no conclusion was arrived at, such as
No. (IV), the gate opening for silent film projectors, No. (V), for
sound film projectors, it will be noticed that an expression of opinion
is requested from the S. M. P. E. It should also be noticed in regard
to No. (VI) that the Society is requested to change its previous
policy by including definite tolerances in their dimensional specifica-
tions.

The chief difference on any one point from the Society's definitions
is in regard to the definition of safety film. In the writer's opinion a
burning test based on the horizontal burning is more reliable than one
based on the test piece in the vertical position. In connection with
these definitions it should be recalled that Dr. Lehmann's proposals
were made in connection with a meeting of the industries interested
on the one side and the governing bodies in Germany on the other

Feb., 1932] INTERNATIONAL CONGRESS 237

side, in connection with the safety regulations for cinema shows, and
in particular a draft act known as a Narrow Film Act. It may be of
interest to quote its first three paragraphs:

NARROW FILM ACT
Paragraph 1

Narrow films, in the sense understood by this act, are film ribbons which are
intended for taking pictures, writings and the like, and whose width is less than
34 mm.
Paragraph 2

Narrow films must not be easily inflammable nor easily combustible. Easily
inflammable and easily combustible narrow films must not be manufactured
at home or be introduced from abroad.
Paragraph 3

Easily inflammable and easily combustible narrow films must not be brought
into the market nor introduced in the trade after the coming into force of this
act. Also, their application in cinema theaters, public buildings, halls, or picture
palaces is prohibited.

It will be seen that the definition given in regard to the specifica-
tions of safety film refer particularly to paragraphs 2 and 3 by way of
actual definition and interpretation of the terms "not easily inflam-
mable" and "not easily combustible," or alternatively, the terms
"slow burning" and "difficultly inflammable."

The conclusions reached by the Committee are necessarily held for
six months for approval by the national committees for the Inter-
national Congress of Photography.

B. SENSITOMETRY

In regard to sensitometric standardization, several important
developments occurred. First, the other national committees on
sensitometric standardization accepted the light source and filter
proposed by the American Committee at Paris, 1925, and accepted by
the British in 1928. In the meantime, no definite agreement had been
reached, nor indeed had very definite proposals been made on the
subjects of sensitometers or exposure meters, development, density
measurement, and methods of expressing sensitometric results, al-
though much discussion and controversy on this subject had taken
place. At the present Congress, a body of recommendations for sen-
sitometric standards was put forward by the Deutschen Normen-
ausschusses fur Phototechnik, which endeavored to cover the latter
questions and bring the subject of sensitometric standardization
into the industrial field. It was stated by the German committee

238 S. E. SHEPPARD [J. S. M. P. E.

that this action had been forced on them by difficulties arising from
indiscriminate and uncontrolled placing of speed numbers on photo-
graphic sensitive goods, a situation which was summarized at the
Congress by the term "Schemer-inflation."

The gist of these recommendations was as follows:

(a) Acceptance of the light source and daylight filter as proposed by the
American commission.

(6) As exposure meter, a density step-wedge combined with a drop shutter
accurate to V»o second.

(c) Brush development in a tray with a prescribed solution of metol-hydro-
quinone according to a so-called "optimal" development.

(d) Expression of the sensitivity by that illumination at which a density of
0.1 in excess of fog is reached.

(e) Density measurement shall be carried out in diffused light according to
details to be discussed later.

These proposals aroused a very lively discussion. The American
and the British delegations criticized the proposals both as a whole
and in detail. As a whole they considered that the time was not
ripe for application of sensitometric standards to industrial usage.
In matters of detail they criticized the proposed employment of a step-
wedge, and the particular sensitivity number proposed. The latter
approaches very roughly the idea of an exposure for minimum gradi-
ent, but even such a number is not adequate for certain photographic
uses of certain materials.

The upshot of the discussion was that the German proposals in
somewhat modified form are to be submitted simply as proposals of
the German committee for sensitometric standardization to the
various national committees for definite expression of opinion within
six months of the expiration of the Congress. Further, in case of
general approval of these recommendations by the other national
committees, that a small International Committee on Sensitometric
Standardization shall, within a further period of six months, work out
a body of sensitometric practices for commercial usage.

In this connection it should be noted that it was agreed that both
the lamps and filters and exposure meters should be certified as within
certain tolerances by the national testing laboratories of the countries
in question.

BRIEF REVIEW OF PAPERS PRESENTED

It is obviously impossible, as it would be undesirable, to review
in extenso the papers of cinematographic interest presented at the

Feb., 1932] INTERNATIONAL CONGRESS 239

Congress. These papers will be published in full in the Proceedings
of the 8th International Congress of Photography, and it is for the
benefit of those who may wish t6 study them more fully there that I
am giving the following references.

The following papers represent those of general interest to cine-
matography on its technical side, although not necessarily cine-
matographic:

A paper by W. Dziobek, of the Physikalisch-Technische Reichsanstalt, dealt
with the use of the tungsten vacuum lamp in sensitometric measurements. It
points out that for this purpose the following data should be known:

(1) The amperage at which the radiation has a color temperature of 2360 °K.

(2) Light intensity in international candles at the amperage determined by 1.
It was concluded that if the color temperature can be reproduced to within

10 ° the resulting error in actinic intensity amounts to only 0.5 per cent. Curves
were given showing alteration of color temperature of a series of tungsten vacuum
lamps over a lengthy period of burning. The constancy was found sufficient
after a period of running of from 80 to 100 hours. If run for at least 80 hours
at a normal load only exceptionally should a falling-off of 1 per cent occur for
100 hours' further running.

Color cinematography was considered in only two communications, and these
both in the nature of semi-popular, general lectures. Professor J. Eggert of
Leipzig gave a special lecture on the present position of color cinematography
illustrated by examples covering two- and three-color additive processes, two-
color subtractive processes, direct and indirect screen processes.

Mr. Thorne-Baker gave a paper illustrated by examples of the Spicer-Dufay
process of color cinematography. This consists in preparing upon a continuous
band of film base a three-color matrix or "screen" having 900 or more colored
rectangular areas per square millimeter. The "screen" is then coated with
emulsion and exposure is made through the support and "screen." It may be
developed as a negative or as a reversed positive. Methods of making copies
at the standard rate of 800 pictures per minute were described.

In connection with sound film and sound pictures papers of both general and
special interest were presented.

Dr. E. Goldberg, of Zeiss-Ikon, Leipzig, gave an extremely well demonstrated
and illustrated popular lecture on "Fundamentals of the Talking Films."

O. Sandvik and L. A. Jones of the Eastman Kodak Company presented a
review of the talking film.

A paper by H. Thirring dealt with "Sound Reproduction by the Selenophon
Process" which has been developed by the Austrian Sound Film Company.
The modulation of the light beam is effected by a string oscillograph in which
a metallized thread stretched in the field of an electromagnet cuts the real image
of a luminous slit at a small angle and in a position of rest covers half of it. The
telephonic currents from the microphone of the taking studio, after suitable
amplification, are conducted through the thread which is thereby set in oscillation
and modulates the length of the free part of the line of light. By the registration
of this line of light of variable length on a film moved perpendicularly to the

240 S. E. SHEPPARD [J. s. M. P. E.

length of the line there results a variable width sound record. It is stated that
the process has been adapted to the reproduction of a photo-phonic gramophone,
the phonograms consisting simply of paper, copies of sound film records.

A paper by C. R. Keith, of the Western Electric Company of London, dealt
with "Distortion Factors in Sound Reproduction by the Intensity Process."
It was pointed out that the differences which exist between the actual gamma
value of a sound record and the gamma of a sensitometer strip developed at the
same time could be traced to (1) the effect of the different light intensity and
reciprocity failure; (2) the Callier effect; and (3) the color effect resulting
from incomplete correspondence of the spectral composition of the light sources
used. The author described methods for overcoming these difficulties.

R. Thun dealt with "Technical Problems of the After-Synchronization of
Films" (Dubbing). He analyzed the problem as follows:

(1) Determining the desired association of sound and picture.

(2) Approaching the sound sequence as closely as possible to the fixed scheme.

(3) Detection of residual defects.

(4) Removal of these by correction of the picture or the sound sequence.

It is claimed that better results are obtained by corrections applied to the
picture rather than to the sound records.

R. Schmidt, of the Agfa Company, discussed "Ultra-Short-Exposure Sensitom-
etry and Reciprocity Failure in Special Relation to the Making of Sound Films
by the Method of Variable Exposure Time." For periods of illumination from
Vioo to 1/8o,ooo second it was observed on decreasing time of exposure that a flatten-
ing of the gamma value of the characteristic curve occurred. It is of special
importance for the intensity process with variable time of exposure to know
the actual gradation curve of the taking exposure in order to compensate for
distortions. The author has applied the form of representation (formerly given
by Arens and Eggert) of the relation of density to light intensity and time of
exposure by means of density-intensity-time surfaces.

New results in x-ray cinematography were described by K. Jacobsohn, scien-
tific editor of "Photographische Industrie." This dealt particularly with ex-
periments made with Dr. V. Gottheimer of the Pankow Hospital, Berlin. They
were made by the indirect method, namely, cinematographing the image on a
fluorescent screen. The improvements discussed consist in:

(1) Taking camera having a special claw mechanism by which the film is
kept longer at a standstill at the gate at the expense of the time of exposure.

(2) A lens of great aperture //1. 25 consisting of two pairs of cemented glasses.

(3) Special fluorescent screen, resembling an intensifying screen.

(4) Ultra-sensitive film.

The value of x-ray cinematography as compared with subjective observation
of movements of internal organs was discussed.

Of papers of more specialized character the following may be mentioned:
"A Micro-Cinematographic Outfit" described by H. Linke, constructed by the
Askania-Werke of Berlin-Friedenau, and which was on view at the exhibition
associated with the Congress.

F. Beck of the same firm described "Cinematographic and Photographic Meth-
ods for Investigating Rapidly Recurring Processes." The operations of high-

Feb., 1932] INTERNATIONAL CONGRESS 241

speed cinematography were described in detail, as well as the use of rotating
cameras and series cameras. The application of these methods to the study of
explosions, operation of explosion motors, combustion processes, spark phe-
nomena, explosive tests, and lightning were discussed.

Another paper on somewhat the same subject was by W. Ende, of the A. E. G.,
Berlin, entitled "New Results in the Application of High-speed Cinematography
to Technical Research." This discussed the special requirements in regard to
speed and registration in the design of high-speed cine cameras for technical and
scientific research. It was considered that the Thun "Zeitdehner" (time
stretcher) was the best instrument for taking a large series of pictures on a running
band of film. Various modifications and accessories of the Thun "Zeitdehner"
were described, such as apparatus for regulating the speed of taking on the film,
an optical indicator, and an automatic release by the camera. A method of in-
creasing the speed of taking was described which allows the number of pictures
to be increased from 6000 to 30,000 per second. The paper was illustrated by
a film showing the high-speed study of mechanical movements, of arcs and spark
phenomena, with exposures ranging from 1000 to 30,000 per second.

9TH INTERNATIONAL CONGRESS OF PHOTOGRAPHY

At the concluding business meeting of the Congress, the writer,
in the names of the Society of Motion Picture Engineers and the
Optical Society of America, offered a provisional invitation to the
Congress to make its next meeting (1934) in North America. This
proposal was received with much appreciation, but with definitely
expressed doubts as to its feasibility. It is hardly to be denied
that a meeting on this side is desirable. Eight of these Congresses
have now taken place in Europe. The last three post bellum Con-
gresses were held in Paris, London, and Dresden. They have ex-
emplified in their own field the unity of science in western culture,
in the face of national and linguistic differences. If the International
Congress of Photography is to be truly international, and not merely
European, it is essential that it should meet before long on this side
of the Atlantic. Our technical societies, directly or indirectly con-
cerned with photography, and the great American industries of cine-
matography and photography, will assuredly honor themselves and
materially assist photographic advance by helping to bring about an
American meeting of the International Congress. I call to your at-
tention that this would be the first meeting of the Congress under its
new name, since at the conclusion of the Congress it was decided to
change the name of the Congress from the International Congress of
Photography to International Congress of Scientific Photography and
Cinematography. It is my sincere hope that this Society will do all
in its power to make the invitation effective.

COMMITTEE ACTIVITIES

REPORT OF THE PROJECTION SCREENS COMMITTEE*

The first report of the Projection Screens Committee was published
in the September issue of the JOURNAL. It was to have been read
at the May Convention of the Society but unfortunately the copies
shipped to Hollywood by air mail were lost in transit. It dealt
with, in some detail, the manufacture, installation, and maintenance
of screens, and their light-reflecting and sound-transmitting proper-
ties. Curves were given to illustrate the reflection characteristics
for the three types: diffusing, metallic, and beaded. Sound re-
quirements and test methods were discussed at length.

It is, of course, our hope to consider screens from every possible
angle of interest to the Society. At the present time we are able
to report further progress on the program we originally formulated.
We have some data on deterioration of screen surfaces, enough to
indicate that a serious condition exists. The troublesome problem
of determining the optimum illumination for screens has been given
considerable attention, and some interesting information on rear
projection screens, and incidentally rear projection, has been ac-
cumulated. This material follows.

LIGHT REFLECTION

That screens lose their reflective power with use is common knowl-
edge. However, reliable data as to the magnitude of this loss have
never been accumulated. We have made a beginning in this direc-
tion. The few results we have had the time to obtain indicate the
range of variation and the really serious extent of the deterioration.

Our measurements were made with equipment constructed by
one of the members of this Committee. The apparatus consisted
of a metal tube 4 inches in diameter, holding a lamp operating at a
color temperature of 2360 °K. Concentric with this first tube,
there was a second narrower one with a viewing aperture at one

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
242

PROJECTION SCREENS COMMITTEE 243

end, in which it was possible to insert the photometer unit of a Mac-
beth illuminometer. The light source was an automobile lamp
placed so that the angle of light incident upon the screen was approxi-
mately 3 degrees with the normal, which was the viewing angle.
Between the viewing aperture and the screen was placed a blue
filter such as to make the color of the light entering the photometer
correspond to 5000 °K. The light in the Macbeth comparison
lamp was corrected with a similar filter in order to eliminate color
difference in making the photometric balance. A battery of five
dry cells made the apparatus entirely portable and independent of
external power.

Measurements were made hi several convenient theaters. The
device was placed against the screen which was observed through
the photometer inserted in the aperture. Obviously, the data are
restricted to only one angle. It is felt that the loss of reflection
which occurs at one angle will indicate approximately what occurs
at other angles. The results are summarized in the following table.
The original values for these screens ranged from 77 to 85 per cent.

Reflection at 3 Degrees from Normal

Sample Per Cent

A . Broadway Theater

(In use 18 months) . 45

B. Auditorium; New York, N. Y.

(In use occasionally for 3 years) 48

C. Broadway Theater

(In use 18 months; lately reprocessed) 70-80

D. Broadway Theater

(In use 9 months; reprocessed 3 months) 64

E. Hoboken Theater

(In use occasionally 9 months) 80

F. Review Room

(In use 9 months) 76

It will be noticed that deterioration is not very consistent. How-
ever, we should expect it to vary widely, depending on the conditions
surrounding the use of screens. The valuable results obtainable
from surface reprocessing are demonstrated by case C. In addi-
tion to a possible degradation of picture quality there will be a finan-
cial loss accompanying deterioration of reflecting ability. Some
idea of the possibilities may be grasped from the following table.

244 PROJECTION SCREENS COMMITTEE [j. s. M. p. E.

The figures are based on a hypothetical decrease of 20 per cent
in reflection, a serious loss. It is assumed as a first approximation
that this corresponds to a 20 per cent waste of electric power. Other
assumptions are: operation, nine hours daily, and energy cost, five
cents per kilowatt hour. These are common conditions.

Type of Lamp

Amperage

Weekly
Current
Cost

Weekly
Loss

Low intensity arc
Hi-Lo intensity arc
High intensity arc

25
75
120

$ 8.75
26.00
42.00

$1.75
5.20
8.40

Obviously, if projection occurs only a few hours daily or weekly,
the loss is not serious. However, it is not difficult to imagine a case
where replacement of a screen would soon pay for itself by the sav-
ing of power required for illumination. One example is the theater
given as case A above, which runs approximately thirteen hours
every day. Another case when difficulty arises occurs when the
projection outfit is operating at the limit of its capacity and is un-
able to supply sufficient light to overcome the loss of reflective
ability of the screen.

SCREEN ILLUMINATION

Upon the appointment of the .Projection Screens Committee,
President Crabtree stressed the need for recommendations on the
amount of screen illumination required for motion picture presen-
tations. This complex subject has received a great deal of atten-
tion in the past, being one of the oldest of projection problems. Much
scattered work has been done without the achievement of standardiza-
tion or a complete realization of the factors involved. Among
these factors are included the simplification of studio lighting and
printing control as well as projection illumination technic. We
will review hastily a few of the facts that are known, before describing
a series of tests conducted at the meeting of the New York Section,
Friday, September 25, 1931.

Factors that must be considered include visual acuity; flicker;
other physiological factors; fidelity of brightness, contrast, and
tone reproduction; and auditorium lighting. A complete unravel-
ing of these is impossible but we may analyze them to some extent
to obtain a better understanding of the problem.

Feb., 1932] PROJECTION SCREENS COMMITTEE 245

A picture projected on a plane surface will be seen to consist
of a grouping of areas of different brightnesses, that is, it is merely
a pattern of contrasts. The relative brightness of the images must
be presented much as the subjects are in actuality. For our pur-
poses it is not necessary to discuss so much the relations involved
among the areas as it is the intensity of light with which the picture
as a whole is to be projected, that is, the absolute values of bright-
ness. We seek to learn how brilliant bright objects must be, how
dull the dark subjects may be and yet be discernible.

Obviously it is not possible to reproduce on the screen values of
brightness as they occur hi nature. To a large extent this is not
necessary and often not even desirable. One purpose of a motion
picture is to create artificially an impression which will be accepted
as a satisfactory illusion of reality. More than that, it aims to con-
vey a story, using its own devices. The brightness element, together
with size, depth, and color, is secondary, being subordinate to the
story and continuity. Hence, it is not necessary that sky scenes be
shown with clouds as lustrous as clouds are, human faces as bright
as they are in every-day life, deep shadows as profound as they often
are. What is essential is not so much faithfulness to actuality as it
is adaptation of illumination to achieve a smooth vivid portrayal
of the story. This is fortunate in as much as we possess no light
sources capable of producing on a screen brilliancies comparable
with those under direct sunlight. Nevertheless there may be some
instinctive demand for reasonable fidelity in brightness reproduc-
tion.

The pictures on the screen should be easy to see under conditions
of illumination existing in theaters. The projected image should
be the brightest area in the theater to facilitate concentration.
However, in addition to this psychological element, there is another
practical requirement. The auditorium should be provided with
as much light as is consistent with preserving satisfactory detail in
the picture, and the intensity should increase with the distance
from the screen. This light should be sufficient to mitigate screen
glare and permit easy finding of seats. There should be no sudden
change at any point, as sharp contrasts are harmful to the eye. Stray
light falling on the screen must be kept to a minimum in order to
preserve picture contrast. Clearly, if the stray light should equal
the illumination on the screen corresponding to a shadow, the shadow
would disappear.

246 PROJECTION SCREENS COMMITTEE [J. S. M. p. E.

The lower limit of screen brightness should therefore be deter-
mined by the light reaching the screen from the auditorium. There
is no criterion for the maximum desirable amount of illumination
corresponding to the highlights. We do know, however, that with
the auditorium in a darkened condition it would not do to have
too bright a screen as this would be physiologically harmful.

Desirable screen brightness is dependent on all these variables.
Only by analysis of judgments drawn from many observers subject
to varied, controlled conditions will it be possible to determine the
optimum relations.

In an endeavor to obtain more information on this subject, we
conducted our tests at the meeting of the New York Section. This
meeting afforded an excellent opportunity in as much as there was
present a body of trained men who would readily understand our
aims. We did not expect conclusive results from our tests, but
regard them as a preliminary step in the investigation. Obviously,
a complete study of all the factors would require the time of many
men over a period of months.

In these tests we used two projectors, one a hi-lo and one a low
intensity arc, setting these to produce known values of screen illumi-
nation. Two types of arc were employed to determine whether
different color characteristics affect the amount of light judged de-
sirable. There was no illumination in the auditorium other than
that supplied by screen reflection. It would have been interesting
to vary the lighting also, but the time at our disposal necessitated
restriction of the variables. Two reels of film were used, one with
a large percentage of brilliant scenes in it, such as outdoor shots,
the other consisting of interiors, emphasizing human features and
shadows. We wished to learn whether different amounts of light
would be found desirable for different types of subject-matter.

The arc light intensity was varied by means of wire filters inserted
in the projection machine behind the condenser lens. Four settings
were used. The first setting was 68 per cent of the maximum,
which was the second setting. The third and fourth settings were
50 and 25 per cent, respectively. The low intensity machine was
first used for both reels and was followed by the hi-lo intensity arc.
There were present 61 observers, most of whom commented on the
projection on questionnaires which were distributed among them.
Their findings are summarized in the following table. The bright-
ness values are without film and with the shutter running.

Feb., 1932]

PROJECTION SCREENS COMMITTEE

247

Reel 1
Interiors

Reel 2
Exteriors

Low Intensity

Brightness

4.7

Foot

Lamberts
3.5

1.7

4.7

Foot

Lamberts
3.5

1.7

Glaring

Bright

Preferred

Acceptable

Dull

Dark

Hi-Lo

11.5

Foot
17

Lamberts
8.5

4.2

11.5

Foot Lamberts
17 8.5

4.2

Glaring

Bright

Preferred

Acceptable

Dull

Dark

Screen reflection factor: 80 per cent.

Screen size: 9 by 12 feet.

Distance from screen: from 27 to 55 feet

Viewing angle* 90 =*= 30 degrees with screen.

Auditorium illumination: 0.02-0.5 foot candle.

Brightness of screen surroundings: 0.1-0.9 foot lambert.

19 Observers expressed a preference for the color of the low intensity lamp;

17 preferred the hi-lo.
A foot lambert is the brightness of a perfectly diffusing surface illuminated

by one foot candle.

Analysis of Results. — Under the circumstances we cannot be too
positive in our conclusions from these tests. It will be sufficient to
point out tendencies and possibilities. To obtain decisive results
it would be necessary to perform repeated and varied experiments
lasting over a period of time. Admitting the limitations, we may
proceed to interpret the data.

With reel 1 and the low intensity lamp the reactions were just
what might be expected. A brightness of 7 foot lamberts was found
to be quite acceptable. This reel consisted of views of a string
orchestra, the players being dressed in dark, formal clothes. The
brightness on the screen was of the same order of magnitude as those
existing at an actual performance of such an orchestra. Obviously,

248 PROJECTION SCREENS COMMITTEE [J. S. M. P. E.

we do not know that this value would have been preferred to a higher
one, which our facilities did not permit.

The results obtained with this reel and the high intensity lamp are
in fair agreement with those for the low intensity. A brightness
of 17 foot lamberts is too great for such an indoor scene projected
in a darkened auditorium. A value between 8.5 and 11.5 is indi-
cated as perhaps the most acceptable.

Reel 2 consisted of comparatively brilliant outdoor scenes. It
was shown after the reel of indoor scenes and it is supposed that the
first reaction of the audience was to pronounce the illumination
bright. After sufficient time had elapsed for ocular accommodation,
a greater brightness was found acceptable and, in the case of the high
intensity lamp, preferred. The light intensities on the screen were
naturally far below those at which the original scenes would have
been viewed.

One conclusion is that it is necessary to vary the light intensity
for different types of prints, although it is theoretically possible to
select one light intensity and maintain it by recording scenes on a
sliding photographic scale, each value of brightness to have a definite
constant position on this scale. The optimum value of brightness
according to these tests should be a compromise between the ex-
tremes of 7 and 17 foot lamberts, the mean of which is 12. This
is somewhat higher brightness than is customary.

REAR PROJECTION

Historical. — Rear projection is not new; it has been used for
fifteen years in Germany, France, and England. In this country
we are all familiar with the small projectors used in public places
for advertising, demonstration, and stock quotations. Application
to the theater was delayed by two difficulties: one, the lack of a
suitable translucent material, and the other, of an efficient distor-
tionless wide angle lens. Within the past six months several small
theaters have opened in New York to show newsreels and sliort
subjects on a rear projection screen.

Mechanics. — There are several possible materials for use as rear
projection screens. The more common are dental rubber, treated
silk, ground glass, celluloid, and a gelatin composition. The last
is one which is being used on a large scale. Glass screens have
a satisfactory transmission characteristic but the large sizes are
heavy and difficult to protect. Celluloid screens would be satisfac-

Feb., 1932] PROJECTION SCREENS COMMITTEE 249

tory if it were not for their fire hazard. All rear projection screens
have the disadvantage that large uniform areas of material must
be used. They differ from front projection screens in this respect,
for the latter are sewed together from strips of standard width.
The process of manufacture of the gelatin screen is as follows:

On a heated table is poured a hot gelatin solution, over which is
stretched smoothly a fine silk fabric which is pressed into the gela-
tin. The combination is allowed to cool slowly about twenty -four
hours, and is then placed on a rack to dry for seventy- two hours.
Care must be taken to keep water from touching the screens as the
composition is soluble in water. The screens may be cleaned with
alcohol. They can be furnished in any desired color but at present
a slight bluish tint is standard.

Installation. — It may be of interest to point out several facts about
the installation of rear projection apparatus as it is done in the new
small theaters. Standard apparatus is used, two machines being
mounted about 8 feet behind the screen at an angle of 45 degrees
with each other and 22 Vz degrees with the screen normal. Each
lens is approximately 7 inches off the screen axis.

The width of screen that is possible is determined by its distance
from the projection lens. The rule is that 1 foot of width is possible
for every foot of separation between the screen and the 1-inch focal
length lens that is employed, 8 feet in this case.

There is a general impression that film as projected over these
machines must be reversed. This is not so, as a prism is employed
to reverse the image on the screen and to bend the light rays through
an angle of 22l/2 degrees. The prism is mounted immediately
ahead of the negative projection lens.

The screen is mounted about 5 inches above the head of an observer
in the first row. This makes possible the installation of a horn or
baffle loud speaker beneath and on a line with the screen. It must
be pointed out that this position for the speaker is not quite correct
for furnishing the proper illusion, which, however, is yet acceptable
in the front rows to the ordinary observer. At the rear of the theater
the effect is quite good, in as much as the auditorium is small and
sound mixing helps create the correct impression.

One advantage of the rear projection installation may be pointed
out. It requires less vertical space and no specially dimensioned
auditorium. Hence it is possible to employ as theaters enclosures
similar to small stores.

250 PROJECTION SCREENS COMMITTEE [j. s. M. p. E.

Light Transmission. — The light transmission may be varied to
meet different requirements. We have already seen that trans-
mission may be made to favor any particular color. It also may
be made to give several different types of distribution. By proper
processing, the distribution is made more uniform, and hence satis-
factory for viewing at wider angles. It must be expected that there
will be an additional loss of contrast as compared with front pro-
jection because of the introduction of another translucent surface,
which adds to the flare effect.

Illumination.- — Since the screen is light transmitting, the light
intensity in the auditorium can be considerably higher than in the
ordinary theater during a performance. It has been stated that the
auditorium is illuminated to about 30 per cent of average theater
full lighting. Nevertheless, it is necessary to take precaution to
keep light from falling on the screen, in as much as there is some
slight reflection from the surfaces. High auditorium illumination
means that confusion in seating is practically eliminated. For types
of theaters where patrons are continually passing in and out, it is
very desirable to have considerable light. However, it must be re-
membered that a partially lighted auditorium tends to prevent
patrons from "living" through a feature presentation, since it makes
one too conscious of his immediate surroundings. In a theater
showing newsreels and short subjects, this is not objectionable.

For much of the above information on rear projection we are
indebted to Mr. W. Mayer and the Trans-Lux Movies Corporation.

•S. K. WOLF, Chairman
D. S. DE'AMICIS W. F. LITTLE

F. M. FALGE A. L. RAVEN

H. GRIFFIN C. TUTTLE

DISCUSSION

PRESIDENT CRABTREE : The work of this Committee points the way in which
a committee can do real research work. They did not have to have a research
laboratory in which to make these tests. They used the available research labora-
tory, which was the membership of the Society. I congratulate the Committee
on this pioneering effort in cooperative research.

It is very interesting to find that it seems to be necessary to have a greater
screen brightness for the outdoor shots than the indoor ones. On second thought,
it is reasonable. Probably the matter could be taken care of by giving a uniform
flash exposure to the interior scenes, or they could be printed a little heavy.

With reference to the brightness test I should like to point out that the figures
show approximately ten foot candles as the minimum desirable brightness of a
picture. In previous years we have made numerous tests of screens, and find that

Feb., 1932] PROJECTION SCREENS COMMITTEE 251

the average lies between three and five foot candles; and with low intensity light
sources, three foot candles. In the studio laboratories where prints are analyzed
by the studio personnel, intensities of about thirty foot candles are used — ten
times the intensities used in the theaters. This is the cause of dark prints and the
troubles that go with them, such as overloading. Dirty screens also require over-
loading, causing additional loss of picture quality.

MR. FARNHAM: In connection with the data on reflection factors of screens,
the figure of eighty per cent appears to be very high. I should like to ask if that
is absolute reflection, i. e., incident light to total reflected light or is it the ratio
of reflected light after a period of use to that of a new screen?

MR. WOLF: The measurements of reflection factor were made as soon as the
process was completed; each time a comparison was made with the standard.

MR. FALGE: It happens to be the brightness at the normal which is hi ques-
tion rather than the total reflection value.

MR. FARNHAM : That is an extraordinarily high value, and that is why I asked.

PRESIDENT CRABTREE: Have the experiments of the Committee, Mr. Wolf,
gone far enough that we can begin to think of recommending a standard of screen
brightness?

MR. WOLF: The data collected at the demonstration at Bell Laboratories
proves more or less conclusively that there are certain limits to be considered.
We cannot say definitely what they are, but they probably lie between seven and
thirteen foot candles. We definitely believe that any picture having a brightness
less than seven foot candles is certainly too dull ; and any picture having a bright-
ness greater than thirteen foot candles is glaring and disagreeable to look at.

PRESIDENT CRABTREE : Can any one explain why a value of thirty is used in
the screening room?

MR. FALGE: In an article published by Mr. Huse some time ago, describing
some tests of Hollywood screening rooms, he gave the value as thirty foot candles.
It is expected that the intensity will always be high in the screening rooms, unless
deliberate attempts are made to keep it within reason, because the picture is
always small and the light intensity is greater than in the average theater.

PRESIDENT CRABTREE: Were Mr. Huse's measurements strictly comparable
with yours? In other words, has the Committee first of all found a method
of getting an absolute measure of this reflection value? Does your figure of ten
correspond with a similar figure in Hollywood?

MR. FALGE: I think you will find considerable variation among the figures
that have been collected; but I think it is sufficiently important, even with an
error as great as twenty-five per cent, to show that the values in the theater differ
considerably from those in the studios.

MR. FARNHAM: As a result of some work that I did a number of years ago on
screen brightness, I found that there is also a relation between the picture size and
the screen brightness. Smaller screens should be brighter for the same projected
picture, so that whatever intensities we recommend for the studio viewing rooms,
they must be corrected for the size of the picture.

PRESIDENT CRABTREE: That was also observed when we were making wide-
film experiments. We did not need as great a brightness as with the smaller pic-
tures.

MR. GAGE: May I ask if the foot candles are measured with a machine sta-

252 PROGRESS COMMITTEE WORK [J. s. M. P. E.

tionary with the shutter open, or with the shutter running and with no film?

MR. WOLF: We have data under all conditions. In preparing the tests we
made measurements both with the shutter standing still and with it in operation.
We made the measurements, also, of auditorium illumination and other quan-
tities. The screen reflection factor was eighty per cent, and the size of screen
was nine by twelve. The auditorium illumination varied from 0.2 to 0.5 foot
candle. The amperage of the high intensity arc varied from 7 to 4.2; of the
lower intensity, from 7 to 4.7.

PRESIDENT CRABTREE: What were the limits of variation due to screen size?
Do you recall, Mr. Farnham?

MR. FARNHAM: The smallest screen used was approximately four feet and
the largest twenty-two feet, a linear ratio of one to five and one-half. However,
the ratio of brightnesses was more nearly two or three to one, the smaller picture
requiring the higher intensity, but it was by no means an inverse ratio.

PRESIDENT CRABTREE: Suppose a value of seven were required for a twenty-
foot screen, what would be the value for a four-foot screen? Would it be greater
than thirteen?

MR. FARNHAM: As near as I can recall, the smaller picture would require
two to three tunes the intensity ratio.

PRESIDENT CRABTREE: Is the Committee considering the effect of screen size?

MR. WOLF: Yes, it is; but sufficient time was not available.

PRESIDENT CRABTREE: If any of you are in New York I would recommend
that you visit one of the Trans Lux theaters where pictures are projected from the
rear of the screen. The most amazing thing is that the brightness level in the
theater is as high as it is in this room, and yet the picture is adequately bright.

MR. GAGE: With a small screen close by or a large screen far off, both sub-
tending the same angle to the eye, and with the same foot candles of illumination
would not this give equally desirable results on both screens? If so, it is necessary
to relate the distance of the observer to the screen size rather than simply say
that a twenty-foot screen requires so many foot candles, and a thirty-foot screen
so many foot candles, etc.

PRESIDENT CRABTREE : That would depend on the opacity of the atmosphere.

MR. WOLF: We did find a difference in the reactions of viewers as they moved
away from the screen. But the brightness is the same whatever the distance may
be.

PRESIDENT CRABTREE : Not if there is absorption, and the air is full of smoke.

MR. WOLF: That effect is not appreciable.

PRESIDENT CRABTREE : I urge the Committee to push forward the experiments
as rapidly as possible, because I am anxious that our Society should be the first
to propose a definite standard of screen brightness with the necessary qualifica-
tions due to the various factors involved.

ORGANIZATION OF PROGRESS COMMITTEE WORK

For three years the past-chairman of the Committee has assisted in
the preparation of the semi-annual report, and it has occurred to him
that a re'sume' of the program of organization may be of some value

Feb., 1932] PROGRESS COMMITTEE WORK 253

to future chairmen. The following notes represent a description of
the plan of organization of the work of the Committee.

Membership of the Committee. — It is very important in selecting
members of the Committee to choose men who are representative of
various departments of the industry. Such phases of the industry
should include: film manufacture, lens design, camera work, and
sound recording technic, studio illumination, laboratory processing,
sound reproduction, theater construction and operation, and applied
cinematography. Besides representatives in the United States, men
should be selected from each country or part of the world where a
well-developed motion picture industry exists, as well as where re-
search on cinematographic problems is in progress.

The widely separated geographical position of the members of the
Committee makes it unfeasible to hold meetings so that all the com-
mittee work must be handled by correspondence. Each member
should be instructed carefully relative to the scope of the field which
he is to cover in his semi-annual report to the chairman. It is very
desirable to distribute the abstracting work of the Committee mem-
bers, and separate journals which are pertinent to the nature of their
own work should be assigned to each member.

The reports from Committee members may be composed of any one
of the following types of information:

(1) Abstracts of journals.

(2) Personal appraisals of conditions in their specific field.

(3) Answers to specific questions asked by the chairman.

A combination of classes (1) and (2) is the most valuable. The
Committee members should realize that information that may sound
commonplace to them because of their nearness to the source may be
of outstanding interest to other branches of the industry.

Work Preliminary to the Preparation of the Report. — The past-chair-
man of the Committee has found a card file to be the most helpful
means of coordinating the many hundreds of details which require
final mention in the report. The contents of this file are assembled
from three sources, namely: (1) clippings from one or more photo-
graphic abstract bulletins such as the Monthly Abstract Bulletin of
the Kodak Research Laboratories, which contains patents as well as
journal abstracts; (2) abstracts and summaries prepared by com-
mittee members; (3) miscellaneous data obtained from sources other
than those mentioned under (1) and (2). One valuable source of

254 PROGRESS COMMITTEE WORK

information on trade news is the weekly report of the Motion Picture
Division of the U. S. Department of Commerce, Washington, D. C.

From past experience it has been found that illustrations comprise
a valuable addition to the Progress Report, particularly during its
presentation. The interest of the audience may be heightened con-
siderably by the judicious use of lantern slides. A special effort
should be made to secure illustrations of new equipment developed
in foreign countries. Short motion picture films of significant develop-
ments may also be used as a valuable adjunct of the report during its
presentation, as was shown at the Washington Meeting of the Society
in May, 1930.

Preparation of Final Report — All reports from the different members
of the Committee should be in the chairman's hands not later than one
month before the date of the meeting at which the report will be pre-
sented. When these are received, every item of value should be card-
indexed and filed so that, as far as possible, all data to be used in the
report is on cards. It is not feasible in some cases to transfer the in-
formation but the reference to it should be prepared so that the data
may be located with the least possible loss of time.

When all available data have been filed according to a definite
classification, the actual writing of the report should be started. All
references may be made most easily at the time that the material is
written up, rather than after the writing has been finished. When the
first rough draft has been typed, the report should be edited for the
principal items of progress or "highlights" which are to be read at the
meeting. These highlights should not comprise more than 20 per
cent of the total report, and sufficient copies (usually about 30)
should be mimeographed for the use of the Publicity Committee.

The general introduction to the report giving a broad summary of
progress should be written last, after a clear impression has been se-
cured of all the significant developments in the report. Courteous
acknowledgment should obviously be made to all sources of informa-
tion and illustrations apart from those actually supplied by Committee
members. Care should be taken that proper credit is given under
each illustration published with the report.

If the work of the Committee is carried out conscientiously and
thoroughly, this report should become an increasingly valuable
compendium of technical information on the motion picture industry
throughout the world.

G. E. MATTHEWS, Past-Chairman

ABSTRACTS

Studio Practice in Noiseless Recording. GEORGE LEWIN. Electronics, Octo-
ber, 1931, p. 146. The theory of noiseless recording by the light-valve method
was discussed in a preceding article (Electronics, September, 1931). Some modi-
fications must be made in adapting the method to studio practice and special in-
struments must be designed to check the characteristics quickly and accurately.
The author points out one very practical advantage of noiseless recording —
namely, that the average level may be kept lower, thereby reducing the danger of
over-shooting. With the introduction of noiseless recording, however, a certain
amount of background noise that had previously been taken for granted has be-
come more noticeable. This includes noises originating on the stage or in the
theater itself due to the ventilating system or projection machines. A. C. H.

Glow-Lamp Noiseless Recording. E. H. HANSEN. Electronics, November,
1931, p. 177. A description of the method of producing "noiseless" records by the
glow-lamp method. A. C. H.

Ideal Camera Blimp in Daily Use. IRA HOKE. Internal. Phot., 3, November,
1931, p. 27. A new and extremely useful camera casing is reported from the
Educational Studios in Hollywood. It is of cast aluminum and sound insulated.
The new feature is the possibility of pumping the air out with a vacuum pump
whenever conditions demand the extreme in noiseless equipment. Only 25
seconds are required in this process and the method interferes in no way with the
operation of the camera or sound apparatus. A. A. C.

A Standard Aperture for Sound Films. JOHN ARNOLD. Amer. Cinemat., 12,
November, 1931, p. 14. Sound on film destroyed the 3X4 proportion of the
motion picture screen, when it was first introduced. Theaters remedied the con-
dition by using a reduced aperture of the old proportion, thus forcing the pro-
ducer to plan his picture to suit, as well as possible, the various sizes that were
being used in the theaters. This has been accomplished by the expedient of
masking the camera aperture accordingly, and confining the action to that portion
of the film. About twenty per cent of the frame area is not used at all, under
these conditions.

A new standard, 0.651 X 0.868 inch, for camera aperture and 0.615 X 0.820 inch,
for projector, is now proposed by the Academy of Motion Picture Arts and
Sciences. A full report of the proposal is being circulated by Lester Cowan,
Executive Secretary of the Academy. A. A. C.

New Photoelectric Cell. Mot. Pict. Proj., 5, November, 1931, p. 37. A de-
scription is given of the Weston Photronic Cell, which employs a light-sensitive
disk to transform light directly into electrical energy without the use of auxiliary
voltage. It delivers about one microampere per foot candle of light intensity
and the response to light variations is said to be instantaneous. The simplicity
and ease of operation of the new unit are advantages that are expected to lead to
its wide use as an indicator in measurements of illumination. A. A. C.

Rectifying Contact Photoelectric Cells. R. SINGER. Technique Cinemat., 2,

255

256 ABSTRACTS [J. S. M. P. E.

November, 1931, p. 18. It has been known for some time that certain devices,
notably those using copper oxide in contact with metal for rectifying alternating
currents, also possessed the property of developing an electrical potential differ-
ence at their electrodes when radiated with light. The characteristics of two
commercial cells of this type are discussed briefly. Another cell is mentioned
which depends on a needle contact with galena crystal. It is stated that these
cells are rugged and simple in use. They require neither vacuum nor a liquid
electrolyte. C. E. I.

How to Determine the Position of the Pick-up Arm. L. LUMIERE. Technique
Cinemat., 2, November, 1931, p. 4. The author proposes a method of determining
a position for the pick-up arm which minimizes variation in the angle made with
the tangent to the record grooves. The geometrical steps are shown. Reference
is made to an article on this subject which appeared in the preceding issue.

C. E. I.

The Panoramic Motion Picture and the Chretien Hypergonar. H. PICARD.
Technique Cinemat., 2, November, 1931, p. 7. A wide-screen picture can be
obtained with film of normal width by compressing the image in width by the use
of an auxiliary cylindrical lens both in making the negative and in projecting the
positive. This method is open to the objection that the graininess of the negative
shows up in the magnified image of the positive. It is proposed to overcome this
fault by using wide negative film and compressing the image by the use of the
auxiliary lens in the process of projection printing to the fine grain positive. The
illustrations with the article show pictures of the French Colonial exposition
buildings made in this manner. Other applications using this scheme are men-
tioned, such as narrow vertical pictures, and color and stereoscopic processes re-
quiring two or more pictures in the standard frame. C. E. I.

New Sound-on-Film Method. Mot. Pict. Herald, 105,October.24, 1931, p. 11.
This process uses a variable density record on 16-mm. film, having the usual
double rows of perforations and 40 frames to the running foot of film. The sound
record is made on a bias which allows greater width of the frequency band, the
over-all width of the track being 0.025 in. It is claimed to be possible to record
not only at the old silent speed of 60 feet per minute but also as slowly as 32 feet
per minute without volume or quality loss. Reduction prints from 35 mm. film
are planned to form the nucleus of a film library for non- theatrical distribution.

G. E. M.

New Photoelectric Cell. Film Daily, 51, November 22, 1931, p. 6. A highly
light-sensitive disk on the face of this photoelectric cell transforms the light
energy directly into electrical energy without the use of auxiliary voltage. The
cell has an instantaneous response to light variations and relays may be operated
directly from the current generated by the cell. About 1 microampere is delivered
per foot candle of light intensity. When exposed to direct sunlight, the output is
about 5 milliamperes. The cell resistance varies from about 1500 ohms for 10
foot candles to 300 ohms for 240 foot candles. A moulded black bakelite case
2x/4 inches in diameter and 1 inch in thickness encloses the cell. G. E. M.

The Screen: A Problem in Exhibition. BENSCHLANGER. Mot. Pict. Herald,
105, Sect. 2, October 24, 1931, p. 14. With the exception of the progress made in
projection engineering, the author claims that the art and science of exhibition
have advanced very little. The position of the screen, for example, is still being

Feb., 1932] ABSTRACTS 257

determined from the stage floor of the drama theater. The average life of a
theater building should be at least 15 years in order to amortize the initial con-
struction cost and to show a reasonable investment profit. Bodily comfort of the
patron is considered of primary importance in theater design. A maximum screen
size having the ratio of 1 to 1.67 is considered preferable to satisfy various re-
quirements. G. E. M.

A Portable Sound Recorder. Kinemat. Weekly, 177, November 19, 1931, p.
56. A very light and portable sound recording apparatus, capable of being car-
ried in a small automobile, has been developed by a British manufacturer. The
recorder may be fitted to almost any modern camera, provided, however, that the
camera has been silenced for sound work. This comprises changing certain gears
to fit construction, enclosing the shutter drive in a sound-proof casing, and pro-
viding more sturdy bearings for the sprockets.

The recording head and amplifier of this new equipment fit underneath the
camera in a casing which consists of two compartments; the front chamber carries
the sound slit and guide rollers while the rear compartment contains a two-valve
amplifier. The glow lamp projects in front of the forward casing and can be
slipped out to protect it from damage. The lamp is made of Pyrex glass, and
special non-spluttering metals are used for the electrodes, thus minimizing the
risk of the glass turning black. The motor is mounted at the rear of the camera
case and has incorporated with it a tachometer of improved design. The micro-
phone used is of the transverse current type. Ear-phones are provided for
monitoring purposes. C. H. S.

BOARD OF ABSTRACTORS

BROWNELL, C. E. MACFARLANE, J. W.

CARRIGAN, J. B. MACNAIR, W. A.

COOK, A. A. MATTHEWS, G. E.

CRABTREE, J. I. McNicoL, D.

FOWELL, F. MEULENDYKE, C. E.

HAAK, A. H. MUEHLER, L. E.

HARDY, A. C. PARKER, H.

HERRIOT, W. SANDVICK, O.

IRBY, F. S. SCHWINGEL, C. H.

IVES, C. E. SEYMOUR, M. W.

LOVELAND, R. P. WEYERTS, W.

ABSTRACTS OF RECENT U. S. PATENTS

The views of the readers of the JOURNAL relative to the usefulness to them of the
patent abstracts regularly published in the JOURNAL will be appreciated. Favorable
views are of particular interest. In the absence of a substantial body of opinion to the
effect that these patent abstracts are desired by the membership, their early discon-
tinuance may be considered.

1,825,663. Film Reel and Spindle. A. G. HAYDEN. Oct. 6, 1931. A reel
and spindle interlock on slight relative rotative adjustment, thereby to give them
a driving connection and prevent the reel from accidental escape from the spindle.
The film reel comprises a pair of plates and a hub between said plates adapted to
have a film wound thereon, one of said plates having a hole therein and the other
of said plates having an opening with tongues therein projecting toward, but not
to, the center of the plate; and a spindle, in said hole and opening, having a groove
for receiving said tongues to prevent movement of the reel axially of the spindle.

1,825,781. Television Scanning Device. L. H. DAWSON. Oct. 6, 1931.
Scanning disk for television systems in which a rotatable disk is provided with a
plurality of conically shaped light conducting and concentrating members extend-
ing through the disk perpendicularly to the plane thereof. The light concentrat-
ing members are constructed from quartz having a high refractive index for in-
creasing the luminous intensity of the image by concentration of the available
light rays.

1 ,825,953. Device for Permitting the Continuous Feeding of the Film in Project-
ing Apparatus. P. G. H. HALLONGREN. Oct. 6, 1931. Projecting apparatus in
which the reflecting members are divided into at least two groups, which suc-
cessively reflect the picture rays and are positively caused to turn synchronously,
during which operation the active surfaces or the surfaces struck by the picture
rays turn in the same direction, and the said rays pass the reflecting surfaces at
the same side of the axis or axes of rotation through which the said reflecting
surface or surfaces extend or with which the surfaces or surface are substantially
parallel, the said axis or axes having an oblique position with relation to the
plane, on which the incoming rays travel (the plane of the wandering picture).

If two axes of rotation are provided the reflecting surfaces may be located
either round the axes or tangentially to cylindrical surfaces enclosing the axes of
rotation and concentric with the same. In practice the two groups of reflecting
members preferably are located around an axis common to the same and the re-
flecting surfaces of the one group located radially, while the reflecting surfaces of
the second group are located tangentially to a cylindrical surface enclosing the
said axis and concentric with the same.

1,825,955. Synchronized Cylinder Record for Talking Picture. E. S. HAY-
FORD. Oct. 6, 1931. Apparatus for synchronizing a sound record with a picture
record comprising a cylinder of conducting material mounted for simultaneous
movement with the picture record, a sleeve of non-conducting material carried
258

PATENT ABSTRACTS 259

upon said cylinder and having an opening, a stylus mounted for movement along
said cylinder and normally engaging said sleeve and adapted to enter the opening
therein and an electrically operated actuating device connected in circuit with
said stylus and said cylinder for operating said sound record. The sound pro-
ducing means may be rendered operative or inoperative at any predetermined
position with respect to the film being projected, thereby permitting the use of a
record having a limited tone groove length in connection with a greater length of
film.

1,826,305. Scanning System for Television. H. P. DONLE. Oct. 6, 1931. A
scanning system having a speed regulating drive interposed between the scanning
disk and the driving motor. The shaft is formed in two parts, and the speed of
rotation of one part is manually controlled by friction means and regularized by
a ball governor. The other part of the shaft carries the scanning disk and the
angular relation between the two parts of the shaft is adjustable by manually
controlled means independent of the speed controlling means and independent of
the speed of the motor.

1,826,332. Drive Mechanism for Scanning Disk. C. O. VERMILLION. As-
signed to Wired Radio, Inc. Oct. 6, 1931. A drive mechanism for a scanning
disk having means for framing the scanning holes of the scanning disk with respect
to the object to be televised or the picture to be reproduced. The scanning disk
driving mechanism is so arranged that constant speed may be obtained at both
the transmitter and receiver even during periods of adjustment for framing the
apertures in the scanning disk with respect to the picture or object.

1,826,522. System for Avoiding Interruptions of Television Program. F. H.
OWENS. Assigned to Owens Development Corp. Oct. 6, 1931. A plurality of
photoelectric cells are arranged in light paths formed through the film. The cells
operate simultaneously for controlling the input circuit of an amplifying system.
The light which is directed through the film is split into diverging paths toward a
plurality of photoelectric cells so that any one of the cells will continue to operate
for controlling the reproduction of sound in the event of failure of the others so
that there will be no interruption to the sound program.

1,826,680. Motion Picture Projector Cabinet. A. STUBER. Assigned to
Eastman Kodak Co. Oct. 6, 1931. A projector is housed with a sound repro-
ducing instrument in the same cabinet, the projector being mounted on a rota table
support for projecting a picture in any desired direction to the most suitable
location on a portable screen. A phonograph or radio apparatus may be housed
in the cabinet, but is so isolated from the projector that the noises of the projector
are muffled and prevented from interfering with the equipment within the cabinet.
The light rays from the projector within the cabinet are directed vertically through
the cabinet and then projected horizontally in any desired angular direction. The
direction of the beam may be selected by shifting the projector to the desired
angular position within the cabinet structure by means of a crank which engages
the rotatable mount for the projector.

1,826,695. Light-Protected Motion Picture Film. P. FAVOUR. Assigned to
Eastman Kodak Co. Oct. 6, 1931. A light protecting covering is interwound
with the film strip and is normally unperforated, but capable of being perforated
as a film moving mechanism advances the film through contact with the film
perforations. Pasters are provided for attaching the supplementary light-

260 PATENT ABSTRACTS [J. S. M. P. E.

protective covering to the perforated film band, the pasters attaching one end
only of each supplementary light-protective covering to the film band.

1,826,754. Method of Making Photophonographic Records. F. EHRENHAFT.
Oct. 13, 1931. A recording lamp is employed having a luminescent gas discharge
controlled by sound waves, which transform said luminescent gas discharge into
a transitional form of discharge intermediate between a glow and arc discharge.

1,826,786. Sound Controlled Still Picture Protector. P. S.HOPKINS. Assigned
by mesne assignments to Agfa Ansco Corp. Oct. 13, 1931. Projecting appara-
tus for still pictures accompanied by a sound program. The still pictures are
shifted automatically to coordinate the picture with the sound program so that a
picture is projected appropriate to the accompanying sound. The apparatus is
capable of use as a projector accompanied by an illustrated lecture without the
attendance of the lecturer.

1,826,812. Electrooptical Transmission Employing Mirrors instead of Light
Valve. H. NYQUIST. Assigned to American Tel. and Tel. Co. Oct. 13, 1931.
A system for transmitting electrical impulses into light impulses of varying in-
tensities, comprising two plane mirrors having their planes intersecting at right
angles and controlled by incoming picture current at the receiving station, which
mirrors take the place of the usual light valve. The term "90-degree mirror" is
used to designate such an arrangement of plane mirrors. This "90-degree
mirror" rotates about an axis at the line of intersection. The surfaces consist of
alternately reflecting and non-reflecting strips which gradually increase in width
from the line of intersection. The rotation of the 90-degree mirror is controlled
jointly by picture currents received from a transmission line, which currents pass
through a movable coil attached to the 90-degree mirror, and by current from a
local source which passes through a stationary coil, the position of the 90-degree
mirror varying in accordance with the amount of current received from the line.
A constant light source is arranged to project a beam of substantially parallel
rays of light toward the surfaces, the axis of the beam being directed toward the
axis, or intersection line, of the surfaces and at an angle thereto. The reflected
beam from these surfaces is directed to a focal point on a light-sensitive surface,
such as a photographic film. The amount of light reflected by the surfaces and,
therefore, the intensity of the light at the focal point will vary directly with the
angular change in position of the surfaces as controlled by the picture currents
received from the sending station. The reflecting strips on the surfaces may be
so designed as to give a non-linear relation between the light intensity and the
received current strength.

1,826,836. Television Scanning Device. M. STACHO. Oct. 13, 1931. A
television scanning system consisting of a pair of rotatably mounted disks having
co-acting intersecting slots therein for the passage of light rays. One of the disks
has an armature member mounted thereon and associated with an electromag-
netic control for retarding the disk at the completion of each revolution in a
manner to cause the same to rotate periodically at a reduced speed as compared
with the other disk.

1,826,858. Photographic printing apparatus. V. K. ZWORYKIN. Assigned to
Westinghouse Electric and Manufacturing Co. Oct. 13, 1931. A concentric
arrangement of drums for aligning a positive film with a negative film for the
printing of positives from the negative. The light source is directed through the

Feb., 1932] PATENT ABSTRACTS 261

drums and through the negative film adjacent to the outside drum to the positive
film adjacent to the inner drum. The light source, when a reduction in film size
is to be made, is positioned exteriorly of the large wheels over which the negative
film is fed, and the light therefrom, passing through the negative film, falls upon
the surface of the unexposed film carried over the smaller wheels. If the device
is to be used for enlarging, the negative film is fed across the small wheels and the
positive film across the large wheels, the light source being so re-positioned that
the negative film passes between it and the positive film.

1,826,970. Television and Telephoto Device. J. L. WALKER. Oct. 13, 1931.
Picture reproducing system in which two separate scanning systems are directed
upon opposite sides of a reproducing screen. A photographic plate or viewing
screen uses light from two separate light sources and projects light from one light
source upon one side and from the other light source upon the other side of said
photographic plate or viewing screen and the illumination from the two separate
light sources combined at one point. The recording lamps of the two scanning
systems are connected in parallel in the output circuit of the receiving apparatus,
and each so positioned on opposite sides of the screen as normally to give equal
illumination upon the screen.

1,827,010. Film Flame Stop. L. D. KOHLMEYER. Oct. 13, 1931. The film
is protected by a fire-proof frame structure forming compartments surrounding
the film reels. The entrances to each of the compartments are provided with
passageways formed between a pair of rollers carried on fixed axes in the passage-
way. A second pair of rollers is mounted adjacent to each passageway for guid-
ing the film through the passageway and at the same time forming a fire stop in
the event of ignition of the film.

1,827,018. Photoelectric Cell. A. JOFFE. Assigned to Industrial Research
Co. Oct. 13, 1931. A photoelectric cell comprising a sheet-like insulating
layer having a thickness not greater than 0.01 mm. having a photoelectrically
active substance distributed through the insulating layer and a pair of electrodes
supporting the layer, at least one of the electrodes being transparent to light.
The invention is based on the discovery that when an ion is initiated or excited
within certain substances of requisite thickness, notably dielectrics or other ma-
terials of low specific conductivity, and further, when the substance is subjected
to considerable electrical stress, the medium through which the ion travels at
high velocity gives rise to an augmentation of the number of charged particles.
The accumulative action effects a general movement of ions toward one of the
electrodes and results in a greatly magnified space current with abrupt reduction
of impedance to produce amplification of the impulse originally exciting the
single ion. The original impulse may be energy derived from any physical phe-
nomenon such as light, heat, electron bombardment, or other electrical effects.

1,827,206. Film Support for Photographic Apparatus. F. H. OWENS. As-
signed to Owens Development Corp. Oct. 13, 1931. A support for traveling
films, comprising a pair of axially aligned movable members, one of said members
being adapted to engage a film and cause the same to travel over the other mem-
ber. A stationary member is disposed between said movable members and
spaced therefrom to permit the passage of light to said film between said stationary
and movable members and on each side of said stationary member.

1,827,282. System of Composite Photography for Motion Pictures. O.

262 PATENT ABSTRACTS [j. s. M. P. E.

CHOUINARD. Assigned to Motion Picture Improvements, Inc. Oct. 13, 1931.
A machine to produce moving pictures of animated objects and scenic or other
effects wherein the scenic or other effects are recorded in positive, direct, and
accurate relation to the moving objects, without the heavy cost of "locating."
The method comprises making duplicate exposures on two films of moving objects
having actinic properties substantially different from those of the background
therefor, developing one of said films, projecting images from the respective
frames of said developed film successively toward an actinic background, suc-
cessively altering the actinic effect of said background complementary to and in
registration with the respective projected images, and doubly exposing said un-
developed film by subjecting its respective frames to said background as suc-
cessively altered in actinic effect and without substantial effect thereon of the
respective projected images.

1,827,588. Film Drive. E. W. KELLOGG. Assigned to General Electric Co.
Oct. 13, 1931. An improved film driving apparatus in which the film is driven
jointly by a sprocket and a roller or drum and in which the speed of one of said
members is varied in accordance with the amount of film moved by the respective
members as determined by the number of film sprocket holes. A free running
sprocket hole counter is provided engaging that portion of the film moved by the
drum and a variable speed driving mechanism for the drum controlled by the rela-
tive movement of a drive sprocket and the sprocket hole counter. There are
means responsive to a difference in speed of those portions of the film moved by
the respective sprocket and drum members, as determined by the sprocket tooth
openings and independently of the length of film between said members, for vary-
ing the speed of one of said members.

1,827,598. Motion Picture Cabinet. A. G. MERRIMAN. Oct. 13, 1931. The
projecting apparatus is mounted within a cabinet structure having a portion at
one side thereof which may be moved away from the cabinet structure for sup-
porting a projecting screen upon which the picture from the projecting apparatus
within the cabinet structure may be displayed. When the apparatus is not in
use the screen is foldable into a position within the cabinet structure, making a
compact article of furniture for the home or a compact advertising apparatus.

1,827,735. Volume Control in Sound Record Reproduction. J. R. BALSLEY.
Assigned to Fox Film Corp. Oct. 13, 1931. The film bearing the sound record
also carries a volume control record driven in synchronism with the sound
record, and adapted to control the volume level of the sound reproduced from the
sound record. This volume control record may be simply a varying density
photographic record, which may be prepared by reference to the volume level of
the sound record as recorded, as may be determined by ordinary reproduction
thereof. The volume control record, which may be printed on the same film that
carries the pictures and sound record, for instance, outside the sprocket perfora-
tions thereof, or on a separate film if more convenient, is operated in conjunction
with a light beam and photoelectric cell to produce a varying electrical current
which is utilized to control the level of reproduction, and to do this irrespective
of the level at which the sound record was recorded. The photoelectric cell
which is acted upon by the volume control record is connected across the grid and
plate of a vacuum tube, whereby a varying plate current corresponding thereto
appears in the plate circuit of the tube with means for modifying the volume

Feb., 1932] PATENT ABSTRACTS 263

level of the reproduced sound in accordance with the variations in said plate
current.

1,827,924. Picture Copying Process. F. D. WILLIAMS. Oct. 20, 1931. A
method of copying pictures which comprises projecting primary component sil-
houette pictures of ultimate composite pictures upon an opaque picture perceptive
screen and light-impressing a sensitized medium with a supplementary compo-
nent, by aid of the light from said screen with the silhouette projected thereon
so as to produce a latent stencil area. The stencil area is then light-impressed
with a regular picture corresponding to the silhouette.

1,827,947. Synchronizing Mechanism for Disk Reproduction. W. R. MOORE,
JR. Assigned to Deca Disk Phonograph Co. .Oct. 20, 1931. Mechanical link-
age for connecting phonograph and a picture projecting machine for taking up
all lost motion between the mechanism for playing the record and that for pro-
jecting the pictures so that the music and the pictures shall perfectly synchronize.
A worm gear connection is provided with an adjusting device which permits the
taking up of lost motion.

1,828,032. Projection Machine with Optical Intermittent. R. DECAUX.
Assigned to SocietS des 6tablissements Gaumont. Oct. 20, 1931. Projector
wherein the film moves in a continuous manner along an arcuate guide, past a
window lighted by a luminous source which is combined with a condenser. The
film occupies the focal plane of an optical system which sends a beam of parallel
rays' on a mirror which is caused to oscillate about an axis located in its plane.
From that mirror, the luminous rays are directed on a stationary mirror disposed
at 45 degrees, caused to pass through an objective, from which they are projected
on the screen. The oscillating movement of the mirror, which is controlled by a
cam, is synchronized with the forward movement of the band in such a way that,
between successive extinctions produced by a rotary blade acting as a shutter, the
image of a determined point on the film is maintained stationary on the screen.
The chief object of the invention is to provide a mechanical arrangement of the
parts owing to which the oscillating mirror, the support of said mirror, and the
control cam for controlling it are caused to cooperate under the best conditions,
account being taken of the inertia of the different pieces and of the play which is
liable to take place as a consequence of wear and tear. The mirror is fixed on a
platform pivoted to a rocking lever of adjustable position and carrying an arm
which receives the oscillations of the cam. The mirror bears at three points on
the platform and is maintained in place by springs, in such a way as to eliminate
all deformation of the reflecting surface.

1,828,199. Toy Talking Movie Device. F. H. OWENS. Oct. 20, 1931. An
inexpensive form of toy talking picture apparatus wherein an intermittent picture
strip may be moved past a viewing window in timed relation to the movement of
a rotatable talking machine record support. The record carries the sound ap-
propriate to the picture and is maintained at proper operating speed by a governor
device.

1,828,236. Method of Producing Visual Effects. A. C. WATSON. Oct. 20,
1931. A neon lamp illuminating device in which substantially instantaneous in-
termittent illuminations are formed in different positions along a periodic path in
rapid succession through repetitive cycles satisfying the critical frequency for
continuous visual sensation. Visual effects of appreciable duration are produced

264 PATENT ABSTRACTS [J. S. M. P. E.

and modified by interposing a mask between the illuminations and the observer.
An instance of usefulness of this method consists in the fact that by combining
the red color of neon with the yellow color obtained from it as in the "Bezold-
Brucke" phenomenon and also with other types of light such as the neon mercury
tube and by placing before the rotating light a rotating mask which may itself be
colored, so as to reflect daylight, it is possible to secure vari-colored visual patterns.
If the mask referred to be rotated at a slightly different speed from that of the
light, then the colored patterns undergo a series of changes of form, as well as of
color and the total effect may be upon such a large scale as to produce exceedingly
attractive and beautiful patterns of various colors.

1,828,364. Film Contact System Employing Air Pressure. F. E. GARBUTT.
Assigned to Paramount Publix Corp. Oct. 20, 1931. The positive and negative
films are pressed into firm contact by an air pressure system in connection with
the printer and a current of air directed against the films in such a manner that
the films are held in perfect contact against the registering means upon which
they are supported.

1,828,399. Photoelectric Cell Light Ray Condenser. C. W. EBELING. As-
signed to General Talking Pictures Corp. Oct. 20, 1931. A photoelectric cell
light ray condenser is provided for condensing the rays of light after the same
have passed through the sound track of the film and before the same impinges
upon the photoelectric cell, thus insuring higher efficiency in the action from the
cell due to the concentration of the beam of light thereon. A condensing lens is
carried in the light slit block in the path of the light rays before they reach the
photoelectric cell.

1,828,444. Method of Dubbing and Printing. W. ROM. Oct. 20, 1931. A
printer for applying a sound record to a previously prepared picture film, which
consists in utilizing two positive films of the same picture and projecting one
positive film on a screen for guidance in applying sound to a negative film made
from the other positive film of the same picture, driving said other positive of said
film in synchronism with the projected film, masking a portion of said other
positive thereby to provide an area for the sound record, driving a negative film
in synchronism and printing relation with said other positive and with the sound
area of said other positive masked as to said negative, and simultaneously record-
ing sound on the sound area of said new negative, the sound record being applied
to the sound area of said negative in accordance with the projected positive of the
same picture.

1,828,569. Film Stopping Apparatus. E. W. KELLOGG. Assigned to General
Electric Co. Oct. 20, 1931. The projector is arranged to stop the film driving
machine before the record film is completely unwound and disengaged from the
reel on which it has been wound. This is the situation, for example, when in
normal operation the film is rewound on the original reel without removal from
the machine, the purpose of rewinding being to leave the film ready for immediate
use, namely, with the beginning part of the record on the outside.

1,828,571. Picture Transmission System. I. LANGMUIR. Assigned to Gen-
eral Electric Co. Oct. 20, 1931. A light source of the flaming arc type is used
at the picture receiver. The current supplied to the arc lamp is modulated in
accordance with the received signal. The picture at the receiver is projected on
a screen. Spots of light from the arc lamp are projected on the screen but light

Feb., 1932] PATENT ABSTRACTS 265

from the electrodes excluded. This is accomplished by a scanning apparatus
comprising a disk having a series of lenses arranged in a spiral therein and ar-
ranged successively to pass between the lamp and the screen when the disk is
rotated with a motor for rotating the disk in synchronism with a sending appara-
tus. An objective lens is provided and a second disk rotatable with the first-
mentioned disk arranged with a series of holes therein corresponding with said
lenses for excluding from the objective all light emanating from the electrodes of
the lamp.

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.)

BOOK REVIEWS

Handbook of the Film Industry, Vol. II, European Films (Handbuch der
Filmwirtschaft, Band II, Film-Europa). Wirtschaft und Politik, Berlin, 1931,
272 pp.

Three volumes of this handbook of film statistics have thus far appeared. The
first volume covered the period 1923 to 1925, giving a cross-indexed register of
information of film productions, authors, directors, cameramen, architects, and
producers. The history and development of the German motion picture pro-
ducing and exhibiting industry were also traced from 1895 to 1923, together with
an outline of the general film situation in Europe.

The second volume gives correspondingly indexed statistics for films produced
and passed by the censors during 1926 to 1929 with indexes of authors, etc.
Statistics also give information as to the size and distribution of theaters in the
various countries of Europe, regulations pertaining to the importation of motion
picture productions into these countries, the general film market in Europe,
division of sales, etc. The book will be of greatest use to executives, film sales
and distributing organizations doing business with Europe.

A third volume of Handbuch der Filmwirtschaft, dealing with the rise of the
sound film industry and covering the period 1929 and 1930, is scheduled to appear
during 1931.

L. E. MUEHLER

Sound Film Reproduction. G. F. JONES. Blackie & Son, Ltd., London &
Glasgow. 1931.

A brief text in simple, non- technical style explaining, primarily for the small
theater manager and projectionist, the principles and details "of construction of
reproduction equipment for both disk and sound-on-film. The principal outfits
available on the British market are described. Sections are devoted to the vari-
ous parts of the equipment as turntables, pickups, sound heads, light-sensitive
cells, amplifiers, etc. A section on home-designed installations mentions the
chief problems to be met but points out that very little saving can be effected by
such assemblies.

H. PARKER

266

SOCIETY OF MOTION PICTURE
ENGINEERS

OFFICERS
1931-1932

President
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE. Fox Film Corp., New York, N. Y.

Secretary
J. H. KURLANDER. Westinghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y.

Board of Governors

York, N. Y.

W. C. HUBBARD, General Electric Vapor Lamp Co., 'Hoboken, N. J.
R. F. MITCHELL, Bell & Howell Co., 1801 Larchmont Ave., Chicago, 111.
J. H. KURLANDER, Westinghouse Lamp Co. Bloomfield, N. J.
W. C. KUNZMANN, National Carbon Co., Cleveland, Ohio

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLE, 277 Park Ave., New York. N. Y.

267

268

COMMITTEES

[J. S M. P. E

COMMITTEES
1931-1932

(The completed list of committees will be published in a later issue]

W. C. HUBBARD

Convention
W. C. KUNZMANN, Chairman

M. W. PALMER

J. I. CRABTREE
E. J. DENISON
T. FAULKNER

Development and Care of Film
R. F. NICHOLSON, Chairman
R. C. HUBBARD

K. MAClLVAIN

D. MACKENZIE

J. S. MACLEOD
H. RUBIN
J. H. SPRAY

H. T. COWLING
W. B. COOK

Finance

L. A. JONES, Chairman
J. I. CRABTREE
W. C. HUBBARD

J. H. KURLANDER

L. C. PORTER

W. CLARK

Historical
C. L. GREGORY, Chairman

N. D. GOLDEN

O. M. GLUNT

Journal and Progress Medal Awards
C. E. K. MEES, Chairman

E. A. WILLIFORD

Membership and Subscription

H. T. COWLING, Chairman
W. H. CARSON, Vice-Chairman

D. M. BALTIMORE C. D. ELMS J. KLENKE
J. R. CAMERON R. EVANS E. E. LAMB

E. J. COUR E. R. GEIB T. NAGASE

B. W. DEPUE 'J. G. T. GILMOUR E. C. SCHMITZ

B. W. DEPUE
O. B. DEPUE

C, L. GREGORY

Museum

W. E. THEISEN, Chairman
C. F. JENKINS
F, H, RICHARDSON

T. RAMSAYE
A. REEVES
A. F. VICTOR

Feb., 1932]

COMMITTEES

269

Non-Theatrical Equipment
R. E. FARNHAM, Chairman

A. A. COOK N. B. GREEN

W. B. COOK H. GRIFFIN

R. F. MITCHELL

A. SHAPIRO
A. F. VICTOR

J. A. BALL
C. DREHER
P. H. EVANS
A. C. HARDY
N. M. LA PORTE

Papers
O. M. GLUNT, Chairman

G. E. MATTHEWS
P. A. McGuiRE
G. A. MITCHELL
D. McNicoL

P. MOLE
K. F. MORGAN
C. N. REIFSTECK
P. H. REISMAN
T. E. SHEA

H. T. COWLING
J. I. CRABTREE

Preservation of Film
W. H. CARSON, Chairman
A. S. DICKINSON
R. EVANS
C. L. GREGORY

T. RAMSAYE
V. B. SEASE

G. A. CHAMBERS
C. DREHER
W. C. HARCUS

Progress

J. G. FRAYNE, Chairman
G. E. MATTHEWS
M. W. PALMER

G. F. RACKETT
H. SINTZENICH
S. K. WOLF

J. O. BAKER
T. BARROWS
W. H. BELTZ
G. C. EDWARDS
S. GLAUBER

Projection Practice
H. RUBIN, Chairman
J. H. GOLDBERG
C. GREENE
H. GRIFFIN
J. HOPKINS

R. H. MCCULLOUGH

P. A. McGuiRE

R. MlEHLING

F. H. RICHARDSON
M. RUBEN
P. T. SHERIDAN
L. M. TOWNSEND

J. L. CASS
H. GRIFFIN

J. H. KURLANDER

Projection Screens

S. K. WOLF, Chairman
W. F. LITTLE
A. L. RAVEN

H. RUBIN

L. T. TROLAND

C. TUTTLE

L. DEL RlCCIO

Projection Theory
A. C. HARDY, Chairman

W. F. LITTLE

270

COMMITTEES

F. C. BADGLEY
B. W. DEPUE

Publicity

W. WHITMORE, Chairman
D. E. HYNDMAN
F. S. IRBY

G. E. MATTHEWS
D. McNicoL

M. C. BATSEL
P. H. EVANS
N. M. LA PORTE

Sound

H. B. SANTEE, Chairman
E. W. KELLOGG
C. L. LOOTENS
W. C. MILLER

H. C. SILENT
R. V. TERRY
S. K. WOLF

L. E. CLARK
L. DE FOREST

J. A. DUBRAY

P. H. EVANS
R. E. FARNHAM
H. GRIFFIN
A. C. HARDY

L. J. BUTTOLPH
R. E. FARNHAM

Standards and Nomenclature
M. C. BATSEL, Chairman
R. C. HUBBARD
L. A. JONES
N. M. LA PORTE
D. MACKENZIE
G. A. MITCHELL
G. F. RACKETT

Studio Lighting
M. W. PALMER, Chairman
C. W. HANDLEY
K. C. D. HICKMAN

W. B. RAYTON
C. N. REIFSTECK
V. B. SEASE
T. E. SHEA
J. L. SPENCE
E. I. SPONABLE
L. T. TROLAND

J. H. KURLANDER

E. C. RICHARDSON

R. S. BURNAP
W. H. CARSON

Ways and Means
D. McNicoL, Chairman
H. GRIFFIN
F. S. IRBY

J. H. KURLANDIiR
J. A. NORLING

Chicago Section

R. F. MITCHELL, Chairman R. P. BURNS, Manager

B. W. DEPUE, Sec.-Treas. O. B. DEPUE, Manager

New York Section

P. H. EVANS, Chairman M. C. BATSEL, Manager

D. E. HYNDMAN, Sec.-Treas. J. L. SPENCE, Manager

Pacific Coast Section

D. MACKENZIE, Chairman C. DREHER, Manager

W. C. HARCUS, Sec.-Treas. H. C. SILENT, Manager

CONTRIBUTORS TO THIS ISSUE

Frederick, H. A.: B.S., E.E., Princeton University; engineering department,
Western Electric Company, 1912-25; transmission instruments director, Bell
Telephone Laboratories, 1925 to date.

Schlanger, B.: See August, 1931, issue of JOURNAL.

Sheppard, S. E.: Born 1882 at Hither Green, Kent, England. D.Sc., Uni-
versity of London, 1906; colloid chemist, Eastman Kodak Company, 1913-26;
chief of department of physical and inorganic chemistry, 1920; acting director
of research, 1922-23; assistant director of research, 1924 to date.

Tuttle, C.: Born March 7, 1898, at Evansville, Wis. B.A., University of
Wisconsin, 1922; graduate assistant at University of Wisconsin, 1922-23;
instructor in physics, University of Georgia, 1923-24; physicist, Kodak Research
Laboratories, Eastman Kodak Company, 1924 to date.

Tuttle, W. N.: A.B., Harvard University, 1924; S.M. in electric communica-
tion engineering, Harvard University, 1926; Ph.D., Harvard University, 1929;
instructor in physics, Harvard University, 1929-30; engineer, General Radio
Company, 1930 to date.

271

SOCIETY ANNOUNCEMENTS

SPRING, 1932, MEETING

By action of the Board of Governors at a meeting held on Decem-
ber 10th at New York, N. Y., and subsequent verification of this
action by the post-card ballot mailed to the membership for de-
termining the location of the Spring, 1932, Meeting, the location of
the latter was determined as Washington, D. C.

The meeting is to be held from May 9th to 12th, inclusive, with
headquarters at the Wardman Park Hotel, in Washington. The
technical meetings will be held in the Little Theater of the Hotel and
the semi-annual banquet in the Gold Room. The Convention
Arrangements Committee, under the chairmanship of Mr. W. C.
Kunzmann, is working on an attractive and interesting program for
the Convention, and the Papers Committee, headed by Mr. O. M.
Glunt, is bending every effort toward making the technical sessions
of outstanding interest.

Mr. N. D. Golden, of Washington, has been appointed Chairman
of the Local Arrangements Committee, and in this capacity is
assisted by Messrs. C. Francis Jenkins, Raymond Evans, C. N.
Nichols, N. Glasser, C. J. North, and N. C. Haefele. As the Con-
vention is to be held at the time of the Washington Bi-Centennial,
there will be much in Washington to attract the members of the
Society to the Convention, in addition to the technical activities of
the Society.

An exhibit will be held of newly developed motion picture appara-
tus, similar to the exhibits held at the Hollywood and Swampscott
Conventions. This exhibit is to be under the direction of Mr. H.
Griffin. As in the past, it will not be of the nature of a
trade exhibit nor will there be booths, but adequate space will be
allotted each exhibitor free of charge. The exhibition rules specify
that equipment be new or have been improved within the past
twelve months. No pamphlets or advertising literature will be
permitted. Each exhibitor will be allowed to display a small card
giving the name of the manufacturing concern, and each piece of
equipment will be labelled with a plain label free of the name of the
272

SOCIETY ANNOUNCEMENTS 273

manufacturer. It is required that a technical expert be present
during the exhibition to explain the technical features of the ap-
paratus.

Requests for space should be made to Mr. Sylvan Harris, editor-
manager of the Society, room 701, 33 W. 42nd Street, New York,
N. Y., stating the number and nature of the items to be exhibited.

STANDARDS COMMITTEE

At a meeting of the Standards and Nomenclature Committee,
held at the General Office of the Society on January 9th, Mr. A. C.
Hardy was appointed chairman of the sub-committee on the glos-
sary. Questions on the standardization of camera motors, aper-
tures, camera mountings, and adapters were discussed, and general
ideas concerning the disposition of these matters were outlined.
The proposal made by the Projection Practice Committee, calling
for the dimensions 0.590 X 0.825 inch for the projector aperture,
was approved by the Committee.

The question of standardization of screen brightness was given
considerable study, and it was finally agreed that the Projection
Practice Committee, in collaboration with the Projection Screens
Committee, should study the problem and recommend to the Stand-
ards Committee, the values of brightness which will indicate the
limits between which a picture may be considered reasonably satis-
factory under existing practical conditions. These values would
not be susceptible of standardization, but would merely represent
recommended good practice.

Mr. L. A. Jones was appointed chairman of a new sub-committee
on sensitometry, and Mr. J. L. Spence was appointed chairman of
the sub-committee to deal, with matters relating to the standardiza-
tion of 16-mm. sound-on-film equipment. Considerable thought was
given to the dimensions of the 16-mm. sound film and the location
of the sound track, etc., and plans were made for enlisting the assis-
tance of the manufacturers in making the study, which requires a
practical knowledge of possible tolerances and practical circum-
stances of manufacture.

In connection with the questions raised by the Cine-Standards
Committee of the International Congress of Photography, recently
held in Dresden, Germany, it was decided that the matter of specifi-
cations of safety film is to be reopened at a later meeting of the
Committee.

274 SOCIETY ANNOUNCEMENTS [J. S. M. p. E.

WAYS AND MEANS COMMITTEE

At a meeting of the Ways and Means Committee, held at New
York on January 9th, under the chairmanship of Mr. D. McNicol,
many of the factors involved in readjusting the rates of subscription
for the JOURNAL, the entrance fees of new members, and dues were
discussed, and recommendations concerning the reduction of the
entrance fees and subscription rates were agreed upon for presenta-
tion to the Board of Governors.

JOURNAL BINDERS

Feb., 1932] SOCIETY ANNOUNCEMENTS 275

MEMBERSHIP CERTIFICATE

Society Motion Picture Engineers

THIS IS TO CERTIFY THAT

Society of Motion Picture Engineers

LAPEL BUTTONS

SUSTAINING MEMBERS

Agfa Ansco Corp.
Bausch & Lomb Optical Co.

Bell & Howell Co.

Bell Telephone Laboratories, Inc.

Carrier Engineering Corp.

Case Research Laboratory

Eastman Kodak Co.

Electrical Research Products, Inc.

Mole-Richardson, Inc.

National Carbon Co.

RCA Photophone, Inc.

Technicolor Motion Picture Corp.

BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS

1924

No.

Price

1917 { I

$0.25
0.25

1918 7

0.25

1920

10
11

1.00
1.00

1921

12
13

1.00
1.00

1922

14
15

1.00
1.00

1923

16
17

2.00
2.00

No.

Price No.

Price

$2.00

$1.25

1.25
1.25

1927

30
31

1.25
1.25

1.25

2.50

23
24

1.25
1.25

1928

34
35

2.50
2.50

1.25

2.50

26
27

1.25
1.25

1929

37
38

3.00
3.00

1.25

1926-

Beginning with the January, 1930, issue, the JOURNAL of the Society has been
issued monthly, in two volumes per year, of six issues each. Back numbers of all
issues are available at the price of $1.50 each, a complete yearly issue totalling
$18.00. Single copies of the current issue may be obtained for $1.50 each.
Orders for back numbers of Transactions and JOURNALS should be placed through
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and
should be accompanied by check or money-order.
276

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR
Volume XVIII MARCH, 1932 Number 3

CONTENTS

Page

Two Special Sensitometers D. R. WHITE 279

The Decibel in the Motion Picture Industry V. C. HALL 292

Optical Instruments and Their Application in the Motion Pic-
ture Industry I. L. NIXON 304

Photographic Sensitometry, Part IV LOYD A. JONES 324

Stroboscopic and Slow-Motion Moving Pictures by Means of

Intermittent Light H. E. EDGERTON 356

Sound in the Los Angeles Theater — Los Angeles, Calif

D. M. COLE 365

The Reducing Action of Fixing Baths on the Silver Image

H. D. RUSSELL AND J. I. CRABTREE 371

Abstracts 398

Patent Abstracts 403

Officers 407

Society Announcements 408

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers.

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc.

Subscription to non-members, $12.00 per annum; to members, $9.00 per annum,
included in their annual membership dues; single copies, $1.50. A discount
on subscriptions or single copies of 15 per cent is allowed to accredited agencies.
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930, at the Post Office at Easton,
Pa., under the Act of March 3, 1879.

TWO SPECIAL SENSITOMETERS*
D. R. WHITE**

Summary. — Design features of two sensitometers are presented. One of the
sensitometers is used to make sensitometric tests on positive film under print-
ing conditions. The other is designed to produce exposures under sound recording
conditions. The results of tests with both of these sensitometers emphasize the im-
portance of making sensitometric tests parallel conditions of film use, and show some
of the errors that occur in judging speed and contrast from sensitometric data obtained
under conditions not corresponding to the actual conditions of use.

Sensitometric workers have found it desirable to test film under
the conditions of its use. The time scale sensitometers frequently
used are not representative of printing conditions where the positive
is always exposed through a negative, nor are they ordinarily arranged
to correspond to the conditions of sound recording where the light
comes to the film through an optical system and has higher intensity
and shorter exposure time than is readily obtained with the usual
form of sector wheel. The two sensitometers herein described were
designed, therefore, to obtain accurate sensitometric information
under actual conditions of printing and sound recording.

H & D PRINTER

This machine makes use of a negative, for instance, an exposed and
developed H & D sensitometric strip, to produce a series of graded
exposures for testing a positive material. In this test, then, the ex-
posure is a photographic printing operation such that the results may
be plotted as a characteristic curve for the material tested. Provision
is also made in this machine for comparison with pictures exposed
and developed under similar conditions.

A schematic view of the mechanism of the machine is shown in
Fig. 1. The exposure timing shutter is driven by a synchronous
motor through a pinion on the motor shaft which meshes with a
large ring gear mounted directly on the rim of the shutter disk. Two

* Presented in the Symposium on Laboratory Practices at the Spring, 1931,
Meeting at Hollywood, Calif.

** Du Pont Film Manufacturing Co., Parlin, N. J.

279

280

D. R. WHITE

[J. S. M. P. E.

sectors are cut from this disk, one having radial sides, 30 degrees
apart, and the other having one straight side and one side stepped
as shown, such that each step has ten per cent greater angular opening
than the preceding, these angles being so adjusted that the center
step of the series has a 30-degree opening. Geared to this exposing
shutter through a countershaft, with a total reduction of 4 to 1, is an
auxiliary shutter which has only one sector cut from it. Two light
houses and lights, two exposing gates, and two hand shutters are pro-
vided, forming two independent exposing systems in which the ex-
posure time is controlled by the one motor and shutter disk mecha-
nism. The hand shutters in each system are not intended to control

Ltghf

^Exposing.
Jhutter

•Step Test Hand
•Shutter and
i Exposing Gate

Hi-2) Print Hand
Shutter and
/Exposing Gate

Shutter-

FIG. 1. Schematic view of H & D printer.

the exposure time at all but are used to prevent undesired double
exposures. The positions of the lights are adjustable by moving the
entire lamp support along rails provided in the light house. This
motion allows a satisfactory inverse square law intensity variation of
8 to 1 in the machine as constructed. The lamp supports can be
removed from the housings intact, and placed on a bar photometer
for color temperature and intensity measurements.

This entire mechanism operates in such a manner that only ex-
posures from the 30-degree exposing shutter opening for sensitometric
tests are produced at one gate, and only exposures from the stepped
sector for picture printing are produced at the other. With the motor
speed and gear ratio used, the 30-degree shutter openings give expo-

Mar., 1932]

Two SPECIAL SENSITOMETERS

281

sures of one-eighteenth of a second, which is representative of printing
exposures.

A photograph of the machine is shown in Fig. 2. The two exposing
gates are visible in the foreground. The light houses are about five
feet long and extend to the rear from the compartment which houses
the shutter disks. The countershaft and gears are also back of this
shutter compartment and between the two light houses. One of the
hand shutters and the signal light used as a guide in its operation ap-
pear on the side of the machine.

FIG. 2. Photograph of H & D printer showing the two exposing gates.

The lamp support, removed from its housing and viewed from the
rear, is shown in Fig. 3. With the adjustments provided, the lamp
chosen for use may be centered, rotated on its own axis, or about an
axis perpendicular thereto. These motions enable the selection of the
lamp position which gives the best uniformity of intensity over the
printing area of each exposing system.

These design features were adopted to permit making sensitometric
tests and picture tests under similar carefully controlled conditions.
The sensitometric tests are made by exposures through a negative,

282

D. R. WHITE

[J. S. M. P. E.

and are so arranged that curves showing both negative and positive
characteristics, as they are effective in the printing operation, are ob-
tained as a result. It is not necessary to know either the positive
or negative characteristic from other tests, for both are obtained from
these exposures and may be compared with the photographic char-
acteristics of the materials as shown by other methods of testing.
The negative used in this H & D print operation may conveniently
be in the form of an exposed and developed sensitometric strip.
Those used in the tests here discussed were the result of time scale

FIG. 3.

Lamp support used in H & D
printer.

sector wheel exposures, with exposure times increasing by factor two
from area to area of the negative.

Consider, now, two exposures made through such a negative on
this machine, in which the intensity of light incident on the negative
is twice as great for the second exposure as for the first. The strips
thus exposed are developed together and the resulting densities, read
photometrically, are plotted against the log E values for the original
negative exposure. Curves drawn through these points are shown
in Fig. 4, where the curves labeled L and H represent the first and
second exposures, respectively. Such curves, called reproduction

Mar., 1932]

Two SPECIAL SENSITOMETERS

283

curves, represent the relation between the density of the positive and
the log E value for the negative. Theoretically perfect tone repro-
duction would require these to be straight lines with slope minus one.
Practically, this condition is neither attained nor desired, since more
pleasing and satisfactory pictures are produced otherwise. How-
ever, this point of tone reproduction, theory and practice, is outside
the scope of the present paper and is only mentioned, since these

/.o

as-

0 .J .6 .7/2 AT /.3 2.1

FIG. 4. Two reproduction curves. Points A and B have received
equal exposure. Point C has received one-half the exposure of B.

reproduction curves are the first and most direct result of the opera-
tion of this H & D printer.

The two exposures represented by L and H had equal exposure
times, hence it follows that any two points of equal density on the
two curves, such as points A and B of the figure, received effectively
equal intensities during the printing exposure. However, these
equal printing intensities were produced under different conditions.
For point A the effective printing intensity /^ is given by

I A = IL X TA

284 D. R. WHITE [j. s. M. P. E.

where IL is the intensity incident on the negative during the ex-
posure represented by curve L, and TA is the transmission of the
negative at a point corresponding to A . • Similarly, for the point B,
the effective printing intensity IB is given by

IB = IH X TB

where IH is the intensity incident on the negative during the second
exposure, and TB is the transmission of the negative at the point
corresponding to B. Since, as has already been pointed out for
these points A and B,

I A = IB
Therefore

IL X TA = IH X TB
or

TL = TB = 2

— was originally made 2

But the difference in the effective printing densities of the negative
between B and A, &DBA is given by

&DBA = log ^ = log 2 = 0.3

1 B

It is also apparent from the method of exposure and plotting used
for these two curves that any two points such as B and C of the figure,
occurring at the same value on the logE scale, have densities produced
by exposure intensities differing by factor two. This may be shown
by the fact that, using notation similar to that above,

IB = IH X TB
Ic = IL X TB

TB occurs in both of the first two equations since it is the same point
in the negative that is effective for both points B and C.

Using these facts in the manner shown in Fig. 5, a 'series of points
having the coordinates (D0, log E0), (A, log Ej), (Dz, log £2), . . .
(Dn, log En) may be found, in which the series of D's, DQ to Dn,
are the densities produced on the positive by printing exposures de-

Mar., 1932]

Two SPECIAL SENSITOMETERS

285

creasing by factor two in intensity from density to density. These
density values may be plotted, then, at uniform spaces, log (EpOS.)

FIG. 5. Reproduction curves, and positive and negative characteristic
curves resulting therefrom.

differences of 0.3, to produce a positive characteristic curve represent-
ing actual printing conditions. This is shown as the curve marked

FIG. 6. Four reproduction curves may be used to increase the precision of

the results.

"positive" in the figure. The characteristic curve for the negative
may be obtained by plotting at the series of values log EQ to log En

888

D. R. WHITE

Q. a M. P. E.

on the log (E^) scale, density values increasing by uniform steps
of 0.3 from one log £ value to the next, as shown in the curve marked
"negative" in the figure. The effective printing density for any
other log E value can be determined by interpolation from this curve.

To improve the precision of these data, it has been more satis-
factory to use four reproduction curves, as shown in Fig. 6, in deter-
mining these characteristics. The results so obtained are less liable
to error, since with the greater amount of data no single point is as
important.

The negative and positive characteristic curves corresponding to

flegrfne Ounxferistic

Visual Diffuse Density
Effect iv* Pinnhng

Density

Densrty

LogE

O .J .6 .9 J2 tf tB 2i

FIG. 7. Comparison of negative characteristics as obtained
visually and by H &

printing conditions have thus been determined without recourse to
photometric readings of negative densities as intermediate data.

The effective printing densities of the negatives tested to date have
always been very close to their diffuse light densities as read on a
photometer. Fig. 7 shows the characteristic curve of a negative as
determined by diffuse density photometric readings and as obtained
by this printing method. In both cases the fog reading is effectively
subtracted by the methods used. No large density difference is
shown here for this negative, which was developed in a metol-borax
developer.

The sensi tome trie characteristics of certain positive emulsions as
determined by the procedure just outlined have been found to be dif-

Mar., 1932]

Two SPECIAL SENSITOMETERS

287

ferent from those shown by tests with the more common form of
sensitometer, the time scale sector wheel. No general conversion

O.S

De-nsi ty

Ttelattve Log I

Sector Wheel
Exposures

0 ,J .6 .? 12 IS /£ 3.1 5-t 3.7 2.0 23 26 21 12 /f 78 01

FIG. 8. Curves showing sensitometric characteristics of a positive emul-
sion as determined under different conditions of exposure. This emulsion
obeys the reciprocity law within the range of the test.

I.S

HtD Printer
Exposures

Density

Relative Log I

Sector Wheel
Exposures

0 3 .6 .9 1.2 I.S 18 S, j.4 37 2.0 5.J 26 21 7.2 U 78 ./

FIG. 9. Curves showing sensitometric characteristics of a positive emul-
sion as determined under different conditions of exposure. This emulsion
shows reciprocity law failure.

factor from one type of exposure to another is possible, since not all
emulsions show the same variation in characteristic.

Fig. 8 shows the characteristics of one positive emulsion as shown

288 D. R. WHITE [J. S. M. P. E.

by a group of tests. The six curves under the heading "sector wheel
exposures" show time scale characteristic curves determined at six
exposure intensity levels, increasing right to left by factor two from
curve to curve in the group. The printing characteristic is shown by
itself, as determined by the H & D printer. No failure of the re-
ciprocity law is observed in the range covered by these data. The
contrast of the material appears independent of the exposure, and
therefore only dependent on the development which the film received.
However, this condition of constancy of contrast does not hold
for the positive emulsion represented by Fig. 9. This emulsion was
prepared under different conditions from the first. Here, there is a
noticeable reciprocity law failure even within the range of time
intensity variations shown in the sector wheel tests. Gamma de-
pends on both the time and intensity used in its determination. The
contrast values shown by the sector wheel curves made with the higher

Objective

V_y Condenser

Hand Shutter

Ribbon nhmenf Lam/3

FIG. 10. Schematic diagram of the optical system of the high
intensity sensitometer.

exposure levels become closely equal to that shown by the H & D
printer, but differ appreciably from the values obtained at lower
intensities.

HIGH INTENSITY SENSITOMETER

The second of the two special machines here discussed was designed
to test film under conditions corresponding to those of sound record-
ing.

The optical system of this machine is shown schematically in Fig.
10. A filament image from a ribbon filament light is formed at the
slit by a condenser lens. The slit in turn is imaged on the photo-
graphic film carried on the film drum by a microscope objective. A
series of fixed slits of various widths is provided to obtain the exposure
range desired. This construction was used in preference to any form
of calibrated light valve as its constancy from time to time is assured

Mar., 1932]

Two SPECIAL SENSITOMETERS

without difficulty. This group of slits is mounted on a circular plate
driven from the film drum shaft through an intermittent motion and
gearing. The final accurate positioning of the slits in the correct
position in the optical system is assured during operation of the
machine by a wedge fitting a V-groove in the rim of the slit carrying
plate. The wedge is disengaged by a cam when the intermittent
motion is about to operate and is re-engaged before exposure. The

FIG. 11.

Photograph of the high intensity sensitometer showing the film
drum, slit carrying plate, and optical system.

mechanism is adjusted to make a complete series of eleven exposures
on a single turn of film fastened to the film drum. The hand shutter
placed in the optical system is used to prevent double exposure and
to permit the machine to come up to speed, 90 feet per minute of the
film, before any exposures are made.

Fig. 11 presents a view of the machine showing particularly the
film drum, slit plate, and the optical system. The disk with the small
sector cut from it is an auxiliary shutter which limits the exposures

290

D. R. WHITE

[J. S. M. P. E.

to the desired portions of the film being exposed and which cuts off
the exposure during the motion of the slit carrying plate.

Fig. 12 is taken from a different viewpoint and shows more clearly
the intermittent motion and slit positioning mechanism.

In making the machine, difficulty was anticipated and experienced
in making the slits to exact spacings. Rather than spend too great
an amount of time on an unnecessary detail, the slits were set at
approximately the values desired and then finally calibrated in place

FIG. 12. A view of the high intensity sensitometer showing some of the

mechanism.

by measuring their light transmission with a photoelectric cell.
These slits vary from approximately 0.00025 to 0.008 inch. With
eleven slits the average step is about factor \/2, but the steps are
not entirely uniform.

A comparison of the results of tests with this machine and with a
time scale sector wheel with much lower intensity and longer expo-
sures is shown in Table I. In this comparison the development was
carried to a point representing variable density recording condi-
tions. Compared emulsion by emulsion, the gammas produced by

Mar., 1932] TWO SPECIAL SENSITOMETERS 291

the two methods of exposure are not widely different, though with
two emulsions the sector wheel gamma is higher. In the relative
emulsion speeds, as determined by each experimental method, larger
differences are found. In the table, emulsion A is arbitrarily taken
as having a speed of unity in both tests, and the others are evaluated
in comparison with it.

TABLE I

Relative Speed and Contrast Values as Obtained by Exposures with a Sector Wheel
and with the High Intensity Sensitometer

Gamma Relative Speed

Emulsion

Type

Developed

H.I.

S.w.

H.I.

S.W.

Positive

21/z min.

0.50

0.49

1.0

Positive

2l/z min.

0.55

0.64

1.7

1.3

Positive

2*/2 min.

0.53

0.51

1.1

1.0

Fast pos.

2l/z min.

0.58

0.66

2.6

2.2

Sp. recording

3 min.

0.49

0.50

3.0

2.2

Sp. recording

3 min.

0.52

0.54

3.5

2.7

In obtaining these data, individual development characteristics
were ignored, but a difference was introduced to compensate for the
lower gamma infinity of two of the special sound recording emulsions.
The residual variations in gamma prevent an accurate statement
of the relative emulsion speeds, but the possible errors due to this
cause are much less than the differences found. Thus, it is evident
that tests made under the one set of conditions furnish no sure guide
to emulsion characteristics effective under the other set of conditions.
The results of all these tests on both types of sensitometers emphasize
again the necessity for care in selecting test conditions in photo-
graphic work. The two machines described have worked out quite
satisfactorily, each meeting the special needs in its own field.

ERRATUM

The following corrections should be made in the paper, A Method of Measuring
Directly the Distortion in Audio Frequency Amplifier Systems, by W. N. Tuttle,
beginning on page 199 of the February, 1932, issue of the JOURNAL:

A square root sign should be placed over the entire numerator of the right-
hand member of the equation on page 200. On page 205, line 2, the symbols
"2-C" should read "a-c."

THE DECIBEL IN THE MOTION PICTURE INDUSTRY*

V. C. HALL**

Summary. — The development of the term "decibel" is outlined, and its convenience
in measuring the characteristics of electrical circuits discussed. The relation of
the decibel to photographic density is pointed out and illustrated by calculations of
the effect of ground noise reduction devices. Finally, the values of acoustic power
of common sources of sound are given as the levels in decibels based on various reference
points in use in the motion picture industry.

The use of the term "decibel" has increased rapidly since its
introduction into the motion picture industry because it is a con-
venient method of handling quantities which might otherwise lead
to cumbersome expressions. Many publications have been written
concerning the decibel and its historical development and its applica-
tion to the problems under consideration. These papers have been
freely drawn from for examples and data for this summary. No
new material is presented but it is hoped that a review of some of
the ways in which the decibel enters into sound motion pictures
may prove of value to those who are not familiar with this unit.

Development of the Term. — The decibel (db.) is the name which
was chosen for the transmission unit (TU), in terms of which a
great deal of telephone and talking motion picture apparatus is
calibrated. Their values are identical, and the name itself was
suggested several years before its final adoption. Units of similar
type had been universal in telegraphy and telephony for many
years, and came into being from the fact that an electric current
representing a certain amount of power loses a certain fraction of
that power for every mile, let us say, of transmission line over which
it travels. The unit of electric power, representing the rate of
doing electric work, is the familiar watt, in terms of which most
electrical equipment is specified. If, then, a power of ten watts
is started out over a telegraph line, it would drop to 0.9 of this in
perhaps two miles. This nine watts would drop to 0.9 of 9, or 0.81,

* Presented at the Spring, 1931, Meeting at Hollywood, Calif.
** Research Laboratories, Eastman Kodak Co., Rochester, N. Y.
292

DECIBEL IN MOTION PICTURES 293

in the next two miles, and so on, the power decreasing the same
fraction every two miles, independently of the actual amount of
power starting over this particular section of line. The multiplying
together of these factors was laborious, so the practice grew up of
expressing a loss of power in terms of the number of miles of telephone
line which would decrease the power by the same amount. Thus,
if the power dropped to 0.81 in one section of a circuit, and to 0.73
in another, the total loss could be found by multiplying 0.81 by
0.73, which equals 0.591. It is much simpler, however, to say that
the first section has a loss of two miles and the second a loss of
three miles, and that the two together have a loss of two plus three
or five miles. The substitution of addition of numbers for the
multiplication of corresponding numbers is a property of logarithms,
and a system which allows this to be done is called a logarithmic
system. Thus the use of the "mile of standard cable" constituted
a logarithmic system, standards for which were set up by those
doing telephone work. It was natural that the properties of cables
should change as improvements were made so that eventually the
old standard became inconvenient. Also, amplifiers were developed
in which the output power was more than the input power so that
the circuit had a "gain" instead of a "loss." These developments
led to the adoption of a simpler unit, based only on the relation
between the power output of the circuit and the power input. The
expression used was

/watts output \

Number of telephone transmission units = logarithm ( : I

V watts input /

The values obtained from the above are positive when the output
is greater than the input, and negative when the output is less than
the input, indicating a loss. This unit as defined above happened
to be a rather large one considering the ratios of power encountered
so that the unit actually adopted was arbitrarily chosen to be 0.1
of this. Accordingly, in practice the equation becomes

Number of transmission units or TU's = 10 X logarithm

The fundamental unit was named the "bel," after Alexander
Graham Bell, with the spelling simplified to avoid confusion. Since
the name is for the larger unit, the prefix "deci" was affixed to the
name to indicate its derivation, and now we have:

294 V. C. HALL [J. S. M. p. E.

Number of transmission units or TU's = number of decibels or number of

/watts output \
db' = 10Xloganthm( watts input)

Method of Calculation. — In order to get some idea of the magnitude
of the decibel, the following short table is presented, giving a list
of power ratios from 1 to 10, with the corresponding number of
decibels. These are found simply by looking up the logarithm of
the ratio in a table of common logarithms, or on a slide rule, and
multiplying by 10.

TABLE I
Power Ratio Decibels Power Ratio Decibels

1 0.0 1.0 -0.0

2 3.0 0.5 -3.0

3 4.8 0.33 -4.8

4 6.0 0.25 -6.0

5 7.0 0.20 -7.0

6 7.8 0.17 -7.8

7 8.4 0.14 -8.4

8 9.0 0.125 -9.0

9 9.6 0.111 -9.6
10 10.0 0.100 -10.0

In the first two columns of Table I the output power is assumed
to be greater than that at the input or, in other words, an ampli-
fication instead of a loss is assumed. When this is true, the output
is said to be so many decibels above the input. When a loss occurs,
as in the last two columns, the input is of course above the output,
and the output is said to be down so many decibels. It will be
noted that the power ratios in the second case are the reciprocals of
the corresponding ratios in the first case, while the number of decibels
is the same, with a minus sign. This means, for example, that

10 x logf = -10 X log^
• °

which in turn follows from the general principle in logarithms that
the logarithm of a number is equal to minus the logarithm of its
reciprocal, which is another way of saying, minus the logarithm of
one divided by the number.

If the amplifications of a group of separate amplifying units are,
respectively, 3, 7, 10, and 4, the total amplification, if they are
connected in series, can be found by multiplying these numbers
together. This gives 3 X 7 X 10 X 4 = 840 times If the ampli-
fications are expressed in decibels, however, this becomes simply

Mar., 1932] DECIBEL IN MOTION PICTURES 295

4.8 + 8.4 + 10 + 6.0 = 28.2 decibels. In this case the result is
about as simple to calculate in one way as the other. If instead of
referring to amplifications, however, the numbers referred to losses,
and the power were reduced to 0.33, 0.14, 0.1, and 0.25 of the original
value, respectively, the multiplication of factors would be more
difficult. Looking up the corresponding decibels in the second part
of Table I gives -4.8, -8.4, -10, and -6.0, respectively, and
the sum of these equals —28.2 decibels, which is the total power
loss of the group. In practice it is rare to find the even power ratios,
while the decibel equivalents are usually expressed to sufficient
accuracy by two or, at the most, three figures, the addition of which
is obviously more quickly done than the corresponding multiplica-
tion, while the chance of error is also greatly reduced.

From Table I other values of either decibels or power ratios can
easily be found. For instance, 26 decibels is made up of 10 + 10 + 6,
corresponding to power ratios of 10, 10, and 4, respectively, and the
product of these, 400, is the power ratio corresponding to 26 decibels.
Similarly, a power ratio of 75 can be divided into its factors 3 X 5 X
5, and we have 4.8 + 7.0 -f 7.0 = 18.8 decibels, which corresponds
to a power ratio of 75.

The decibel is a convenient unit as the ear can just recognize a
change in the volume of sound corresponding to an attenuation or
gain of one decibel, equivalent to about 12 per cent. For this
reason, volume controls and faders calibrated in decibels give a very
uniform increase in the volume as the ear recognizes the changes as
equal steps. Although the ear can detect a change of one decibel,
most volume controls are set to change the volume by about 3 db., as
this step is not so great that a change of one step will raise the volume
from too low to too high, and the reduction of the number of steps
simplifies the apparatus. For the control of volume in sound re-
cording where a smooth stepless change is wanted, a potentiometer
or volume control similar to that on a radio receiver is used, and the
scale is graduated in decibels for certain intervals on the dial.

Methods of Measurement. — So far there has been no mention of
the method by which the measurements are made. In the electric
circuits used in talking motion pictures it is usually too difficult
to measure the power, in watts, directly, since the frequency of the
alternating currents which must be measured varies all the way
from 30 to 10,000 cycles per second, and the power levels are of the
order of thousandths of watts. Both of these factors tend to make

296 V. C. HALL [j. s. M. p. E.

the use of wattmeters impracticable, so most results are obtained
by measuring either the current through a known resistance, or the
voltage across a known resistance. If it is necessary to know the
power it may be found from the following relations which exist
between the various electric quantities.

Power (watts) = current X voltage (amperes X volts)
Voltage (volts) = current (amperes) X impedance (ohms)

From these two are derived :

Power (watts) = (current)2 X impedance
= (voltage)2/impedance

In order further to simplify the circuits electrically, most sound
motion picture apparatus, and a great deal of telephone equipment,
is designed so that the impedances in both the input and output are
equal, the actual value usually being approximately 500 ohms.
This is true of amplifiers, niters, equalizers, volume controls, etc.,
and further simplifies the calculation of losses or gains in the various
units as follows: If we substitute for the power input and output
in the formula for the decibel the expression for the power in terms
of the current and resistance, it becomes:

. ,. (current output)2 X impedance

Number of db. = 10 X log ; —

(current input)2 X impedance

and the resistance cancels out, leaving the formula

. ,. (current output)2 (A)2

Number of db. = 10 X log -. — . ' ; = 10 X log ^f,

(current input)2 6 (72)2

letting 72 stand for the current input, and /i for the current output.
From the principle that the logarithm of the square of a number is
equal to twice the log of the number it is possible to write, instead
of the relationship above,

Number of decibels = 10 X 2 X log ^ = 20 log®

U2J (1-1)

It is important to note that while the power is independent of
the impedance, the above formula is true only when the impedances
through which the currents /i and 72 are flowing are equal. If it
is more convenient to measure the voltage across a known impedance
than the current through it, as is often the case, we can write, again
assuming that the impedances are equal in the two places the mea-
surements are made,

Number of db. = 10 log (.VO'tage °"tput)Vimpedance
(voltage mput)2/impedance

Mar., 1932] DECIBEL IN MOTION PICTURES 297

The impedance cancels out as before, and the expression becomes
Number of db. = 10 log = 20 l°S

which is exactly the same as if the current output and input were
measured. The most usual methods of measuring electric currents
of audio frequency as found in sound motion picture work are by
the hot wire ammeter and the vacuum thermojunction. The
thermojunction is the more accurate, and in it the current heats a
length of resistance wire in a vacuum, while near the center of the
wire is fastened a junction of two wires of different composition.
When this junction is heated a voltage is developed which depends
on the nature of the two metals and the temperature to which it is

TABLE II

Voltage or

Power Ratio Current Ratio Decibels

1 1.0 0.0

2 1.4 3.0

3 1.7 4.8

4 2.0 6.0

5 2.2 7.0

6 2.45 7.8

7 2.64 8.4

8 2.83 9.0

9 3.00 9.6
10 3.16 10.0
20 4.47 13.0
40 6.33 16.0

100 10.0 20.0

heated. A meter connected across the other ends of the two wires
will deflect in accordance with the current variations, and if cali-
brated in terms of a standard meter can be used to measure the
current accurately. The volume indicator, which is essentially of
the vacuum tube voltmeter type, although it may draw a small
current, due to its ruggedness is the most popular of audio frequency
measuring instruments. In this the voltage across a resistance
changes the grid voltage on a vacuum tube which is so connected
that the change causes a variation in the steady plate current of
the tube. A milliammeter in the plate circuit then reads in ac-
cordance with the changes in voltage across the resistance. Most
volume indicators are built on this principle although, since the
entire scale of the meter corresponds to a relatively small number

298 V. C. HALL [J. s. M. p. E.

of db., most of them have a resistance device connected so that the
readings may be cut down by fixed amounts, say, 2 db. per step.
By using this device the needle of the meter can be kept at one
position on the scale as the voltage varies, and the db. change noted
by the change in the setting of the control mechanism.

In Table II are shown the power ratios, voltage or current ratios,
and the decibel changes which correspond to them. Thus, if the
measured current ratios between output and input of three amplifiers
are 2.45, 4.5, and 6.3, the total gain is

7.8
13.0
16.0

36.8db.

If the composite voltage or current gain is desired it is found by
36.8 = 20 log (voltage ratio)

from which log (voltage ratio) = 36.8/20 = 1.84, and voltage
ratio = 69.1

Reference Levels for Calibration of Apparatus. — While from its
definition the difference in decibels between any two amounts of
power is calculated without reference to any standard power, it is
convenient to have some value of power which may be considered
as a reference level. Power is always expressed in watts, so that
at first thought it might seem obvious that one watt of power would
be the correct unit to choose. The amount of power encountered
in either electrical or acoustic measurements in sound motion pictures,
however, is nearly always much smaller than one watt, and the
expression of levels when the ratio of power is less than unity is
negative. This unit would involve the use of negative numbers
for the expression of nearly all powers measured, and would prove
inconvenient.

In acoustic measurements, the power of a sound wave in air is
usually very small and is spread throughout a considerable volume
of space. To simplify calculations it is generally assumed that the
sound waves radiate uniformly in a hemisphere from the source.
Results indicate that the assumption is justified, provided that a
reasonable distance from the source is allowed and that no difficulty
is encountered from reflections from walls, ceilings, etc.

Mar., 1932] DECIBEL IN MOTION PICTURES 299

The energy in the sound wave may be measured either in its
entirety, or as the amount passing through a unit area (usually
a square centimeter) at certain distances from the source. Since
the smallest amount of energy which can excite the sensation
of hearing is the smallest amount of useful energy a sound wave can
have, this value is taken as the "audibility threshold" or "acoustic
level," and is usually considered to be about 4 X 10~16 watts per
square centimeter.

Another value sometimes used is the "phonic level," which is
simply one microwatt per square centimeter of cross-section of the
air through which the wave is traveling.

In the electric circuits associated with sound motion pictures,
the powers vary in value from as low as the audibility threshold
up to several watts, as the power necessary to operate loud speakers
in theaters satisfactorily may in extreme cases be as great as 15
watts or more. The general levels at which measurements can be
made easily correspond to a few milliwatts, and volume indicators
are usually calibrated to read "0" level at about 6 milliwatts.

Relation between the Decibel and Photographic Density. — The
amount by which the silver deposit on a photographic film reduces
the amount of light transmitted by the film is expressed by a logarith-
mic unit called density. The first measurements of the decrease of
light intensity were made by observing the percentage of the incident
light which the film transmitted. Various considerations led to
the adoption of a logarithmic unit which is defined as the logarithm
of the reciprocal of the fractional transmission, thus density =
log 1/r, where T is the fraction of the light transmitted by the silver
deposit. Since the value of the light transmitted is always less
than that which is incident, this fraction is always greater than
unity, all photographic densities being positive, and varying in
practice from 0.04, the ratio in which clear film base reduces the
light transmitted to about 6.0, representing opacity for all practical
purposes.

When a variable density sound record passes through a projector
the changes of density cause the light of the exciting lamp which is
incident on the photoelectric cell to vary in intensity. This causes
the photoelectric current to vary, but as it is proportional to the
light striking it, the change of current is proportional to the trans-
mission of the film, and not to the density. Thus, if a transmission
TI corresponds to a current /i, and if the transmission should change

300 V. C. HALL [J. S. M. P. E.

to T2, the photoelectric current would change proportionally to J2, so

that it can be written:

7\ =/i

This can be changed to log ~ = log — by taking logarithms of both

J. 2 -*2

sides of the equation, and can be multiplied by 20 also without chang-
ing the validity of the statement. This leaves

20 log ~ = 20 log £

2-2 12

The left-hand side of this equation is identical in form with the ex-
pression for the power reckoned in decibels when two currents act
through equal resistances; and since the photoelectric currents in
this case both pass through the same resistance (the amplifier input
resistance) we can substitute the decibel for this part of the expres-
sion. We have

Number of db. = 20 log -=^

The right-hand side contains the logarithm of the ratio of two photo-
graphic transmissions. These can be written as the product of
V TI X TI, and since the logarithm of a product is equal to the sum
of the logarithms, it becomes

Number of db. = 20 Aog ^ + log

\ •* 2

Number of db. = 20 ( loe^r -

The logarithm of l/7\ is no more than the density corresponding
to this transmission (Di), and log l/r2 equals the corresponding
density (D2). We may therefore substitute these values and reach
the expression

Number of db. = 20 (D2 - A)

showing that any change in photographic density, multiplied by
20, gives the corresponding change in electrical power in decibels.
From these statements it is possible to calculate the efficiency of
the noise reduction units now in use in light valve recording studios.
The "ground noise" arising in sound motion pictures is due partly
to slight irregularities in the current which are inherent in the photo-
electric cell, but chiefly to the changes in current caused by dirt

Mar., 1932] DECIBEL IN MOTION PICTURES 301

and scratches in the photographic film. The effect of a scratch or
particle of dirt is to cut the light down by a certain fraction, so that
its effect on the photoelectric current will be less in proportion to
the amount of light which is left to be affected. Therefore, whatever
can be added to the average density of the positive sound track will
help to reduce both types of ground noise in an amount equal in
decibels to 20 times the difference between the two densities. It
must be noted that an increase in the average density of an ordinary
sound print does not cut down this noise, as the volume it is possible
to get also is cut down, so that the amplification must be raised,
restoring not only the signal, but also the noise, to its former level.
It is only by cutting down the light while the sound volume is low,
or during silent passages, that any effect is found, and if the density
of the film can be decreased to its normal value when the sound
volume increases, the amplification does not have to be increased
to keep the proper level in the theater. The amount by which
the noise may be decreased depends fundamentally on the amount
by which the valve may be closed in recording. This narrowing of
the light valve slit is accomplished by sending a direct current
through the valve in addition to the amplified signal coming from
the microphone. This decreases the density of the negative during
sound passages of low volume, increasing the density of the positive
during the same sequences. In following through the theory of
sound recording and reproducing by the light valve method, proper
sound reproduction depends on the proportionality of the movement
of the light valve strings to the sound pressure at the microphone
in recording, and the photographic processing must be such that
the transmission of the positive is also proportional to the sound
pressure. Therefore the transmission of the positive must be pro-
portional to the valve opening. If the normal slit width is 1.0
mil (0.001 inch), and it is biased in noiseless recording to 0.3 mil,
the positive will have a transmission TI for the 1.0 mil slit and a
transmission Tz for the 0.3 mil slit. The proportionality equation is

1.0/0.3 = 7yr2

taking logarithms of both sides

log 3.3 = log 7yr2.

It has been shown that log Ti/Tz = (Dz — A), so the above becomes
DZ — DI = 0.52, and since the reduction in noise is 20 times the

302 V. C. HALL [J. S. M. P. E.

change in density, in this case it becomes 10.4 decibels. As has
been stated this would be about three steps on an ordinary fader
and would be very appreciable.

In the variable width method of recording, during silent passages
one-half the sound track has a high density and one-half is clear.
In order to reduce the noise due to transmission of light through the
clear area, a mask is arranged in making the negative to cut off the
light incident to this area. During printing this area is printed
to a high density, leaving only a very narrow unexposed line in the
center of the record. As the sound intensity increases, the mask
is moved farther and farther over, leaving more and more of the
sound track available for the making of the record. In such a case,
the intensity of noise is again dependent on the amount of light
transmitted, but since the film is either clear, letting through all the
light, or so dense as to allow practically none, the amount of light
transmitted is proportional to the width of the clear portion of the
track. The decrease in the noise intensity, following the same line
of reasoning as before, will depend on the width of track it is neces-
sary to leave in the center during silent passages. This is at least
5 mils, and since in the variable width recording system the width
of the whole sound track is 70 mils, the clear portion is normally
35 mils. The reduction in the intensity of the noise can be cal-
culated from the change in the photoelectric current, which depends
on the width of the clear track. Thus from equation

Number of db. = 20 log 35/5 = 20 log 7 = 20 X 0.845
Number of db. = 16.9 or approximately 17 db.

Conclusion. — It is hoped that the foregoing explanation of the
various ways in which the decibel enters into the sound motion
picture may prove of value to those who find that the literature of
the art includes many statements which depend, for a complete
understanding, on an accurate conception of exactly what the
function of the unit is, and the reasons why its use is convenient.
For reference a tabulation of the various levels occurring in parts of
sound motion picture systems is added. These levels are in some
cases only approximate, owing to their nature, but they indicate
the order of magnitude to be expected.

Distances as given refer to columns A, B, and C. Radiation is
assumed to be in the form of a hemisphere with the power given in the
first column generated at the center. (A), decibels above audibility

Mar., 1932] DECIBEL IN MOTION PICTURES 303

threshold (acoustic level, assuming 4 X 10 ~10 microwatts per sq.
cm.); (B), decibels above one microwatt per sq. cm. (phonic level);
(C), decibels above 0.006 watts (electrical level) of output of con-
denser microphone into 25 megohms input.

TABLE III

Source of Sound
Total Power Microwatts ABC Distance

Soft whisper 0.001 17 -77 -144 3 feet

Average speech 10 57 -37 -104 3 feet

Very loud speech 1000 77 -17 -84 3 feet

Peak of speech 5000 84 -10 -77 3 feet

Peak of singing 30000 91.8 -2.2 -69.2 3 feet

Soft violin in orchestra 4 43-51 -118 10 feet

Piano — average 4000 73 -21 - 88 10 feet

Piano — highest peak 2 X 106 100 +6 - 61 10 feet

Bass drum peak 25 X 106 107 +13 - 54 15 feet

75 piece orchestra peak 66 X 106 113 +19 -48 15 feet

Pipe organ— peak 13 X 106 105 +11 - 66 15 feet

No attempt has been made to quote sources of data given in the
course of the paper. The data for Table III were derived from the
first two references given and further references will be found in
the following list.

REFERENCES

FLETCHER, H.: "Speech and Hearing," D. Van Nostrand Co., Inc., New
York, N. Y. (1929).

SIVIAN, L. F., DUNN, H. K., AND WHITE, S. D.: "Amplitudes and Spectra
of Certain Musical Instruments and Orchestras," /. Acoustical Soc. Amer.,
2 (January, 1931), p. 330.

WOLF, S. K., AND SETTE, W. J.: "Acoustic Power Levels in Sound Re-
production," /. Acoustical Soc, Amer., 2 (January, 1931), p. 384.

DREHER, C.: "Progress in Sound Picture Recording," Electronics, 2 (March,
1931), p. 542.

MARTIN, W. H.: "Decibel — the Name for the Transmission Unit," Bell
System Tech. Jour., 8 (January, 1929), p. 1.

SHEA, T. E.: "Transmission Networks and Wave Filters," D. Van Nostrand
Co., Inc., New York, N. Y. (1929), p. 43.

OPTICAL INSTRUMENTS AND THEIR APPLICATION IN
THE MOTION PICTURE INDUSTRY*

I. L. NIXON**

Summary. — This paper deals not with the optics of the photographic lens, motion
picture projector, or studio illuminator, but rather with those instruments such as
microscopes, photometers, etc., the use of which has contributed greatly to the advance
of the motion picture art of today. A simple explanation is given of the different
types of instruments and the general optical principles involved, and some of their
specific applications, which indicate the debt which industry owes to optical science.

When speaking of optics in the motion picture industry, it is but
natural for those of us who are most intimately connected with the
industry, to think of optics as applied to the photographic lens as
used on the motion picture camera, or to the optics of the projector, or
for illumination in the studio, but I purpose to outline briefly some of
the different types of optical instruments that have been used or
might be used in the development of new materials, control of proc-
esses, and control of accuracy of parts. Because of the fact that
many of these devices have been used largely in a research way and
not in sufficient numbers to attract attention, they might be classed
as the modest group of silent workers that have made the high perfec-
tion of the present art possible and that will play an important part
in the achievements of the future.

The microscope of one form or another is probably the most widely
used optical instrument in the motion picture industry. It hardly
seems necessary to define a microscope, but it might be described
generally as a device having a system of lenses, suitably supported by
mechanical arrangements, which will produce a magnified image of a
small object so that the eye may distinguish between details of
structure not otherwise discernible.

A simple magnifying glass might be considered as qualifying as a
microscope under this definition, but this paper will deal with what
may be termed a compound microscope, a typical one being repre-

* Presented at the Spring, 1931, Meeting at Hollywood, Calif.
** Bausch and Lomb Optical Co., Rochester, N. Y.
304

OPTICAL INSTRUMENTS

305

sented by Fig. 1, where a system of lenses, mounted together and
known as an objective, is attached to the lower end of what is known as
a body tube and another system known as the eyepiece is mounted
in the upper end of the tube. The objective acts as a photographic
lens would act, and forms a magnified image of the object in the focal
plane of the eye lens which, in turn, magnifies that image. Hence we
have compound magnification and in turn a compound microscope.
By varying the power of one or both of these units the magnification
is accordingly changed, the range of magnification being from IOX
to approximately 2000X.

FIG. 1. A typical compound microscope.

The design and accuracy of the mechanical parts of such an instru-
ment are quite as essential to its functioning as are the optical parts.
It must be substantially constructed, and yet a certain symmetry
of design is demanded and its movable parts must be accurately fitted
and free from any lost motion, yet immediately responsive to adjust-
ment.

The mechanical part of the compound microscope is referred to as
the stand and consists of the following general parts:

(A} Base, of a design that will have sufficient spread and weight to assure the
stability of the instrument in either an upright or inclined position.

306 I. L. NIXON [j. s. M. P. E.

(5) Stage, on which is placed the object to be observed. This may be either
plain, rectangular, or circular revolving, and both styles may be fitted
with mechanical devices for moving the object in two directions at 90
degrees to each other for easy searching of the specimen. These ad-
justments may also be provided with scales for relocation of the specimen
if desired.

ments.

(D) Body tube.
(£) Objective.

(F) Eyepiece.

(G) Coarse adjustment, by rack and pinion, which must move easily and yet

be free from lost motion.
(H) Fine adjustment.
(/) The substage with condenser or illuminating lens system, which functions

either in conjunction with daylight or a suitable artificial light source to

illuminate the specimen efficiently, if it be one with which transmitted

light may be used.

In Fig. 2 we have shown diagrammatically the path of light of such
a microscope, which seems to need no further explanation except to
point out that when looking into the microscope the image appears as
though it was being viewed at a point 10 inches below the equipment.
If a screen is held 10 inches above the eyepiece an image will be
formed at that plane equal in magnification to the image observed in
the eyepiece and the magnification would increase proportionately
as the distance was increased beyond 10 inches.

This represents the typical biological or medical type of microscope,
large numbers of which are manufactured annually for use in the
schools and colleges, but which are being used more extensively each
year in industrial laboratories where transmitted light may be used.

A number of deviations from this typical instrument in the way of
special illuminating devices and accessories of one sort or another
make the equipment particularly suited for some specialized work.
Before passing on to these, however, it will be interesting to note the
similarity of the optical system of the microscope to that of the motion
picture projector. A light source with the substage condenser cor-
responds to the light source and the condensing lens system; the
stage on which the specimen is placed may be compared to that of
the film gate supporting the film, and the objective lens and eyepiece
may be considered as one unit corresponding to the projecting lens.

In addition to using a microscope as a device for studying the
structure of materials it may also be used for measuring the size of

Mar., 1932]

OPTICAL INSTRUMENTS

307

particles or parts, or their separation, by the use of a ruled disk to be
placed in the eyepiece at what is known as the diaphragm plane or in
the same plane as that of the image formed by the objective so that
both the scale and the image of the object will be in the focus of the
eye lens. (Fig. 3.)

Such a scale may be ruled with divisions to represent a definite
value on the specimen (0.001 of an inch, for instance) for use with
definite combination of eyepiece and objective producing a fixed
magnification, or the eyepiece disk may be ruled in definite values,

FIG. 2. Diagram showing path of
light of a typical microscope.

and a stage micrometer used to evaluate the rulings on the eyepiece
disk, according to the combination of objective and eyepiece being
used. The use of such a device was probably first used by the doctor
in counting the number of blood corpuscles per given quantity of
solution, but has been adopted by the industry as a means of mea-
suring and determining the distribution of silver grains in emulsion.

As evidence that the microscope is being recognized as one of the
most important tools in modern industry is the fact that a number of
the leading universities are introducing as a division in the chemical
engineering courses one known as "Chemical Microscopy" in which

308

I. L. NIXON

[J. S. M. P. E.

the principles of the microscope, its applications, and the interpreta-
tion of the results are taught.

There is no industry that I know of in which a microscope could

FIG. 3. Photomicrograph of sil-
ver grains with micrometer scale.

FIG. 4. Path of light of special
microscope for examination of
paper surfaces.

not be used to advantage in the control of its raw materials and
finished product.

The number of ways in which a microscope is used in the laboratory

FIG. 5. Special paper microscope.

of a manufacturer of film is amazing. An almost constant study is
made in the size, shape, and distribution of silver grains in the emulsion
both before and after development. Photomicrographs are fre-

Mar., 1932]

OPTICAL INSTRUMENTS

309

quently made for record and control purposes from which frequency
curves may be plotted if desired.

We do not ordinarily think of the finished film being built up of a
series of layers, but it is; and each one of the processes contributing to
this building up must be carefully controlled. When something goes
wrong, they send for the trouble shooter, the man with the micro-
scope. A cross-section of the film will probably be made with a
microtome, a device for making sections only a few microns in thick-
ness; and when this is observed one clearly sees these layers of differ-
ent materials, and the trouble can usually be traced to a certain opera-
tion or to impurities that are causing the trouble.

Standards of surface finish of both film and paper may be set up

FIG. 6. Photomicrograph of a
paper surface at 40 magnifications.

FIG. 7. Path of light when il-
luminating opaque objects.

according to microscopical specifications and in the event of trouble it
is fairly easy to trace back against the standard and locate the source.

Dr. L. A. Jones, of the Eastman Kodak Co., in his study of paper
surfaces decided that there was needed a special type of illumination
with provision that the exact illumination could be duplicated at any
time, because surface appearance of paper depends so much on the
amount and angle of illumination. There was developed a micro-
scope with an illuminating system as indicated in Fig. 4.

A light source and condensing system produces a beam of light
passing through a ground glass at G, which strikes the 45-degree
annular reflector M , and the light is reflected upon the specimen at 0.
It is obvious that this is annular illumination which illuminates equally
from all directions, and with no direct top illumination. By means of
the movable tube D the amount of illumination and the angle of

310

I. L. NIXON

[J. S. M. P. E.

incidence may be regulated. A very smooth surface will be best
illuminated by light from a small angle while a rough surface requires

FIG. 8. Metallurgical microscope.

higher angular illumination. The length of fibers, how they are ar-
ranged, and how the filler and the coating has been applied may all
be easily studied under this kind of illumination.

FIG. 9. Complete metallographic equipment.

Furthermore, since the tube is graduated it is possible to record the
exact setting and to return time and again to the same illumination.

Mar., 1932]

OPTICAL INSTRUMENTS

311

Fig. 5 shows this microscope as it is now commercially made, with
an observation tube which may be withdrawn so the light passes on
through the other tube to the camera for making photomicrographs.
Fig. 6 shows such a photomicrograph. Provision is made for the
making of stereophotographs if desired.

A number of other modifications of the standard microscope or

FIG. 10. Path of light for large
metallographic equipment.

special accessories are made use of in the film laboratories in more or
less highly specialized investigations, among which is a device known
as a dark ground illuminator for the illumination of colloidal particles,
a microscope with accessories for producing polarized light by means
of which strains may be detected in crystals and film base, and a

FIG .11. Photomicrograph of steel .

FIG. 12. Photomicrograph of steel.

micromanipulator, by means of which individual crystals or particles
of impurities may be isolated and submitted to all kinds of treatment.
While many materials may be satisfactorily illuminated with
transmitted light there are many that are opaque and, consequently,
must be illuminated from the top. In the case of low-power equip-
ment this may be by means of light directed downward, striking the

312

I. L. NIXON

[J. S. M. P. E.

object at an angle, in other words, flood lighted ; while in the case of
high-power equipment the objective works so close to the object
that it is no longer possible to illuminate in such a manner and then
one must resort to what is generally known as a vertical illuminator.

Fig. 7 is a diagram showing the path of light of such a device. This
vertical illuminator is inserted between the end of the body tube and
the objective. In one side of the mounting is an opening usually
fitted with a small condensing lens. A small concentrated beam of
light from a suitable light source enters through this aperture and is

FIG. 13. Wide field binocular
microscope.

FIG. 14. Comparison microscope.

reflected 90 degrees downward, either by a clear glass reflector or by a
prism, through the objective lens onto the specimen; and since the
rays of light are striking normally to the surface of the specimen they
will be reflected directly back along their original path. This is as-
suming that the specimen is fairly well polished. If a clear glass
reflector is used, a portion of the returning light passes through the
glass to the eyepiece. If using a prism as the reflecting medium, it
must be mounted off the center of the optical axis; the light then
passes down through one side of the objective and back through the
other, past the prism and on through to the eyepiece.

Mar., 1932] OPTICAL INSTRUMENTS 313

This kind of illuminator is a part of all metallurgical microscopes
of which there are two general types. The first one, Fig. 8, is essenti-
ally the same as the regular microscope except that it is fitted with the
vertical illuminator and usually is without substage condenser, but it
has the stage movable vertically by rack and pinion. This is neces-
sary to bring the object into focus without changing the position of
the vertical illuminator with relation to the light source after it has
once been aligned and centered. Such a microscope as this has wide
use as a routine instrument in the industrial laboratory. The other
type of metallurgical microscope is that shown in Fig. 9.

This is known as an inverted microscope, the stage being at the
top, the specimen placed with its polished side down and the illumina-

FIG . 1 5 . Photomicrograph of two
pieces of textile as seen through the
comparison microscope.

tion coming up from underneath by means of a vertical illuminator,
as will be seen from the diagram of the path of light in Fig. 10.

The illuminating system of this instrument is mounted on a base
with the microscope, so that both units may be very carefully and
permanently aligned and centered, a point which is very essential
when working at high magnifications. This microscope is almost
always sold in conjunction with the camera as shown, the combination
then being known as a metallographic equipment. It is common
practice with such equipment to make photographs at 2000 or 3000
diameters, and they have been made as high as 15,000 diameters.

The value of such equipment to the manufacturer and user of
steel, brass, copper, and all metal alloys is beyond estimation. Suffice
it to say that if it were not for the microscope we probably would not

314 1. L. NIXON [J. S. M. p. E.

have the high-speed motor cars, aeroplanes, etc., that we have today.
The chemical analysis will determine whether the correct percentage
of the different constituents has been maintained or not, But that
does not tell whether a piece of steel will be suitable for the purpose
for which it has been intended or not ; such can be told only by deter-
mining the crystalline formation after the various heat treatments,
rolling, drawing, etc.

A piece of steel, for instance, of a given mixture will have a very
definite crystalline structure following certain treatments which the

FIG. 16. Special microscope for wax records. Photograph by
courtesy of Bell Telephone Laboratories.

trained metallurgist can at once recognize. So he takes a small
sample, grinds, polishes, and etches one surface, checks it on the micro-
scope and many times will photograph it for record purposes. Such
photographs might look like Figs. 11 and 12.

The first shows a steel heated to a certain point of the treatment and
the other a steel carried beyond this point in the hardening process.

Probably one of the most useful instruments for general use around
the laboratory or the factory is a wide field binocular microscope.
(Fig. 13.)

Mar., 1932]

OPTICAL INSTRUMENTS

315

This, as its name implies, is arranged for binocular vision with a
large field, and the image produced is an erect one, so that in working
with materials, dissecting, etc., all movements maybe naturally made.
Furthermore, one sees naturally, that is, stereoscopically. Its most
useful range is at from magnifications of approximately IX to 3QX.
Because of the large field and third dimension it is most useful in
examining small parts, machine surfaces, and raw materials.

FIG. 17. Special microscope for setting
width of light valve. Photograph by
courtesy of Bell Telephone Laboratories.

In a paper presented by O. E. Conklin1 at the 1930 Fall Meeting of
the Society, the applications of the comparison microscope in the film
industry are set forth in detail. Such a microscope has two objectives
which focus on two objects to be compared, and a prism which brings
their images together so that they can be seen in a single eyepiece side
by side. (Fig. 14.)

In a film laboratory, either manufacturing or processing, a compari-
son microscope is invaluable because one is always wanting to com-

316

I. L. NIXON

[J. S. M. P. E.

pare two things, one of which may be a standard. Two films may be
compared for general graininess, or tint, papers for surface finish, etc.
(Fig. 15.)

For the details of its application I refer you to Mr. Conklin's paper
in which he also describes how this instrument had been used as the
basic unit in the construction of a picture comparator, a sound track
photometer, a graininess comparator, and a perforation comparator.

This paper of Mr. Conklin's shows most conclusively how with a

FIG. 18. Toolmaker's microscope.

little ingenuity a single basic microscope may be made to serve in a
number of important control steps — leading to the uniformity and
general efficiency of the film.

In making sound records on wax a microscope becomes indispen-
sable to check the performance of the cutting needle and for that
purpose there has been employed a body tube with objective and eye-
piece mounted on an arm to be swung over the record. A special
vertical illuminator has been employed for proper illumination and
an inclined eyepiece for greater convenience in viewing. (Fig. 16.)

Mar., 1932]

OPTICAL INSTRUMENTS

317

The setting of the width of the slit on the light valve of the Western
Electric sound recording unit necessitates the use of a microscope
to which a special holder is attached for holding the valve unit.
(Fig. 17.) This equipment is provided with an optical system having
a magnification of 100 and a scale in the eyepiece with ten spaces,
each space representing 0.001" on the specimen.

In modern shop practice it becomes essential to work to much closer
limits in making screw threads, gears, cams, cutting and forming tools,
and the mechanic is often confronted with checking curved forms
whose contour is almost impossible to control without recourse to one

FIG. 19. Contour measuring projector.

or the other of two optical devices, one known as a toolmaker's
microscope and the other as a contour projector. (Fig. 18.)

The toolmaker's microscope is provided with an illuminating system
producing a parallel beam of light passing the specimen and an
objective specially corrected for use with parallel pencils of light.
The stage has a micrometer movement of one inch in two directions at
90 degrees to each other with graduated drums reading at a tenth of a
thousandth of an inch.

The eyepiece may be with a simple cross-hair against which settings
of the image may be made, or it may be what is known as a protractor
eyepiece with cross-hairs adjustable for angles of from 40 to 70
degrees.

318

I. L. NIXON

[J. S. M. P. E.

With such an instrument the angle, lead, and pitch diameter of
screw threads may be easily and accurately measured, and in addition
it has been used particularly in the motion picture industry for check-
ing the spacing and size of perforations, checking the tools with which
the perforations are made and the slit on the Western Electric light
valve.

The contour projector, as its name implies, is a projector of contour
forms, screw threads, gears, cutting tools, etc. (Fig. 19.)

There are a light source, a special object holder, and a projection lens
producing the image either on a distant screen or upon the chart
attached to the stand depending on the size of object being in-

FIG. 20. Path of light for contour measuring projector.

spected. (Fig. 20.) Here again we have an optical system resem-
bling the typical motion picture projector, except that the beam il-
luminating the object is made up of parallel pencils.

A five-ampere arc lamp is usually used as the light source with an
aspheric condenser in an adjustable mount which may be set to il-
luminate approximately a 2-inch area for large objects, and a supple-
mentary condenser with diaphragm for use with small objects and
high magnification objectives.

The object may be held either between centers, in V blocks or, in
the case of gears, on studs and special holders for special forms. The
object holder has a forward and backward movement for focusing, and
a vertical and transverse movement* for moving the object into the

Mar., 1932]

OPTICAL INSTRUMENTS

319

field of view. These two movements may be fitted with micrometer
screw and drum for measuring distance, lead of screw threads, etc.

In checking screw threads it is customary to use a special chart
against which the thread outline is readily checked for angle, size, and
general correctness of form. (Fig. 21.)

A plate holder or special paper holder may be substituted for the
chart and photographs made for record purposes.

The contour of gear tooth form may be checked against a master

FIG. 21. Photograph of commercial thread and thread chart.

drawing for size and exact form or two gears may be mounted en-
meshed and slowly rotated, and their exact rolling action carefully
studied. The silent transmission gear systems on the present-day
automobile are a result of the study of gear action with such a device.
Many tool departments are using such a device for checking the
contour of cutting tools against a master template before turning them
over to the operating departments.

Several manufacturers in the motion picture industry are using such
equipment for controlling the accuracy of their mechanical parts, and

320

I. L. NIXON

[J. S. M. P. E.

no doubt the silent mechanisms of today are largely the result of a
study of the parts involved on such a projector. Spacing and shape
of film perforations are also being controlled by such a device. In
measuring the spacing, one edge of a perforation is carefully lined up
with a target on the chart, then the carrier is moved over until the edge
of the next perforation is in line with the target. Then, with the
micrometer screw and drum, the amount of movement or the spacing
can be easily checked to a tenth of a thousandth of an inch. The
radii of the corners can be easily checked against a master outline so

FIG. 22. Densitometer.

that the accuracy of the die and the amount of wear may be quickly
determined.

Up to this point we have been considering the type of equipment by
means of which material things might be examined, measured, etc.,
but there is another group of optical instruments that are quite as
important to the industry and their use marks in many instances the
high state of achievement. Such instruments are photometers, and
spectrometers in a broad sense.

A photometer is an instrument which has for its purpose the
measurement of light intensity. There are many kinds ranging
from the portable kind, known as a luminometer for approximate

Mar., 1932] OPTICAL INSTRUMENTS 321

measuring of lumens or foot candles for screen illumination, to the
highly specialized type for close measurement of film density.

In designing optical systems for projector or lighting units the
amount of illumination and its distribution are checked by a photome-
ter, and the highly efficient systems are largely due to the ability
to record accurately their performance with some type of photometer.
The same thing applies to the study and development of light sources.

We are indebted to Martens, a German physicist, for the conception
and development of the polarization type of photometer, which has
been embodied in a number of special instruments, an example of

eye Poinf \

Scale Lens Scale Lens

> /

Prism ~
noiyzet

Scale

/
Bi- Prism

Apertures

opal
GIQS.S —
DISC

FIG. 23. Path of light of densitometer.

which is one known as a densitometer. Its purpose is, as its name
implies, to measure the density of the exposed portions of film.

Fig. 22 illustrates a densitometer developed along lines suggested
by the Bell Telephone Laboratories, with which extensive study of
sound track density has been made. While this incorporates the
general principle of the Martens photometer there are a number of
modifications and the general design of the instrument has been
around the requirements necessary for easy and accurate study of
sound track densities.

Fig. 23 shows the path of light of this densitometer. In this de-
sign the two entrance pupils have been removed from the body of the

322 I. L. NIXON [J. S. M. P. E.

instrument and placed on the stage over which the film travels and
are diffusely illuminated.

The upper portion of the photometer consists of a Wollaston prism,
a bi-prism, and an analyzing prism, with suitable lenses to produce a
photometric field, one-half of which is illuminated by light from the
aperture over which the film travels and the other by light from the
clear aperture. The rotatable analyzing prism is provided with a six-
inch divided circle, on which are engraved both density and trans-
mission scales. These scales occupy two oppositely located 45-de-
gree sectors, and are designed so that either transmission or density
may be read from the same setting, the former on the left and the
latter on the right. The scales are read by means of two magnifiers
mounted on the eyepiece tube of the photometer. The accuracy of

FIG. 24. Spectrophotometer.

reading is within 3 per cent at any point in the scale and the maximum
density which can be read is approximately 3.5.

Other still more highly specialized instruments, known as micro-
densitometers, have been made in very limited number by European
manufacturers for examination of very small areas; usually used in
the study of spectral lines, but have been used in the sound research
laboratories for the study of individual vibrations on the sound track.

Another specialized piece of photometric apparatus which has been
of decidedly important usefulness to the industrialist has been the
Spectrophotometer. The Spectrophotometer is an instrument by
virtue of which any controllable beam of light may be split up into
its spectral components; that is, into the component colors of light
tending to compose it, and to measure the comparative proportion of
each wavelength or color present in this light. It is possible by this

Mar., 1932] OPTICAL INSTRUMENTS 323

instrument to identify or to make a record of the color characteristics
of a substance whether it be a reflector of colored light or, let us say, a
piece of colored fabric, or whether it transmits colored light, for
instance, a piece of colored glass. (Fig. 24.)

The spectrophotometer consists of three essential parts: first, the
source of light, in which must reside all possible colors or wavelengths
within the visible spectrum. This condition is admirably met by the
incandescent lamp. The second element is a spectroscope-like piece
of apparatus, by virtue of which light entering it may be broken up
into its component colors; and to do this quantitatively, that is, so
that one may be able accurately to identify which color the instru-
ment is transmitting at a particular time. The third element is a
photometer, the function of which is to measure the intensity of the
one-colored light being transmitted at the instant the measurement is
taken.

The spectrophotometer is fundamentally a kind of comparator as
differentiated from an instrument which makes absolute measure-
ments, but since the intensity of light originating in the incandescent
lamp can be held practically constant by controlling its impressed
electric current, and because of certain characteristics inherent in the
design of the apparatus, the basis of comparison is always unity, so
that the reading of the instrument may be reduced directly to per-
centage of transmission of light of any given particular wavelength.

Such an instrument is widely used by dye makers, paint manu-
facturers, and all people who have anything to do with the specifica-
tion of color. The dyes used in tinting film must be or should be
checked spectrophotometrically during development to ascertain
positively the desired color, and during manufacture to guarantee
uniformity of product. Of course, the filters used in all processes of
colored photography must be carefully and painstakingly studied with
such apparatus if satisfactory results are to be obtained in their use.

It is safe to say, therefore, that the achievement of color photog-
raphy up to the present time and its further perfection will be due to
the availability of spectrophotometric apparatus.

REFERENCE

^ONKLIN, O. E.: "Some Applications of the Comparison Microscope in the
Film Industry," /. Soc. Mot. Pict. Eng., XVI (Feb., 1931), No. 2, p. 159.

PHOTOGRAPHIC SENSITOMETRY, PART IV*
LOYD A. JONES**

The following is the fourth and final installment of Mr. Jones' paper on sensi-
tometry, which, due to its length, was presented in part on three consecutive days at
the Spring, 1932, Meeting of the Society at Hollywood, Calif. The preceding
installments appeared in the JOURNALS of October and November, 1931, and January,
1932. The paper deals in a tutorial manner with the general subject of sensitometry ,
its theory and practice.

OUTLINE

I. Introduction.

04) Definition.

(5) Scope of field.

(Z>) The characteristic ZMog £ curve.

II. Sensitometers.

(A) Light sources.

(1) Historical resume.

(a) Natural light (sunlight, skylight, etc.).

(6) Activated phosphorescent plate.

(d) The Hefner lamp.

(e) The Harcourt pentane standard.
(/) The acetylene flame.

(g) Electric incandescent lamps.

(2) Spectral composition of radiation.

(a) The spectral emission curve.
(6) The complete radiator.

(d) Effect of color temperature on sensitivity values.

(3) Modern standards of intensity and quality.

(a) Acetylene flame plus dyed gelatin filter.
(6) Acetylene flame plus colored glass filter.

(d) Electric incandescent plus colored filters.

(4) The international unit of photographic intensity.

(B) Exposure modulators.

(1) Intensity scale instruments.

* Presented at the Spring, 1931, Meeting at Hollywood, Calif.
** Kodak Research Laboratories, Eastman Kodak Co., Rochester, N. Y.
324

PHOTOGRAPHIC SENSITOMETRY 325

(a) Step tablets (7 variable by finite increments).
(6) Wedge tablets (/ variable by infinitesimal incre-
ments).

(d) Tube sensitometer.

(e) Optical systems with step diaphragms.

(/) Optical systems with continuously variable dia-
phragms.
(2) Time scale instruments.

(a) Exposure intermittent.

Finite exposure steps (discontinuous gradations).
Infinitesimal exposure steps (continuous grada-
tions).

(b) Exposure non-intermittent.

Finite exposure steps (discontinuous gradations).
Infinitesimal exposure steps (continuous grada-
tions).
HI. Development.

04) Developers.

(1) Standards for sensitometry.

(a) Ferrous oxalate.

(b) Pyro-soda.

(2) Standards for control of processing operations.
(B} Temperature control.

(1) For standardized sensitometry.

(2) For control of processing operations.

IV. The measurement of density.

04) Optical characteristics of the image.

(1) Partial scattering of transmitted light.

(2) Diffuse density.

(3) Specular density.

(4) Intermediate density.

(5) Relation between diffuse and specular values.

(6) Effective density for contact printing.

(7) Effective density for projection.

(8) Color index.

(B) Fog and fog correction.

(1) Source of fog.

(a) Inherent fog.
(6) Processing fog.

(2) Fog correction formulas.

(1) Bench photometer,
(a) Rumford.
(6) Bunsen.

326 LOYD A. JONES IJ. S. M. p. E.

(2) Martens polarization photometer.

(a) Simple illuminator.

(6) Split beam illuminator.

(3) Integrating sphere.

(a) For diffuse density.

(6) For diffuse and specular density.

(4) Completely diffused illumination.

(a) For diffuse density.

(5) Specialized forms.

(a) Furgeson, Renwick, and Benson.
(V) Capstaff-Green.

(d) Density comparators.

(6) Physical densitometers.

(a) Thermoelectric.
(6) Photoelectric.
(c) Photovoltaic.

V. Interpretation of Results.

04) Speed or sensitivity.

(1) Threshold speed.

(a) Scheiner speed numbers.

(b) Eder-Hecht.

(2) Inertia speeds.

(a) H & D scale.
(&) Watkins scale.

(3) Luther's crossed wedge method.

(4) Minimum useful gradient.

(B) Gamma infinity, y^.

(D) Time of development for specified gamma.

(1) Td (y = 1.0).

(E) Latitude, L.
(/O Fog, F.

VI. Spectral Sensitivity.

(A) Dispersed radiation methods.

(1) Monochromatic sensitometers

(2) Spectrographs.

(a) Ordinary.
(6) Glass wedge.
(c) Optical wedge.

(B) Selective absorption methods.

(1) Tricolor.

(2) Monochromatic filters.

(3) Progressive cut filters.

Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 327

VI. SPECTRAL SENSITIVITY

Any treatment of the subject of sensitometry, especially in these
days of great popularity of panchromatic materials, cannot be con-
sidered as complete without a discussion of the various methods used
for the measurement and specification of the sensitivity of photo-
graphic materials to radiation of different wavelengths. This
characteristic is usually referred to as spectral sensitivity. A knowl-
edge of the way in which sensitivity is distributed throughout the
spectrum, including the ultra-violet and the "infra-red as well as the
visual regions, is of great importance from both the practical and
the theoretical points of view. The rendition of color by a photo-
graphic process is determined largely by the spectral sensitivity of the
negative material. For instance, it is well known that the ordinary
blue-sensitive materials, red, orange, and yellow, are rendered in about
the same tone value as black, while in the case of some panchromatic
materials, which have been rendered very sensitive to the longer wave-
lengths of visible radiation, these same colors may be rendered "as
almost white. The correct rendering of colored objects on the black
to white tone scale, which represents the entire discrimination gamut
of the photographic process, is conditioned almost entirely by the
spectral sensitivity of the material. It is evident, therefore, that a
knowledge of this characteristic of photographic materials is of great
importance wherever orthochromatic rendering of colored objects is to
be considered.

Many workers in the field of photography have studied this problem
of spectral sensitivity and, in fact, its investigation dates back almost
as far as the beginning of sensitometry. As early as 1882 Abney106
studied the spectral sensitivity of photographic materials by exposing
them to dispersed radiation in a spectroscope and by plotting densi-
ties, as determined by visual estimation, against wavelength ob-
tained a curve of spectral sensitivity. Abney later improved the
method until finally, in 1888, 107 he suggested a technic which involved
impressing on the same plate an exposure to a spectrum and a series of
known exposures to white light. The opacities of the spectrum ex-
posures were then measured and interpolated between those of the
white light exposures. He thus obtained a curve showing the equiva-
lent spectral intensities for various wavelengths.

The exposures of the photographic material to dispersed radiation
afforded by an instrument of the spectroscopic type gives information

328 LOYD A. JONES [J. S. M. p. E.

. which, while it is undeniably complete, nevertheless is not conveni-
ently expressible by simple numerical values but must be shown in
graphic form. Many methods have been developed, therefore, which
employ not spectroscopically dispersed radiation but spectral bands,
more or less narrow, isolated by selective absorption. This may be
accomplished either by using transmitting materials such as colored
glass, dyed gelatin, etc., or reflecting materials such as pigment coated
surfaces. One of the earliest of such color sensitometers was also
devised by Abney.108 Since that time almost numberless methods
have been proposed for the measurement and expression of color
sensitivity. No attempt will be made at this time to give a complete
bibliography of the subject but a few references to some of the more
recent work may be of interest. Leimbach109 has made a systematic
study of the spectral energy distribution for five different emulsions
in the region between 450 and 700 imx. He found the maximum
sensitivity to occur in the spectral region corresponding to 450 mju.
Luckiesh, Holladay, and Taylor110 have published sensitivity curves of
four emulsions indicating a maximum sensitivity near 450 mju.
Otashiro111 found maximum sensitivity at about 465 m/z, the sensi-
tivity decreasing uniformly through the blue and violet. Helmick112
(using an emulsion of the ordinary blue- sensitive type) has measured
the average number of quanta required to make a silver bromide grain
developable by radiation at various isolated wavelengths ranging
from 253.7 to 549.0 m/x. He found that the least number of quanta
per grain are required at wavelength 549 and the maximum number at
wavelength 253.7 imx. More recently Harrison113 has published
results showing the relation between sensitivity and wavelength for
six different photographic plates in the region between 200 and
450 mju. His results indicate that sensitivity is practically constant
for wavelengths greater than 250 imx, decreasing rapidly for wave-
lengths less than 250 m^i. He also shows the relation between
contrast (gamma) and wavelength.

All methods used for obtaining information as to the spectral sensi-
tivity of photographic materials involve the isolation of more or less
narrow spectral bands and then observing either qualitatively or
quantitatively the response produced when the material is exposed
to these more or less homogeneous radiations. For this purpose, a
wide variety of spectral instruments — monochromatic sensitometers,
spectrographs, tricolor tablets, ratiometers, color charts, and filter
assemblies — have been devised.

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY
DISPERSED RADIATION METHODS

329

The more refined and elegant methods involve the dispersion of
radiation by some suitable element, such as a prism or diffraction
grating. In this way radiation of high spectral purity may be ob-
tained to which the photographic material may be exposed. Instru-
ments of this type may be divided, for the sake of convenience, into

FIG. 51.

Diagram of optical system of monochromatic
sensitometer.

two general classes: the first including those instruments in which
the photographic material is exposed to the entire spectrum at the
same time; the second including those where a single very narrow
band of practically homogeneous radiation is isolated, to which the
photographic material is exposed in a manner similar to that used in
the sensitometers already described in the earlier sections of this
paper. While it is impossible to draw a strict line of demarcation,

330

LOYD A. JONES

[J. S. M. P. E.

the term "monochromatic sensitometer" is usually applied to instru-
ments of the second class, and the term "spectrograph" is used in
reference to those of the first class.

Monochromatic Sensitometer s. — The optical system employed for
isolating monochromatic radiation of high spectral purity as used in a
monochromatic sensitometer described by Jones and Sandvik114 is
shown in Fig. 51. This consists essentially of two quartz mono-
chromatic illuminators, A and B. The radiation emerging from the
exit slit of the first monochromatic illuminator, A, passes into the

FIG. 52. The sector disk of monochromatic sensitometer.

entrance slit of the second monochromatic illuminator, B. In this
way practically all the stray radiation may be eliminated so that the
radiation emerging from the exit slit of the second monochromatic
illuminator consists entirely of the wavelength as indicated by the
wavelength drums of the two instruments. Great emphasis must be
laid on the necessity of obtaining high purity for work of this kind,
especially if it is desired to work in those spectral regions where the
amount of energy obtainable from the light source used is relatively
low as compared with that present in other spectral regions. For a

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

331

complete discussion of this subject and of the relatively great errors
which may be introduced by failure to eliminate all stray radiation,
the reader is referred to the more complete discussion of the subject
by Jones and Sandvik (loc. cit.).

After having been isolated, the homogeneous radiation is allowed to
fall on the photographic plate and, by means of a suitable mechanism,
the time of exposure is varied in a known manner. It is usually
difficult to obtain a quantity of monochromatic radiation so that it
may be spread over a sufficient area of the photographic material to
permit the use of the conventional type of exposure time controlling
elements. It is usually necessary, therefore, to illuminate a relatively

FIG. 53. Schematic diagram of monochromatic sensitometer.

small area of the photographic plate and to make the exposures of the
various steps on the sensitometric strip one after the other, varying
the time factor of exposure in the desired manner. The method
adopted by Jones and Sandvik (loc. cit.) employs a sector disk of
special type in which the apertures are arranged spirally around the
axis of rotation, the entire disk being moved laterally, while rotating
at a uniform angular velocity. The structure of this disk is shown
in Fig. 52 and the arrangement of the essential parts of the exposing
mechanism is shown in Fig. 53. The shutter disk as shown is keyed
to the shaft carried by a movable bearing sliding between the ways
MM. The rotation of the lead screw H (driven by the same shaft
which imparts rotational motion to the shutter disk) moves the

332

LOYD A. JONES

[J. S. M. P. E.

shutter disk laterally while it is being rotated. Mounted on the shaft
carrying the shutter disk is a cam disk which carries a series of thirteen
cam elements. As these cam elements rotate with the cam disk they
close the electrical contact, 7, at definitely predetermined intervals.
The closing of this contact energizes the solenoid, Q, which, through
a suitable escapement, moves the photographic plate forward one
step during the time when the opaque element of the shutter disk
occupies a position in the path of the exposing radiation. By utilizing
the spiral arrangement of the apertures the maximum exposure time

0.8

0.4

"2.0 1A

I.&

O.Z.
I_OG>

O.fe

1/2.

\.Q>

FIG. 54.

0.0 04 O
/cm*)

D-log E curves obtained by exposing to monochromatic radiation of
wavelength 350 m/z.

corresponds to an angular rotation of 720 degrees of the shutter disk.
In this way twelve exposures increasing by consecutive powers of two
are obtained, thus giving a range of exposure times from 1 to 2048.
In this way an exposed sensitometric strip of the conventional type is
obtained.

By placing a thermopile in the exposure plane of the sensitometer,
that is, in the same position as that occupied by the photographic
material during exposure, the energy flux density of the monochro-
matic radiation may be measured. Since this energy value is usually
relatively small, it is necessary to use a thermopile-galvanometer

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

333

combination of high sensitivity, and great care must be exercised in
making the energy measurements in order to obtain reliable and
precise results. The determination of these energy values with the
required precision is perhaps the most difficult step in the process of
obtaining absolute values of spectral sensitivity.

The monochromatic sensitometer thus gives an exposed strip of the
conventional type except that the exposure values are expressed in
terms of energy (ergs) per unit area. This is developed under fixed
conditions, the densities read, and a curve plotted which, of course, is

O.O O.4

LOG E (ERGS. /cm1)

1.2.

FIG. 55. D-log E curves obtained by exposing to monochromatic radiation of

wavelength 450 m/z.

the density-log E characteristic for that particular wavelength of
radiation. In making a complete study of the spectral sensitivity
of the material it is necessary to expose several sensitometric strips at
each wavelength. These strips are then subjected to a series of
increasing development times, and in this way a complete family of
characteristic curves is obtained for each wavelength. These may
then be interpreted in the usual manner yielding the usual time of
development-gamma curve, time of development-fog curve, etc., for
each wavelength. By proceeding in this manner throughout the
spectrum, exposing a set of strips at a sufficient number of different

334

LOYD A. JONES

[J. S. M. P. E.

wavelengths, a complete set of data is obtained from which the various
spectral response curves may be plotted.

In Figs. 54, 55, 56, and 57 are shown typical families of D-log E
characteristic curves obtained by exposing a panchromatic material at
wavelengths 350, 450, 600, and 700 m//, respectively. Careful
comparison of these curves will show that the wavelength of radiation
used in making the exposures has a profound influence upon the
general shape of these curves. The development times used were 2.5,
3.5, 5, and 7.5 minutes, and it is quite evident from an inspection of

2.4

2.0

0.8

0.4

Z.O 1.4

O.TL O-Q> O.O 0.4 O.fc 1.7. i.«
LOG I

FIG. 56. D-log E curves obtained by exposing to monochromatic radiation of

wavelength 600 m/*.

these figures that gamma rises to a higher value in case of the exposures
made to the longer wavelengths than in case of those made to the
shorter wavelengths. While photographic materials differ to a certain
extent among themselves in their response to radiation of different
wavelengths, the effect mentioned is rather typical, although, of
course, there may be some exceptions.

In Fig. 58 are plotted the complete gamma-wavelength curves for
the four times of development as mentioned previously. It is
interesting to note that gamma increases from a minimum value at
the short wavelength end to a maximum at approximately 550 m/i,

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

335

decreasing from this point to a minimum at about 650 m/*, after
which it rises again as wavelength increases. There is relatively little
variation in gamma for the shortest time of development but if time
of development is increased the dependence of gamma upon wave-
length of the exposing radiation becomes much more marked.

In Fig. 59 is plotted a group of curves, one for each of the different
wavelengths as indicated, all of these being obtained by a single time
of development, namely, 5 minutes. It will be noted that there are
characteristic differences depending upon the wavelength of the

Z.4

t\&

ifl

0.8

0.4-

FIG. 57. ZMog E curves obtained by exposing to monochromatic radiation
of wavelength 700 m/z.

exposing radiation. For instance, the curves resulting from ex-
posures to the shorter wavelengths show lower maximum densities
and appreciably greater latitude than those obtained by exposures to
the longer wavelengths. In plotting these curves their relative
positions with respect to the log exposure axis must not be taken as
indicative of the sensitivity of the material to the various wavelengths
of radiation. They are assembled from left to right with increasing
values of wavelength in a convenient manner to show the relative
shapes. Their actual relationships to the log exposure scale are given
by the values of log exposure as indicated at the intersection of the

336

LOYD A. JONES

[J. S. M. P. E.

straight line portion of the characteristic curve with the log exposure
axis. These values are the log exposures corresponding to the respec-
tive inertia points.

The problem of expressing sensitivity must now be considered, and
it is obvious that the shape of the wavelength sensitivity curve will
depend profoundly upon the manner in which sensitivity is defined.
In expressing spectral sensitivity it is necessary to depart from the
method which has already been defined for expressing the speed or
sensitivity of a photographic material. It will be recalled that speed

06
0.4
OZ

feftO

FIG. 58. Gamma-wavelength curves for various times of development as

indicated.

value to heterogeneous radiation (white light) is defined, for ordinary
sensitometric purposes, in terms of exposure where exposure is
expressed in meter candle seconds. Now, the meter candle is a unit of
illumination and is measured visually or, even if measured by some
radiometric or physical method of photometry, is referred to as the
unit of luminous intensity, the international candle. For the expres-
sion of monochromatic sensitometric results it is obvious that this
method is quite useless. For instance, let it be assumed that mono-
chromatic radiation of wavelength 350 m/i is used. The eye is
entirely insensitive to radiation of this wavelength and hence the

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

337

luminous intensity, that is, candle power, of such radiation will be zero
no matter how great the radiant flux or radiant intensity (expressible
in units of radiant flux or radiant flux density) may be. For the
purpose of monochromatic sensitometry, it is necessary, therefore, to
express exposure in terms of energy units, and the unit most usually
used is the erg. Since the photographic material integrates more or
less perfectly the energy which falls upon it over a period of time, it is
necessary of course to include the time factor, and in expressing
photographic exposure in energy units it is necessary to multiply the
rate at which energy falls upon the surface (radiant flux density) by the
time during which the exposure persists. Exposure, therefore, must
be expressed in terms of ergs (or other suitable energy units) per unit

t 1.6

«•*

I4O 1^4 I 40 1.40 ZOO

LOG EXPOSURE.

FIG. 59. .D-log E curves for the various wavelengths as indicated; develop-
ment time 5 minutes.

area. The abscissa values in Figs. 54 to 57, inclusive, are therefore
in terms of log exposure where exposure is expressed in terms of ergs
per sq. cm. We may not express the sensitivity of the photographic
material for any particular wavelength in a manner analogous to that
used in white light (heterogeneous radiation) sensitometry. Thus we
may use the value of inertia, which now must be expressed in terms of
ergs per sq. cm. as a means of deriving a sensitivity number which of
course must be proportional to the reciprocal of the inertia value. Or
if it is desired to use any other method of speed expression, such, for
instance, as the exposure required to give a just perceptible density
(threshold speed) or the exposure required to give some minimum
gradient (gradient speed), this can be done; but it must be kept in

338 LOYD A. JONES [J. S. M. p. E.

mind at all times that exposure is not expressed in terms of meter
candle seconds.

Having available now information as to the sensitivity of the photo-
graphic material to radiations of different wavelengths, it remains to
consider a suitable method of expressing the spectral sensitivity. It
is quite possible of course to compute the sensitivity at various wave-
lengths in terms of reciprocal inertia. By plotting these reciprocal
inertia values as a function of wavelength a spectral sensitivity curve
will be obtained, and for many purposes such a method of expressing
spectral sensitivity seems to be quite satisfactory. It should be
pointed out, however, that spectral sensitivity may be expressed in
other ways, and it is possible that some of these may be somewhat
more useful from the practical point of view.

Referring now to Fig. 59 it is apparent that if spectral sensitivity
be defined in terms of the energy required to give a density of unity
for a fixed time of development, the shape of the spectral sensitivity
curve will be quite different from that based upon inertia. Moreover,
if a higher density value were chosen a still further modification in
the shape of the spectral sensitivity curve will be obtained. There
does not seem to be any means of deciding just what mode of express-
ing spectral sensitivity will be found most desirable from all points
of view and, in fact, it seems very probable that the method chosen
must depend upon the particular problem in hand. For theoretical
purposes there is considerable argument for defining spectral sensi-
tivity in terms of the energy required to give a density of unity when
development for all wavelengths is carried to a gamma of unity. For
practical purposes, however, it seems that the evaluation of spectral
sensitivity in terms of a fixed development time is more suitable and,
in order to discount somewhat the misleading effects of gamma varia-
tion, it seems probable that the determination of the energy per unit
area required to give a density of unity for a fixed time of development
is most satisfactory as a mode of expressing spectral sensitivity. The
most suitable development time is probably that which produces on a
sensitometric strip exposed to white light a gamma approximately
equal to that at which the material is usually developed in practice.
In Fig. 60 is shown a spectral sensitivity curve determined in this
manner. This is for high speed panchromatic motion picture film,
the development time used being that which gives a gamma of 0.7 on
a white light sensitometric strip.

It should be borne in mind that the spectral sensitivity curve, when

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

339

plotted in accordance with the specifications given in the last para-
graph, represents the characteristics of the photographic material
itself, quite apart from any consideration of the energy distribution in
the light source used. The curve as shown in Fig. 60 shows the
response of this material when used with a light source emitting the
same amount of energy at all wavelengths. In practice, light sources
used for photography depart appreciably from this condition of an
"equal energy" spectrum. Where it is desired to determine the
effective response of a photographic material when used with a light

I4O

izo

>ioo

TOO

FIG. 60. Spectral sensitivity curve for a high speed panchromatic motion

picture film.

source which is not emitting equal energy at all wavelengths, it is of
course necessary to compute a new relationship which may be termed
the photicity of the particular photographic material-light source com-
bination. In considering the rendition of colored objects in practice
it is very important to consider this effective response curve.

In order to illustrate the profound influence which the spectral
composition of the radiation may exert on color rendition, the curves
in Fig. 61 are shown. The dotted curve, A, represents the distribu-
tion of energy in the radiation emitted by an incandescent tungsten
filament operating at a color temperature of 3000 degrees, which is an

340

LOYD A. JONES

[J. S. M. P. E.

efficiency frequently met with in practice. The curve B is obtained
by multiplying, wavelength by wavelength, the ordinates of curve A
in Fig. 61 by the ordinates of the curve shown in Fig. 60. This then
becomes the photicity curve for a 3000-degree tungsten lamp as
evaluated by a high speed panchromatic material. It should be
noted that the relatively small amount of energy present in this radia-
tion at the shorter wavelengths produces a marked decrement in the
effective response in the longer wavelength part of the spectrum. This
accounts for the fact that with this combination of photographic

noo

FIG. 61. Curve A, spectral energy curve of incandescent tungsten at color
temperature 3000° K. Curve B, photicity curve for a 3000° tungsten lamp as
evaluated by a high speed panchromatic material.

material and light source, colors such as red, orange, and yellow are
rendered on the neutral tone scale by brightnesses which are relatively
too high as compared with their true positions on the visual brightness
scale.

Where a complete analysis of the spectral sensitivity characteristics
of a photographic material is required, the foregoing methods are
undoubtedly superior to any of the less perfect analyses which may be
obtained by the use of spectrographic records or by the use of the
various test chart methods relying upon selective absorption of dyes
or pigments. These latter methods, however, are frequently much

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

341

simpler and more rapid. In many cases they give results which are
quite significant and for some practical purposes entirely adequate.

Spectrographs. — In instruments usually referred to as spectrographs
the radiation from some suitable source is dispersed by means of a
diffraction grating or prism, and the spectrum thus produced is
allowed to fall directly upon the photographic material. The resul-
tant spectrogram gives considerable information relative to the spectral
sensitivity of the material. The method has the advantage of
simplicity and rapidity. These results usually are inspected directly
and estimates are made of the amount of sensitivity at various wave-
lengths. It is quite possible, of course, to obtain quantitative data by

Condenser

FIG. 62. Optical system of wedge spectrograph.

making densitometric measurements of the silver deposits, and under
certain conditions this method may yield data of a high order of pre-
cision. The usual forms of densitometers are not adapted for reading
the densities in these spectrograms and it is generally necessary to use
microdensitometers which are designed to measure the density of rela-
tively small or at least very narrow elements of the spectrogram.

It is necessary in work of this kind to know definitely the distribu-
tion of energy incident at various points on the photographic material.
This may be measured directly by means of a thermopile so arranged
as to pick up a relatively narrow line element in the spectrum plane.
The distribution of energy may in some cases be computed. This of
course presupposes a precise knowledge of the spectral emission

342 LOYD A. JONES [J. S. M. P. E.

characteristics of the light source and, furthermore, a complete
knowledge of the optical characteristics of the dispersing system, such,
for instance, as slit width, dispersion, losses due to reflection and
scattering within the optical system, etc.

As stated previously, the spectrographic method is of particular
utility where a graphic record meets all of the requirements of the
problem. By controlling the distribution of energy incident upon
the entrance slit, the spectrograph may be made to give directly a
graphic representation of the effective spectral response curve of the
photographic material and light source used. Instruments of this
kind are usually referred to as wedge spectrographs. The distribu-
tion of radiation on the entrance slit may be controlled by a rotating
sector of logarithmic form placed between the light source and the
slit of the instrument. In this way the energy incident upon the slit
can be made to decrease from one end of the slit to the other according
to a logarithmic law. Such rotating sectors, since they must be
quite small, are rather difficult and expensive to manufacture, and a
better solution of the problem is obtained by using a neutral gray
wedge placed directly over the slit of the spectrograph as proposed by
Mees and Wratten.115 The construction of such an instrument is
shown in Fig. 62. In this a diffraction grating is used which gives
normal dispersion and it is therefore considerably more convenient
than prism instruments which compress into a relatively small space
the long wavelength end of the spectrum and stretch out unduly the
short wavelength end. A suitably engraved scale plate is placed in the
plate holder so that during exposure the sensitive surface of the photo-
graphic material is in direct contact with this scale plate. In this
way the wavelength scale is printed directly on the spectrogram, thus
facilitating the reading of the results.

In Fig. 63 are shown examples of records obtained in the wedge
spectrograph with photographic materials having various spectral
sensitivities. The envelopes of the light portions constitute the
spectral response curves for the various photographic materials as
used with a particular light source. In the case illustrated the
quality of radiation used was approximately equivalent to that of
noon sunlight. It should be remembered in the interpretation of
these spectrograms that the wedge used over the slit has a linear
density gradient and therefore the distribution of radiation along the
slit decreases logarithmically from one end to the other. These
envelope curves therefore are in logarithmic form, and cannot be

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

343

compared directly with spectral sensitivity curves such as are illus-
trated in Fig. 60 where the ordinates are relative sensitivity, not the
logarithms of sensitivity. .

One other fact should be kept in mind in judging these spectro-
grams. The apparent decrease in sensitivity at the short wavelength
end is due to selective absorption in the neutral gray glass of which the
wedge is manufactured. So far as the author knows, all of the so-
called neutral gray glasses depart appreciably from neutrality in the
wavelength region below 440 imz, the absorption there being consider-

ORDiMKRV

ORTHOCHROMArriC

FIG. 63. Wedge spectrograms obtained with instrument illustrated in Fig. 62.

ably greater than throughout the rest of the visible spectrum. While
this is inconvenient it is not particularly serious since interest is
usually centered on the distribution of sensitivity for wavelengths
longer than 440 mju. It is well established also that photographic
materials differ relatively little among themselves in the distribution
of sensitivity in the region of shorter wavelengths. The reader should
be cautioned again to remember at all times that wedge spectrograms
made in this manner include not only the spectral characteristic of the
material but also the spectral emission characteristic of the source
used in illuminating the spectrograph.

344 LOYD A. JONES [J. S. M. P. E.

It is possible to avoid the undesirable selective absorption charac-
teristics of a neutral gray glass wedge by the use of a specially de-
signed non-spherical condensing system for the illumination of the
slit of the spectrograph. Such a condenser was proposed by Callier116
and a modified form of the Callier condenser has been used by
Hansen.117 More recently Miller118 has published a paper in which
an improved form of this condenser is described. Such a condenser

FIG. 64. Wedge spectrograms obtained with Miller's optical wedge spectro-
graph.

involves the use of a diaphragm which may be cut so as to give any
desired distribution of illumination on the slit of the instrument.
For photographic purposes a logarithmic distribution is usually most
desirable and it may be made either continuous or in the form of steps
as required. The form adopted by Miller is that of a stepped logarith-
mic diaphragm giving results as illustrated in Fig. 64 which represents
the spectral sensitivity of a panchromatic, an orthochromatic, and an
ordinary (blue-sensitive) photographic material. An inspection of
the spectrograms in Fig. 64 will show that they carry out into the

Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 345

short wavelength region much better than those obtained with the
neutral gray glass wedge instrument as illustrated in Fig. 63.

The use of a stepped wedge or diaphragm is particularly
advantageous where it is desired to make actual density measure-
ments and to obtain quantitative results from these wedge spectro-
grams. The short wavelength cut-off obtainable with an instrument
of this type is determined by the absorption of the glass lenses and by
the distribution of energy in the source used as an illuminant. In
case it is desired to extend the work into the ultra-violet region an
instrument with quartz optics may be used. In order to take ad-
vantage of this transmission, of course, it is necessary to use a light
source emitting radiation throughout the ultra-violet.

An inspection of wedge spectrograms yields a great deal of informa-
tion as to the distribution of sensitivity and also some qualitative idea
of the variation in gamma with wavelength of the exposing radiation.
They cannot be considered as satisfactory as the determinations
made by methods of monochromatic sensitometry, described in a
previous section, but where it is desired to have permanent compara-
tive records which can be obtained easily and without undue labor the
wedge spectrogram has much to commend it.

SELECTIVE ABSORPTION METHODS

The spectral sensitivity of a photographic material as determined by
the methods of monochromatic sensitometry and by the usual spectro-
graphic technic is most conveniently and almost necessarily expressed
graphically, the usual mode being a curve showing sensitivity as a
function of wavelength. It is almost impossible to express the
information relative to spectral sensitivity as derived by these methods
in brief numerical terms. The results of course can be shown in
tabular form in which the sensitivity values are given for certain
selected wavelengths, but in general such a tabulation is not particu-
larly convenient and is too complex and voluminous for practical
purposes of classification and record. In order to obtain a more
simple specification of color sensitivity which can be expressed by a
few numerical values, it is frequently convenient to depart from the
monochromatic method and determine the response of photographic
materials to relatively broad spectral bands. This may be ac-
complished with apparatus of the spectrographic type using, instead
of very narrow spectral regions, broad bands, each embracing a
relatively large proportion of the entire spectral range. If

346 LOYD A. JONES [J. S. M. P. E.

results of this type are desired it is usually much more convenient to
resort to methods of selective absorption for the isolation of the de-
sired spectral regions. Incidentally, the instrumental equipment
required for this work is much less expensive than that for the spectro-
graphic type. As mentioned previously, this method of obtaining a
numerical expression for spectral sensitivity is very old, being used
first probably by Abney in about 1895. 119 The test as devised by
Abney consisted of a tablet made of a series of colored glasses, each
transmitting a relatively broad spectral band and adjusted in such a
way as to give equal illuminations. A similar method was used by
Eder for testing the relative spectral sensitivity of orthochromatic
plates. In his earlier work the spectrum was divided into two parts,
one containing all wavelengths longer than approximately 495 m/z
and the other all wavelengths shorter than this value. This wave-
length represents approximately the long wavelength limit of the
sensitivity of an ordinary non-color sensitized material. The values
obtained by this method, therefore, give the ratio of the sensitivity
due to optical sensitizing as compared to that due to the inherent
sensitivity of the unsensitized silver halide. Later, Eder120 divided
the spectrum into three regions: orange-red, green, and blue- violet.
E. J. Wall121 also employed three selectively absorbing niters dividing
the spectrum into three parts similar to those used by Eder. Since
this early work almost numberless devices have been constructed and
used, employing colored glasses, dyed gelatin, colored solutions
or pigment coated surfaces for the isolation of more or less narrow
spectral bands. No attempt will be made to give a complete bibli-
ography of this subject but one or two of those methods which have
been most extensively used will be described and discussed briefly.

Tricolor Filters. — Probably the most widely used method of this
type involves the use of three filters having selective absorption so
adjusted as to divide the visible spectrum into three approximately
equal wavelength bands. As typical of such filters, the Wratten tri-
color sets may be mentioned, and, in fact, since this set of filters has
become almost standard throughout the world for three-color photo-
mechanical processes, the expression of color sensitivity in terms of
these three filters has become quite universal. In Fig. 65 are plotted
the spectral transmission curves of the three filters in question,
namely, Wratten No. 25 (A), red; Wratten No. 58 (B), green; Wratten
No. 49 (C4), blue. The red filter (No. 25) transmits quite freely all
radiation of wavelength greater than 600 m/z. It has a maximum

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

347

transmission of 80 per cent, and hence is quite efficient as a means of
isolating the third of the visible spectrum lying between 600 and 700
m/z. The green filter (No. 58) has a maximum transmission at wave-
length 520 m/z but at this wavelength it transmits only 60 per cent of
the incident radiation. Its transmission falls rapidly on both sides of
this point, the lower transmission limit being approximately 480 m/x
and the upper limit 600 imx. While its total transmission computed
on the basis of energy is relatively low, its transmission for white light
as measured visually is relatively high since the maximum of the
visual sensitivity lies at approximately 550 rmx. The blue filter (No.

140

IZO
£

h 40

/T\

300

AOO ?>OO <bOO

WA V E l_EI NGTTK

TOO

FIG. 65. Spectral transmission curves of Wratten tricolor filters.

49) has its maximum transmission at 455 m/i at which point
the transmission value is only 27 per cent. For both longer and
shorter wavelengths transmission decreases rapidly, the short wave-
length limit being approximately 360 m/x and the long wave-
length limit 500 imz. Evaluated in terms of visual transmission
for white light its efficiency is low, its transmission value determined
in this manner being only 0.5 per cent. It is relatively inefficient as a
means of isolating the third of the visible spectrum lying between 400
and 500 m/z, but since photographic materials in general are very
sensitive in this region its photographic transmission is quite high.

348

LOYD A. JONES

[J. S. M. P. E.

For the panchromatic materials which were in use up to one or two
years ago, this filter had a multiplying factor of 8 which is fairly
comparable to the multiplying factors of the green and blue filters as
measured in terms of these older panchromatic materials. For the
panchromatic materials recently introduced,which have a much higher
proportion of their total sensitivity concentrated in the green and red
regions, the multiplying factor for this filter is appreciably higher,
being of the order of 16.

(a)

(b)

E1/\R

u
o

D
U

0
1

(J)

FIG. 66.

Exposure = a 8a 8a 8a

(a) Tricolor tablet; (b) result obtained with tricolor filter method.

In practice, sensitometric results are obtained by using a tricolor
tablet, the structure of which is illustrated at (a) in Fig. 66. Strips
of these filters, which are manufactured in the form of dyed gelatin,
together with a strip ef plain (undyed) gelatin film No. 0, are
cemented between two sheets of glass. The dimensions of this
tablet are such that it just fits into the plate holder of a sensitometer
of the ordinary white light type, and the strips are of such width that
one sensitometric exposure is made through each of the four filters.

Mar., 1932] PHOTOGRAPHIC SfiNSITOMETRY 349

In order to obtain exposures through the tricolor filters, which balance
fairly well with that made through the clear filter, it is customary to
increase the time of the tricolor filter exposures so that it is eight
times as great as that made through the clear area.

At (&) Fig. 66 is shown a reproduction of the result obtained by
application of this method to a panchromatic material. Since the
relative exposures acting on each step of the sensitometric strip are
known, it is possible to estimate by inspection the relative exposures
required through each of the three tricolor filters to give the same
result as that obtained by the known exposure through the clear
filter. This is usually and most conveniently expressed in terms of
the filter factor for each of the tricolor filters, the filter factor being
defined as the number by which the exposure incident upon the clear
filter must be multiplied in order to obtain the value of the exposure
which must be incident upon the filter in question so that the photo-
graphic effect on the material exposed through that filter shall be
equal to that resulting from the exposure through the clear filter.
In the case illustrated, the exposure increases by consecutive powers
of 2 from step to step ; that is, it doubles for each successive step of
the sensitometric scale. The estimation of the filter factors from the
sensitometric exposure made in this manner will depend to some
extent on whether the equilization of density be made in the region
of extremely low densities, in the region of the half-tones, or in the
region of high densities. This, of course, is due to the gamma wave-
length effect which, while not great in this particular case, is sufficient
to affect the values of the estimated filter factors. While it may not
be possible in the half-tone reproduction of this tricolor exposure to
detect the small differences that were present in the original, it is quite
evident from an inspection of the original that it would be necessary
to multiply the exposure given through the blue filter by a factor of
approximately 2 in order to make it balance the exposure through the
clear filter. Since the exposure through the blue filter is already eight
times that given through the clear, it follows that the exposures which
would be required to produce balance between the blue and clear filter
strips must be in the ratio of approximately 1 to 16. The filter factor
for the tricolor blue filter, therefore, is 16. Likewise, judging the
green and red exposures, also in the low density (shadow) regions,
filter factor values of 8 and 6, respectively, are obtained. Now, if
equilization be transferred to the half-tone regions, a somewhat
different result is found. For instance, the factors for the blue, green,

350

LOYD A. JONES

[J. S. M. P. E.

and red filters are 32, 6, and 4, respectively. Equilizing the response
in the high density (highlight) region, values of 40, 6, and 3 for the
blue, green, and red are obtained. It is customary to make this
estimation of filter factor for the half-tone region, thus balancing up
to a certain extent the gamma-wavelength effect.

Using this method three numbers are obtained which are a fair
index of the spectral sensitivity of the material. It is possible, of
course, actually to read the densities resulting from these tablet
exposures and to plot the usual density-log E characteristic curves for

FIG. 67.

O.Z

LO& EXPOSURE:

Density-log E curves for the sensitometric strips illustrated in Fig.
66(6).

the four filters. This has been done for the particular sample of film
from which the illustration in Fig. 66 was made. The curves obtained
are shown in Fig. 67. In plotting these curves account has been
taken of the fact that the A, B, and C4 filters had exposures eight
times as long as those given through the clear filter. The curves are,
therefore, placed correctly in relation to the log exposure scale. Now
it is possible to determine by measurement the filter factors in terms
of the inertia values, or, by drawing a horizontal line through the
region of half-tones (D = 1.00 is used) and reading off the log E values
where this horizontal line cuts the four characteristic curves, the

Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 351

evaluation of filter factor can be made in terms of the half-tone
region. Results actually obtained in this manner are as follows:

D = i.o

Filter

Log E

Factor

Log E

Factor

No. 0

2.4

0.0025

. .

T.66

0.0457

No. 49

1.5

0.0316

12.5

1.04

1.097

24.

No. 58

T.3

0.0200

8.0

0.50

0.316

6.9

No. 25

1.2

0.0159

6.3

0.24

0.176

3.8

It will be noted that these do not check precisely the estimated
values already given but are of approximately the same order. There
is little doubt that greater precision can be obtained by reading the
densities and plotting the curves as illustrated in Fig. 67, but for all
practical purposes satisfactory values may be obtained by the estima-
tion process, especially if the observer has had some experience.

In Fig. 65 the dotted curve, D, is the spectral sensitivity curve of
the panchromatic material which was used in making the tricolor
exposure reproduced at (b) in Fig. 66. The tricolor ratio for this
material, as estimated by use of the densities lying in the half-tone
region, is 16-8-6, these numbers being, as mentioned previously, the
multiplying factors for the green, red, and blue filters, respectively.
This conveys some idea as to the correlation existing between the
spectral sensitivity of a photographic material, expressed in terms of
the sensitivity-wavelength function, as derived by the methods of
monochromatic sensitometry, and the tricolor ratio values, as derived
by the methods of selective absorption.

In order to illustrate further the significance of these tricolor ratio
values and to enable the reader to obtain a somewhat more definite
correlation of these values with the spectral sensitivity functions, the
data in Table XVI are given. These are the tricolor ratios obtained

TABLE XVI

Tricolor Ratios for Materials Differing in Spectral Sensitivity

Filter Factors
Material No. 49 No. 58 No. 25

Ordinary 4

Orthochromatic 6 20

Panchromatic (Type A) 8 14 12

Panchromatic (Type B) 10 6 10

Panchromatic (Type C) 16 8 5

by estimation in the half-tone region of tricolor exposures made on the
photographic materials of which the spectral sensitivity is shown

352 LOYD A. JONES [j. s. M. P. E.

graphically in terms of wedge spectrograms in Fig. 63. For the first
material which is sensitive only to blue radiation, the filter factors for
the green and red filters are extremely high and of no practical interest.
In the case of the second material which is sensitive only to blue and
green, the filter factor for the red is of no particular interest since it is
exceedingly great. For the three panchromatic materials the filter
factors are as shown and a little study of these and the wedge spectro-
grams in Fig. 63 will show the correlation between the two modes of
expressing spectral sensitivity.

Monochromatic Filters. — The general method of isolating spectral
bands by means of selectively absorbing materials, as described in the
previous section, may be elaborated considerably and thus provide a
more detailed analysis of spectral sensitivity. For instance, it is
possible to obtain filters transmitting very much narrower spectral
bands than those isolated by the tricolor filters already described. The
rather misleading name of monochromatic filters is sometimes applied
to filters which transmit relatively narrow spectral bands. It is very
difficult to obtain filters which transmit spectral bands less than
50 mju wide and which at the same time have sufficiently high trans-
missions at the wavelengths of maximum transmission to be of use for
practical purposes. By exercising great care, however, the visible
spectrum extending from 400 to 700 mju may be split into five or six
non-overlapping parts and it is possible in addition to isolate one or
perhaps two additional sections in the near ultra-violet between 300
and 400 m/z. Using selectively absorbing filters of this type and
applying the same general sensitometric procedure which has been
described under the tricolor method, six or eight numbers may be ob-
tained, which are of course the multiplying factors for these narrow
band transmitting filters. It is obvious that this gives a more
complete analysis of the spectral sensitivity characteristic and from
these values it is possible to obtain a fairly precise idea of the shape of
the sensitivity- wavelength function. The value of such a method
is somewhat doubtful since it lacks the convenience and brevity of the
tricolor method, by which the result is expressed in terms of three
values, and, furthermore, lacks the precision and completeness which
can be obtained by the methods of monochromatic sensitometry. So
far as the author is aware this method has not been applied to any
great extent and it would seem wise, if high precision and complete
data are required, to adopt directly the methods of monochromatic
sensitometry; while, if the more convenient and simple specification

Mar., 1932]

PHOTOGRAPHIC SENSITOMETRY

353

in terms of filter factors is adequate for the occasion, the tricolor
method seems preferable.

Progressive Cut Filters. — One other modification of the selective
absorption method has been used and advocated by some workers in
this field and has some merits. This involves the use of a series of
filters which cut progressively at shorter and shorter wavelength
values. As mentioned in the previous section, it is very difficult to
obtain filters which transmit narrow spectral bands efficiently. The
total transmission of such filters is usually relatively low; hence the

500
WAVELENGTH

TOO

FIG. 68. Curves 1 to 8, inclusive, spectral transmission curves of a set
of progressive cut niters. Curve D, spectral sensitivity curve of panchro-
matic material.

illumination which can be applied to the surface of the sensitive
material is low and, therefore, exposure times are relatively long.
This difficulty can be avoided by adopting the filters of the progressive
cut type. Such a set is illustrated in Fig. 68. The filter represented
by curve No. 1 transmits quite freely all radiations of wavelength
greater than approximately 660 rmz. Exposures made through such a
filter, therefore, utilize only that portion of the spectral sensitivity
of the photographic material which lies on the long wavelength side of
the filter cut. The dotted curve D (Fig. 68) again represents the
spectral sensitivity curve of a highly panchromatic material, and an

354 LOYD A. JONES [J. S. M. P. E.

inspection of the figure will show to what portion of this sensitivity
any exposure made through filter No. 1 is due. Filter No. 1, of course,
is a deep red color. The cut of filter No. 2 moves over to approxi-
mately 610 m/x, that of No. 3 to 590 m/x, No. 4 to 540 m/z, No. 5 to
500 mju, No. 6 to 450 imx, No. 7 to 390 m/x, No. 8 to 350 imz. Each
filter, therefore, includes a somewhat greater portion of the spectral
sensitivity of the photographic material. Exposures made through
such a set of filters result in a series of sensitometric curves similar to
those shown in Fig. 67. The integrated sensitivity for each succes-
sively widened spectral transmission band may be determined as
previously, either in terms of the exposure corresponding to the inertia
point or in terms of exposure required to give some constant density
value, for instance, D = 1.0. By setting up a series of simultaneous
equations and inserting the sensitivity values derived from the
individual density-log E curves, a series of numbers can be obtained
which represent the integrated sensitivity within a computed wave-
length region, which, of course, depends upon the transmission
characteristics of the filters. In this way a fairly good approximation
to the spectral sensitivity curve may be obtained, although it is
quite impossible to hope to locate all the maxima and minima which
may actually occur in a function of this type.

In general, the same criticism applies to this method as to the
monochromatic filter method. The results are expressed in terms of
a relatively large number of numerical values and hence lack the
convenience and brevity of the tricolor filter method. On the other
hand, the results obtained are not as precise or complete as those
derivable by means of monochromatic sensitometry. There are cases
of course where the expense of the equipment required for mono-
chromatic sensitometry is prohibitive, and in such cases the pro-
gressive cut filter method may be very desirable since it offers a
somewhat more complete analysis of spectral sensitivity than can be
obtained by the tricolor method.

REFERENCES

108 ABNEY, W. DEW.: Phot. J., 6 (1882), pp. 136, 154.

107 ABNEY, W. DEW.: Phot. J., 13 (1888), p. 2.

108 ABNEY, W. DEW.: Phot. J., 19 (1895), p. 328.

109 LEIMBACH, G.: Zeit. Wiss. Phot., 7 (1909), p. 157.

110 LUCKIESH, M., HOLLADAY, L. L., AND TAYLOR, A. H.: J. Frank. Inst., 196
(1923), p. 495.

111 OTASHIRO, T.: Bull. Kiryu Tech. College, Japan (Aug., 1923), No. 2.

Mar., 1932] PHOTOGRAPHIC SfiNSITOMETRY 355

112 HELMICK, P. S.: J. Opt. Soc. Amer., 6 (1922), p. 998; 9 (1924), p. 521.

113 HARRISON, G. R.: J. Opt. Soc. Amer., 11 (1925), p. 341.

114 JONES, L. A., AND SANDVIK, OTTO: /. Opt. Soc. Amer., 12 (1926), p. 401.

115 MEES, C. E. K., AND WRATTEN, S. H.: Brit. J. Phot., 54 (1907), p. 384;
Phot. J., 49 (n. s. 33) (1909), p. 235.

116 CALLIER, A.: Brit. J. Phot., 60 (1913), p. 972.

117 HANSEN, G.: Z. Physik, 29 (1924), p. 356.

118 MILLER, O. E.: Rev. Sci. Instr., 3 (1932), p. 30.

119 ABNEY, W. DEW.: Phot. J., 19 (1895), p. 328.

120 EDER, J. M.: Phot. Korr. 39, (1903), p. 426.

121 WALL, E. J.: Brit. J. Phot., 51 (1904), p. 926.

ERRATUM

The author regrets having to point out an error in the wording of the paragraph
of the paper Photographic Sensitometry , Part I, on page 519 of the October, 1931,
issue of the JOURNAL, beginning "A very ingenious device. . . " to and includ-
ing "...approximately 1 to 1,000,000 for each wedge." The paragraph in
question should read as follows:

"An ingenious method of obtaining directly the characteristic curve of a
photographic material was suggested by R. Luther20 in 1910, using a square
neutral gray wedge behind which the photographic material under test is exposed.
The resultant negative, developed preferably to high contrast, after having been
rotated through 90 degrees with respect to its original position, is placed in
register with the tablet through which the exposure was made so that the lines
of equal density on the negative are perpendicular to the lines of equal density
on the tablet. By direct observation of this tablet-negative combination the
density-log E characteristic may be seen. Or, by making a print, preferably on
a high contrast material, a permanent record may be obtained."

STROBOSCOPIC AND SLOW-MOTION MOVING PICTURES
BY MEANS OF INTERMITTENT LIGHT*

H. E. EDGERTON**

Summary. — In a paper published in the June issue of the JOURNAL the author
showed that mercury-arc lamps when excited by quick violent electrical transients
make a practical source of intermittent light which is very actinic and has a short
duration of flash. The timing of the flashes is easily controlled.

In this present paper, further information regarding the duration and the quality
of the light are givsn. Also improvements upon the mercury-arc tubes are described
which simplify the construction of the light-pulse tube and the electrical circuits.

Uses of intermittent light for taking motion pictures are described and illustrated
by examples. There are in general two methods of using the intermittent light.
One method is used to take pictures where the light is caused to flash for each frame
and the film runs at a continuous speed. The second is used to take stroboscopic
moving pictures of rapidly moving objects by causing the light frequency to approach
the frequency of the motion of the object. Examples of the later method are shown,
these being stroboscopic motion pictures of a crude motion picture claw mechanism
operating at 30 fps. and of the surges in the valve springs of a gasoline engine
running at 1930 rpm.

In a paper published in the June issue of the JOURNAL, the author
discussed briefly some of the possibilities of the use of intermittent
light in taking motion pictures and described how the mercury-arc
tube could be used to produce intense light of short duration. It
was shown how the method was used to study the angular swinging of
synchronous machines. Since then considerable improvement has
been made in the source of intermittent light and several uses for the
application of the light have been made which are of interest to motion
picture engineers.

Before discussing the applications of the intermittent light to
motion pictures, some data will be given regarding the mercury-arc
tube as a source of intermittent light.

The circuit for producing intermittent light from mercury-arc
tubes has been modified so that nearly any type of mercury-arc lamp

* Presented at the Fall, 1931, Meeting at Swampscott, Mess.
** Massachusetts Institute of Technology, Cambridge, Massachusetts.
356

STROBOSCOPIC PICTURES

357

can be used. This circuit eliminates the necessity for the holding-arc
electrodes, the auxiliary anodes, the grid around the main anode, and
the mercury condensation chamber, all of which were required for
tubes to be used with the circuit which was described before. Tubes
now need an anode at one end, a small pool of mercury at the other,
and an external electrode around the mercury pool. The shape of the
tube may have practically any form. A very convenient shape to use
is a long slender tube which may be placed at the focus of a parabolic
reflector in order to concentrate the light.

The necessary elements and arrangement of a variable frequency
source of intermittent light are shown in Fig. 1. A source of d-c.
power is connected to the light-pulse tube through a choke and a
resistor which are large enough to hold back the current until the tube
has de-ionized, and still are small enough to allow the condenser to

FIG. 1 . Elementary wiring diagram showing an arrange-
ment to produce intermittent light from a mercury-arc
light-pulse tube.

charge in time for the next flash. The condenser is connected in
parallel with the light-pulse lamp, and before a flash the condenser is
charged so that the anode is positive. The tube is started by applying
a sudden high voltage to the external connection around the mercury
pool. This is conveniently accomplished by using a step-up trans-
former through which a small condenser is discharged by means of a
switch or a small thyratron.

TIME DURATION OF THE FLASH OF LIGHT

One of the most important qualities of the light is the quickness of
its flash. From a practical consideration the exact time of discharge
is not of importance except that it must be less than a certain mini-
mum. This minimum allowable time of flash depends upon the
specific use which is being made of the tube.

358 H. E. EDGERTON [j. S. M. p. E.

If photographs of a moving object are being taken, the minimum
allowable time of flash must be such that there is no appreciable
motion. If motion -picture photographs are being taken or projected
by means of intermittent light, the minimum allowable time of flash
must be such that there is no appreciable blur on the film or screen.

Many factors influence the time of flash. Before enumerating
these a brief discussion of the electrical transients, which are the source
of the stroboscopic light, will be given. A condenser is charged to a
certain voltage and then is discharged at the desired instant through
the mercury-arc tube. This discharge is quite violent and quick and
causes a pulse of light to be emitted from the tube.

When a condenser is discharged through a constant resistance the
current rises to a value determined by the voltage across the con-
denser divided by the resistance. The current then decreases
exponentially at a rate determined by the time-constant of the circuit,
which is smallest for a small resistance. The mercury-arc tube acts
somewhat like a small resistance and thus the condenser is quickly
discharged. The light is some function of the current and thus of
time. Another influencing factor in addition to the resistance of
the tube and the leads is the inductance of the leads, and this tends to
make the discharge oscillatory, helping to de-ionize the mercury-arc
tube so that it will not conduct while the condenser is being charged
for the next flash.

Factors that influence the time of discharge and which are believed
to tend to increase the time are:

1. Resistance of the leads to the tube.

2. Inductance of the leads to the tube.

3. High temperature of the tube.

4. Long tube dimensions.

5. Large discharge capacity. .

6. Low voltage on the condenser.

The first two do not contribute very much to the time of discharge
for the usual arrangement. The third — temperature — may increase
the discharge time materially but, on the other hand, a hot tube gives
out much more light than a cold one for the same amount of electrical
energy input to the circuit. The exact effect of tube dimensions for
such transients has not been investigated thoroughly to the author's
knowledge. The remaining two factors — capacity and voltage — are
somewhat related since the energy of a flash is proportional to the
energy stored in the condenser, that is, E2C/2 joules. A large capacity

Mar., 1932] STROBOSCOPIC PICTURES 359

requires a longer time to be discharged in a circuit of linear resistance,
and it is believed that somewhat the same phenomena are involved
with the mercury-arc tube. A high voltage tends to speed up the
discharge especially since it aids in ionizing the gas in the tube so that
it will start more quickly.

A revolving drum camera, built by Mr. C. S. Draper of the
Aeronautics Department of the Massachusetts Institute of Tech-
nology, was used for the following experiments to determine the time
of flash. A drum one foot in diameter was rotated at a speed of
1800 rpm. The stroboscope tube was covered, except for a narrow
slit, by a piece of cardboard. The linear velocity of the periphery of
the drum was 1800/60 X 12 X TT = 1130 inches per second; or one
inch = 1/1130 sec. = 0.00088 sec.

The data from several runs are tabulated in Table I. These data
are approximate but do show the influence of temperature.

TABLE I

Tabulation of Data Giving the Time Duration of the Light Flashes

Length of 63% Time Constant

Temperature of Capacity in Condenser of Exposure on of Flash in

Tube Estimated Microfarads Voltage Film in Inches Microseconds

125 °C 2 1350 0.06 53

40° C 2 1350 less than 0.01 less than 9

75° C 4 1350 0.02 18

Tube dimensions: 2 ft. long, 20 mm. in diameter. Leads from condenser to
tube consist of 12 ft. of lamp cord.

QUALITY OF THE LIGHT FROM THE TUBE

The spectrum of the light from a light-pulse tube is radically
different from the spectrum of the light from the same tube excited
with direct current. The violent electrical discharge excites many of
the enhanced or spark lines in the spectrum. As a result there are ten
or so additional lines in the red and a great many additional ones in
the green besides other lines. The appearance of the light to the eye
is yellow- white, which is quite pleasing when compared to the ghastly
blue of the ordinary mercury-arc lamp.

The two spectrograms shown in Fig. 2 were taken by Mr. W. E.
Albertson through the courtesy of Professor G. R. Harrison of the
Physics Department of the Massachusetts Institute of Technology.
These photographs show the spectrum of an ordinary mercury-arc
lamp (lower) and the spectrum of the intermittent mercury-arc tube

360

H. E. EDGERTON

[J. S. M. P. E.

(above). The additional lines may be observed by comparing these
two spectrograms. The exposures of these two spectrograms was

Red Green

Blue Violet

Stroboscope arc Nonex tube

A 5461 4359 3653 3132 2804

D-C. arc in quartz tube

FIG. 2. Spectrum of the light from a quartz mercury-arc lamp and from
a light-pulse stroboscope tube of nanex glass.

made so that the main arc lines of the two would have approximately
the same intensity.

MOTION PICTURES WITH INTERMITTENT LIGHT

In general, there are two methods of taking motion pictures with
intermittent light. One method is to synchronize the light with the
position of the film so that the exposure is properly placed on continu-
ously moving film, or so that one flash of light will occur when the
shutter is open for the ordinary method of taking pictures. The
second method is to synchronize the light with some rotating or
vibrating object which is to be photographed. Exposures are then
obtained by random coincidence of a flash of light and an open
shutter, it being possible to get both more than one exposure on one
frame or none, depending upon the frequency of the light flashes, the
exposure angle of the shutter, and the speed of the framing mechanism .
This will be discussed more completely later.

Little needs to be said of the first method. For this the intermit-
tent light is caused to flash at the proper time by the camera mecha-
nism so that the frames are properly spaced if they are to be projected.
The flash of light needs to be short enough so that the film does not
move an appreciable distance while it is on, say, for instance, one
thousandth of a frame.

The upper limit of film speed for 16-mm. film with a light whose
duration is ten microseconds, allowing a motion of one-thousandth
of a frame, is calculated below:

The film moves 7.5/1000 mm. in 10 microseconds, whence its velocity
is (7.5 X 106)/(1000 X 10) = 750 mm. per second, which corresponds
to 100 frames per second. Allowing the film to move one-hundredth

Mar., 1932]

STROBOSCOPIC PICTURES

361

of a frame while the light is on increases the maximum allowable
speed to 1000 frames per second.

The second method — that of synchronizing the light with a moving
or vibrating object — is very useful in taking slow motion pictures
(stroboscopic) of rapidly moving mechanisms. Say, for instance, it is
desired to take a moving picture of the claw mechanism of a motion
picture projector or camera while it is operating at normal speed.
Obviously, it is impossible to get such a picture without a camera that
will take at least eight frames while the claw mechanism completes its
cycle of operations. This calls for a 192-frame-per-second camera if
the speed of the claw is at 24 frames per second. The pictures would
not be very clear since the claw would move quite a distance during
the exposure.

FIG. 3. Diagram showing the relation between the
lowest light frequency and the camera speed and shutter
exposure angle to prevent blank frames.

The motion of such claw mechanism is easily photographed with a
motion picture camera if the claw is illuminated with intermittent
light which has a frequency slightly different from that which cor-
responds to the speed of the claw mechanism. The claw is seen once
each revolution in a slightly different position and as a result it ap-
pears to be moving at a slow speed. This is the well-known strobo-
scopic effect which has been used for studying moving mechanisms of
all sorts. The mercury-arc tube is powerful enough to produce
sufficient light to take motion pictures by this means.

As has been mentioned previously, the exposure of a film in an
ordinary motion picture camera does not depend upon the shutter
angle or speed of framing when intermittent light is used as an
illumination source. The exposure is entirely determined by the

362

H. E. EDGERTON

[J. S. M. P. E.

\
\

JL.

amount of light in the flash from the tube. There are several possi-
bilities which must be kept in mind regarding film speed and light
frequency when taking stroboscopic motion pictures. The ideal
method is to control the speed of the camera so that only one exposure
occurs on each frame, but this is not possible with the constant-speed,

spring-driven cameras. The exposure
ratio between one and two flashes is not
objectionable in a projected picture, but a
blank frame causes a flicker which disturbs
the continuity of the events. Fig. 3 shows
the relation that exists between the light
frequency and that of the camera mecha-
nism to prevent blank frames, this relation
being that the lowest frequency of the
light should be equal to or greater than
the frequency of the camera (frames per
second) multiplied by the percentage of
time that the shutter is open. This limit-
ing condition spaces two light flashes so
that if one occurs when the shutter has just
opened, the other will occur when the
shutter has just closed.

Since the shutter does not open and
close instantaneously and because the light
is practically instantaneous, it is possible
to get an incomplete picture. Frame No.
12 in Fig. 4 is one of these.

EXAMPLES OF STROBOSCOPIC MOTION PICTURES
(STROBOGRAMS)

The motion of a crude 35-mm. claw
mechanism was photographed with a 16-
mm. cine kodak, using intermittent light
which was of a slightly different frequency
than the claw mechanism. The moving

pictures were taken with an //1. 9 lens on the standard film at 16
frames per second. The claw mechanism which was photographed
was operating at about 30 frames per second. Fig. 4 shows an en-
larged section of this film.

The first three frames of Fig. 4 show the claw as it pulls the film
down. For the frame numbered 4, the mechanism has started to pull

FIG. 4. Stroboscopic
motion pictures of a claw
mechanism operating at 30
frames per second, taken
with a 16 fps. camera.

Mar., 1932J

STROBOSCOPIC PICTURES

363

away from the film. Frames numbered 5 to 12 show this drawing
back in its various stages. The lag of the spring due to its inertia is
easily observed. The next frames (13 to 21) show the return stroke
of the claw. The inertia of the spring here causes it to be bent back
the other way. Frame 20 shows the spring just after it has touched
the guide and the end of it has bounced back. The remaining frames
complete the cycle of events, showing the claw as it pulls the film
down.

Twenty-four frames are shown in Fig. 4 of a phenomenon that

FIG. 5. Photograph of the stroboscope
arranged to take motion pictures of the
surges of the valve springs of an experi-
mental engine.

occurs in one-thirtieth of a second, and so the apparent speed of the
camera is 24 X 30 or 720 frames per second. This stroboscopic
method is good only for mechanisms which are periodic, but it is very
useful for this purpose.

As a second example, the oscillations or surges of a pair of valve
springs were photographed in the Aeronautical Power Laboratory at
the Massachusetts Institute of Technology with the cooperation of
Mr. C. S. Draper and Mr. Towner. The stroboscope with its para-
bolic reflector is shown in Fig. 5 together with the experimental
engine whose valve mechanism was photographed. The pictures

364 H. E, EDGERTON

show the rocker arm slowly going up and down, followed by compres-
sion waves traveling back and forth through the spring. Three
enlarged pictures which were selected from a 35-mm. motion picture
film are shown in Fig. 6. The top picture of the left spring shows
the coils open at the top and compressed at the bottom. The

FIG. 6. Three enlarged strobo
scopic photographs from a 35-mm.
film of surges in a valve spring.

middle picture shows the spring still opened more between the top
coils than the bottom. The lower picture shows the spring coils
widely separated at the bottom and compressed at the top. The
time of exposure for each of these pictures was about 0.00001
second.

SOUND IN THE LOS ANGELES THEATER— LOS ANGELES,

CALIF.*

D. M. COLE**

Summary. — The sound reproducing equipment used in the Los Angeles Theater
is described in a general manner. Many refinements have been used in this in-
stallation, including aids for the hard of hearing, broadcast pick-up, and a public
address system, which enable the exhibitor to furnish better entertainment and more
comfort to the patrons. Means are provided for reproducing the picture and the
accompanying sound in the lounge, and provision is also made for disk reproduc-
tion, in addition to film reproduction. A reproducer set is also provided for the
reproduction of non-synchronous commercial records, making possible the running
of continuous programs for entrance music, exit music, and sound effects.

The trend in modern theater construction is toward larger and
better equipped theaters. Mechanical and electrical devices, which
enable the exhibitor to furnish better entertainment and more comfort
to patrons, are being used increasingly in new theaters, refinements
being added as they become available.

The Los Angeles Theater is an example which included in its
construction and furnishings all available refinements. The acoustic
properties of the theater were given careful consideration and, hand in
hand with good sound equipment, excellent results are being achieved.
In addition to the sound picture equipment, various attachments and
special features have been provided. The sound facilities include
sound picture reproduction, both film and disk for three projectors,
hard-of -hearing aids, non-synchronous attachment, broadcast pick-up,
and public address systems. Fig. 1 is a view of the equipment installed
in the projection room. The amplifiers and control panels are mounted
on five racks, centralizing all the panels, with the exception of the
public address control equipment, which is located in a room adjacent
to the projection room. Two sets of amplifiers are provided, permit-
ting simultaneous reproduction of two programs; i. e., while sound
pictures are being shown in the theater auditorium, announcements

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Electrical Research Products, Inc., New York, N. Y.

365

366

D. M. COLE

[J. S. M. P. E.

can be made to other parts of the theater, if required. The duplicate
set of amplifiers insures sound in the theater auditorium at all times.
Switches have been used throughout in this installation, with the
exception of the inputs connecting the microphones to the mixing
panel, where jacks are used. Monitoring facilities for both systems
are provided. Loud speakers of various types to fit the particular
purpose are installed about the theater to care for the distribution of
programs.

Sound Picture Equipment. — The sound picture equipment is of
the largest type of Western Electric equipment supplied for de luxe

FIG. 1. View of equipment installed in projection booth.

theaters. The amplifier equipment consists of a voltage amplifier, a
medium power amplifier, and two high power amplifiers. The ampli-
fiers, with the exception of the voltage amplifiers, have "built-in"
rectifiers and filters which furnish plate supply from alternating
current. The plate current of the voltage amplifiers is obtained
from the rectifier of the medium power amplifier with which it is
associated. The filament supply for the medium and high power
amplifiers is obtained from 110 volts a-c. stepped down to the proper
voltage. The filament supply for the voltage amplifier is obtained
from a motor generator set. Horn control panels are provided for
impedance matching and testing of the horn receivers. Pick-up

Mar., 1932]

SOUND IN Los ANGELES THEATER

367

equipment is provided to permit the reproduction of either film or
disk records on any one of three projectors. This equipment is of
the universal base type.

Three shallow type stage horns, each equipped with two receivers
are used behind the screen for obtaining correct illusion and distribu-
tion of sound. A large sound screen 60 by 40 feet, having a good

FIG. 2. View of the miniature screen which enables patrons
to see in the lounge the picture simultaneously projected in the
theater auditorium.

frequency transmission characteristic and good light reflecting quali-
ties, is installed.

The volume of sound is normally controlled in the projection room
but an auxiliary fader is available for use in various locations in the
auditorium. The auxiliary fader is used for previews and premiere
openings where special attention to volume is essential.

In the Grand Salon a miniature screen is provided which enables
patrons to view the picture which is being shown simultaneously in

368 D. M. COLE [j. s. M. P. E.

the theater auditorium. This is shown in Fig. 2. The accompany-
ing sound is reproduced by a loud speaker which is located above
the screen behind the grille work.

Loud speakers are provided in two "cry rooms," enabling those
viewing the picture from this point to hear the accompanying sound.

Hard-oj f -Hearing Aids. — Hard-of -hearing aids enable partially deaf
patrons to hear both the sound picture reproduction and stage pro-
grams. Single receivers, provided with head bands, are employed.
A regulating device in the cord permits the patron to adjust the
volume to suit his need. The cords are equipped with plugs which
are plugged into receptables installed on the arms of the seats. An
a-c. operated amplifier, which obtains a small speech input voltage
from one of the system amplifiers, furnishes the power for these
receivers and precludes the possibility of short circuits in the hard- of -
hearing aid attachment from interfering with the operation of the
system with which it is associated.

N on- Synchronous Attachment. — For the reproduction of incidental
music recorded at 78 rpm., a reproducer set is installed in the projec-
tion room. Two turntables with a fader make possible the running
of continuous programs for entrance music, exit music, and sound
effects.

Radio Broadcasting Feature. — Two amplifiers are provided to
furnish programs over telephone lines to radio broadcasting stations.
Programs from any of the microphone pick-up points, including the
broadcasting studio, can be transmitted. The amplifiers are all a-c.
operated and the necessary impedance matching and isolating trans-
formers are provided.

Public Address. — The public address portion of this installation
consists of high quality microphones of the condenser type, with their
associated amplifiers, control equipment, voltage and power ampli-
fiers, switching panels, and loud speakers of various types. Micro-
phone outlets are provided for pick-up from the footlights, stage, the
orchestra pit, broadcasting studio, foyer, check-room, and lobby.
Provision is also made for a hanging type microphone over the
orchestra pit. Suitable mountings are provided, depending on the
location in which the microphones are used and the function which
they perform. The microphones are of the same type as those used
in field and studio recording. A 200-volt dry battery is provided to
furnish the polarizing voltage for the condenser microphones and the
plate supply for their associated amplifiers. The low voltage supply

Mar., 1932] SOUND IN LOS ANGELES THEATER 369

for filament currents and grid potentials is obtained from the filtered
output of a motor generator set. The amplifier associated with each
condenser microphone is so constructed that it is not disturbed by
shocks, this being accomplished by means of spring suspension
construction. The microphone pick-up control panel is located in a
room adjacent to the motion picture room. From this point, the
operator can observe the results of amplifying speech or music in the
auditorium. The mixing facilities enable the operator of the public
address equipment to blend the output of any three microphones, as
required. Standard studio equipment is provided for this purpose.
The public address amplifying equipment consists of two voltage
amplifiers, a medium power amplifier, and two high power amplifiers.
It should be noted here that additional voltage amplification over that
needed for sound picture reproduction is required for public address
work. The power required to operate the public address amplifiers is
obtained in the same way that the power for the sound picture
amplifiers is obtained. The medium power amplifier is capable of
furnishing the plate supply for two voltage amplifiers.

For general reinforcement work in the theater, large horns equipped
with high quality receivers, are located over the proscenium arch and
in the right and left organ grilles. During operation, the volume is
maintained at a point which creates the illusion to the patron that
the reinforced sound is coming from the real source. The relation of
the horns to the pick-up source is very important, and in general it is
essential that the horns be located directly over and a little forward of
this point. The dynamic loud speakers, installed in the "cry rooms,"
the Grand Salon, the Main Lounge, and the foyer, furnish incidental
music to patrons entering and leaving the theater and to those waiting
about for one reason or another.

Power Supply. — The low voltage power supply for the entire
installation is obtained from two motor generator sets with associated
filters. The motor generator sets can be used interchangeably and, in
emergency, either would handle any load which might be required to
keep the show running. They furnish low voltage to the condenser
microphone amplifiers, the voltage amplifiers, the film reproducer
amplifiers, signal circuits, and the fields of the horn receivers. The
remainder of the equipment, including the power amplifiers, operates
from the standard power supply. A voltage control cabinet is pro-
vided to care for fluctuations in the line voltage.

Conclusion. — All the equipment is of the very highest quality,

370 D. M. COLE

from a mechanical and voice and music transmission standpoint.
With the service rendered by the supplier of the equipment and
the excellent work of the theater personnel, the system has been kept
in operation with a minimum of trouble, in spite of the fact that the
theater operates during long hours. Close cooperation between the
theater management and the manufacturer of the equipment insures
maximum use of the equipment, particularly the public address and
special features.

There are indications that the larger first-class motion picture and
legitimate theaters will soon be equipped with facilities similar to
those enjoyed by the patrons of the Los Angeles Theater.

THE REDUCING ACTION OF FIXING BATHS ON THE
SILVER IMAGE*

H. D. RUSSELL AND J. I. CRABTREE**

Summary. — The extent of the reducing effect of fixing baths on the silver image
during the progress of fixation is greater than has been generally supposed. For
example, in sensitometric work it is dangerous to prolong the fixation of motion picture
positive film in the average fresh potassium alum fixing bath beyond 5 minutes at
65° F. and with certain highly acid chrome alum baths a measurable degree of re-
duction occurs even in this short space of time.

In regular laboratory work the degree of reduction which takes place in the normal
time for fixation is usually of no practical importance with the baths in common
use. In any given bath the rate of reduction increases with the acidity, the tempera-
ture of the bath, and degree of agitation of the film.

During use, the reducing action of a fixing bath falls off because it becomes more
alkaline and accumulates silver thiosulfate which tends to retard the reduction.

In order to insure the minimum degree of reduction, baths having a minimum
degree of acidity should be used although such baths have a short life and often do
not harden satisfactorily. It is therefore necessary to revive such baths either by
adding further quantities of acid or hardening solution at intervals during use,
otherwise if the film is not rinsed in water before fixing an objectionable sludge will
form in the fixing bath.

The nature of the reduction with the negative emulsions tested was found to be
almost strictly proportional, and some of the more active baths enumerated could
therefore be used advantageously for reducing the contrast of photographic images.

OUTLINE

I. Experimental methods.

II. Degree of reduction of the silver image in various fixing baths.

III. Effect of reduction on shape of characteristic curve.

IV. Factors affecting the rate of reduction.

(A) Composition of fixing bath.

1. Acidity (/>H) of bath.

2. Sulfite concentration.

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Communica-
tion No. 484 from the Kodak Research Laboratories.

** Kodak Research Laboratories, Eastman Kodak Co., Rochester, N. Y.

371

372 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

3. Hypo concentration.

4. Hardener concentration.

5. Nature of hardening agent.

(B) Temperature of bath.

(D) Age of bath before use.

(£) Nature of developer and degree of development of image.

(F) Concentration of exhaustion products.

(G) Concentration of various addition agents.
(IT) The presence of oxygen and oxidizing agents.

V. Factors affecting the rate of reduction in solutions of plain hypo.

VI. Theoretical discussion.

VII. Summary.

VIII. Practical recommendations.

Although it is well known that under certain conditions a fixing
bath may exert an appreciable reducing action on the silver image of
negatives and prints, no precise data have been available on the
magnitude of this effect with present-day motion picture emulsions.
However, with the widespread application of sensitometry to every
branch of photography and especially to the photographic recording
of sound, the question of the extent of this reaction under practical
conditions is of increasing importance.

I. EXPERIMENTAL METHODS

The emulsions tested are tabulated below.

Emulsion

Nature of Motion Picture Film Number
Panchromatic Negative, Type 2 1218

Supersensitive Panchromatic Negative, Type 2 1217

Negative 1201

Duplicating Negative 1505

Duplicating Negative 1503

Duplicating Positive 1355

Positive 1301

In the majority of the tests only a relative measure of the degree of
reduction in a stipulated time was obtained when the film to be bathed
was developed, fixed in the F-2 fixing bath, washed, and dried before
treatment. In the other tests the film was developed, fixed in the
bath under test for twice the time required to clear it, and then
treated for a further period. A separate test strip of the film treated
for twice the time to clear it was washed, and the density of the dried

Mar., 1932] REDUCING ACTION OF FIXING BATHS 373

strip taken as the density before treatment. Film treated in this
manner is termed "wet film."

The film was exposed on the Eastman sensitometer, type 2-B, and
after processing, was bathed in the various solutions, contained in
250-cc. cylinders, for a given period of time with little or no agitation.
The positive film was developed in the D-16* formula to a gamma
between 1.0 and 1.2, and the negative film in the £>-76** formula to a
gamma between 0.6 and 0.7. All the tests were made with fresh fixing
baths containing 30.0 per cent hypo unless otherwise stated.

In several experiments in which the film was agitated continuously
during bathing, the film was pinned to a small drum immersed in the
solution to be tested and rotated at a peripheral speed of approxi-
mately 100 feet per minute.

The progress of the reduction was determined by measuring the
density removed in a given time from a step having a known density.
This degree of reduction in a given time was considered as a relative
measure of the rate of reduction.

The pH values of the solutions were determined with organic
indicators in a manner similar to that described in a previous publica-
tion by the authors.1

H. DEGREE OF REDUCTION OF THE SILVER IMAGE IN VARIOUS
FIXING BATHS

The degree of reduction of the silver image in the various fixing
bath formulas published by the Eastman Kodak Company was
determined with the emulsions listed previously. The constituents
of the fixing baths are given in Table I in terms of grams or cubic
centimeters per liter.

The degree of reduction at 70°F. with dried processed film in the
fixing baths given in Table I is shown in Table II for various times of
bathing. The results show that in general the rate of reduction in a

*D-16

** D-76

Elon

0 . 3 gram

2 . 0 grams

Hydroquinone

6 . 0 grams

5 . 0 grams

Sodium sulfite (desiccated)

40 . 0 grams

100.0 grams

Sodium carbonate (desiccated)

19 . 0 grams

Borax

. . .

2 . 0 grams

Citric acid

0 . 7 gram

Potassium metabisulfite

1 . 5 grams

Potassium bromide

0.9 gram

Water to make

1.0 liter

374 H. D. Ru JSELL AND J. I. CRABTREE [j. s. M. p. E.

TABLE I

Constituents of Fixing Baths Used for Determining Degree of Reduction

Constituents

F-l

F-2

F-14

F-16

F-23

Hypo 300 grams 300 grams 300 grams 300 grams 300 grams
Sodium sulfite

(desiccated) 15 grams 3 grams 7. 5 grams 15 grams 17. 5 grams

Acetic acid (glacial) 13 cc. 5 cc. 13 cc.
Sulfuric acid

(concentrated) ... ... ... 2 cc. 2 cc.

Potassium alum 15 grams 6 grams 15 grams

Potassium chrome ... ... ... 15 grams 32 grams

alum

Water to make 1 liter 1 liter 1 liter 1 liter 1 liter

FIG. 1. Effect of reduction on the shape of the characteris-
tic curve.

Mar., 1932]

REDUCING ACTION OF FIXING BATHS

375

given fixing bath is dependent upon the state of division of the image.
In the case of images from fine grained emulsions, the rate of reduction
is much greater than with coarser grained materials such as motion
picture panchromatic negative film type 2. It is also seen that the
reducing action is a minimum in the case of certain potassium alum-
acetic acid baths, while the maximum effect is obtained with chrome
alum baths containing sulfuric acid.

III. EFFECT OF REDUCTION ON THE SHAPE OF THE CHARACTERISTIC CURVE

The effect of the F-23 fixing bath on the shape of the characteristic
curve is shown in Fig. 1. With motion picture negative film the
reducing action is almost truly proportional, while with positive film
the behavior is between that of a cutting and a proportional reducer.

IV. FACTORS AFFECTING THE RATE OF REDUCTION OF THE SILVER
IMAGE IN FIXING BATHS

At the outset it was considered that the rate of reduction might
depend upon the following factors which were investigated.

2.0 -

(DESICCATED)
WATER. I UATER

Z.O 4.O 6.O 8.O Z.O 4.O fe.O & O

FIG. 2. Effect of />H on the degree of reduction in various fixing baths.

376 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

A. Composition of Fixing Bath. — 1. Effect of pH on the Rate of
Reduction. — The effect of the £H of the bath on the rate of reduction
in the F-l and several experimental formulas is shown in Fig. 2. The
bath numbered F-x was an experimental bath employed in several
of the tests throughout the paper and contained 30 per cent hypo, 3
per cent chrome alum, and 3 per cent sodium bisulfite. The pH
values of the baths were varied by the addition of either sulfuric acid
or sodium hydroxide. The graphs show that the rate of reduction
increases rapidly for a given solution as the pH value is decreased
below a value of 4.0. As the £H increases to values greater than 4.0
the rate decreases more or less rapidly depending upon the composi-
tion of the fixing bath.

TABLE II

The Degree of Reduction of the Silver Image from Different Emulsions in Various

Fixing Baths

Density Removed for Different Times of Bathing at 70 °F.
Fixing Bath Original Density 30 Min. 60 Min. 3 Hours 6 Hours

Positive Film, Emulsion

1301

F-2

1.60

0.04

0.20

0.34

0.64

F-16

1.60

0.16

0.36

0.64

1.24

F-23

1.60

0.20

0.46

1.42

1.58

F-l

1.60

0.10

0.24

0.40

0.68

F-14

1.60

0.10

0.30

0.50

0.70

Super sensitive

Panchromatic Negative Film

Type 2, Emulsion

1217

F-2

1.36

0.06

0.16

0.30

F-16

1.36

0.06

0.16

0.24

0.36

F-23

1.36

0.12

0.44

0.60

0.98

F-l

1.36

0 . 08

0.10

0.20

0.40

F-14

1.36

0.08

0.10

0.22

0.42

Panchromatic Negative Film

Type 2,

Emulsion 12L8

F-2

1.50

0.06

0.08

0.20

0.30

F-1Q

1.50

0.06

0.18

0.30

0.40

F-23

1.50

0.10

0.22

0.70

1.18

F-l

1.50

0.08

0.10

0.16

0.30

F-14

1.50

0.08

0.10

0.20

0.40

Negative Film, Emulsion

1201

F-2

1.50

0.08

0.10

0.16

0.26

F-16

1.50

0.10

0.30

0.40

0.50

F-23

1.50

0.16

0.30

0.70

1.28

F-l

1.50

0.08

0.10

0.20

0.30

F-14

1.50

0.10

0.18

0.28

0.40

Mar., 1932] REDUCING ACTION OF FIXING BATHS 377

TABLE II (continued)

The Degree of Reduction of the Silver Image from Different Emulsions in Various

Fixing Baths

Density Removed for Different Times of Bathing at 70 °F.
Fixing Bath Original Density 30 Min. 60 Min. 3 Hours 6 Hours

Duplicating Negative

Film, Emulsion

1505

F-2

1.24

0.04

0.12

0.20

0.42

F-W

1.24

0.08

0.20

0.32

0.96

F-23

1.24

0.10

0.30

0.82

1.14

F-\

1.24

0.04

0.14

0.24

0.52

F-14

1.24

0.10

0.18

0.34

0.60

Duplicating Negative

Film, Emulsion

1503

F-2

1.40

0.04

0.12

0.20

0.38

F-W

1.40

0.10

0.20

0.42

0.98

F-23

1.40

0.16

0.34

0.90

1.26

F-l

1.40

0.06

0.10

0.16

0.36

F-14

1.40

0.10

0.16

0.26

0.50

Duplicating Positive

Film, Emulsion

1355

F-2

1.64

0.08

0.14

0.24

0.40

F-W

1.64

0.14

0.24

0.40

0.84

F-23

1.64

0.14

0.34

0.86

1.40

F-l

1.64

0.06

0.14

0.34

0.50

F-14

1.64

0.10

0.24

0.40

0.64

2. Sulfite Concentration. — The effect of the concentration of sulfite
was determined by the addition of increasing quantities of sodium
bisulfite to a solution containing 30 per cent hypo. The results in
Fig. 3 are given for two pH values. The value of 4.4 was that of the
plain solutions, while the value of 3.0 was chosen arbitrarily and was
obtained by the addition of sulfuric acid. The data indicate that for
a solution containing 300 grams of hypo per liter an increase in the
sulfite concentration at a constant />H value increases the rate of
reduction of the silver image. The rate of reduction was very much
greater at a pR value of 3.0 than at 4.4. Similar results were ob-
tained when equivalent quantities of sodium sulfite were substituted
for sodium bisulfite.

3. Hypo. Concentration. — A decrease in the concentration of hypo
for a given sulfite concentration and pH value decreased the rate of
reduction as is shown by experiments 3, 4, 5, and 6 in Table III-A.
Tests were also made which indicated that with concentrations of
hypo greater than 30 grams per liter the rate of reduction decreases.

378 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E.

TABLE IH-A

The Effect of Various Reagents on the Degree of Reduction of the Silver Image in

Fixing Baths

Sodium Sulfuric
Bisulfite Acid Time of
Nature Hypo (Grams 10% Bathing
of (Grams per (Cc. per at 70 °F.
No. Bath per Liter) Liter) Liter) (Hrs.)

Density Color
Original Re- of

Density moved Image

Remarks

300

0.0

4.4

3.08

0.45

Black

300

40.0

3.2

3.08

1.94

Brown

300

0.0

4.4

3.08

0.17

Black

300

40.0

3.2

3.08

1.20

Brown

10.

0.0

4.4

3.08

0.10

Black

10.

40.0

<3.0

3.08

0.12

Black

Sulfurized

300

40.0

3.2

3.06

2.00

Brown

75.0

12.5

40.0

3.2

3.06

1.40

Black

37.

6.25

20.0

3.0

3.06

0.70

Black

10*

18.

3.12

20.0

3.0

3.06

0.46

Brown

11 F-l

3.8

3.08

0.16

12 F-l

10.0

3.6

3.08

0.20

13 F-l

20.0

3.4

3.08

0.35

14 F-l

40.0

3.0

3.08

1.58

Sulfurized

Equal ratio of sulfite and hypo.

3.0

z.c.

1.4

\.TL

1.0
Oft

POSITIVE: ni_Ni
ORV

NO A.G \TACT\OM

BISULPHITE.

FIG. 3. Effect of sulfite concentration on the degree of reduction.

Mar., 1932]

REDUCING ACTION OF FIXING BATHS

379

TABLE III-B

The Effect of Various Reagents on the Degree of Reduction of the Silver Image in

Fixing Baths

No.

Nature
of Bath

Cone, of

Substance
Added
Nature of Substance (Grams
Added per Liter)

Time of
Bathing
at 70°F.
(Hrs.)

Original
Density

Color
Density of
Re moved I mage

F-X

3.0

.08

1.94

Brown

F-X

Silver Iodide

3.0

.08

1.55

F-x

Silver Iodide

3.0

.08

0.30

F-x

Silver Iodide

100

3.0

,08

0.10

F-x

Silver Bromide

3.0

1.90

F-x

Silver Bromide

10,

3.0

1.20

F-x

Silver Bromide

100.

3.0

0.10

F-x

Potassium Bromide

3.0

3.08

2.00

Brown

F-x

Potassium Bromide

10.

3.0

2.30

Brown

F-x

Potassium Bromide

100.

3.0

2.55

Brown

F-x

Potassium Iodide

3.0

1.94

Brown

F-x

Potassium Iodide

10.

3.0

2.50

Brown

F-x

Potassium Iodide

100.

3.0

3.08

Brown

F-x

Ammonium Chloride

10.

3.0

2.08

F-x

Ammonium Chloride

100.

3.0

2.50

F-x

Ammonium Sulf ate

100.

3.0

1.00

F-x

Sodium Chloride

100.

3.0

1.10

F-x

Methylene Blue

10.

3.0

1.20

Further tests were made to determine if the rate of reduction was
dependent upon the concentration or ratio of sulfite to hypo. The
results of experiments 7, 8, 9, and 10 in Table III-A indicated that the
rate of reduction decreased as the concentration of hypo and sulfite
were decreased in equal proportions.

4. Hardener Concentration. — The effect of hardener concentration
on the rate of reduction, degree of hardening, and £H value of the
F-l* and F-2** formulas is shown in Fig. 4 from which it is seen that,
if the hardener concentration is decreased to one-half its normal value,
the degree of reduction is also decreased approximately one-half,

* F-l. — For use add 125 cc. .F-la hardener to 1.0 liter of hypo solution.
** F-2. — For use add 50 cc. F-2a hardener to 1.0 liter of hypo solution.

Hardener Formulas

Sodium sulfite (desiccated)
Acetic acid (glacial)
Potassium aluminum alum
Water to make

F-la

120 grams
105 cc.
120 grams
1 liter

F-2a

60 grams
100 cc.
120 grams
1 liter

380

H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E.

while the degree of hardening is not seriously affected. A further
decrease in the hardener concentration does not produce a correspond-
ing decrease in the rate of reduction and lowers the degree of harden-
ing to a value which is too low for practical purposes.

5. Nature of Hardening Agent. — In Table IV are given figures
comparing the extent of the reduction, obtained with the F-l and F-IQ
formulas, for equal pH values. The pH values were adjusted by the
addition of either sodium hydroxide or sulfuric acid.

The results in Table IV indicate that the rates of reduction were

0.4

ZLOCf

no'

140*

5.0

4.0

3.0

F--I FIXING.
POSITIVE. FIL.M
10° F-.

REDUCTION
WET FIUN'V
ISO AO»TACTIO«H
4.O

0.4

o.z

-bl

F-Z. F-lXtlHQi BATH
POSITIVE

RCDUCTIOK
W
MO

4.0 HOURS

FIG. 4. Effect of the hardener concentration on the degree of reduction, degree
of hardening, and />H of the F-l and F-2 formulas.

similar for potassium alum and chrome alum baths for equal sulfite
concentrations and pH values.

B. Effect of Temperature on the Rate of Reduction. — The effect of
temperature on the rate of reduction of the silver image with dry
motion picture positive film is shown in Fig. 5 for the F-2 and 7^-23
fixing baths. The results indicate that as the temperature is increased
from 50° to 90 °F., the rate of reduction also increases and the effect of
temperature was much greater with the F-23 formula than with the
F-2 formula. In the F-23 bath the degree of reduction in a given
time was increased approximately ten times for an increase in tempera-

Mar., 1932]

REDUCING ACTION OF FIXING BATHS

381

TABLE III-C

The Effect of Various Reagents on the Degree of Reduction of the Silver Image in

Fixing Baths

No.

Cone, of
Substance Time of
Nature of Added Bathing
Nature Treatment Substance (Grams at 70 °F. Original
of Bath with Gas Added per Liter) (Hrs.) pH Density

Color
Density of
Removed Image

F-X

None

3.0

.08

2.30

F-x

Air

3.0

.08

2.54

F-x

Carbon

Dioxide

3.0

.08

2.38

F-l

None

3.8

.08

0.80

F-l

Air

3.8

1.80

F-l

Carbon

Dioxide

3.8

1.40

F-l

3.8

0.30

Black

F-l

Sodium

Perborate

3.8

0.50

Black

F-l

Hydrogen

100 cc.

3.8

0.45

Black

F-x,

Peroxide

3.0

,10

0.90

Brown

F-x

Sodium Sulfate

3.0

0.70

Brown

F-x

Sodium Sulfate

100

3.0

0.36

Black

F-x

Sugar

3.0

0.90

Brown

F-x

Sugar

100

3.0

0.70

Brown

F-x

Glycerin

3.0

0.80

Brown

F-x

Glycerin

200

3.0

0.50

Brown

TABLE IV

A Comparison of the Degrees of Reduction in Potassium Alum and Chrome Alum

Fixing Baths

Concentration
of Alum
Fixing (Grams
Bath per Liter)

Concentration
of Sulfite
(Grams
per Liter) pH

* Time of
Bathing
(Hours)

Original
Density

Density
Removed

F-l

3.8

4.0

1.60

0.30

F-l

3.6

4.0

1.60

0.40

F-l

3.4

4.0

1.60

0.94

F-l

4.0

1.60

0.22

F-l

4.8

4.0

1.60

0.16

F-IQ

3.4

4.0

1.60

0.96

F-IQ

3.2

4.0

1.60

1.40

F-IQ

3.0

4.0

1.60

1.50

F-IQ

4.0

1.60

0.30

F-IQ

4.4

4.0

1.60

0.18

* Positive film (wet).

382 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

ture from 65° to 95 °F., while with the F-2 formula the increase was
only about four times.

C. Effect of Agitation on the Rate of Reduction. — -The effect of
agitation on the rate of reduction with wet and dry film is shown in
Table V.

TABLE V

Effect of Agitation on the Rate of Reduction at 70° F.

Fixing
Bath

Emulsion

Original
Density

Time of
Bathing

Constant
Wet Film

Density Removed
Agitation No Agitation
Dry Film Wet Film Dry Film

F-2

1301

1.70

30 Min.

0.14

0.10

0.00

F-2

1301

1.70

4Hrs.

1.40

1.16

0.60

0.30

Hypo 30%

1301

1.70

1 Hr.

0.50

0.60

0.06

0.08

Hypo 30%

1301

1.70

4Hrs.

1.30

0.16

0.19

F-23

1218

1.34

30 Min.

0.34

0.18

0.12

0.06

F-23

1218

1.34

1 Hr.

0.52

0.36

0.32

0.14

The results indicate that with constant agitation, as compared with
no agitation, the amount of reduction in the case of positive film was
increased about ten times, while with negative film the rate was ap-
proximately doubled. Also, the degree of reduction obtained was
much greater with the wet film previous to drying than with the dry
film.

D. Effect of Age before Use. — The effect of age of the fixing baths
before use on the rate of reduction is shown in Table VI, from which it
is seen that: (1) the rate of reduction in potassium alum fixing baths
did not change appreciably on storage for 10 days before use, and
(2) the reducing action of the chrome alum fixing baths decreased with
age owing to an increase in the pJI value of the solutions.

TABLE VI

Effect of Age before Use on Rate of Reduction

Fixing
Bath

Age

(70°F.)

*Time of
Bathing

Original
Density

Density
Removed

F-l

Fresh

3.8

4.0 Hrs.

1.54

0.30

F-l

10 days

3.8

4.0 Hrs.

1.54

0.28

F-2

Fresh

3.6

4.0 Hrs.

1.54

0.26

F-2

10 days

3.6

4.0 Hrs.

1.54

0.26

F-1Q

Fresh

3.4

4.0 Hrs.

1.54

0.98

F-IQ

10 days

3.8

4.0 Hrs.

1.54

0.60

F-23

Fresh

3.2

4.0 Hrs.

1.54

1.22

F-23

10 days

3.6

4.0 Hrs.

1.54

0.80

Positive film (dry).

Mar., 1932] REDUCING ACTION OF FIXING BATHS 383

E. Effect of Nature of Developer and Degree of Development. — At
the outset it was considered that the rate of reduction under any
given conditions would depend on the size of the silver grains which,
in turn, is determined by (a) the nature of the emulsion, (b) the nature
of the developer, and (c) the degree of development or "gamma."

Motion picture panchromatic negative film was developed in
formulas D-16 and D-76 to equal gammas and then bathed in the
F-23 fixing bath. From the results in Table VII it is seen that equal
degrees of reduction were obtained for equal densities regardless of
the degree of development or the nature of the developer.

TABLE VII

Effect of Degree of Development and Nature of Developer on the Degree of Reduction

Developer

Fixing
Bath

Time of
Dev.
(Min.)

Gamma

Original
Density

Density
Removed

* Time of
Bathing

ZM6

7^-23

2.5

0.30

0.60

0.10

1.0 Hr.

D-7Q

7^-23

4.0

0.30

0.60

0.08

1.0 Hr.

D-1Q

77-23

5.0

0.80

0.60

0.10

1 . 0 Hr.

D-7Q

.F-23

15.0

0.80

0.60

0.11

1 . 0 Hr.

* Negative film (emulsion 1218) wet.

F. Effect of Exhaustion Products on Rate of Reduction. — With use
the chemical nature of the fixing bath changes. The undeveloped
silver halide grains are dissolved from the emulsion and accumulate
in the bath as complex silver thiosulf ates and sodium halides. Experi-
ments 15 to 21, inclusive (Table III-B), indicate that the addition of
silver bromide or silver iodide to the F-x fixing bath decreases the rate
of reduction.

Practical exhaustion tests were made with the F-2 and F-23 formula
in order to determine the effect of exhaustion with developed and
undeveloped positive film on the rate of reduction. The results are
shown in Fig. 6 from which it is seen that the rate of reduction is less
in a bath exhausted with developed film than in one exhausted with
undeveloped film. In the case of the bath exhausted with
undeveloped film, the pH value remained practically constant, while
with the developed film the pH of the bath gradually increased during
exhaustion, which may have caused a decrease in the rate of reduction.

The effect of removing the silver from an exhausted fixing bath on
the rate of reduction was investigated. The silver was removed by
an electrolytic method similar to that used in actual practice.2 The

384

H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

F-2 fixing bath was exhausted with undeveloped motion picture
panchromatic negative film (type 2) to the extent of 250 feet per
gallon when the silver content was 7 grams per liter. The silver was
then removed by electrolysis and the bath exhausted further to 200
feet per gallon or a total footage of 450 feet per gallon which is equiva-
lent to 13 grams of silver per liter. The pH value and sulfite

O.fc

14
I.I
1.0
0.8
O.Q,
0.4
O.Z

-t FIXINO BATH

HO Afe \TI\T\O IX
POSITIVE. F»\_|v\

FIG. 5. Effect of temperature on the degree of reduction
with the F-2 and .F-23 fixing baths.

concentration of the solution changed during the electrolysis but
were maintained constant by additions of sulfite and alkali. The
reduction tests were made with wet motion picture positive film dur-
ing the last stage of the electrolysis, that is, when the solution con-
tained less than 3 grams of silver per liter and also after all the silver
was removed. The tests in every case indicated that the degree of
reduction obtained in an exhausted F-2 fixing bath from which the

Mar., 1932]

REDUCING ACTION OF FIXING BATHS

385

silver had been removed was less than that obtained in the fresh
solution.

The fixing bath also becomes contaminated during use with
partially exhausted developer, which in the case of a hydroquinone
developer consists of sodium halides, sodium sulfite, hydroquinone
sulfonates, and alkali. The hydroquinone sulfonates are the result

UtMOEV/E-UOPEO
ATM e.XHAAJ*TEO WITH

FIG. 6. Effect of exhaustion on the degree of reduction
with the F-2 and 7^-23 fixing baths.

of the reaction between quinone and sulfite, quinone being an end-
product of the reduction of the exposed silver halide by hydroquinone.
The alkali in the developer decreases the acidity of the fixing bath,
in which case the rate of reduction would also decrease.

The effect of the addition of an exhausted developer on the rate of
reduction was tested by the addition of: (1) an oxidized D-IQ de-

386 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

veloper, and (2) quinone to the F-x fixing bath. The D-IQ developer
was oxidized by bubbling air through the solution until it would no
longer develop. Two hundred cubic centimeters of such a developer
were evaporated by boiling to a volume of 20 cc., and added to 250
cc. of the fixing bath. A comparison made between this solution and
one containing 1 per cent quinone for equal pH values indicated that
these products have very little, if any, effect on the rate of reduction.

G. Effect of Miscellaneous Addition Agents. — Various chemicals
which are not usually considered as oxidizing agents for silver were
added to the F-x fixing bath as described below. Potassium bromide
and potassium iodide were added in concentrations ranging from 0.1
per cent to 10 per cent (experiments 22-27, Table III-B). Both
chemicals increased the rate of reduction, the potassium iodide being
more effective for a given concentration than the potassium bromide.

Ammonium chloride increased the rate of reduction for concentra-
tions between 1 per cent and 10 per cent while the addition of 10 per
cent sodium chloride or 10 per cent ammonium sulfate had little or no
effect on the reaction (experiments 28-30, Table III-B). Since
ammonium chloride and ammonium sulfate both tend to decrease the
clearing time in a fixing bath, further tests were made. The effect of
these salts on the rate of reduction and clearing times of undeveloped
positive and negative film is given in Table VIII for the F-2 and F-23
formulas, from which it is seen that the addition of either ammonium
chloride or ammonium sulfate increased the clearing time of positive
film, while in the case of negative film the clearing time was decreased.
A concentration of either salt between 2.5 per cent and 5.0 per cent
produced the greatest decrease in the clearing times, the chloride
being more effective than the sulfate. The above concentrations of
ammonium chloride also increased the rate of reduction to the greatest
extent, while an equal quantity of the sulfate did not affect the reac-
tion. With the 7^-23 formula, the rate of reduction increased up to a
concentration of 300 grams per liter, but beyond this concentration
the rate began to decrease.

This critical point does not correspond with the concentration of 400
grams per liter of hypo which gives a minimum clearing time with
motion picture panchromatic negative film.

The addition of restraining agents such as sodium sulfate, sugar, and
glycerin decreased the rate of reduction (experiments 43-48, Table
III-C). Sodium sulfate in this respect was more effective than either
of the other chemicals for equal concentrations.

Mar., 1932]

REDUCING ACTION OF FIXING BATHS

387

TABLE VIII

Effect of Ammonium Chloride and Ammonium Sulfate on the Degree of Reduction
in the F-2 and F-23 Formulas

Bath

Per Cent * Time of
Ammonium Bathing
Chloride (Hours)

Original
Density

Density
Removed

Time to
Clear
Positive

(Sec.)

Time to
Clear
Negative
(Sec.)

F-23

1.60

0.76

240

F-23

1.0

1.60

0.86

115

F-23

2.5

1.60

1.10

F-23

5.0

1.60

0.96

F-23

10.0

1.60

0.90

100

Per Cent
Ammonium
Sulfate

F-23

1.60

0.76

240

F-23

1.0

1.60

0.60

125

F-23

2.5

1.60

0.70

105

F-23

5.0

1.60

0.60

130

F-23

10.0

1.60

0.48

170

Per Cent
Hypo

F-23

1.60

0.75

240

F-23

1.60

0.60

120

F-23

1.60

0.56

220

F-23

1.60

0.40

>300

Per Cent
Ammonium
Chloride

F-2

1.60

0.30

240

F-2

1.0

1.60

0.30

115

F-2

2.5

1.60

0.40

F-2

5.0

1.60

0.40

F-2

10.0

1.60

0.20

100

Per Cent
Ammonium
Sulfate

F-2

1.60

0.30

240

F-2

1.0

1.60

0.30

125

F-2

2.5

1.60

0.30

105

F-2

5.0

1.60

0.22

130

F-2

10.0

1.60

0.10

170

* Positive film (dry).

388 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E.

H. Effect of Oxygen and Oxidizing Agents. — The effect of oxygen
on the rate of reduction in the F-x and F-l fixing baths is shown in
Table III-C (experiments 33 to 38, inclusive). In the tests, air and
carbon dioxide were bubbled through the fixing baths for two hours.
The results indicated that the effect of bubbling air is probably a
result of the increased agitation. The rate of reduction, however,
was slightly greater with air than with carbon dioxide.

Further tests were made in which wet positive film, which was
flashed to a uniform density and developed in the D-IQ formula, was
bathed in the F-2 and F-x formulas. The strips were suspended above
the baths in such a manner that only part of the film was totally
immersed. The part above the solution was moistened with the
solution at 1 -minute intervals throughout the time of bathing. With
the F-2 formula the density above the solution was decreased to a
greater degree than that which was immersed in the bath, while with
the F-x formula the reverse effect was obtained.

The addition of oxidizing agents such as hydrogen peroxide and
sodium perborate slightly increased the rate of reduction, as shown by
experiments 39-41, Table III-C. The effect on the silver image of the
addition of methylene blue to hypo solutions has been determined
by one of the authors,3 who found that under certain conditions
methylene blue produced reversed dye images. The effect of methy-
lene blue on the rate of reduction of the silver image was determined
when added to the F-x formula and the fixing solution recommended in
the above publication. In each case the low densities of the sensito-
metric strips were reduced very rapidly, while with the high densities,
the rate was similar to that of the bath without methylene blue.
Methylene blue produced the greatest effect when strips were bathed
in a solution of the dye previous to immersion in the fixing solution.

To Summarize: — From the above tests it was concluded that for a
given fixing bath containing alum, sulfite, acid, and hypo, (1) the rate
of reduction increased rapidly if the />H value of the bath was reduced
below 4.0, (2) for pR values greater than 4.0 the degree of reduction
was of a much lower order of magnitude, (3) for a given hypo
concentration and a />H value less than 4.0, an increase in the sulfite con-
centration increased the rate of reduction, (4) for a given sulfite
concentration and a />H value less than 4.0 an increase in the hypo con-
centration up to 300 grams per liter increased the rate of reduction
but with concentrations of hypo greater than 300 grams per liter, the
rate of reduction decreased.

Mar., 1932] REDUCING ACTION OF FIXING BATHS 389

V. FACTORS WHICH INFLUENCE THE RATE OF REDUCTION IN SOLUTIONS

OF PLAIN HYPO

The chemicals and reagents listed in Table III-A, -B, and -C were
added to a solution containing 300 grams of hypo per liter, but no
noticeable increase in the rate of reduction was observed. From
these experiments it was concluded that the nature of the reducing
action in plain hypo solutions was different from that in acid sulfite
fixing baths.

The greatest increase in the rate of reduction was obtained when
oxygen or air was bubbled through the solution. The effect of
bubbling various gases through plain hypo solutions is shown in
Table IX.

TABLE IX

Effect of Various Gases on the Rate of Reduction of the Silver Image in Hypo

Solutions

Gas

* Time of
Bathing

Original
Density

Density
Removed

Before
Treatment

After
Treatment

None

Hrs.

.10

Air

Hrs.

Air

Hrs.

11.

Oxygen

Hrs.

6.0

Nitrogen

Hrs.

Carbon Dioxide

Hrs.

Sulfur Dioxide

Hrs.

* Positive film (dry).

All the gases were bubbled through 250 cc. of the solution at a rate
equal to 200 cc. per minute. Throughout the period of bathing
sulfur dioxide was bubbled through the solution until the hypo
sulfurized, which required about 15 minutes. The slight increase in
the rate of reduction with carbon dioxide and sulfur dioxide is possibly
due to the decrease in pH value.

Further experiments were made in which the air and other gases
were removed from a 30 per cent solution of plain hypo by means of a
vacuum pump. An image on positive film bathed in this solution was
not reduced in 10 hours, while a density of 0.44 was removed from a
density of 1.10 in a similar solution from which the air had not been
removed.

The above tests indicate that air or oxygen is a very important
factor in the bleaching action of solutions of plain hypo. It was also
observed that when air or oxygen is bubbled through a solution of

390 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

plain hypo in which a silver image is being reduced, the solution be-
comes more alkaline. The changes in alkalinity take place only in
the presence of a silver image. When air or oxygen was bubbled
through a solution of hypo without a silver image, no increase in
alkalinity occurred. The change in alkalinity probably results from
the oxidation by oxygen of the finely divided silver to silver oxide,
which is dissolved by the hypo, forming complex silver thiosulfates
and sodium oxide. The sodium oxide would exist in such a solution as
sodium hydroxide, which is very alkaline.

In experiment 3, Table IX, the hypo solution was made alkaline by
the addition of sodium hydroxide. A comparison of the rate of
reduction with that in experiment 2 indicates that the reduction in
alkaline hypo is less than that in plain hypo.

The effect of the concentration of hypo on the rate of reduction of
the silver image in solutions of plain hypo is shown in Table X.

TABLE x

Effect of Concentration of Hypo on Rate of Reduction

Concentration of
Hypo
(Grams per Liter)

Original
Density

Density
18 Hours

Removed
36 Hours

800

2.06

0.00

400

2.06

0.18

0.34

300

2.06

0.26

0.62

200

2.06

0.36

0.70

100

2.06

0.36

1.00

2.06

0.12

0.16

2.06

0.00

The results indicate that the rate of reduction of the silver image in
plain hypo solutions increases as the concentration of hypo is de-
creased from 800 grams per liter to 100, and then decreases for a
further decrease in the hypo concentration.

VI. THEORETICAL DISCUSSION

The chemical reaction involved in the reduction of the photographic
image in an acid fixing bath is probably one of oxidation of the silver
to a soluble compound. The reaction may be represented by the
following equation:

(7.) 2Ag + H2S203 + O ^^ Ag2S20? + H2O

Silver Thiosulfuric Oxidizing Silver Water

Acid Agent Thiosulfate

The silver thiosulf ate formed readily dissolved in the excess hypo.

Mar., 1932] REDUCING ACTION OF FIXING BATHS 391

Although the exact chemical nature of the oxidizing agent is un-
known, F. Foerster4'6'6'7-8 and his colleagues have shown that addition
compounds of certain sulfur acids with sulfur dioxide can exist in a
fixing bath and H. Bassett and R. G. Durrant9 have suggested that
these probably react as oxidizing agents.

Foerster and Vogel8 have prepared the yellow addition compound
(K^jOa.SC^) by the action of sulfur dioxide on a potassium thiosulfate
solution. They claim that the yellow color of an acidified sulfite and
hypo solution is due to such compounds rather than colloidal sulfur.

Other yellowish colored addition compounds of sulfur dioxide are
recorded in the literature such as:

1. H2S203(S02)x

2. H2S03.(S02)x

3. HCNS.(SO2)x

4. (HO)2S.(S02)x

5. HI.(SO2)x

The effect of these compounds on the silver image was investigated.
Various concentrations of the potassium salts were added to a 5 per
cent solution of sodium sulfite acidified with sulfuric acid. The re-
sults are given in Table XI from which it is seen that an acidified solu-
tion of sulfite and iodide reduced the silver image very rapidly, while a
similar solution containing potassium bromide did not affect the image
to any great extent. From the standpoint of chemical composition
bromide forms an addition compound with sulfur dioxide similar to
that of the iodide, which is colorless.

The effect of the iodide-sulfur dioxide compound on the silver image
explains the increase in the rate of reduction obtained when potassium
iodide was added to the F-x fixing bath. A similar increase, although
not as great, was obtained when potassium bromide was added to the
F-x bath, which cannot be explained on the basis of the formation of
these addition compounds.

The rate of reduction with the addition of potassium thiocyanate to
an acidified solution of sulfite was considerably less than that with the
addition of potassium iodide. The mixture of sulfite and thiocyanate
without acid attacked the gelatin and removed the emulsion from the
support. The solution of 5 per cent sodium sulfite with acid did not
attack the silver image and even a highly concentrated yellow solution
of metabisulfite (experiment 3) did not reduce the silver image, which
indicates that this addition compound is not an oxidizing agent for
silver.

392

H. D. RUSSELL AND J. I. CRABTREE [j. b. M. P. E.

TABLE XI

The Effect of Addition Compounds with Sulfur Dioxide on the Reduction of the

Silver Image

Exp
No.

Sub-
. stance
Added

Sodium Sulfuric Time of
Sulfite Acid Bathing
Grams (Grams 10% at
per per (Cc. per 70° F. Original
Liter Liter) Liter) (Hrs.) £H Density

Color
Density of
Removed Solution

Color
of
Image

200 1

3.0

2.84

Colorless

Black

9.0

2.84

Colorless

Black

K2S2O5

400

50 1

3.0

2.84

Yellow

Black

KBr

200 1

3.0

2.84

.12

Colorless

Black

KBr

100

200 1

3.0

2.84

.14

Colorless

Black

KBr

4.0 1

3.0

2.84

Colorless

Black

KBr

9.0

2.84

Colorless

Black

KBr

100

9.0

2.84

Colorless

Black

200 1

3.0

2.84

.50

Yellow

100

200 1

3.0

2.84

2.74

Yellow

9.0

2.84

.36

Colorless

Black

100

9.0

2.84

.02

Colorless

Black

4.0 1

3.0

2.84

.00

Colorless

Black

100

. .

4.0 1

3.0

2.84

Gelatin was removed

Na2S2O4

250

200 1

2.84

.42

Yellow

Brown

Hypo

300

200 1

3.0

2.84

.70

Yellow

Brown

Hypo

300

9.0

2.84

.20

Colorless

Black

Hypo

100

200 1

3.0

2.84

.32

Yellow

Brown

KCNS

200 1

3.0

2.84

0.00

Yellow

Black

KCNS

100

200 1

3.0

2.84

.30

Yellow

Black

KCNS

100

9.0

2.84

Gelatin was removed

KCNS

100

4.0 1

3.0

2.84

Gelatin was removed

K2S2Os = Potassium

Metabisulfite

KBr

= Potassium Bromide

= Potassium

Iodide

Na2S2O4 = Sodium Hydrosulfite

KCNS

= Potassium

Thiocyanate

Sodium hydrosulfite forms a permanent yellow solution when acidi-
fied in the presence of sulfite, and this is due to a thiosulfate-sulfur
dioxide complex, according to Bassett and Durrant.9 The presence of
this compound probably accounts for the reduction of the silver
image in a solution of sodium hydrosulfite (experiment 15, Table XI).

If the addition compounds between sulfur dioxide and hypo are
oxidizing agents for silver, the reaction represented by the following
equation II readily explains why the factors previously mentioned
control the rate of reduction of the silver image in a solution of acid,
sulfite, and hypo.

(77.) H,S,0, + H2SO, 3=± H2S203.S02 + H2O

Mar., 1932] REDUCING ACTION OF FIXING BATHS 393

The application of the mass action law to the equation indicates
that the formation of H^CMSC^) depends upon the acidity, and the
concentrations of sulfite and hypo. The tendency to form this
compound increases with an increase in the acidity, and the concentra-
tion of sulfite and hypo, and hence causes an increase in the rate of
reduction. A corresponding decrease in the acidity, or the concentra-
tion of sulfite and hypo, decreases the concentration of the compound
which also decreases the rate of reduction, which is in accord with
experimental evidence.

Bassett and Durrant9 have shown that methylene blue acts as an
oxidizing agent in the presence of hypo solutions which explains the
fact that the dye increases the rate of reduction of the silver image.
The reaction may be represented by the following equation:

(7/7.) 2 Ag + C16H18N3SC1 + H2S2O3;=— C16Hi9N3S + Ag2S2O3 + HC1
Methylene Leuco

Blue Base

The silver thiosulfate formed in the reaction readily dissolves in
the excess hypo present.

Silver halides decrease the rate of reduction of the silver image in
the fixing baths, probably owing to the formation of complex silver
anions with the thiosulfate ion, thereby saturating the solution with
silver.

The reaction involved in the reduction of the silver image in a solu-
tion of plain hypo is probably different from that in acid fixing baths,
since experimental evidence indicates that the rate of reduction for a
given hypo concentration is only affected by oxidizing agents and
oxygen.

Oxygen possibly converts the silver image into silver oxide which is
readily dissolved by the hypo.

The reactions may be represented by the following equa-
tions:

(77.) 4Ag + 02 ;=± 2Ag20

(F.) Ag2O + Na2S2O3 + H2O ?=i 2 NaOH + Ag^Os

The silver thiosulfate formed in equation V readily dissolves in the
excess hypo present. Equation V also indicates that the solution be-
comes alkaline when silver oxide is dissolved by hypo which is in ac-
cord with the experimental facts.

394 H. D. RUSSELL AND J. I. CRABTREE [j. S. M. p. E.

VII. SUMMARY

The object of this investigation was to determine some of the factors
which control the rate of reduction of the silver image in fixing baths.

(1) The degree of reduction in a given time was determined for
images obtained from various emulsions bathed in different fixing
baths. The emulsions tested included motion picture panchromatic
negative film type 2, emulsion 1218, and supersensitive panchromatic
negative film, emulsion 1217, motion picture positive film, emulsion
1301, motion picture negative film, emulsion 1201, motion picture
duplicating negative films, emulsions 1505 and 1503, and motion
picture duplicating positive film emulsion 1355.

The fixing baths tested were the F-l, the F-2, F-U, F-16, and 7^-23,
and several experimental formulas.

(2) The rate of reduction in a given fixing bath was greater with
images from fine grained emulsions than with coarser grained materials.
The fixing bath having the lowest rate of reduction was the F-2
formula, while the highest rates of reduction were obtained with fixing
baths containing a relatively high concentration of sulfite and acid.

(3) The rate of reduction increased with an increase in temperature.

(4) The factor which affected the rate of reduction to the greatest
degree in an ordinary acid fixing bath was the acidity of the bath.
For a given bath the rate increases rapidly for pH values below 4.0.

(5) The rate of reduction was increased for pH values less than 4.0
with an increase in either the sulfite or hypo concentration. The rate
of reduction was decreased with concentrations of hypo greater than
30.0 per cent.

(6) The exhaustion products which accumulate in a fixing bath
such as silver halides and developer decreased the rate of reduction.
Developer oxidation products which also accumulate to a small ex-
tent did not affect the rate of reduction.

The rate of reduction in an exhausted F-2 fixing bath from which
the silver had been removed by an electrolytic process was less than
in a fresh bath.

(7) Ammonium chloride, potassium bromide, and potassium iodide
increased the rate of reduction, while ammonium sulfate, sodium
chloride, sodium sulfate, glycerin, and sugar produced the opposite
effect.

(8) Oxygen and oxidizing agents such as the peroxides have no
apparent effect on the rate of reduction in highly acid fixing baths.
The tests indicated, however, that the presence of oxygen increased

Mar., 1932] REDUCING ACTION OF FIXING BATHS 395

the rate of reduction in fixing baths containing a low concentration of
sulfite and acid such as the F-2 formula and was largely responsible
for the reduction in solutions of plain hypo.

(9) From a theoretical standpoint most of the factors which control
the rate of reduction in an acid fixing bath can be accounted for by
assuming that an oxidizing agent for silver is formed by reaction of
the hypo and the sulfite. The general formula for such compounds
is represented by H^OaCSOs)^ and they have been shown9 to exist in
an acid solution of sulfite and hypo.

Other sulfur compounds as well as the halides formed similar addi-
tion compounds which did not attack the silver image, with the excep-
tion of the iodide and the hydrosulfite compound. In solutions of these
compounds, however, the reduction might have been due to the H2S2O3-
(80)2) # complex present as an impurity, or as a decomposition product.

VIII. PRACTICAL RECOMMENDATIONS

The extent of the reducing effect of fixing baths on the silver image
during the progress of fixation is greater than has generally been
supposed. For example, in sensitometric work it is inadvisable to
prolong the fixation of motion picture positive film in the average
fresh potassium alum fixing bath beyond 5 minutes at 65°F. and
with certain highly acid chrome alum baths a measurable degree of
reduction occurs even in this short space of time.

Since little or no reduction of the image occurs in an alkaline hypo
solution, sensitometric tests should be checked against images fixed
in a 25 per cent solution of hypo containing 1 per cent of sodium
carbonate (anhydrous). The film should be rinsed in water and
agitated on first immersing in the bath in order to prevent the forma-
tion of dichroic fog.10

In regular laboratory work the degree of reduction which takes
place in the normal time for fixation is usually of no practical impor-
tance with the baths in common use. In any given bath the rate of
reduction increases with the acidity, the temperature of the bath, and
degree of agitation of the film, so that with certain chrome alum baths
used under tropical conditions, the decree of reduction is excessive,
especially with fine grained emulsions. For high temperature
processing, if a minimum of reduction is required the use of a chrome
alum hardening stop bath after development, followed by a fixing bath
consisting of plain hypo containing 1 per cent sodium bisulfite, is
recommended.1

396 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E.

During use, the reducing action of a fixing bath falls off because it
becomes more alkaline and accumulates silver thiosulfate which tends
to retard the reduction.

In order to insure the minimum degree of reduction, therefore,
baths having a minimum degree of acidity should be used though such
baths have a short life and often do not harden satisfactory. It is
therefore necessary to revive such baths either by adding further
quantities of acid or hardening solution at intervals during use;
otherwise, if the film is not rinsed in water before fixing an objection-
able sludge will form in the fixing bath.11

The desirable range of acidity lies between pH values of 4.0 and 4.5.
At higher values the bath does not harden, and below this there is
danger of reduction of the image.

Exposure of the film to air during fixation has little or no effect with
acid baths, except those containing a relatively low concentration of
sulfite and acid, in which case the rate of reduction is greatly increased.
Air also accelerates the rate of reduction in solutions of plain hypo
which are seldom used in practice.

The addition of restraining agents such as sodium sulfate, glycerin,
and sugar to the acid fixing bath decreases the degree of reduction but
their use is not recommended because they also decrease the rate of
fixation.

In some laboratories the acidity of the fixing bath is maintained by
passing sulfur dioxide gas into the bath. Under these conditions, if
an excess of the gas is used, a strongly reducing fixing bath is pro-
duced.

The nature of the reduction with the negative emulsions tested was
found to be almost strictly proportional and some of the more active
baths enumerated could therefore be used advantageously for reduc-
ing the contrast of photographic images.

REFERENCES

1 CRABTREE, J. I., AND RUSSELL, H. D.: "Some Properties of Chrome Alum
Stop Baths and Fixing Baths," Parts I and II, /. Soc. Mot. Pict. Eng., 14 (May,
1930), p. 483; (June, 1930), p. 667.

2 HICKMAN, K. C. D., SANFORD, C., AND WEYERTS, W.: "The Electrolytic
Regeneration of Fixing Baths," /. Soc. Mot. Pict. Eng., 17 (Oct., 1931), p. 568.

3 CRABTREE, J. I.: "A Method of Producing Reversed Dye Images," Photo.
Era, 46 (1921), p. 10.

4 FOERSTER, F., AND HoRNic, A.: "The Polythionic Acids," Z. anorg. Chem.,
125 (1922), p. 86.

Mar., 1932] REDUCING ACTION OF FIXING BATHS 397

5 FOERSTER, F.: "The Formation and Decomposition of Polythionates,"
Z. anorg. Chem., 139 (1924), p. 246; Z. anorg. Chem., 144 (1924), p. 337.

6 FOERSTER, F., AND MOMMSEN, E. T.: "Thiosulfates," Ber.,57 (B) (1924),
p. 258.

7 FOERSTER, F., BROSCHE, A., AND NORBERG-SCHULTZ, C.: "Sodium and
Potassium Salts of Sulfurous Acid," Z. physik. Chem., 110 (1924), p. 435.

8 FOERSTER, F., LANGE, F., DROSSBACH, O., AND SEIDEL, W.: "The
Decomposition of Sulfurous Acid and Its Salts in Aqueous Solutions," Z. anorg.
Chem., 128 (1923), p. 245; FOERSTER, F., AND KUBEL, K.: "The Decomposition
of Sulfites at Red Heat," Z. anorg. Chem., 139 (1924), p. 261; FOERSTER, F.,
AND VOGEL, R.: "The Behavior of Sulfurous Acid toward Thiosulfuric Acid,"
Z. anorg. Chem., 155 (1926), p. 161; FOERSTER, F., AND CENTNER, R.: "The
Action of Sulfites on Polythionates," Z. anorg. Chem., 157 (1926), p. 45; FOERS-
TER, F., AND HAMPRECHT, G.: Z. anorg. Chem., 158 (1926), p. 277; FOERSTER,
F., AND HAUFE, E.: "The Auto- Decomposition of Aqueous Hydrogen Sulfite
Solutions," Z. anorg. Chem. 177 (1928-29), p. 17; FOERSTER, F., AND KIRCHEISEN,
E.: "The Interaction of Hydrogen Sulfite and Hydrosulfite," Z. anorg. Chem.,
177 (1928), p. 42; FOERSTER, F.: "The Inter-Relationship of the Sulfur Acids,"
Z. anorg. Chem., 177 (1928-29), p. 61.

9 BASSETT, H., AND DURRANT, R. G.: "The Inter-Relationships of Sulfur
Acids," /. Chem. Soc. (1927), p. 1401.

10 CRABTREE, J. I.: "Stains on Negatives and Prints," Amer. Ann. Phot., 35
(1921), p. 204.

11 CRABTREE, J. I., AND HARTT, H. A.: "Some Properties of Fixing Baths,"
Trans. Soc. Mot. Pict Eng., 13 (1929), No. 38, p. 364.

ABSTRACTS

The dews of the readers of the JOURNAL relative to the usefulness to them of the
abstracts regularly published in the JOURNAL will be appreciated. Favorable views
are of particular interest. In the absence of a substantial body of opinion to the
effect that these abstracts are desired by the membership, their discontinuance may be
considered.

Experiments with Visual Aids in High School Classes. W. LEWIN. Visual
Instr. News, 5, Nov., 1931, p. 9. Another quite independent experiment to test
the efficacy of motion pictures in teaching, the subject being high school physics.
Preliminary intelligence, reading, and physics tests showed the control groups to
have a very slight advantage. Motion pictures were presented to the experi-
mental group during their preparatory period while the control group met for
supervised study. It was concluded that motion pictures impart more informa-
tion in a given time and also contribute to retention of information. The gain in
the test grades of the experimental group over the control group was three times
the standard error while at the end of the term, 50 per cent more pupils of the ex-
perimental group passed the course. R. P. L.

A Modern Theater for the Classics. N. BEL GEDDES. Theater Management,
26, Nov., 1931, p. 8. A theater specially designed for the staging of Dante's
Divine Comedy at the Chicago World's Fair has a seating capacity of 5000 and is
similar to the ancient Greek theater. Its plan is a half -circle facing the stage
without balconies or galleries. No proscenium or curtain divides the auditorium
from the stage. The absence of balconies and galleries allows a steeper ramp and
better vision from all seats. The stage is circular and composed of steps. In the
center is a pit, at the far side of which the slope rises to a height of 50 feet. On the
near side, the slope terminates in a ledge only one-fourth as high, which steps
down toward the audience in a series of terraces until it reaches the level of the
bottom of the pit where it terminates in a valley running half-way around the
circle. A 7-foot wall separates the valley from the audience. Mention is made
of two other theaters also planned for the World's Fair in which the absence of
transverse aisles is notable, the rows of seats being given liberal spacing instead.

L. E. M.

Room Noise Reduction for Improved Sound Reception. V. A. SCHLENKER.
Theater Management, 26, Nov., 1931, p. 3. A study of the relations of speech,
music, and room noise in the theater indicates that the noise level should be re-
duced below 30 decibels for the speech, and music must be uncomfortably loud to
be heard above the noise level of 40 to 50 decibels. Excessive treatment of the
theater proper should be avoided in view of a possible interference with the proper
reverberation period which is considered essential to the proper diffusion of sound
to all parts. The room noise can generally be controlled to suitable value by de-
creasing street and lobby noise through maximum treatment in the lobby and
foyer. L. E. M.

398

ABSTRACTS 399

A Clockwork Driven Slow-Motion Camera. Kinemat. Weekly, 178, Dec. 17,
1931, p. 38. A new type of slow-motion picture camera which is actuated by
clockwork is claimed to expose 100 feet of 35-mm. film with one winding of the
mechanism. The speed can be varied from 40 to 120 frames per second and a
reversed fitting allows dissolving to be carried out while the film is being exposed.
A pick-up speed has been developed which permits only 18 inches of film passage
before full rate is obtained. Stopping and starting can be accomplished with a
loss of less than 2 feet of film. -A reflex focusing device permits accurate focusing
when taking close-ups, and the enclosed view finder is fitted with a device to allow
for parallax when the object is near the camera.

A standard speed camera designed similarly to the slow-motion model, but
capable of exposing 200 feet of film at speeds from 10 to 24 frames, has also been
introduced. A special tripod is used with these models. C. H. S.

Effect Lighting. J. H. KURLANDER. Theater Management, 27, Jan., 1932, p.
10. Suitable lighting effects are proposed for the theaters having a straight sound
picture program so as to relieve the show of monotony. A description of equip-
ment required for effect lighting is given. The uses of effect projectors, shutters,
framing devices, masks, slides, special screens, etc., for producing different effects
are discussed. Color effects, animated scenic effects, silhouettes, trick effects,
and others may be used as the occasion suggests. W. J. W.

Diminishing the Fire Hazard. J. J. GREILSHEIMER. Theater Management, 27,
Jan., 1932, p. 16. The use of concrete vaults or sheet metal lockers, even though
equipped with sprinkler systems and vents, is deemed inefficient in preventing
film fires because of the large quantity of film concentrated in one compartment.
Several requirements for a safe and efficient film storage cabinet are enumerated.
A description is given of a cabinet designed to meet these rigid requirements. The
cabinet is constructed in sections featuring individually insulated and ventilated
compartments of 10 pounds capacity which are sealed tightly with automatically
closing and latching doors. A number of fire tests were carried out on the cabinet
filled with film to determine its safety. Detailed results of the tests are given.

W. J. W.

Advances in Sound Reproduction Demonstrated to Motion Picture Engineers.
Theater Management, 27, Jan., 1932, p. 5. Reproductions of organ, orchestral,
and vocal music, which closely approached the quality and volume of the original,
were effected by the use of disk records cut by the vertical method. This method
employs grooves which vary in depth instead of wavering back and forth along the
spiral path as in the commonly used lateral method. The moving element of the
electrical reproducer is made of light-weight materials so that it is able to follow
vibrations up to 10,000 per second with fidelity. A tiny permanent sapphire
point is used which rides smoothly up and down in the grooves. Finished
records are pressed in cellulose acetate which has a surface of extremely fine tex-
ture. Mr. H. A. Frederick of the Bell Telephone Laboratories made the demon-
stration. W. J. W.

Television Talkiola. Theater Management 26, Nov., 1931, p. 34. This ap-
paratus incorporates mechanisms for producing six different types of entertain-
ment within a single cabinet, namely, television with synchronized sound, talking
motion pictures (16-mm. or silent pictures), phonograph, short wave radio, and

400 ABSTRACTS [j. S. M. p. E.

standard broadcast radio. A Vis-horsepower synchronous motor operates the
perforated scanning disk used for television, giving a 6- by 8-inch picture. Rear
projection is used for the 16 mm.-projector. G. E. M.

New Type Record. Theater Management, 26, Nov., 1931, p. 34. This new disk
record is made of much thinner material and is much less easily broken than the
old type shellac record. Although only 12 inches in diameter, as compared with
the older 16-inch record, the new disk will record sufficient sound for 1000 feet of
film. This has been accomplished by employing a lower amplitude of recording,
smaller grooves, and by placing the grooves nearer together. G. E. M.

Novel Loud Speaker. R. H. CRICKS. Kinemat. Weekly, 173, July 9, 1931, p.
69. New principles are claimed in the construction of a novel loud speaker which
has recently been demonstrated in London. Known as the Cinemavox, it is
stated to combine the principles of the piano and violin by providing a large
tuned area for the dissemination of sound. A number of speaker armatures are
distributed at the back of a sounding board some 5 feet square, and are connected
to struts, which are parts of various wooden sections, each having its own natural
resonance frequency. A frequency range of from 13 cycles to 17,000 cycles with
extremely even response is claimed. The sound output is stated to be almost
non-directional. Kodak Abstract Bulletin

New "Jofa" Studio. P. HATSCHEK. FUmtechnik, 7, Sept. 19, 1931, p. 6. A
description is given of the new "Jofa" sound film studio of Jahannisthal, Berlin,
which is the most up-to-date in the city. There are three large studios, 840,
1155, and 840 square meters in area, each associated with a smaller studio, res-
pectively, 480, 480, and 450 square meters in size, and a large number of dressing-
rooms, and smaller rooms for operators, technicians, actors, etc. There are two
studios for re-recording, dubbing, and synchronizing, four projection rooms, two
cutting rooms, and a number of work-shops. Thirty thousand square meters of
land are available for outdoor work, and an additional seventy thousand meters
(the local aerodrome) are at hand if required. The three large studios have
enormous sliding doors opening on the outside lots. This provides a natural
background for studio sets, if desired, and permits a continuation of the studio
action outdoors. For sound-proofing, air spaces are provided between studios,
the floors are insulated from the walls by coke-ash, walls and doors are all double,
and are packed with sound-absorbing material. Doors are provided with a
novel "double-fold system" which is described, and there is a new treatment of
the roof. The electrical supply and the projection and cutting rooms are also
described. A pool, 35 by 15 meters wide and 2.5 meters deep, is provided.

Kodak Abstract Bulletin

Modern Effect Lighting. J. H. KURLANDER. Mot. Pict. Proj., 5, Jan.,
1932, p. 18. A descriptive article on the production of stage and screen light-
ing effects, including information on lamp and lens equipment, types of screen
and screen materials, and the use of color filters, slides, and design glasses.

A. A. C.

Projected Background Cinematography. R. G. FEAR. Amer. Cinemat., 12,
Jan., 1932, p. 11. A method of composite photography is described in which
the foreground action takes place in front of a screen placed so as to receive from
a projector an image of the background desired. Translucent screens in back of

Mar., 1932] ABSTRACTS 401

the action are now often used for this purpose with a standard camera and pro-
jector. The background picture must be absolutely steady on the screen,
illuminated to the highest possible extent, and must be synchronized with a
camera shutter if good results are to be secured. After a discussion of means of
fulfilling these requirements, the author suggests modifications that may prove
useful, and gives a list of patents relating to the process. A. A. C.

New Filters for Exterior Photography with Super- Sensitive Film. EMERY
HUSE AND GORDON A. CHAMBERS. Amer. CinemaL, 12, Dec. 1931, p. 13. Two
new filters, the 3 N5 and 5 N5, are combinations of yellow dyes with a neutral
density filter of 32 per cent transmission. They combine, in a single unit, a
means of decreasing exposure and a color filter suited to the super-sensitive emul-
sion. This means of reducing light intensity has been found preferable to using
a lens diaphragm or a change in shutter opening A. A. C.

Projector Drive Motors. ALBERT PREISMAN. Mot. Pict. Proj., 5, Jan.,
1932, p. 10. Since the advent of sound, the projector drive motor has assumed a
greater importance than ever before. Ease and precision of control, affording a
constant and definite speed, are imperative. The article discusses the underlying
principles of the common types of projector motors and explains how the new
demands are met in modern motor design. A. A. C.

Reverberation Time Measurements in Coupled Rooms. CARL F. EYRING.
/. Acoust. Soc. Amer., Ill, No. 2, Part I, Oct., 1931, p. 181. The paper pre-
sents experimental data on the decay of sound intensity level in acoustically
coupled rooms, together with a theoretical study of the subject.

The type of problem investigated is illustrated by one of the experiments,
which was a study of the sound decay in an enclosure which consisted of a small
live room connecting through an open window into a large dead room. Data
were taken with the sound source in the large room and microphone in the small
room, and vice versa, and with both source and micrpohone in each room. Com-
binations of other types of rooms are included.

Theoretical equations of decay for acoustically coupled rooms are developed,
and are applied to describe the data. The application of these equations to an
idealized theater is shown. W. A. M.

Audible Frequency Ranges of Music, Speech, and Noise. W. B. SNOW. /.
Acoust. Soc. Amer., Ill, No. 1, Part 1, July, 1931, p. 155. "The program of
listening tests described in this paper was undertaken primarily to establish the
audible frequency ranges of the sounds most often encountered in sound repro-
duction. ..." The sound sources studied included twenty separate musical in-
struments, an orchestra, male and female speech, and certain noises.

Qualitative observations by the crew of listeners are tabulated for each sound
source. Quantitative results are given in a table. Two general conclusions are
as follows: "An upper cut-off of 10,000 cycles did not affect the tone of most of
the instruments to a marked extent, but every instrument except the bass drum
and tympani was affected by the 5000 cycle cut-off. A frequency range of 100 to
10,000 cycles was shown to be entirely satisfactory for speech." ". . . . trans-
mission of the entire audible range would seem much more important for noise
reproduction than for reproduction of musical sounds."

The paper contains a great amount of experimental data. W. A. M.

402 ABSTRACTS

Plane Sound Waves of Finite Amplitude. R. D. FAY. /. Acoust. Soc. Amer.,
Ill, No. 2, Part I, Oct., 1931, p. 222. The principal object of the analysis is
to find the change in type of periodic plane waves of sound of finite amplitude
propagated in free air.

A solution of the exact equation of motion is obtained as a Fourier series. Due
to the non-linear relation between pressure and specific volume there is found to
be a gradual transfer of energy from components of lower frequency to those of
higher frequency. Since the effect of viscosity is to attenuate the higher fre-
quency components more than the lower, there is always a wave form having the
harmonic components in a stable relation such that the decrease in relative mag-
nitude of any component due to viscosity is compensated by the relative increase
due to non-linearity. The conditions for stability vary with intensity. There is
therefore no permanent wave form, but the stable wave will change its form more
gradually than any other wave of the same intensity and wavelength. The
change in type of any wave is toward this stable form. There is a marked de-
parture from the sinusoidal in the stable type even for waves of very moderate
amplitude. AUTHOR

A Planetary Reduction Gear System for Recording Turntables. A. V. BED-
FORD. /. Acoust. Soc. Amer., Ill, No. 2, Part I, Oct., 1931, p. 207. "The
present paper has two objects: to present an example justifying the use of a de-
tailed numerical application of electrical circuit analysis to mechanical rotational
systems, and to describe a new planetary turntable drive system that promises
increased steadiness."

The conclusion of an analysis of a simple gear system is that, "... the error of
the turntable position at any moment is about as great as the fundamental error
in the angular tooth pitch in the lowest speed gear."

In the planetary gear system described no gear runs as slow as 33 Vs rpm. with
respect to its meshed mate, and also no gear in the system runs at a speed lower
than 375 rpm. Therefore, disturbances due to errors in gears and irregularities in
bearing friction are of a relatively higher frequency than in a simple gear system
and consequently can be more easily filtered out.

An experimental model of a planetary gear system drive "exhibited less than
0.03 per cent variation in turntable speed at turntable revolution frequency."

W. A. M.

BOARD OF ABSTRACTORS

BROWNELL, C. E. MACFARLANE, J. W.

CARRIGAN, J. B. MACNAIR, W. A.

COOK, A. A. MATTHEWS, G. E.

CRABTREE, J. I. McNicoL, D.

HAAK, A. H. MEULENDYKE, C. E.

HARDY, A. C. MUEHLER, L. E.

HERRIOT, W. PARKER, H.

IRBY, F. S. SANDVICK, O.

IVES, C. E. SCHWINGEL, C. H.

LOVELAND, R. P. SEYMOUR, M. W.

WEYERTS, W.

ABSTRACTS OF RECENT U. S. PATENTS

The views of the readers of the JOURNAL relative to the usefulness to them of the
Patent Abstracts regularly published in the JOURNAL will be appreciated. Favorable
views are of particular interest . In the absence of a substantial body of opinion to
the effect that these Patent Abstracts are desired by the membership, their early dis-
continuance may be considered. If, after two weeks from the date of mailing the
March issue of the JOURNAL, no letters concerning the continuance of the depart-
ment will have been received, the Patent Abstracts will be discontinued.

1,828,798. Film Treating Apparatus. G. C. BEIDLER. Oct. 27, 1931. The
film is delivered edgewise to means for removing the film from the receptacle in
which a submerging device is located and other guiding means operate to prevent
lateral movement of the film as it is being moved. Means are provided for regu-
lating tension or pressure on the film by coacting rollers which operate to move
the film and at the same time exert pressure upon the film to remove fluid, in
order to prevent film from carrying an excess amount of fluid from the receptacle
in which the film was treated. At the bottom of the coils where they coact,
means are provided for moving the film to eject it from a receptacle, an assembly
of rollers and conveying bands being provided for continuously directing the film.

1,828,749. Motion Picture Screen. A. L. RAVEN. Oct. 27, 1931. The pro-
jection screen comprises a plurality of wavy horizontal strips arranged in overlap-
ping relation with the hollows of the waves of adjacent strips opposite one another
and forming sound passages extending upwardly from the rear toward the front of
the screen between the strips. The sound from the sound reproducer behind the
screen freely passes through the screen at the same time that a proper reflection
surface is provided for the screen.

1,828,768. Film Guide. A. DINA. Assigned to International Projector
Corp. Oct. 27, 1931. One set of guide members is rigidly mounted for positively
locating the film edge with respect to the projection aperture and comprises a
plurality of sections spaced longitudinally of the film for permitting dust and
accumulations of foreign material to escape therebetween. The other set of
guide members comprises a plurality of disks rotatably mounted with their axes
transverse to the film and held in firm engagement therewith by means of suitable
spring members. The disks are capable of rotating as the film is moved through
the projection head thereby eliminating sliding friction and reducing the wear on
the film.

1,828,867. Scanning Device. C. FRANCIS JENKINS. Assigned to Jenkins
Laboratories. Oct. 27, 1931. The film image is enlarged by projection and
directed through a scanning disk thereby permitting (1) the apertures in the
scanning disk to be larger, so that diffraction bears a lesser relation to the aper-
ture area; (2) the disk may be positioned in a free air, removed from the proximity
of the film, and, therefore, does not clog up with dirt and/or oil; and (3) the
apertures may be made square, increasing the light efficiency.

1,828,875. Electrooptical Translation System. C. H. W. NASON. Assigned

403

404 PATENT ABSTRACTS [J. S. M. p. E.

to Jenkins Television Corp. Oct. 27, 1931. A method of employing photo-
electric variations to control the resonance characteristic of an oscillatory circuit
supplied from a source of carrier current. The frequency spectrum of the trans-
mitted carrier waves is substantially independent of the frequency variations of
the light impulses incident upon the light-sensitive device under control of a film.
The light passing through each elemental area of the film is projected upon the
photoelectric cell, preferably of the Elster-Geitel type, comprising a light-sensitive
electrode and another electrode. The electrostatic capacity of such a cell under-
goes variations in value as the coating is subjected to different degrees of illumi-
nation.

1.828.940. System for Correcting Sound Records. R. J. POMEROY. Oct. 27,
1931. A method and system for making a distortion corrected record, by intro-
ducing to the original record correction distortions that are compensatory of, or
have a neutralizing effect on, the distortions which are introduced by reproduc-
tion. This is done by recording the distorted reproduced sound, and utilizing
this distorted record to modify the original record in such a manner that the dis-
tortive effects of the system are compensated in the modified record, and accurate
reproduction is thus obtainable.

1.828.941. System for Correcting Sound Records. R. J. POMEROY. Oct. 27,
1931. A method and system for making a distortion corrected record, and this is
done by introducing, to the record, correction distortions that are compensatory
of, or have a neutralizing effect on, the distortions which are introduced by repro-
duction. A sound current representing the distortion record is combined with a
sound current representing the original undistorted record, and this combination
is so effected that the resultant current carries variations which represent only the
difference between the two records or, in other words, the distortion. A record
of this current may be made upon a film and subsequently printed above an ori-
ginal record. In either case, the result is a distortion corrected record from which
sound may be finally reproduced without the distortions of recording and repro-
duction.

1,828, 942. Production of Corrected Sound Records. R. J. POMEROY. Oct.
27, 1931. A method and system for making a distortion corrected record, and
this is done by introducing to the record correction distortions that are compensa-
tory of, or have a neutralizing effect on, the distortions which are introduced in
recording and reproduction. This is accomplished in the present instance by
making a photographic distortion corrected sound record, or photographic sound
record compensated for distortions, and from this making a distortion corrected
mechanical record from which distortionless reproduction is obtainable.

1 ,828,974. Photographic Film with Visible Reproducible Inscriptions. H. LUM-
MERZHEIM AND E. ScHNiTZLER. Assigned to Agfa Ansco Corp. Oct. 27, 1931.
An ink is provided for continuous printing on a photographic film, the ink com-
prising a mixture of cerasine-red in glycol acetate. A photographic film provided
on the rear side with inscriptions by means of the said dye-ink may be polished as
usual in the photographic film industry; it may be exposed, developed, and fin-
ished in the usual manner without fading of the impressed symbols.

1,829,095. Film Reel. W. G. KING AND M. E. KRAUSE. Oct. 27, 1931. An
endless film may move continuously or intermittently in a continuous path from
the inner convolution of a roll of film revolving about the circularly grouped rollers

Mar., 1932] PATENT ABSTRACTS 405

onward through the mechanism of a projector and past the lens and back onto
the outer convolution of a roll of film without undue strain or intricate twists or
loops in the film by the provision of yieldable film guides.

1,829,103. Loading Device for View Taking Cinematographic Apparatus.
A. N. MERLE. Assigned to Pathe Cinema, Anciens Etablissements Pathe Frdres.
Oct. 27, 1931. The cover is provided with a bevelled part corresponding to that
of the cover and is formed on the face of the loading case coacting with the cover.
This bevelled part is situated outwardly of the film-holding chamber. The said
bevelled part may extend upon the whole periphery of the said chamber or upon
only a certain portion thereof. The loading case may be readily opened to allow
access to the hollow interior of the box for the insertion or removal of the film.

1,829,121. Sound Recording Apparatus. E. R. VINSON. Oct. 27, 1931. An
electromagnetic vibratile device for moving a light valve in the form of a V-
shaped notch in the path of a beam of light for varying the exposure of the film
according to impressed sound vibrations.

1,829,359. Picture Projecting Machine Cabinet. R. W. KITTREDGE. Oct.
27, 1931. Cabinet for a motion picture projecting machine, and a projection
screen and stand therefor which is removably stored on the cabinet in such a
manner as not to decrease materially the space afforded in the cabinet for the re-
ception or storage of other articles such as a projecting machine, related appara-
tus, and film, which permits the quick and convenient storing of the screen and
stand on the cabinet and removal of the same therefrom, and which does not de-
tract from the appearance of the cabinet or require an unattractive shape thereof
for use in the home to form an attractive and convenient article of furniture.

1,829,475. Projection Lamp Holder. G. H. CUSHING. Oct. 27, 1931. A
tubular holder for an incandescent lamp by means of which a standard electric
light bulb may be positioned in an accurately designed reflector, so that the fila-
ment of the bulb will be located at the focus of the reflector.

1,829,482. Motion Picture Film Reel. Oct. 27, 1931. A. C. Hayden. A
film reel for motion picture apparatus comprising a pair of plates and a hub be-
tween said plates adapted to have a film wound thereon, one of said plates having
a hole therein and the other of said plates having a cup integral therewith, pressed
thereform and in axial alignment with the hole, said hole and cup being adapted
to receive a spindle of motion picture apparatus, said hole being formed for driving
engagement with the spindle, and said cup insuring application of the reel to the
spindle with the hole plate in advance of the cup plate.

1,829,633. Taking or Projecting Panoramic Views or Views Extending in
Height. H. CHRETIEN. Assigned to Societe Anonyme Francaise Dite Societe
Technique D'Optique et de Photographic. Oct. 27, 1931. Method of photo-
graphing or projecting which consists of reducing optically the space occupied by
the images on a sensitized surface, by compressing them in one single direction,
either in height, or in width, or in any inclined direction selected, this result being
obtained by disposing, in front of the photographing objective, a special optical
combination, referred to as a local anamorphoser, suitably oriented about the
optical axis of the objective. The process also consists in restoring or projecting
these images through an optical combination similar to that which has served for
obtaining them and similarly directed, which has the result of reestablishing the

406 PATENT ABSTRACTS

images in their exact proportions on a screen of suitable dimensions and arrange-
ment.

1,829, 634. Optical Compression of Film Pictures. H. CHRETIEN. Oct. 27,
1931. A film which includes a series of pictures thereon of uniform dimensions
and proportions which are optically compressed, some in one dimension and some
in another, so as to obtain when projected and restored views which are consider-
ably extended but only in the one dimension or the other.

1,829,791. Device for Recording Sound on Film. H. A. DEVRY. Assigned to
Q. R. S.-De Vry Corp. Nov. 3, 1931. Incandescent lamp having a bulb, part of
which is opaque except for a minute slit in the tip end thereof. The lamp is
adapted to be positioned with respect to feeding or winding mechanism for the
film so that the beam of light emanating from the slit strikes against one of the
side margins of the film and forms on the film, as the latter is driven by the feed
mechanism, an exposed portion of strip-like conformation.

1,829,912. Sound Picture Film and Method of Making the Same. D. G.
SHEARER. Assigned to Metro-Goldwyn-Mayer Corp. Nov. 3, 1931. Con-
tinuous picture film and sound record comprising a strip of film bearing pictures
between rows of sprocket holes made therein, and a continuous photographic
sound record on film stock attached to one longitudinal edge of the picture film,
one edge of the picture film being stepped and the stepped edges cemented to-
gether whereby the combined picture film and sound record are of substantially
equal thickness transversely thereof.

1,830,082. Color Attachment for Cinema Projectors. W. R. BECKLEY, A. E.
CHURCH AND J. F. MERKEL. Assigned to Beckley and Church, Inc. Nov. 2,
1931. A rotatable disk is placed upon the front of the projector and arranged in
front of the lens and adapted to present various differently colored transparent
segments thereof in the axis of the lens, selectively, so that the projected rays will
be colored or filtered in a manner such as will protect the eye of the observer from
the glare of the image as projected upon the screen and also when desired to im-
part a colorful effect simulating, for instance, moonlight, twilight, etc.

1,830,121. Color Attachment for Cinema Projectors. J. F. MERKEL. As-
signed to Beckley & Church, Inc. Nov. 3, 1931. A carrier for a lens plate is
mounted on the projector and a vari-colored ray screen rotatably and reversibly
mounted on the shaft in operative relation to the lens for the reproduction of
images in color.

1,830,158. Film Trap and Film Trap Door. A. DINA. Assigned to The Pre-
cision Machine Co., Inc. Nov. 3, 1931. Construction of film guide and film
trap door in which the door is closed upon the film against impact absorbing
means which prevents transmission of shocks to parts of the projector. A resilient
contacting pad is provided against which the door is moved to closed position.

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.)

SOCIETY OF MOTION PICTURE
ENGINEERS

OFFICERS
1931-1932

President
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester. N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE, Fox Film Corp., New York. N. Y.

Secretary
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y.

Board of Governors

York, N. Y.

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLE, 277 Park Ave., New York, N. Y.

407

SOCIETY ANNOUNCEMENTS

SPRING, 1932, MEETING

The Spring, 1932, Convention of the Society is to be held at Wash-
ington, D. C., with headquarters at the Wardman Park Hotel.
Excellent service is assured and plenty of space is available for
accommodating the members without crowding. The Congressional
Country Club and the Indian Spring Country Club are both available
to the visiting members. In addition, the four tennis courts main-
tained by the hotel and riding facilities provide additional recreation.

As the Convention is to be held at the height of the activities of
the Washington Bi-Centennial celebration, there will be much to
attract members to Washington in addition to the technical and
social activities of the Society. Sight-seeing tours will be provided
for visiting points of historic and diplomatic interest, such as the
Capitol, the Treasury, the Smithsonian Institution, the Congres-
sional Library, the Pan-American Building, the new Museum,
Scottish Rite Temple, Tomb of the Unknown Soldier, the Amphi-
theater, and other points of interest at Arlington, Mount Vernon, and
Annapolis.

An especially attractive program of technical papers is being
prepared by the Papers Committee, under the chairmanship of
Mr. O. M. Glunt; and Mr. W. C. Kunzmann and his Convention
Arrangements Committee are sparing no efforts to make the social
aspects of the Convention a success. The technical sessions will
be held in the Little Theater of the Wardman Park Hotel, and special
film programs for the evenings are being arranged by Mr. J. I.
Crabtree.

The semi-annual banquet of the Society will be held in the Gold
Room of the hotel on Wednesday, May llth, at 7:30 P.M. In
addition to an attractive and entertaining program, an unusually
interesting group of speakers is expected to address the members.

NEW YORK SECTION

On February llth, the members of the New York Section were in-
vited by the Illuminating Engineering Society to attend its Febru-
408

SOCIETY ANNOUNCEMENTS 409

ary meeting held at the plant of the Sperry Gyroscope Company,
Brooklyn, N. Y. A paper entitled "The Theory of the Arc, and the
Carbon Arc as a Projection Source" was presented by Mr. Bassett,
of that company.

The members of the Section were also invited by the New York
Section of the American Institute of Electrical Engineers to attend
its meeting held on February 26th, at the Engineering Societies
Building, New York, N. Y. The meeting was devoted to "The
New Music of Electrical Oscillations," and included demonstrations
of the electronic organ-piano, developed by Mr. Benjamin Miessner
and the Ranger tone electric organ, developed by Captain Richard
Ranger. Professor Leon Theremin demonstrated the three types
of theremin — the space theremin, the new keyboard theremin, and
the new fingerboard theremin.

CHICAGO SECTION

At a meeting of the Section, held on January 7th, papers were
presented by Mr. R. W. Fenimore, entitled "Educational and Com-
mercial Films with Sound on Disk," and by Mr. L. D. Minkler, on
"Disk Recording for Motion Pictures."

At another meeting held on February llth, Mr. R. F. Mitchell
presented a paper entitled "New Improvements in Camera Con-
struction."

The next meeting of the Section will be held March 3rd at the
Electric Association, Chicago, 111. Mr. H. Shotwell will present a
paper on the subject of "Portable A-C. Amplifiers."

STANDARDS COMMITTEE

At two meetings of the subcommittee of the Committee on
Standards and Nomenclature, which deals with the establishment
of dimensional standards for 16-millimeter sound film, on January
28th and February 8th, two lay-outs were made, which are to be
submitted to the entire Standards Committee for consideration
and appropriate action. The one lay-out, providing for a single
row of perforations, is to be submitted for adoption as a recommended
standard of film lay-out; the other, providing for two rows of per-
forations, is also to be submitted to the Standards Committee, with
the suggestion that this be published (somewhat as in the nature of
a minority report) as a non-recommended standard, to be followed
if future developments of the art so indicate.

410 SOCIETY ANNOUNCEMENTS [J. S. M. P. E.

Drawings of the two lay-outs are being prepared, showing all
details and tolerances, which will be submitted to the Standards
Committee at its next meeting, to be held in the near future. Upon
ratification of these lay-outs, as submitted or modified, they will be
published in the next succeeding issue of the JOURNAL.

SOUND COMMITTEE

At a meeting held on December 10, 1931, an outline of the work to
be prosecuted by the Committee during the current year was formu-
lated, and included a considerable amount of study of the acoustical
properties of auditoriums and studios, with particular reference to
the influence these properties exert in the recording and reproducing
of sound. An attempt will be made to define an optimum theater,
that is, one whose properties may be regarded as reference standards
which will indicate the factors to be considered in making audi-
toriums acceptable for the reproduction of sound. Among the other
items included in the agenda are: (1) the accuracy and application
of testing methods and formulas; (2) absorption data of acoustic
materials; (3) wide-range recording and reproducing of sound;

(4) sources of ambient or interfering noises, and their correction;

(5) the relation between the acoustical properties of studios and
theaters; (6) the influence of the light slit and of the methods of
processing film on the frequency characteristic of reproduction;
(7) the desirability of increasing the range of volume of reproduction ;
and (8) variations in negative exposures.

Various subcommittees have been appointed to study these
several subjects outlined, the reports of which subcommittees are
to be submitted at a meeting of the entire Committee in the near
future.

JOURNAL AND PROGRESS AWARDS

At a meeting of the Board of Governors held May 24, 1931,
it was decided that the following actions of the Board, relating to
the Journal Award and the Progress Medal, should be published
annually in the JOURNAL.

JOURNAL AWARD

The motion was made and passed that "an award of $100.00 shall
be made annually, at the Fall Convention of the Society, for the

Mar., 1932] SOCIETY ANNOUNCEMENTS 411

most outstanding paper published in the JOURNAL of the Society
during the preceding calendar year. An appropriate certificate
shall accompany the presentation.

"The Journal Award Committee shall consist of not less than six
Active members of the Society, to be appointed by the President sub-
ject to ratification by the Board of Governors. The Chairman of the
Committee shall be named by the President and a two-thirds vote is
necessary for election to the award. (Proxies are permitted.)

"The Committee shall be required to make its report to the Board
of Governors at least one month prior to the Fall Meeting of the
Society, and the award must be ratified by the Board. A list of
five papers shall also be recommended for honorable mention by
the Committee. These rules, together with the titles and authors'
names, shall be published annually in the JOURNAL of the Society."

PROGRESS MEDAL

"The Board of Governors may consider annually the award of a
Progress Medal in recognition of any invention, research, or develop-
ment, which in the opinion of the Progress Award Committee shall
have resulted in a significant advance in the development of motion
picture technology.

"The Committee shall consist of not less than six Active members
of the Society, to be appointed by the President subject to ratifica-
tion by the Board of Governors. Names of persons deemed worthy
of the award may be proposed and seconded, in writing, by any
two Active members of the Society and shall be considered by the
Committee during the month of June; a written statement of ac-
complishments shall accompany each proposal.

"Notice of the meeting of the Progress Award Committee must
appear in the March and April issues of the JOURNAL. All names
shall reach the Chairman not later than April 20th.

"A two-thirds vote of the entire Committee shall be required to
constitute an award of the Progress Medal. Absent members may
vote in writing. The report of the Committee shall be presented
to the Board of Governors for ratification at least one month before
the Fall Meeting of the Society.

412 SOCIETY ANNOUNCEMENTS [J. S. M. p. E.

who have received the medal, the year of each award, and a state-
ment of the reason for the award shall be published annually in the
JOURNAL of the Society."

Active members of the Society are invited, according to the above,
to propose names of those deemed worthy of receiving the Progress
Medal Award, which proposals should be seconded by another Ac-
tive member and forwarded to the Chairman of the Committee,
Dr. C. E. K. Mees, addressed to the General Office of the Society.
A written statement of accomplishments should accompany each
proposal, which should reach the Chairman not later than April 20th.

The two committees have this year been amalgamated into a single
committee known as the "Committee on Journal and Progress Medal
Awards."

MEMBERSHIP CERTIFICATE

Society )Joti<Hi Picture Engineers

THIS IS TO CERTIFY THAT

Society of Motion Picture Engineers

Mar., 1932] SOCIETY ANNOUNCEMENTS 413

LAPEL BUTTONS

SUSTAINING MEMBERS

Agfa Ansco Corp.
Bausch & Lomb Optical Co.

Bell & Howell Co.

Bell Telephone Laboratories, Inc.

Carrier Engineering Corp.

Case Research Laboratory

Du Pont Film Manufacturing Co.

Eastman Kodak Co.

Electrical Research Products, Inc.

Mole-Richardson, Inc.

National Carbon Co.

RCA Photophone, Inc.

Technicolor Motion Picture Corp.

BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS

No. Price No.

Price

1Q17r 3 $0.25

917 \ 4 0.25

1924-

18
19

$2.00
1.25

1918 7 0.25

1.25

1920
1921

10 1.00
11 1.00
12 1.00
13 1.00

1925

21
22
23
24

1.25
1.25
1.25
1.25

1922
1923

14 1.00
15 1.00
16 2.00
17 2.00

1926'

25
26
27
28

1.25
1.25
1.25
1.25

1927

1928

1929 ^

No.

29
30
31
32
33
34
35
36
37
38

Price

$1.25

1.25
1.25
1.25
2.50
2.50
2.50
2.50
3.00
3.00

Beginning with the January, 1930, issue, the JOURNAL of the Society has been
issued monthly, in two volumes per year, of six issues each. Back numbers of all
issues are available at the price of $1.50 each, a complete yearly issue totalling
$18.00. Single copies of the current issue may be obtained for $1.50 each.
Orders for back numbers of Transactions and JOURNALS should be placed through
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and
should be accompanied by check or money-order.
414

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR
Volume XVIII APRIL, 1932 Number 4

CONTENTS

Page

The Problem of Projecting Motion Pictures in Relief

H. E. IVES 417

The European Film Market — Then and Now

C. J. NORTH AND N. D. GOLDEN 442

Victrolac Motion Picture Records F. C. BARTON 452

Optics of Projectors for 16 Mm. Film A. A. COOK 461

Silica Gel Air Conditioning for Film Processing

E. C. HOLDEN 471

Measurements with a Reverberation Meter

V. L. CHRISLER AND W. F. SNYDER 479

16 Mm. Sound Film Dimensions R. P. MAY 488

Proposed Change in the Present Standards of 35 Mm. Film

Perforations A. S. HOWELL AND J. A. DUBRAY 503

The Animatophone — A New Type 16 Mm. Synchronous Disk

Reproducer A. F. VICTOR 512

The Acoustics of Large Auditoriums S. K. WOLF 517

Committee Activities:

Report of the Sound Committee } 526

Abstracts .. 530

Patent Abstracts • 533

Book Reviews 536

Officers 537

Society Announcements 538

Spring Convention, Arrangements Program 540

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Board of Editors

J. I. CRABTREE, Chairman

L. DE FOREST A. C. HARDY F. F. RENWICK

O. M. GLUNT E. LEHMANN P. E. SABINE

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers.

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc.

Subscription to non-members, $12.00 per annum; to members, $9.00 per annum,
included in their annual membership dues; single copies, $1.50. A discount
on subscriptions or single copies of 15 per cent is allowed to accredited agencies.
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930, at the Post Office at Easton
Pa., under the Act of March 3, 1879.

THE PROBLEM OF PROJECTING MOTION PICTURES

IN RELIEF*

HERBERT E. IVES**

Summary. — The essential conditions for producing pictures in stereoscopic
relief are two: First, separate pictures must be made from different points of view,
corresponding to the two eyes; second, each eye of the observer must receive its appro-
priate view. No compromise with these fundamental requirements appears possible.

If stereoscopic projection is to be achieved in such a form that a large group of
observers may simultaneously see the projected picture in relief, the distribution of
the appropriate views to the two eyes must be accomplished for each observer. There
are two places where the distribution may be made: the first is at the observers' eyes;
the second is at the screen on which the picture is projected.

If the first method be employed, two separate images must be provided on the screen,
and every observer must have means for directing one image to the right eye and one
to the left eye.

If distribution of the images is to be made at the screen, two images are no longer
sufficient. Theoretically an extremely large number must be provided, a separate
one for each position that can be occupied by any eye in the audience.

Several methods of utilizing the parallax panoramagram method are discussed.
It appears that from the theoretical standpoint the problem of relief projection is
entirely soluble, and experimental tests of still picture projection have been success-
fully made. Practically, the solution of relief projection of motion pictures will
depend upon the use of apparatus involving excessive speeds of operation, great
multiplicity of taking or projecting units, projection screens containing minute
ridged reflecting or refracting elements of extreme optical perfection, projection lenses
of extraordinary defining power, microscopic accuracy of film positioning, and photo-
graphic emulsions of speeds at present unknown.

The perception of relief in vision, that is, the location of different
objects in the field of view at their proper relative distances from the
eyes, is contributed to by a number of factors. We may list among
these: geometrical perspective, according to which objects decrease in
angular extent with the distance from the eyes; aerial perspective, by
which distant objects are more or less veiled by intervening atmos-
pheric haze; the effort of focusing or accommodating the eyes to

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Bell Telephone Laboratories, New York, N. Y.

417

418 HERBERT E. IVES [J. S. M. P. E.

objects at different distances; and, when the observer can move, by
the different relative angular motion of near and distant objects. All
these factors have been utilized to stimulate relief by makers of pic-
tures both still and moving. The most important factor, however,
and the only one that needs discussion as a problem still awaiting
practical solution is binocular vision, which is peculiar to man and
certain of the higher animals, because of the location of the eyes side
by side, both receiving images of the same objects. I shall, therefore,
in this discussion, proceed at once to binocular or stereoscopic relief,
and our problem will be to consider the ways and means by which
motion pictures might be projected so as to exhibit relief of this
character.

BINOCULAR OR STEREOSCOPIC RELIEF

While the complete explanation of the process by which we ap-
preciate relief when the two eyes receive images which are somewhat,
but not too different, in character, has not been worked out to the
satisfaction of psychologists, the essential physical conditions of
stereoscopic relief are simply stated. They are as follows: (1)
Separate pictures must be available, made from different points of
view, corresponding to the two views that are seen by the right and
left eyes. (2) Each eye of the observer must receive its appropriate
view. These conditions are essential and inescapable. No com-
promise with them appears possible. No scheme which calls for a
single picture or series of pictures taken from one point of view will
meet the first requirement. No scheme which does not provide means
for distributing the appropriate views to the two eyes will meet the
second requirement. Once stated, these requirements appear ob-
vious, and they have indeed been clearly understood by students of
optics for approximately 100 years. In spite of this, however, would-
be inventors continue with surprising regularity to announce schemes
for projection in relief which they claim require no special camera or
form of picture, or, if they propose taking two pictures in order to meet
the first requirement, evade the provision of means for separating
these pictures in the process of viewing.

Having now cleared the ground, we are prepared for a straight-
forward discussion of our problem. For purposes of presentation, we
may conveniently discuss it in three steps : The first step will be the
production of relief pictures by processes which do not involve pro-
jection. The second step will take up relief pictures produced by

April, 1932] MOTION PICTURES IN RELIEF 419

projection processes, but in the form of "stills," that is, not embodying
motion. The third step will be to consider the projection of relief
pictures in motion.

METHODS OF MAKING RELIEF PICTURES

In accordance with the requirements as stated above, the first
piece of special apparatus which is needed in order to produce a
picture in relief is some form of camera (we shall, of course, assume
that the process of producing pictures is photographic), which can
produce pictures from a number of points of view. In the simplest
case, the number of points of view will be two, one for each eye, the
apparatus consisting of a pair of similar cameras whose lenses may be
separated by approximately the distance between the two eyes.

Pursuing this simplest method of making relief pictures, that is,
simple stereoscopic pictures of the old and well-known form, we may
now go over to the viewing end and consider means of meeting the
second requirement: namely, the distribution of the two pictures to
the appropriate two eyes. The simplest apparatus for viewing two
pictures, one at each eye, consists of no apparatus at all, but lies in the
proper directing of the two eyes. Holding up a pair of stereoscopic
prints in front of the eyes, with the right eye view at the right and the
left eye view at the left, one can, by practice, learn to diverge the
optic axes and see one picture with each eye ; or, if the two pictures
are mounted side by side, but in the reversed relative positions to
those just considered, one can, by converging the optic axes to a point
between the eyes and the pictures, again see one picture with each eye,
and thus produce a picture in stereoscopic relief.

Next in order of complexity of viewing device is some form of
stereoscope. This may consist of mirrors or prisms placed one over
each eye, and so directed or of such angle as to present one view to
each eye, the eyes being in their normal unconverged or undiverged
position. The stereoscope is an instrument very familiar to students
of optics, and in a previous generation achieved wide popularity as a
form of entertainment. In our present more feverish age, the appeal
of pictures without action, even though possessing another aspect of
naturalness, is so slight that it is now not unusual to find people who
have never looked through a stereoscope.

Another means of distributing the pictures to the appropriate eyes
is provided by utilizing color. In the anaglyph, the two elements of
the stereoscopic pair are printed in complementary colors, and special

420

HERBERT E. IVES

[J. S. M. P. E.

spectacles are provided for the observer with a screen of different color
for each eye, whereby only one picture is seen through either element
of the spectacles.

The revolutionary idea that the distribution of the different views to
the two eyes might be made, not at the eyes of the observer, but at the
picture itself, was introduced by Frederic E. Ives about thirty years
ago in the invention of the parallax stereogram. This device, since it
is the direct ancestor of the most interesting projection methods which
I shall describe, demands careful description and comprehension.
According to requirement (1), as stated above, two pictures are taken,

FIG. 1. The principle of the parallax stereogram.

from two points of view. Instead, however, of being mounted side by
side as in the ordinary stereogram, these pictures are divided into
very narrow strips, these strips being juxtaposed so that the left-hand
strip of a pair is from the right eye view, and the right-hand strip from
the left eye view. Close to this picture of alternate strips, which is in
the form of a transparency, is mounted an opaque line grating with its
clear spaces approximately half the width of its opaque spaces. This
grating is mounted at such a distance in front of the stripped picture
and in such relative lateral positioning of its lines that at a certain
distance from the observer's face, the right eye strips are entirely con-

April, 1932]

MOTION PICTURES IN RELIEF

421

cealed from the left eye and the left eye strips are entirely concealed
from the right eye. Each eye then sees only a single view composed of
a series of strips which, however, are made of such fineness (say, 100 to
the inch) as to be invisible or unobjectionable at the viewing distance.
This parallax stereogram, when held directly in front of the face,
parallel to the two eyes and at the proper distance, exhibits stereo-
scopic relief without the interposition of any viewing device located at
the observer's eyes. The principle of the parallax stereogram is
illustrated in Fig. 1, and Fig. 2 is a photomicrograph of a small portion
of an actual parallax stereogram transparency.

FIG. 2. Photomicrograph of portion of parallax stereogram showing alter-
nating juxtaposed strips from right and left eye images.

A limitation of the parallax stereogram is that it must be viewed
from a single definite direction and distance. While this detracts but
little from the appeal of the picture if only one observer is to be con-
sidered, it is a serious defect if, as must be the case when we come to
discuss means for projecting pictures visible to an audience, a large
number of people, variously placed, must observe the relief picture
simultaneously. In order to achieve a relief picture which shall be
visible at any distance from any direction of observation, it is neces-
sary to break away from the idea that stereoscopic relief is essentially
a matter of two images. Consider that the picture is to be viewed not

422

HERBERT E. IVES

[J. S. M. P. E.

by one person in one position, but by any number of people in any
possible positions. It is obvious at once that while each of these
observers needs only two images to satisfy his two eyes, the total
number of eyes to be satisfied may be very great. This demands at
the taking end that some camera arrangement be adopted which will
make the pictures from a very large number of points of view. At the
receiving end it demands that the grating, or its equivalent, have
relatively extremely narrow clear spaces so that, as an observer's eye
takes up different angular positions, an entirely new composite view
will be seen. In short, in place of the two strips which are behind
each grating of the stereogram, there must be an extremely large
number of minute strips behind each very narrow grating opening,
and since these strips are (in the horizontal direction) little panoramas>
I have proposed the name of "parallax panoramagram" for this kind

L ' L 'R

FIG. 3. The principle of the paral-
lax panoramagram.

of picture which shall exhibit relief from any angle or direction of
observation. The principle of the parallax panoramagram is illus-
trated in Fig. 3. Fig. 4 shows, greatly enlarged, a portion of a
parallax panoramagram positive suitable for viewing through a grat-
ing with very narrow clear spaces.

It is evident that the problem of making parallax panoramagrams
with their large number of points of view, must inevitably call for
bulky or complicated apparatus. Several methods have been pro-
posed. The most obvious is to provide a battery of cameras, ar-
ranged, say, in an arc about the object, with their lenses in close
juxtaposition. If these cameras are then subsequently used as
projectors for the pictures made in them, and are all directed to a
sensitive plate placed behind a grating having very narrow clear
spaces, the resultant photographic print will, with its grating, consti-

April, 1932]

MOTION PICTURES IN RELIEF

423

tute a parallax panoramagram. In order to avoid the very large
number of cameras and printing projectors required by this ele-
mentary scheme, the alternative has been proposed of using a motion
picture camera which is moved about the object at a slow rate, while
the requisite large number of views are taken in succession upon
a motion picture film. Upon projecting the developed film from a
projector similarly moved, on a sensitive plate behind a grating, a
parallax panoramagram is obtained with considerable simplification
of apparatus, but at the cost of the greater time required for the

FIG. 4. Photomicrograph of portion of parallax panoramagram showing
panoramic strips which are placed opposite the narrow spaces of the viewing
grating.

successive as contrasted with the simultaneous exposures of the first
scheme.

Another method of making parallax panoramagram negatives
consists once more of a moving camera, but uses a grating in front of
the sensitive plate and develops the minute panoramas behind the
grating as the camera is moved relatively to the object, either by
moving the grating during exposure by the width of its spacing (a
method due to C. W. Kanolt) or by separating the grating and plate,
and depending on the sweeping of the beam of light through the grat-
ing slit across the plate behind it as the relative positions of lens,

424

HERBERT E. IVES

[J. S. M. P. E.

grating, and plate are altered during the exposure. This method, like
the one using a motion picture camera, requires a sufficient time for
exposure for the camera to be moved through an arc or other suitable
path about the object.

A third, optically ideally simple, method of making parallax
panoramagram negatives consists in using a single very large diameter
lens or concave mirror for providing the different points of view.
This method requires that the lens or mirror subtend an angle from
the object as large as it is desired that the final picture be visible in
relief. For an angle of 60 degrees this requires that the lens or mirror
have a diameter as great as the distance from which the object is
photographed. Practically, in order to obtain such angles as this, a
concave mirror is the only feasible device. An arrangement which

FIG. 5.

Method of using a large concave mirror for making parallax pano-
ramagrams.

has been used successfully for this purpose is shown in Fig. 5. It
consists of a strip from a 4-foot diameter concave mirror, in front of
which is placed a half -silvered plane mirror at 45 degrees. The light
passing from the object to the concave mirror is reflected back to the
45-degree mirror and then downward to the sensitive plate, which is
placed slightly behind a grating having clear spaces Yso the width of
the opaque. Each element of the concave mirror sees the object from
a different point of view, and reflects an image in a definite direction
through the grating lines. By using this scheme, a parallax panorama-
gram negative may.be made at a single exposure. A certain price
must be paid for such simplification, which is that the perspective
relations are disturbed; infinitely distant objects are imaged at the
focus of the mirror, which lies a relatively short distance behind the
picture plane, thus restricting the method practically to objects near

April, 1932]

MOTION PICTURES IN RELIEF

425

that plane. This restriction is, however, already present in any
practical parallax panoramagram since the definition in the panoramic
strips necessary to differentiate clearly objects far away from the
picture plane is much beyond that possible by the "pinhole" action of
the grating spaces.

Before going on to the question of projection, a few points with
regard to still relief pictures of the parallax panoramagram type may
be noted. As above described, the pictures are transparencies viewed
through an opaque line grating. The form of grating described with

FIG. 6. Section of parallax panoramagram structure
suited for viewing by reflected light. The same struc-
ture is used for several forms of screen for projecting
parallax panoramagrams.

its extremely narrow clear spaces is quite wasteful of light. In its
place may be substituted a grating composed of convex ridges of such
curvature as accurately to focus parallel rays on the panoramic strips.
In order to realize the full advantages of such convex ridges, however,
it is necessary that the strip picture be printed, not on a flat surface,
but on a series of surfaces which are concave with respect to the
ridges already considered. This means that the parallax panorama-
gram should consist of a sheet provided with front and back convex
ridges, each of different curvatures, as shown in Fig. 6. The curva-
tures for this purpose are easily computed, and if the technical difficul-
ties of preparation are overcome, will provide parallax panoramagrams
which are not wasteful of light, and in which the panoramagrams are

426 HERBERT E. IVES [J. S. M. P. E.

visible equally well from all directions of observation. Another point
to be mentioned in passing is that while only transparencies have been
considered, the form of picture just described with its ridged structure
may be made up as a picture for viewing by reflected light, provided
the photographic emulsion be backed by some white reflecting mate-
rial, the picture being printed, of course, to low density. Light inci-
dent on this doubly ridged structure can only come off from any given
narrow element of a panoramic strip in a certain definite direction,
thus meeting the essential conditions.

One further point must be touched upon as presenting an ever-
present technical problem. In making pictures for the ordinary
stereoscope, the photographic lenses, of course, invert each element of
the stereoscopic pair. It is accordingly necessary when stereoscopic
pictures are made on a single plate, that the prints be cut in two, and
each separately inverted. If this be not done, the pictures will
exhibit in the stereoscope, not stereoscopic, but pseudoscopic relief,
that is, solid objects sink in instead of stand out. Now, in the prepara-
tion of parallax stereograms and panoramagrams are involved similar
inverting operations which must be done by some optical inverting
device. As an illustration, the pictures made by means of a large
lens or mirror show pseudoscopic relief if the picture is viewed through
the grating. In order to obtain stereoscopic relief, the expedient is
adopted in this case of viewing the grating through the picture. In
every form of taking and viewing device used for parallax pano-
ramagrams, a close watch must be kept in the inversions due to the
optical elements, and means must be adopted for assuring that the re-
lief is stereoscopic instead of pseudoscopic.

PROJECTION IN RELIEF

Taking up now the problem of projecting pictures in relief, the
logical order is first to study projection of still pictures, leaving until
the end a discussion of the peculiar difficulties introduced by motion.
In general, all the methods which we have discussed for producing
relief pictures are available, with certain modifications for projection.
The essential feature of projection is, of course, that in place of a
picture fixed in the plane which is observed, the actual picture used is
placed in a lantern or other projecting device, and an image, usually
enlarged, is thrown upon the observing plane, which for convenience
may be spoken of as the screen.

Following the same outline as that used in the previous section, we

April, 1932] MOTION PICTURES IN RELIEF 427

note, first of all, that the simplest method of projecting pictures in
relief is to throw upon the screen the two elements of a stereoscopic
pair, and to look at them directly without interposing an optical
instrument, diverging or converging the optic axes so that each eye
appreciates only one picture. All that is necessary, therefore, to
achieve projection in relief is to project pairs of pictures, and to train
our audiences to control their optic axes by making themselves
temporarily cross-eyed, or the reverse, during the projection period.
While this method of stereoscopic projection is entirely feasible for an
audience of optical experts who have had a little training and prac-
tice, it does not appear promising for popular use.

Proceeding next to apparatus to be placed before the eyes of each
observer, we note that each person in the audience may wear the
equivalent of a stereoscope of either the mirror or prism form. Next
in order is the anaglyph scheme, in which the two pictures are pro-
jected in different colors, and each member of the audience wears
colored spectacles. This scheme has been used with success in numer-
ous demonstrations; it suffers from the limitation that it is not
applicable to projection in natural colors. Two other schemes, which
might conceivably be used for non-projected pictures, are nevertheless
specially feasible with projection and are to be ranked among the
practical methods of this sort. These are, respectively, projection of
the two images with polarized light, and projection of the two
images in quick alternation. In the first of these methods, the
two images are projected by two projectors, one with light
polarized, say, in the horizontal plane; and the other with light
polarized in the vertical plane. Each observer is then provided with a
pair of polarizing prisms, the prisms being mounted in front of the
eyes, one vertical and the other horizontal, with respect to its plane of
polarization. By this means, perfect separation of the two images is
obtained. In the alternate projection method, the two images are
thrown on the screen alternately in such rapid succession that they
appear continuous by persistence of vision. In front of each observer's
eyes are then placed shutters which expose the two eyes alternately,
operated in such phase that each eye sees its appropriate image as
projected. This method of relief picture projection has been success-
fully demonstrated to a full theater audience.

These methods of relief projection, which call for separate viewing
apparatus for each member of the audience, are, optically speaking,
simple and reasonably satisfactory, and are easily adapted to motion

428 HERBERT E. IVES [J. S. M. P. E.

pictures. However, the goal of speculation in relief picture projection
has always been some means of achieving relief without subjecting the
observers to the inconvenience of special individual spectacles or the
picture producer to the expense of the multiple viewing apparatus
demanded. While it is at present doubtful whether schemes which
provide the distribution of images to the different observers at the
screen can approach, in simplicity and feasibility, these methods which
divide the images at the eyes, they are of great optical interest, and
I shall proceed forthwith to a discussion of them.

In discussing projection schemes of this general type, I shall adopt
an order of presentation which is not perhaps logical, but which ties in
most closely with the results of our study of non-projected relief
pictures. I shall proceed at once to the problem of projecting parallax
panoramagrams in their most fully developed form. Let us imagine
that instead of putting behind the opaque line grating a transparency
print from a parallax panoramagram negative (made with its pano-
ramic strips properly oriented to be placed behind the grating), we put

FIG. 7. Perspective view of glass or celluloid
rod from which a translucent projection screen can
be built up for projecting parallax panoramagrams.

a translucent screen, and that we remove our parallax panoramagram
print to a projection lantern placed at an appropriate distance behind
the grating and screen; we then project this parallax panoramagram
print upon the screen in exquisite focus and in accurate registration as
to size, position, and inclination of the panoramic strips behind the slits
of the grating. If this operation can be performed with the requisite
accuracy, an observer stationed anywhere in front of the grating will
see a relief picture which will be indistinguishable from the ordinary
parallax panoramagram.

The opaque line grating which we have assumed will, of course, be
very wasteful of light, and in its place it is preferable to use a ridged
structure such as has already been discussed. In the case of projection
we are, of course, interested in much larger pictures than, for instance,
in show window transparencies. In a screen several feet across con-
taining 200 or 300 ridges, the individual ridges may be as large as a
quarter-inch in diameter. This relatively large size makes it feasible

April, 1932] MOTION PICTURES IN RELIEF 429

to consider building up the screen of separate rods of transparent
material, such as glass or celluloid. These rod3 will have a cross-
section consisting of two flat sides, a front surface of one radius of
curvature, and a back surface of another radius of curvature such that
all points of the real surface are in the sharp focus of the lens formed
by the front surface and the body of the rod. This rear surface must

FIG. 8. Experimental arrangement for projecting parallax panoramagrams
upon a translucent screen.

then be given a frosted or other diffusing finish. When a large
number of these rods are clamped together they form a screen of the
desired type, on the back of which the parallax panoramagram print
can be projected. A single rod for such a screen is shown in Fig. 7.
An experimental screen, built up of 200 rods, of this form is shown in
Fig. 8, together with the projection lantern used in an experimental
demonstration of relief projection by this method.

430 HERBERT E. IVES [J. S. M. P. E.

Postponing for the present a discussion of how the slide containing
the several hundred panoramic strip images is to be made, we can
discuss the practical difficulties which must be faced in projection of
this sort. Assuming that the picture to be projected is of ordinary
lantern slide size and that the picture is to be divided into 500 narrow
panoramic strips, which would correspond to a screen 10 feet across
with Y^inch rod elements, we must have on our lantern slide some-
thing like 150 panoramic strips per inch. Each one of these strips
must be a complete little panorama containing enough sharply denned
elements to provide separate images for each pair of eyes in an
audience spread out through at least 60-degrees angular position in
front of the screen. As a working figure, if we assume 100 differentiable
strip elements in each panoramic strip (this corresponds to a separate
view for each eye 20 feet from the screen, lying within 10 feet from the
center line of the auditorium), we must have a lantern slide in which
the resolving power is of the order of magnitude of Vi5,ooo of an inch,
approximating a wavelength of visible light. Proceeding now to the
projection lens, thia must, of course, give an accurately rectilinear
image, in order that the panoramic strips on the slide may be ac-
curately positioned on the back of the projection screen. Next, the
defining power of this lens must be such that it images the panoramic
strips on the backs of the screen rods with exquisite fidelity. Proceed-
ing now to the rod screen, it is obvious that the individual rods must be
figured with an accuracy comparable with that found in good optical
lens work if line elements of approximately one-hundredth the width
of the rod are to be focused from the back diffusing surface into parallel
beams to be passed into the observing space. It may be mentioned in
passing that in place of the transmission screen which has been dis-
cussed, forms of reflecting screen are also possible in which concave or
convex cylindrical rods are used. In every case, however, the require-
ments as to extraordinary perfection of all the optical parts obtain.

From this rough discussion of the requirements, it is obvious that
the projection of a parallax panoramagram by this method calls for
most extraordinary refinement of all the elements concerned. On a
crude scale, however, it has been found by experiment that the
procedure can be carried through, and relief projection has been
accomplished experimentally in this way.

Taking up now the problem of how to produce the "lantern slides"
for projection by this method, it may be said in general that any of the
methods which have been described, such as those employing multiple

April, 1932] MOTION PICTURES IN RELIEF 431

lenses, moving lenses, and so on, may be used. However, looking
ahead toward a procedure which might be applicable to motion
pictures, the most desirable method would be one in which the pictures
are made by a single exposure on a single plate. The one method
which is now available for this is to use a large concave mirror as al-
ready described. When it comes to making pictures for projection,
however, a complication is introduced, which is, briefly, that the
mirror method produces pictures which are too large for insertion
into an ordinary projection lantern. Due to the physical impossi-

FIG. 9. Arrangement for produc-
ing parallax panoramagram nega-
tives of convenient size.

bility of producing a lens or mirror which shall be both of such large
size as to subtend a large angle with ordinary objects, and at the same
time of such short focus as to produce images as small as a lantern slide,
the mirror method is, generally speaking, only successful in making
pictures of natural size. Thus, for making portraits, a mirror having
a radius of curvature of four feet, with the face and the sensitive plate
each placed four feet from the mirror, is a practical arrangement. If
larger or more distant objects are to be photographed, the size of the
mirror must increase in proportion, as well as the size of the picture
which is obtained.

432

HERBERT E. IVES

[J. S. M. P. E.

To overcome this difficulty, the parallax panoramagram negative
made with a large mirror must be reduced in size, by some photographic
procedure. The preferred way to do this is to re-photograph the strip
images, formed behind the grating upon a diffusing glass, directly, in
the first picture- taking operation. A more satisfactory method of
obtaining the strip images is to substitute for the grating and diffusing
glass, a transparent ridged screen. The ridges, in order to assure that
the final picture shall be stereoscopic instead of pseudoscopic, must be
of the correct direction of curvature for the kind of projection screen

Cf)

FIG. 10. Method of projecting pictures in relief
using a large number of projections and two gratings
at the screen.

which is to be used. For a projection screen of the type we have been
discussing, the screen in the camera should have concave cylindrical
ridges, which form minute virtual panoramic images. When a photo-
graphic lens is placed behind this screen at the proper distance with
respect to its focal length, as illustrated in Fig. 9, a second reduced
image is formed which may be made of any desired size, such as that of
a lantern slide. Prints made from this negative are then suitable for
projection. It is again obvious that the optical quality of the concave
ridged screen and of the photographic lens just described must be of

April, 1932] MOTION PICTURES IN RELIEF . 433

extraordinary perfection. Also, that the photographic emulsion used
must be of exceedingly high resolving power.

The system of relief projection which has just been described is,
from the strictly scientific standpoint, bearing in mind its limitation to
objects near the picture plane, a complete solution of the problem of
projecting still pictures in relief.

Before going on to discuss the peculiar problems of motion picture
projection, we may consider some suggestions for evading the severe
requirements which the ideally complete method just described in-
volves; in particular, the great practical difficulties of exact registra-
tion of the projected panoramic strips on the screen elements, and the
necessity for extraordinary resolving power in the photographic
emulsion. There are several ways of escaping from these requirements
which, however, demand giving up the single projector or the single
image. One method which has been experimentally demonstrated
consists in projecting images from a battery of projectors. If, for in-
stance, a translucent screen is mounted with an opaque line grating
both in front of and in back of it, and a multiplicity of images are
projected from different directions through the rear grating upon the
translucent screen, the space in front of the screen will present relief
pictures from any position or direction. (Fig. 10.) The registration of
these multiple images upon the screen is a matter of relatively in-
significant difficulty compared with the registration problem above
considered. A more practical form of screen is one of the reflection
type which is exactly similar to the translucent screen above described,
except that the back surface of the rods is given a diffuse reflecting
finish as, for instance, with aluminum paint. Such rods have the
property of reflecting light exactly in the direction from which it is
incident. A screen built of such rods, therefore, will exhibit to each
eye in an audience only that picture which originates at the projector
lying in its line of sight produced backward. With a battery of
juxtaposed projectors in front of the screen, observers in the space
between the projectors and the screen see the pictures in relief. An
experimental apparatus for demonstrating relief projection of this
sort is shown in Fig. 11.

Another means of avoiding the accurate registration problem and
also of avoiding the necessity for a large number of projectors is to use
a single projector, projecting a rapid succession of images, as from a
motion picture film, but with the projector arranged to move rapidly
from side to side through a sufficient distance to sweep through the

434

HERBERT E. IVES

[J. S. M. P. E.

whole observation angle. This scheme is obviously not very practic-
able because of the mechanical difficulties involved. It is possible to
imagine some optical means by which the beam of light from the pro-
jector could fall upon the screen from different angles without the
projector itself moving, but these again demand very large rapidly
moving parts, and are of little promise.

Still another scheme, which may be described as a hybrid method,
may be mentioned. Suppose that we again use a single motion
picture projector, projecting in rapid succession a series of views

FIG. 11. Experimental apparatus for projecting pictures in relief, using a
battery of projectors and a ridged reflecting screen.

which have been taken from different directions as, for instance, with
a moving lens camera. Suppose that we place immediately behind
the translucent rod screen, an opaque line grating with very narrow
clear spaces, and that we move this grating laterally back and forth
so that each succeeding projected image falls in a series of extremely
narrow bright lines upon the rear surfaces of the rod elements of the
screen. (Fig. 12.) If the pictures are projected with sufficient rapid-
ity, and if the opaque line grating oscillates in exactly the phase rela-
tions required, we shall, by persistence of vision, again have a projected

April, 1932] MOTION PICTURES IN RELIEF 435

motion picture in relief. A modification of this scheme consists in
removing the opaque line grating from immediately behind the screen
to the projection lantern, placing it immediately in front of the motion
picture film, and imaging it accurately upon the back of the rod screen.
We have here again a problem of very perfect image registration, but
the problem has to be faced only once with a built-in element instead
of with every picture.

There are probably other combinations of apparatus which might be
devised, but the point which I wish to make is, I think, sufficiently
clear — that our fundamental problem is one of providing a vast
number of images, and that in order to do this we are inevitably forced
either to make these images of excessively small size or to resort to a
multiplicity of apparatus or to a multiplicity of projections in time.
The whole problem is, philosophically speaking, a manipulation of

sF.

FIG. 12. A "hybrid" projection scheme, using a large
number of images projected in rapid succession.

space and time elements comparable in many ways with the problems
presented in television.

PROJECTION OF MOTION PICTURES IN RELIEF

Because of what has gone before, the discussion of the specific
problem of projecting motion pictures in relief can be made quite
brief. All that is necessary is to take one of the methods which have
been outlined for still picture projection in relief, and to increase
its speed to the point where the required number, say, 20, complete
pictures are projected per second. This calls for all the multi-
plicity of apparatus, all the accuracy of the constituent parts and
other features which have been discussed, together with the additional
difficulties of obtaining greater sensitiveness of the photographic ma-
terials and of performing the accurate registration operations at high
speed.

The specific case of the most scientifically complete method may be
gone into in some detail in order to illustrate what these requirements

436 HERBERT E. IVES [J. S. M. P. E.

amount to. Let us assume that the original film is to be made by
means of the large concave mirror in conjunction with the transparent
ridged screen and photographic lens. (Fig. 9.) With this apparatus
as set up for making still pictures for projection of lantern slide size, a
ridged screen of approximately 200 ridges has been found by experi-
ment to photograph down to lantern slide size with some success, pro-
vided the objects to be photographed lie very closely in the plane of the
picture. With an intense but practicable illumination, the exposure
necessary with this apparatus is about one minute, due to the small
photographic aperture of the mirror and the loss of light occasioned
by the semi-transparent mirror used for throwing the image to one
side. For the process to be applicable to motion pictures at 20 frames
per second, the speed of the photographic emulsion would have to be
increased by approximately a thousand times. If the picture were
photographed down to ordinary motion picture frame size, the resolv-
ing power of the film would, without going into figures, have to be far
better than anything now available; and, if the number of panoramic
strips were increased from 200 to 500 (Y^inch strips on a 10-foot
screen), the resolving power necessary to present individual views to
100 eyes in a row at a 20-foot distance would simply be impossible of
attainment because it would demand film images smaller than the
wavelength of light. Much larger film than that now used would then
be another special requirement.

Going over now to the projection of the picture so obtained, the
problem of the exact registration of the strip images upon the projec-
tion screen is made excessively difficult by the motion of the film.
Each image must, in turn, be so accurately positioned that no waver-
ing of the picture occurs. This means that the film must not shift
laterally by as much as one-hundredth of the width of the strip image
or approximately, for the case just considered, Yao,ooo part of an inch at
the projector. When, in addition to these requirements, we remember
that warping, expansion, and contraction of the film material would
injuriously affect the registration, it is sufficiently obvious that pro-
jection of motion pictures in relief by this method calls for a perfection
of apparatus and materials quite beyond anything now in sight.
The alternative methods which were noted in the last section, while
avoiding the chief difficulties of registration, call, as already noted in
the case of still projection, either for a multiplicity of projectors or for
high speeds of projection, which in the case of motion pictures would
mean some hundreds of times present projection speeds.

April, 1932] MOTION PICTURES IN RELIEF 437

CONCLUSION

As we have reviewed the problem of projecting pictures in relief, it
appears that there are two clearly differentiated methods:

(1) Involves the distribution of the images to the observers' eyes by
means of apparatus individual to each observer.

(2) Calls for means of producing this distribution at the projection
screen.

The practical disadvantage of the first scheme is that it involves
multiplication of viewing apparatus, and some effort and inconve-
nience on the part of the observers. The disadvantages of the second
method, as they have appeared from this analysis, are the excessive
refinement of all the apparatus parts, which could be avoided only in
part by having recourse to a multiplicity of projecting units or
excessive speeds of projection. In their present experimental state of
development, the special screens and other devices called for by the
second method of projection are too crude for projecting pictures
visible to audiences of any great size, and the relief images can be
produced with any satisfactory degree of definition only if the objects
of interest lie close to the plane of the screen. This latter objection
is entirely lacking in the first method of projection. In fact, it may
be said that in the present state of art, the only good quality stereo-
scopic projection which is now possible is accomplished by means
of alternate projection, complementary color filters, polarizing devices,
or other means operating at the eyes of the observers. The means
involving distribution of the images at the screen are of great optical
interest, and may be said to be completely postulated theoretically,
but their practical realization on anything like a commercial basis
appears remote.

It has been tacitly assumed throughout this discussion that, if the
various projection schemes were worked out to perfection, the
resultant relief motion pictures would possess qualities of naturalness
which would add to the appeal of the motion picture. There is, how-
ever, one general consideration which must be recognized: namely,
that it is not, speaking broadly, possible to project a picture in relief
which will be "correct," and at the same time exhibit noteworthy
relief to all members of an audience of any size, stationed at greatly
different distances from the screen. Striking stereoscopic relief is
observed in real life only for relatively close objects, and the amount
of relief varies with the distance of the observer from the object. If,
therefore, a scene were projected in relief to natural size in the average

438 HERBERT E. IVES [J. S. M. P. E.

auditorium using the "parallax panoramagram" method, in which the
relief changes as it should with the distance of observers, only those
members of the audience who were in the front rows would find the
relief quality much of an addition to the picture. If the first method
of projection be used, in which only two pictures are taken, it is true
that all members of a large audience will perceive striking relief,
since the two eyes of each observer will see definitely different
pictures; but it is at only one observing distance, namely, that from
which the original object was photographed, that the relief will be
correct. At other distances, the two pictures will correspond to
points of view greater or smaller than the normal distance between the
eyes, giving exaggerated or diminished relief.

In ordinary projection, in particular motion picture projection,
objects are rarely reproduced in their natural sizes; usually the
screen picture is very greatly magnified. In relief projection,
magnification presents a difficult problem. In the absence of relief,
gigantic close-ups produce little or no impression of unnaturalness. If,
however, a typical close-up were presented in relief, the appearance of
the picture would inevitably be strange and unnatural to many in the
audience. For instance, if the relief picture be produced by one of
the first methods, involving the projection of two images to be sepa-
rated at the eyes of the observers, all the observers, as just noted, will
have the same two points of view, which will correspond to eyes
separated by various distances, according to the viewing position.
The observers from nearby, for whom the pictured object subtends a
much larger angle than normal, will be virtually seeing the object as
though their eyes were separated by several feet. In the case of the
second kind of relief projection, enlarged images are, strictly speaking,
ruled out. A magnified image will actually appear magnified; a
face, for instance, will appear as a giant's face, larger than natural, and
exhibiting the decreased stereoscopic relief that a large object does as
compared with a small one of similar shape. (To put it another way,
if the screen image be magnified, the separation of the eyes of the
observers should be increased in the same proportion.) Close-ups for
this kind of projection should be shown in natural size, but should be
so photographed as to appear located in space in front of the screen at
such a distance from the observer as to give the desired degree of
intimacy. This introduces the interesting complication in this kind
of projection that observers nearer than the point where the image is
formed in space will be between the image and the screen, and will

April, 1932] MOTION PICTURES IN RELIEF 439

get no picture. Practically, it means that no image in space should be
very far in front of the screen.

I do not purpose at this time to enter into a detailed discussion of
these complications, but merely to draw attention to the fact that the
attainment of entirely correct relief projection would carry with it an
inevitable restriction in the size of the audience which would get much
benefit from the added factor of relief. If the relief effects are to be
entirely natural, the motion picture would have to return to a close
simulation of the dimensions of the regular stage, abandoning one of
its unique advances over the stage, namely, the "close-up." Doubt-
less, were relief projection to become feasible and commonplace, a
special art would be developed, which would strike some workable
compromise between the appealing qualities of relief and the unnatural
distortions which great magnification would introduce.

REFERENCES

IVES, H. E.: "A Camera for Making Parallax Panorarnagrams," Jour. Opt.
Soc. of America, 17 (Dec., 1928), No. 6, p. 435.

IVES, H. E.: "Motion Pictures in Relief," Ibid., 18 (Feb., 1929), No. 2,
p. 118.

IVES, H. E.: "Parallax Panoramagrams with a Large Diameter Lens," Ibid.,
20 (June, 1930), No. 6, p. 332.

IVES, H. E.: "Parallax Panoramagrams for Viewing by Reflected Light,"
Ibid., 20 (Oct., 1930), No. 10, p. 585.

IVES, H. E.: "Parallax Panoramagrams with a Large Diameter Concave
Mirror," Ibid., 20 (Nov., 1930), No. 11, p. 597.

IVES, H. E.: "Reflecting Screens for Relief Picture Projection,' Ibid., 21
(Feb., 1931), No. 2, p. 109.

IVES, H. E.: "Optical Properties of a Lippmann Lenticulated Sheet," Ibid., 21
(March, 1931), No. 3, p. 171.

IVES, H. E.: "The Projection of Parallax Panoramagrams," Ibid., 21 (July,
1931), No. 7, p. 397.

DISCUSSION

MR. KELLOGG: When the two views of the Capitol at Washington were
shown, I wondered how far to the side one could go before the actual difference
in the size of the pictures makes it practically impossible to merge them.

MR. IVES: I do not know. Stereoscopic relief can be obtained with one of
the two images in very poor shape. I imagine one could tolerate a lot of dis-
tortion and yet get the effect.

MR. GAGE: These parallax panoramagrams, particularly those that are
colored, are very pleasing. However, I wonder, although they may be inter-
esting as novelties, whether they would be desirable as a regular thing in the
theater. This is a question that ought to be considered here.

PAST-PRESIDENT CRABTREE : I believe that the mind can imagine a lot of things.

440 HERBERT E. IVES [J. S. M. P. E.

If the necessary willingness to believe is created in the observers, they will un-
doubtedly imagine a certain amount of relief, both in the sound and in the
picture.

MR. KURLANDER i In view of the many difficulties in getting true stereoscopic
effects, is there not a simpler method of creating a pseudo effect that would be
better than the present flat effect?

MR. IVES: A great deal can be done by lighting, and by poor depth of focus.
And where the nature of the subject will permit, the relative motion of the
camera and object provides a beautiful relief. But this is a subject to which
I have not given much attention. I was talking about binocular or stereoscopic
relief.

MR. VICTOR: May I offer, as my personal belief, that we will eventually
find the solution to stereoscopic relief in colors? Every artist knows the value
of color perspective, and I think that some day we shall have a color projection
system that will give us very nearly the effect of stereoscopic pictures.

MR. IVES: There is a beautiful painting in the Gardiner Museum, in Boston,
which I would cordially recommend to every one interested in this line of specu-
lation. It is a picture by Sargent — a group of dancers with musicians in the
background. It is lighted with spotlights, in a rather long room, and as one
enters through a door at the back, he obtains a remarkable illusion of relief.

MR. KELLOGG: Concerning the true stereoscopic effect, when two pictures
are merged, there is only one plane in registration. Everything else is out of
registration, and in so far as the images of the two observing eyes fuse in the
consciousness, there must be a blur. Now, a number of efforts have been made
to produce three dimensional pictures, I believe, by printing two photographs
or otherwise combining them into a single picture which both eyes see, and in
which everything in one plane is shjarp and everything in any other plane is
double. I should like to know whether, as long as one is willing to center his
attention on that one plane, it provides any better sense of depth than can be
gotten from any ordinary sharp picture.

MR. IVES: My comment on that would simply be that it does not conform
to the second of my series of points— that appropriate images are not distributed
to the appropriate eyes.

MR. GREGORY: Has cylindrically lenticulated film been used for the taking
of parallax panoramagrams?

MR. IVES: Yes, it has. It is more efficient in utilizing the light.

MR. MAXFIELD: There is an effect I have noticed quite frequently on en-
tering a motion picture theater ; someone coming out swings the door open, and
I see a close-up on the screen standing out in beautiful third dimension. On
measuring approximately the distance from the place where I had noticed that
effect, to the screen, it looked as though the close-ups had been taken with ap-
proximately a four-inch lens. I was viewing the picture from my position, at
the same angle subtended by the lens when the scene was photographed, assum-
ing that a long focus lens had been used on the close-up. I should like to ask
Mr. Ives if he has any information regarding the relative importance of really
correct perspective versus binocular, where one views the picture from a rela-
tively long distance.

April, 1932] MOTION PICTURES IN RELIEF 441

MR. IVES: I do not know anything about viewing from a long distance.
It is well known in ordinary photography that for a picture to present correct
perspective it should be viewed from the distance which held between the lens
and the plate. And by looking at such a picture at that distance with one eye,
I have heard that one gets a sensation of relief. What happens is, that he does
not miss the other eye so much. Personally I do not get a sensation of relief
by looking at a picture with one eye at that distance. But people tell me they
do.

MR. MAXFIELD: I do.

THE EUROPEAN FILM MARKET— THEN AND NOW
C. J. NORTH AND N. D. GOLDEN**

Summary. — European producers will offer some 450 feature pictures during
1931. Leading producing countries are Germany, offering over 150 German dialog
pictures, England some 140 sound features, and France over 100 Frenchdia log films
for the year 1931. Europe is rapidly wiring its theaters, as indicated in the 10,400
wired theaters during 1931 as compared with the 4950 theaters wired during 1930,
over a 100 per cent increase in the short space of one year.

Elimination of legislation detrimental to American interests occurred in France
during 1931, and a tightening up of quota legislation was continued in Germany.
Agitation to increase the quota percentage to 50 per cent in the United Kingdom
gained very little headway, while other countries that have become picture conscious
are trying to encourage their own production through subsidies, contingents, or taxes,
as the case may be. Coupled with the problem of European production competition
and the artificial trade barriers set up by European governments, is that of supplying
European countries with dialog pictures in their native tongue.

While the above obstacles appear difficult to surmount at the present time, the
ingenuity of American technicians and producers will find a way to solve these
problems to the extent of producing a sufficient quantity of foreign dialog films at a
cost that will bring a fair return on their investment.

The financial destinies of most of our film companies — certainly
all the leading ones — come unpleasantly close to depending on the
state of our film revenues from abroad, of which Europe supplies
nearly 70 per cent. Back in the days of the silent film, approximately
30 or 40 per cent of our entire rentals came from overseas. After
foreign audiences stopped going to see pictures merely because they
had sound — in other words, in 1930 — our foreign revenues dropped to
about 25 per cent. As of today, many authorities consider that they
have fallen below 20 per cent. Obviously the slack must be taken
up somewhere, and it is therefore no coincidence that the economies
forced on the various motion picture companies this year have oc-
curred just at the time of the curtailment of our foreign revenues.

Perhaps the most effective way in which to apprehend the condi-

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Chief and Assistant Chief, Motion Picture Division, Bureau Foreign and
Domestic Commerce, Washington, D. C.
442

EUROPEAN FILM MARKET

443

tion in Europe, and the changes that have taken place, is to con-
sider briefly the situation that exists in a few of our major European
markets, after which we can possibly reconstruct a picture of the
scene as a whole.

The United Kingdom— our most important customer, not only in
Europe but in the world — began to find itself film-minded during
1930, and by the end of that year was quite strongly entrenched as a
competitive factor in its own market. Prior to that time, it suffered
an overwhelming dependence on American pictures to the extent of
about 80 per cent, even as late as 1929. However, the final establish-
ment of talking pictures gave Great Britain a medium for the exploita-
tion of its fine stage traditions, and although slow to realize how quickly
sound films would dominate the scene, once under way the momentum
acquired was fairly great. Thus, British production that accounted
for less than 90 films in 1929, increased to a total of 135 in 1930, and
will reach better than 140 by the close of this year. It should be
noted that of this 140, about 108 are the products of six companies, led
by British International and Gaumont-British with 35 each, all of
which companies have well equipped plants, extensive distribution
facilities, and theater outlets aggregating a capital investment accord-
ing to unofficial estimates of not far from two hundred millions of
dollars.

TABLE I

Production Schedules of Foreign Producers

Germany

France

England

1930

1931

1930

1931

1930

1931

Ufa

X-Y-Z*

British Int.

Suedfilm

Gaumont-

Pictures

D. L. S.

Franco

Gaumont-

Emelka

Film-

British

Terra

Aubert

Corp.

Others

136

P a t h e -

Gains-

Natan

borough

Jacques

British &

Haik

Dominion

B r a u n -

British

berger-

Lion As-

Richebe

ssociated

Others

X-Y-Z*

Miscellane-

ous

TOTAL

174

164

TOTAL

.107

TOTAL

135

134

* American Producing Co.

444 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E.

In spite of England's financial difficulties, the past year has been
relatively prosperous so far as motion picture receipts are concerned.
For, while theater attendance has declined considerably from the
novelty days of 1929, the outlets for sound film showings are much
greater. For instance, at the end of 1930, about 2600 houses were
wired. As of today's date, this number has increased to about 4100
out of less than 5000 theaters, a gain of nearly 65 per cent. The only
difficulty here from the American point of view is that, even though the
quantity of films imported from the United States is still decidedly in
the majority and may amount to as much as 70 per cent of the total for
the year, British films are gradually getting more advantageous play
dates, and in a striking number of cases are grossing up to 50 per cent
higher. In other words, the British public is at last getting something
that pleases them, out of their own studios, and even though there is no
language barrier as in Continental Europe, many of the present run
of the British-made product have local features, whether of voice,
setting, plot, theme, or the like, from which they acquire greater
audience value than those imported from the United States.

This is a new situation, and is one that supplies food for thought.
For now that the snowball has gathered momentum, there is no know-
ing how great it may become. We may see the time when England
will produce not only enough for the greater proportion of its own
needs, but will also supply the bulk of the pictures shown in its
dominions. This is probably a pessimistic outlook for us, but un-
doubtedly revenues from this market are due for a steady decline.

The recent recommendation of the Federation of British Industries
and of the Trades Union Council, to increase the quota to 50 per cent,
made little headway. The English government decreed, in 1927, that
a certain proportion of all the films distributed and shown in England
must be British made. This proportion last year was 10 per cent for
the distributor and 7.5 per cent for the exhibitor; and this year is 12.5
and 10 per cent, respectively. That the advantages of the quota
outweigh its initial disadvantages through the organization of mush-
room companies offering quota films is a question. In any event, it
now seems to be recognized that nature should be allowed to take its
own course, and that no further attempt should be made to legislate an
increase in film output.

When we come to Continental Europe, the language factor im-
mediately appears, and must constantly be borne in mind in any
consideration of France and Germany. In the former country great

April, 1932] EUROPEAN FILM MARKET 445

strides have been made in production. In 1929, about 52 films were
produced (mostly silent); in 1930 this was increased to 76 films in
sound alone, as well as 18 silent films and a number in foreign dialog.
This number will be further increased this year to 107 sound and
dialog pictures, exclusive of foreign versions. Of those pictures
produced, Pathe-Natan and Gaumont-Franco-Film Aubert are the
leaders; but in France production in general is considerably more

TABLE II

EUROPE
Increase in Number of Wired Theaters, 1930-1931

October, 1930 October, 1931

United Kingdom 2600 4100

Germany 940 2000

France 350 850

Sweden 90 550

Italy 120 450

Czechoslovakia 75 300

Spain 145 290

Austria 55 295

Denmark 45 200

Netherlands 95 180

Hungary 70 175

Switzerland 65 140

Rumania 50 135

Belgium 30 ... 125

Yugoslavia 35 110

Poland 60 105

Finland 15 70

Greece 20 70

Norway 30 60

Turkey 10 40

Bulgaria 10 35

Portugal 15 30

Other European countries 25 90

TOTAL 4950 10,400

decentralized than in this country, no less than 42 companies engaging
in the production of pictures in 1930, of which 27 produced only 1
film each.

French films have in general great popularity, some indeed such as
Rene Glair's Sous les Toite de Paris and Le Million grossing sums be-
yond what all but a very few American films have grossed in the most
prosperous era of silent pictures. Here again, French stage tradition
and the opportunity of hearing the French language spoken by French

446 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E.

actors has brought about a strongly nationalistic attitude on the part
of French audiences, with the result that American companies must
supply French language pictures in order to do business in France.

And yet the French have to contend with one of the most difficult of
film problems, namely, product shortage and an insufficiency of play
dates to enable them to secure sufficient revenue on the average run
films to expand their production to any marked degree. Obviously,
their own production is insufficient to meet their demands. And
outside of that, they have only the French versions of American made
films and an additional few supplied by Germany. Of the former,
only about 25 were made in Hollywood so far this year, with possibly
an equal number in Germany. The result was that at the end of last
June the French abolished their quota system in favor of limitations
against only those countries which themselves have restrictions.
Thus, free entry of American product into France is assured. And
there is no question that the French market can absorb as many films
from the United States as our companies are likely to put out, the
understanding being that these must be in the French language.

I referred a moment or two ago to France's insufficiency of play
dates. There are about 3000 theaters in France. At the end of 1930
some 350 theaters were wired, and even now the number is only 850.
With less than one-third of the theaters in France adapted to sound
reproduction it can readily be seen how far the French market is from
realizing its true potentialities. Obviously, the silent exhibitors
must wire or go to the wall.

When we turn to Germany we find that the film business is at an
exceedingly low ebb. The economic depression is having a retarding
influence on theater attendance ; and many exhibitors are on the verge
of bankruptcy which has its painful repercussion on both distribution
and production. In addition, high taxes threaten to take what small
profits the few exhibitors are making. This situation incidentally is a
hold-over, more or less, from last year. Germany has, furthermore,
continued her rigid policy of import restriction against foreign sound
pictures, only 105 of these being permitted entrance between July 1,
1931, and June 30, 1932. Incidentally, this official trade barrier has so
far been more or less meaningless to the American trade, which has
not produced German dialog films in quantity beyond what they
could get permits for — and it must be remembered that German
audiences insist on German dialog — but it definitely limits our future
chances of deriving much revenue from Germany in the near future.

April, 1932] EUROPEAN FlLM MARKET 447

In spite of the rather pessimistic picture just drawn of German film
conditions, it must not be understood that no money at all is being
made in Germany. As a matter of fact, the bankruptcies are most
numerous among the smaller and weaker elements in the industry.
Ufa and Emelka, the latter now being reorganized, are the two largest
producer-distributor-exhibitors in Germany. The former produces
approximately 25 features a year and spends upward of $80,000 on
each of them. In addition, it controls 170 theaters, including many
of the best locations in Germany. The Emelka chain consists of 50
first-run houses, and produces 10 to 15 films a year. These and two or
three other companies are said to be making money — Ufa just de-
clared a six per cent dividend — and taken together, they will account
for a large proportion of the 164 films to be produced in Germany dur-
ing 1931. This, with such foreign versions as may be secured from
French and American sources, will come fairly close to filling the needs
of the German market though there is the possibility of a product
shortage. It is to be noted that German films are designed not only
for the needs of the domestic market but to compete actively
throughout all Central Europe where German, even if not the primary
language, is generally understood. Special agreements have been
made with Austria, Czechoslovakia, and other countries, by which
German pictures gain easy entrance. Germany is also concentrating
on foreign versions, particularly French, and may soon have as many
as 40 films for that market. The intensity of this competition and the
headway it has made must be considered by American film interests as
at least a subsidiary factor in our diminishing film revenues from
Europe.

As to German exhibition outlets, at the end of 1930 about 940
German theaters were wired. This number has now increased to
about 2000. The equipment used is German made, mostly Tobis-
Klang film. There still remain more than 3000 theaters not wired
which exhibit only old silent films and which must be wired or pass
out of existence.

The rest of Europe can be covered in a few words. Little is to be
expected from Italy, where the ban on all foreign dialog films has
created such an acute product shortage that fewer than 50 features are
available for a market that requires over 250 a year. Nearly half of
these are being produced by Pittaluga, also one of the largest ex-
hibitors, but the product is not sufficient even for his own houses.
Obviously, American features in Italian dialog must be very cheap

448 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E.

to show a profit, and with English dialog banned, even when explained
by Italian sub-titles, there will be very little of the film product of this
country seen on Italian screens. I might add that the Italian situa-
tion is merely an intensification of what has been going on since the
early days of talking pictures.

Central Europe, meaning Austria, Hungary, Czechoslovakia,
Poland, and the Baltic States, as implied above, is being drawn
somewhat into the German sphere of picture influence. They have
become picture conscious, however, and are trying to encourage their
own production through subsidies, contingents, or taxes, as the case
may be. The Scandinavian states, and especially Sweden, are also
trying production in their own language. American films are, of
course, being shown in all these countries, but the language obstacle is
difficult to overcome. The outlet for sound pictures is gradually

TABLE III
FOREIGN THEATERS

Approximate Number of Theaters Approximate Number of Theaters

Wired

1930 1931 1930 1931

Europe 27,000 29,535 4,950 10,400

Far East 4,000 5,350 900 1,900

Latin America 4,000 4,700 450 1,575

Canada 1,100 1,100 450 700

Africa 750 770 40 116

Near East 50 85 10 25

TOTAL 36,900 41,540 6,800 14,716

being extended in all of them through increased wirings, with the
result that Europe is far more overwhelmingly committed to sound
pictures than even figures would seem to indicate.

As a summary, the charts show, on the one hand, European produc-
tion, and, on the other, the expansion in European play dates through
increases in wiring. They provide an illuminating picture, espe-
cially on the production side, with well over 400 films offered in com-
petition to our own.

All told, one must remember that the American trade is faced with
two important obstacles in Europe. The first is the language ques-
tion and its subsidiary competition, the latter being almost the direct
result of the former. The various European countries must have
films they can understand, and until we can devise a method economi-
cally profitable to give them such films, with the additional factor that
they must be of a quality to compete with locally produced films

April, 1932] EUROPEAN FlLM MARKET 449

molded on native stage traditions, this problem will not be solved.
In fact, it is doubtful whether the correct solution to this has been
given either by that school of thought which advocates production
of foreign versions abroad, or those that believe production can best be
done at home. Perhaps the newest types of dubbing, if not too costly,
will come closest of all to the solution, particularly when applied to
films in which action predominates. In addition, and this is the second
obstacle, we have to run the gauntlet of contingents, subsidies, and
other forms of government protection designed to foster the develop-
ment of the home product. These may tend to decrease when it is
comprehended that an industry cannot be legislated into existence;
but at present they, in combination with high taxes, are doing
much to make the European film field a series of pit-falls for the un-
wary.

In order to brighten the picture, it may be well to state that these
somewhat pessimistic observations on the decline of our European
revenues do not necessarily imply that these revenues will reach the
vanishing point. Far from it. This is a period of adjustment. If
competition is increasing, so also are film outlets through an increase
in the number of wired theaters. Europe is going through a profound
depression which is keeping many people out of the theaters, and is
impeding theater construction. When things pick up, and with
better theaters, the chance of increased revenue from an individual
picture will be greater. In other words, we can make more money on
fewer pictures. And finally it might not be presumptuous to believe
that the ingenuity of our producers will find a way to solve the
language difficulty to the extent that we shall be able to turn out
foreign language films in sufficient quantity and quality, at a cost that
will bring us a fair return on the investment. The easy-money
Europe of silent picture days is gone, but as a market offering better
returns than now, it holds possibilities.

DISCUSSION

MR. RICHARDSON: My reports from France indicate that the projection
of pictures in France, both as regards sound and the picture itself, is nothing less
than terrible as compared with our own. The same is largely true in Germany.
And there is no question, gentlemen, but what that very largely decreases the
revenue of theaters. I believe that the reason why the photoplay theaters in
North America are so well patronized is that the picture and sound are repro-
duced by expert men.

I believe that the producers might well call the attention of European ex-
hibitors to the fact that they cannot possibly obtain the requisite revenue if

450 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E.

they put on the screen a very poor picture, and radiate from the horns sound of
very unsatisfactory quality.

MR. McGuiRE: While Mr. Rubin, chairman of the Projection Practice
Committee, was in France about a year ago, he reorganized the entire projection
staff of Publix Theaters in that country. That program included raising the
compensation of the men and improving their standing. If these methods were
more generally adopted in foreign countries much better projection would be
secured. The importance of projection and of the projectionist is now fully
realized in the United States, and other countries would do well to follow our
example along these lines.

PAST-PRESIDENT CRABTREE : In connection with the matter of producing films
with foreign dialog, in Hollywood I saw a synchronization of the dialog of a
foreign actress with the lip movements of an American actress. When the
picture was projected on the screen, Italian actors and actresses equipped with
ear-phones, were arranged in front of the screen before a number of music stands.
By watching the screen and listening to the sound coming from the horn, re-
markable synchronization was effected. This method of synchronizing is beyond
the experimental stage now, and the films are now being supplied to the trade.

MR. GOLDEN: It is true that our technicians at the studio have been able
to produce a fine result by synchronizing the foreign language and the lip mo-
tions of our American actors. However, there is one obstacle that they have
not yet been able to overcome, and that is the question of the proper language
as used in the country for which the version is made.

In New York the other day I had the pleasure of talking with a man con-
nected with the foreign department of one of our large producing units, and his
complaint was that regardless of how short a time a foreigner has been in this
country, even as short as a six months' period, there is something that creeps
into his language that is offensive to the native foreigner in his own country.
The producer, to secure the true speaking language of a given country, must
bring the cast from their native country and use them for a certain number of
pictures, release them and send them back to the country from which they came,
and then bring over other native actors. The synchronization part of it is all
right, but idioms of expression, and a certain amount of slang, get into the
foreigner's speech that are not acceptable abroad.

MR. KELLOGG: How nearly universal is the standard film track location and
offset?

MR. GOLDEN: It is practically the same as used in this country. From
reports, and samples of film we receive from foreign countries, it is practically
the same. I am quite sure that Klangfilm-Tobis is about the same as Western
Electric or any one of our recording systems.

MR. MONOSSON: Does Europe include the U. S. S. R.?

MR. GOLDEN: No. In Table II Soviet Russia is excluded because this
country does not maintain diplomatic relations with Russia, and we are in no
position to receive authentic information from our own offices. Foreign audiences
insist on our American stars. It is going to be some years before the foreign
producer can establish his stars to the point where our American stars will be
rivaled, and since the foreigner likes our American stars, he must like our

April, 1932] EUROPEAN FILM MARKET 451

technic in the production of motion pictures. And as long as the foreign pro-
ducer, therefore, puts out pictures of the type that he is putting out today, he
is not going to get very far.

True, most of this production abroad, as a matter of fact, has been sponsored,
not by the movie goer, but by the business man of the country in question, and
even reaches so far as the governmental heads. You have the finest example
in the British quota system. The British quota system was not instituted by
the producers or exhibitors of the country. It was originally pushed forward
by the manufacturing interests of England. They felt our American pictures
were carrying a propaganda into England which resulted in the sale of our
American goods in England and its dominions.

The revenues received by the exhibitor abroad are much smaller than those
received by the exhibitor in this country. He does not get the patronage at the
box-office that we do. Therefore, he cannot pay the salaries that our American
projectionists get. He must accept inferior workmanship.

But, with many of the theaters having an average seating capacity of two
hundred, passing out of the picture and being supplanted by the de luxe houses,
I am sure that projection, theater construction, and entertainment values will be
improved and placed upon a higher plane.

VICTROLAC MOTION PICTURE RECORDS*
F. C. BARTON**

Summary. — A new type of disk record, known as the Victrolac record, is described.
The material of which it is made is a thermoplastic resin, which must be cooled before
being removed from the mold. The paper discusses briefly the characteristics of this
material, the time of playing of records made of it, operating features of the tone
arm and pick-up system, resonance characteristics of the tone arm, and the char-
acteristics of the chromium needle.

A new type of disk record has recently been made available to the
motion picture industry, which record presents a number of ad-
vantages over previous types in that it has better reproducing quali-
ties, better wearing properties, is non-inflammable, practically
unbreakable, water resistant, smaller for equivalent playing time, has
much lower surface noise, is lighter and flexible.

The development of the new record has come about through the
constant search being made by research engineers for new materials
which would advance the art of record making and would bring about
an improvement in an art that has remained almost stationary, except
for minor changes, for a period of twenty years. In the last few years,
chemical engineers have given a great deal of time and attention to the
development of synthetic resins, and the outcome of this work has
resulted in the production of a great many variations of two general
groups of these resins. The groups comprise those resins which
polymerize or cure and become infusible and insoluble after the
application of heat, and those which are thermoplastic but non-curing ;
in other words, which will flow and mold under heat but which must
be cooled before removal from the mold.

A great number from each group have been used experimentally in
the hope of finding a material which would be modified to give the
looked-for improvement in record quality and an almost equal
number of disappointments have been encountered. Approximately

* Presented at the Spring, 1931, Meeting at Swampscott, Mass
** RCA Victor Co., Camden, N. J.
452

VICTROLAC RECORDS 453

a year ago, however, a resin of the second group, that seemed to hold
promise of having the desired characteristics was found. The
chemical engineers responsible for the development of this resin were
called into conference with the engineers of the record manufacturers
and a cooperative program was laid out in which the technics of the
two groups were combined to further the development of the resin and
to combine it successfully with other materials, to the end that a satis-
factory record material might be evolved. A number of months of
concentrated effort on the part of these two groups of engineers
resulted in the production of the compound now known as Victrolac.
This compounded resin has very remarkable properties, and the
records made from it have many points of superiority over former
products.

Among the principal advantages is the greatly reduced inherent
background or surface noise as compared with former types of record
material. In the past it has been found necessary to use a large
groove and to record sounds of great amplitude so that the recorded
amplitude would be large compared with the amplitude of the surface
or scratch noise ; and that by this means the music would mask the
surface noise, or at least make it less noticeable. Advantage has been
taken of the improved surface conditions of the new material by
employing a lower amplitude of recording, smaller grooves, and by
placing the grooves closer together, thus increasing the playing time
per inch of recorded radius of the record in direct proportion to the
increased number of grooves per inch, which in this case is from 90
lines per inch on the old records to 120 or 130 lines per inch on the new
records. This represents an increase of from 2.7 minutes per inch of
recorded radius on the old records to 3.9 minutes per inch on the new,
and since a film 1000 feet long projected at 24 frames per second re-
quires about 11 minutes to run, the recorded radius of one of the new
records corresponding to 1000 feet of film will be 2.82 inches. Allow-
ing y4 inch °f radius or Y2 inch of diameter for margin and 2.82 inches
of radius or 5.64 inches of diameter for recording, we have left 5.86
inches as the center diameter for a 12-inch record which is satisfactory
as regards frequency response for the width of groove used. The
decrease in the amplitude of the recording for the case of the smaller
groove is about the difference between +9 db. for the old records and
+5 db. for the new records, or the new recording amplitude is about
60 per cent of the old. The decrease of surface noise is propor-
tionately much greater than the decrease of recording level. The

454

F. C. BARTON

[J. S. M. P. E.

surface of the new material is only about 43 per cent of the old, which
leaves a net gain of approximately 1.4 to 1 in apparent surface noise-to-
signal ratio in favor of the new material. In other words, if the
scratch and the recording noises were reduced by equal percentages
there would be no change in the noise-to-signal ratio, but in this case
the surface noise level has been reduced much more than the signal
level; therefore, there is a net gain in performance.

Fig. 1 illustrates the relative size of the standard groove on the 16-
inch record as compared with the new groove on the 12-inch record.
It will be noted that the curvature in the bottom of the groove is the
same in each case, and that the groove of the new record is merely a
little narrower and shallower.

.0065-Hg

///////////////

121 GROOVES PER JNC

k- .0070-+ 1

//////////////////y/'
'-SECTIONAL VIEW OF 9O GROOVES PER INCH,

////////////////////////////////////////////I///

FIG. 1. Relative size of standard groove on 16 -inch
record, as compared with the new groove on the 12-inch
record.

It is well known that the response characteristic, or the ability to
reproduce from the record certain frequencies, is directly associated
with both the linear speed and the width of the groove. In other
words, the higher the linear speed or the narrower the groove, the
greater is the possibility of reproducing from the record the higher
frequencies. The narrowing of the groove on the new record ac-
counts for the fact that smaller center diameters down to 5x/2 inches
may be used in the new records, while still maintaining a frequency
response characteristic equal or superior to that obtained from the 16-
inch records with the larger center diameter.

Another advantage of the material used for these records lies in its
strength and flexibility. On account of these features it has been
found possible to produce a 12-inch record for motion picture work

April, 1932] VlCTROLAC RECORDS 455

weighing approximately 4 ounces, as compared with 24 ounces for the
16-inch record. In addition to the reduction in weight, the record is
practically unbreakable.

These two features make possible a very considerable saving for the
producer or distributor in shipping the records. Extremely careful
and cumbersome packing of records is no longer necessary, and ship-
ments may be made by mail or express without other protection than
a couple of sheets of corrugated board on either side of the record so as
to prevent damaging the record surface by allowing it to come into

Steel Chromium Point

After playing one 16 -inch shellac record

Steel Chromium point

After playing two 12-inch Victrolac records

FIG. 2. Showing comparative wearing of needles used on
shellac and Victrolac records.

contact with other packages. The new record is approximately
0.040 inch thick.

Another but possibly less important advantage lies in the decrease
of abrasion of the needle. An ordinary full-tone steel needle will
show much less wear after playing one of the new records than after an
equivalent amount of playing one of the old. (Fig. 2.)

It would now be of interest to present a few points in which the
manufacturers of reproducing equipment, and operators of the equip-
ment, can assist in the full realization of these advantages. The
inherent strength of the resin itself is relied upon to give the record the

456

F. C. BARTON

[J. S. M. P. E.

required solidity. This permits using a soft filler which assists in
reducing the surface noise. None of the hard, highly abrasive fillers
commonly used in manufacturing records are used. But the strength
of the material and its ability to withstand abuse do not necessarily go
hand in hand, and it may therefore be stated that the new material is
susceptible to injury through improper use. The records have been
designed to be operated under the same average conditions as the old
records ; that is, a standard full-tone needle with a pick-up pressure of
approximately 5 ounces and a needle placement which will bring the

- C£Mr£# Of TONC ARM

FIG. 3. Illustrating proper adjust-
ment of needle on arc tangent to tone
arm radius 1 inch from center.

needle within iy4 inches of the center of the turntable when the tone
arm is swung to position directly in line with the center. This place-
ment will make the needle tangent to a circle approximately 1 1 inches
in diameter when using a tone arm IP/2 inches long. (Fig. 3.)

A more desirable set of conditions, with particular reference to the
new record, would first require a pressure on the needle of 3 ounces, a
pressure which can be maintained by additional counterweighing of
the reproducing tone arm. Such a simple correction can be made by
the operator by allowing the pick-up to rest upon the platform of a

April, 1932]

VICTROLAC RECORDS

457

small postal scale, placing the tone arm in a horizontal position and
adjusting or adding to the counterweight to get a reading of 3 ounces.
Second, a displacement of the needle from the center equal to 1
inch would be required, making it tangent to a circle approximately
93/4 inches in diameter, which would lie approximately at the middle
of the recorded area of a 12-inch record. (Fig. 3.) A change of this
nature may be difficult to make in existing equipment, but if the dis-
tance between the needle and the center of the turntable is not more
than I1/ 4 inches, no difficulty will be experienced. In designing new

0.2

1.8 2.0

FIG. 4.

0.4 0.6 0.8 1.0 1.2 1.4- 1.6
A/EEDLE POSIT/ ON BEHIND CENTER P/M
/N /NCHES

Relation between needle position and circle of tangency for a tone
arm Iiy2 inches long and a tone arm 93/4 inches long.

equipment, however, this point should be considered, and the place-
ment should be made so as to get the best results out of the 12-inch
record. (Fig. 4.)

The decreased level of the recording which, as I have said, is about 4
db. below that of the 16-inch record, will make it necessary for the
operator to increase the gain by a point or two in order to raise the
volume in the theater up to the level formerly obtained with the 16-
inch disk. No change will be necessary in the needle, provided a
normal full-tone needle that is not excessively sharp is used.

Relatively little consideration has been given in the past to the

458 F. C. BARTON [J. s. M. P. E.

shape of the needle point. Although the record material itself has
been sufficiently abrasive to wear down the needle point rather quickly
to fit the groove, with the new material this process takes place much
more slowly; and with the slightly softer record stock, cutting of the
record may result from either too fine a point or too high a pressure.
Assume that the combined weight of the pick-up and the tone arm
is 5 ounces, or roughly 1/3 of a pound, and that the area of the
point in contact with the record is 0.003 inch in each direction, or
approximately 9 square mils. Under such conditions, if the pick-up
weighs 1 pound, the pressure under the needle would be 110,000
pounds per square inch; but since it weighs only l/3 of a pound, the
pressure will be of the order of 37,000 pounds per square inch, a fairly
high stress even for metals. When we consider the nature of the
record compounds it is remarkable that such a stress can be withstood
even for a single playing. A reduction in weight from 5 to 3 ounces
will cause a corresponding reduction of stress from 37,000 to 22,000
pounds per square inch, a value still quite high for an ordinary thermo-
plastic molding compound to stand. The existing standard of 5
ounces was selected to insure tracking of the needle, or following of the
sound wave, on the very heavily recorded 16-inch picture records;
but since the amplitude of recording of the new records is considerably
reduced, there is no longer the need for so great a weight to insure
tracking, and 3 ounces have been found ample.

The new records, if used under the conditions recommended above,
will have a life much longer than any records that have been previously
produced for the motion picture industry.

Needle development has been carried on in parallel with record
development, and there is now available a new type of needle admir-
ably adapted to the new record, although its use is in no way restricted
to this record. It is a full- tone steel needle having a chromium tip.
When used under a 3-ounce load this needle will successfully play at
least twenty-five of these 12-inch records. A number of playings
greatly exceeding twenty-five have been successfully made in the
laboratory, but this number is recommended as representing good
practice. Assume a 12-reel feature motion picture show running four
times a day. Twenty-four records would be played on each projector
each day, requiring a change of needle only once a day per projector.

Before closing, a short statement referring particularly to the design
of tone arms and their effect on the performance of a record might be
appropriate. Judging from the characteristics shown by some of the

April, 1932]

VICTROLAC RECORDS

tone arms that have been tested, their designers apparently have
considered them as merely means for holding the pick-up in its proper
position on the record. True, this is one of its functions, but another
and equally important function is that of controlling the tendency of
the pick-up as a whole to rotate around the natural longitudinal
axes of the arm, the impulse causing this tendency being furnished by
the lateral motion of the needle during the recording. Some tone
arms, instead of exerting a corrective influence against this tendency,
by their construction actually tend to aggravate the tendency to
rotate. The increase of this tendency will, of course, occur at or near

70N£. ARM WITH "U" CROSS SECTION AND LOW
IMPEDANCE PICK-UP

600

500

400-

300

200

100

HEEDLE PRESSURE * /7o GR.

2 4 & Q 100 2 4 6 8 1000 2 4.

FREQUENCY
FIG. 5. Resonance characteristic of an undesirable tone arm.

the frequency at which the tone arm and pick-up would vibrate if they
were placed under torsional stress and suddenly released; in other
words, at the period of natural resonance. If this resonance fre-
quency occurs, as it frequently does, in the lower musical register,
then a severe load will be imposed on the record; and the needle will
tend to leave the groove each time the arm is shocked into vibration
by a passage in the record of a frequency corresponding to the natural
period of the tone arm system. A curve plotted from data obtained
from a particularly bad tone arm is here shown in Fig. 5. The con-
clusion reached from this is that, if it were not possible to design an
arm free of natural periods, the arm should be designed so that the

460 F. C. BARTON

period will occur at a frequency well below 100 cycles, or when the
recording has been so attenuated that the shocks produced will not be
large. In general, long straight [/-section channels should be avoided.
In reviewing the performance possibilities of this record it is the
firm opinion of the developers and manufacturers of the record that an
outstanding advance has been made and that with a small amount of
cooperation by designers and operators of the equipment, the full
advantages of the new development may be realized.

OPTICS OF PROJECTORS FOR 16 MM. FILM*
A. A. COOK**

Summary. — The limits of illumination available in a projector are fixed by three
factors: the size and brilliance of the light source, the effective aperture of the optical
system and the design of the condensing lenses. In modern 16 mm. machines of the
standard type, about 100 to 120 lumens are available through an f/2 optical system;
these values, which are not corrected for shutter and film losses, mean that 1.6 to 2.0
per cent of the total radiation is being used. The use of low voltage lamps has not
changed this ratio to any extent. The effect of varying each of the above factors is
discussed, and the increase in screen brightness that is likely to be obtained is estimated.

The fundamental requirements of apparatus designed to project
motion pictures from 16 mm. film are too well known to need any
detailed description. The apparatus must be compact and light, and
the number of adjustments necessary to operate it should be
reduced to a minimum. As an optical instrument it ought to produce
a clearly defined image on the screen. It is also obvious that the
location of the optical elements and their relation to the light source
must be exactly maintained if maximum illumination is to be con-
sistently secured.

Projection optical systems consist of a source of light, a collective
system for directing the light through the film gate, and an objective
lens for imaging the film upon the screen. Let us first consider the
light source. The advantages of tungsten lamps are evident from the
requirements already outlined. They are small in size, easily located
in a fixed position, and require a minimum of adjustment during
operation. Several filament designs of high efficiency have been
developed with parallel coils arranged to fill a rectangular space about
two-thirds the size of the film gate opening. The spaces between the
coils are of approximately the same width as the coil, this arrange-
ment permitting the use of a spherical mirror behind the lamp to
image each coil in the adjacent space. This adds to the efficiency by
heating the filament and gives the unit nearly the appearance of a solid

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Bausch & Lomb Optical Co., Rochester, N. Y.

461

462 A. A. COOK [J. S. M. P. E.

source. By doubling the useful angle of radiation in this way an
increase in illumination of 50 to 75 per cent is obtained. The exact
amount depends on the quality of the mirror and the position of the
filament supporting wires.

The filament housing is a tubular bulb iy4 inches in diameter.
This size has been adopted as standard for 16 mm. equipment, al-
though it may not prove sufficient for the continual demands for
higher wattage.1 Bulb diameter is an important dimension from the
optical point of view. The efficiency of the condenser and reflector
depend on the angular size of the cone of light that they can take in
from the source and transmit through the system. A shorter distance
between filament and condenser would be helpful, therefore, in that it
would permit a larger angle to be used by a condenser of given di-
ameter. Lamp manufacturers have been working on this problem, as
is shown by the fact that in some of their recent designs the filament
has been offset to a position well forward of the center of the bulb.
This change provides a mechanical advantage which can be especially
useful in the 16 mm. projector. Condenser design has often been
handicapped here by the limited space available. An increase in the
diameter of the mirror will be necessary, of course, for its distance
from the filament has been increased. There is more room behind the
lamp, however, and this slight change can be easily made.

The collective system may be either a condenser or a reflector.
Both methods have been applied to the illumination problem in
projection, but more space is required by a reflector, for the same
useful angle of radiation, than by a condenser with rear mirror.
Therefore, the condenser has been the preferred form in 16 mm.
machines.

The function of the condenser is a subject that has been thoroughly
analyzed and presented before this Society.2 Only an outline will be
given here of the working of this element of the optical system as it
applies in this special case. If a solid source of light of sufficient size
and uniform distribution could be placed at the film gate, no con-
denser would be needed. A tungsten filament is not solid, however,
nor can a lamp bulb be placed at that point. By using a condenser
a source image is substituted for the source itself; by locating the
image in front of the film plane the unevenness of the source can be
equalized. Fig. 1 is a sketch showing the condenser in its relation to
the other parts of the system. The condenser, L\, produces a magnified
image of the filament of such size as to fill the projection lens, L2. In

April, 1932]

OPTICS FOR 16 MM. FILM

463

doing this it takes in the large angle of radiation marked a, and forms
the image at a smaller angle a'. The radiation can now be trans-
mitted through the projection lens L2, as a result of this change in its
direction. In this way the condenser makes useful the radiation
from a small source through a large solid angle in space. Otherwise,
a very large source would be needed to produce the same effect.

SOURCE IMAGE

PROJECTION LENS

MIRROR

FILM GATE

FIG. 1. Projection optical system for 16 mm. film.

There is a very definite relation here between the size of the source,
the size of the projection lens, and the focal length of the condenser.
All the parts of the optical system are interdependent in this way, and
proper proportions must be maintained to obtain maximum efficiency

SOURCE IMAGE

PROJECTION LENS
L2

FILM GATE

MIRROR

FIG. la. 16 Mm. optical system, showing illumination at margin
of film.

of the whole unit. The conditions determining the diameter of the
condensing lenses are shown in Fig. la. Two solid lines drawn from
the extreme edge of the effective lens opening to the center of the film
aperture form an angle a'. The broken lines in the same way deter-
mine angle b' at the margin of the picture. These two solid angles, a'
and bf, must be equal in size and must be filled with light in order to

464

A. A. COOK

[J. S. M. P. E.

get the best possible illumination at the corners of the screen. This
means that the condenser should be large enough to furnish light
through all of the angle b'. This condition is usually not perfectly
fulfilled in practice. A 15 per cent decrease of illumination at the
margin is commonly accepted as satisfactory.

Condensers constructed according to these specifications are still
found to differ considerably in efficiency, due to differences in their
correction for spherical aberration. This is a well-known defect,
found in all simple lenses, that causes in this instance a loss from the
marginal portion of the light beam as it is converged to the image point
by the condenser. The loss is not so serious in 16 mm. projection
systems as in cases where the source image is located at the film gate.
It can be corrected to a large extent by proper condenser design. The

FIG. 2.

Relay condenser. Conjugate images are connected by
brackets.

use of aspheric surfaces is one effective method, this kind of correction
having been found to result in screen illumination 15 per cent greater
than that obtained with the ordinary plano-convex condenser lenses.

The relay condenser is a more complex device that may prove useful
with 16 mm. equipment. Its use in motion picture work is not new.3
But it produces uniform illumination from a tungsten source with so
little loss that it ought to be included in any discussion that deals with
projection from filament lamps. As shown in Fig. 2, it is a compound
lens system composed of three units. There is a condenser system,
LI, of large angular aperture to image the source upon a relay lens,
LZ, placed a short distance in back of the film gate. The third element,
L3, serves to form a second image of the source in the projection lens.
The relay lens must be large enough to receive all of the source image,
and of such focal length as to form a reduced image of the condenser

April, 1932]

OPTICS FOR 16 MM. FILM

465

at the film gate. Note that it is the evenly illuminated condenser
surface, not the source, that is imaged on the film. This accounts for
the uniform screen illumination produced by the system. It is 40
per cent more efficient than plano-convex condensers. The extra
length of the unit, amounting to six inches over all for a 16 mm. outfit,
is a decided disadvantage. But if it ever becomes necessary to build
a special type of projector for school or auditorium use, this method of
illumination should be of great service. It can be constructed to
work with a small source, and provide sufficient magnification to fill
larger projection lenses than any that are now used in 16 mm. work.

The projection objective is the third important part of the optical
system. Two-inch focus lenses of //2.0 are standard equipment at the
present time on practically all projectors except those designed for
use in cabinets. They must be well corrected for this large aperture,
but the field to be covered is so small that the requirements can be met

FRONT

BACK

FILM

FIG. 3. Projection objective of Petzval type.

without difficulty. There are many types of lenses that could be
used. In any such situation the cost element is bound to be a
decisive factor, and it has operated in this case to select the least
expensive lens that can be made to do the work. Before discussing
the details of this particular lens construction, it would be well to
consider the original from which it was derived. This lens form,
shown in Fig. 3, is Petzval's portrait objective. It has undergone
modification many times, but is still the formula most often used for
projection work. It can be very precisely corrected for the small
field required, and has a light transmission, in short focal lengths, of
73 per cent.

Fig. 4 shows the modified form that is now used in so many 16 mm.
projectors. Note that the two rear elements have been cemented,
and that the spacing between front and back has been increased to
nearly twice the length of the original construction. The first
change, by eliminating two air-glass surfaces, increases the light

466

A. A. COOK

[J. S. M. P. E.

transmission to 81 per cent; the increase in length has the effect of
shortening the back-focus of the objective. This means that the rear
element can be made smaller in diameter without sacrificing in light
transmission, and that it has more space in which to converge the
beam of light from the film gate. The rear element thus acts as a
collective lens for the system, which results in the practical advantage
that objectives of this construction, of any focal length, can be used
interchangeably on a projector without alteration or adjustment of
the condensing system. The only disadvantage of this short back-
focus objective is that it has a slightly curved field. This defect is
noticeable only in critical tests, however, and would be difficult to
detect in practical use on a projector, with moving film as a test
object.

The final screen illumination produced by a 16 mm. projection
system depends on the effectiveness of the four elements that have
been described: the light source, the rear mirror, the condenser

FRONT

FILM
FIG. 4. Projection objective with short back focus.

system, and the projection objective. Increases can be obtained by
using a brighter source, by improving the condenser correction, and
by increasing the aperture ratio of the entire optical unit. Recent at-
tempts at improvement in the 16 mm. field have been mainly directed
toward the light source, and this choice is a logical one for the equip-
ment manufacturer because it involves the least amount of redesign
on his part. To meet this demand lamps of greater brightness have
been developed, the increase being due to the use of larger wire size in
the filaments operated at a lower voltage than previously used.1 The
possibilities here are beyond the field of optics, and must be left to the
electrical engineer.

There are two points about lamp filaments, however, which are of
optical interest. One is the fact that filament supporting wires cause
illumination losses unless they are placed outside the angular field of
both the condenser and the rear mirror. The second concerns the
filament itself. The aperture of a projection system must be filled

April, 1932] OPTICS FOR 16 MM. FILM 467

with light if it is to work at its best efficiency. With a filament lamp,
the source acts as a discontinuous surface, and the openings in its
area cause a real loss of light. This effect is shown in Fig. 5, which
is a photograph of a 4-coil tungsten filament and its mirror image, as
they appear at the aperture of a projection lens. Any change that
would help fill up these spaces and thus make the source more solid
would mean an increase in illumination.

Improved condenser design offers a small field for improvement
which is applicable, perhaps, to many of the
commercial machines. Even with a perfect
condenser, however, one can do no more
than to fill the projections lens with an image
of the light source. The brightness of the
source and the effective aperture of the
system then determine the illumination.
Increasing the aperture offers interesting
possibilities that are yet to be considered. FIG. 5. The filament
An //1. 5 optical system should give 75 per image as it appears in
cent more light on the screen than the //2.0 the projection lens,
lenses now used; experience indicates that

these theoretical increases are seldom attained, however, and that a
figure of 50 per cent is much nearer the probable increase. The
cost element enters into this situation to such an extent that an
increase in aperture is not likely to be attempted in commercial
practice until all possibilities of the light source have been realized.

REFERENCES

1 ROPER, V. J., AND WOOD, H. I.: "Trend of Lamp Development and Opera-
tion in Motion Picture Projectors Employing 16 Mm. Film," /. Soc. Mot. Pict.
Eng., 15 (Dec., 1930), No. 6, p. 824.

2 KELLNER, HERMANN: "The Function of the Condenser in the Projection
Apparatus," Trans. Soc. Mot. Pict. Eng. (Nov., 1918), No. 7, p. 44.

3 KELLNER, HERMANN: "Can the Efficiency of the Present Condensing Sys-
tems Be Increased?" Trans. Soc. Mot. Pict. Eng. (Oct., 1923), No. 17, p. 136.

DISCUSSION

MR. PALMER: It has seemed to me, from casual observation, that the propor-
tions of the filament should be one to three-quarters— three-quarters as high
as it is wide, in order to conform to the dimensions of the picture aperture. Am I
correct in assuming that?

MR. COOK: In the case of 16 mm. projectors, we can not get uniform illumina-
tion when the filament is imaged on the aperture itself. For that reason the

468 A. A. COOK [J. S. M. P. E.

image of the filament is moved forward enough to produce the desired effect of
a uniformly illuminated screen. It actually amounts to imaging the source be-
tween the projection lens and the aperture. The projection lens is round, and it
seems to me that a nearly square source would be as valuable as one that is oblong.
The effect of the aperture in stopping down the light is, of course, noticeable as
soon as we get the image in front of the aperture. But in order to follow out that
line of reasoning we should use square condensers and a square projection lens. It
seems to me that a round source would be more nearly the ideal, from the present
set-up we are using in sixteen millimeter work. There is no doubt that if we
image a square source to fill a round projection lens, we waste the light coming
from the corners of the filament. But the illumination obtained depends on the
brightness of the source and the effective aperture of the system.

MR. HICKMAN: It seems to me that it makes no difference how much light
is spilled over the edge, provided a little more can be obtained in the center.
No one is really concerned with what is lost around the side.

MR. KURLANDER: The shape and size of the filament are also governed by the
desire of the projector manufacturer. Of course, the projector manufacturer
is susceptible to some influence by the lamp manufacturer, but sometimes he is
not, and he has his own ideas. I believe that the present trend is toward the
square shape, the size being dependent upon the inscribed circle determined by
the back element of the projection lens. Also, with some special forms of optical
systems, special shapes and sources are required, and those special shapes im-
mediately give rise to new lamps. Sometimes the new lamps are placed on the
schedules and are available to other manufacturers who do not know the history
of their development, and choose from them at random to meet their con-
ditions.

So there are a number of reasons for the different shapes of light sources, and
while theoretically, a solid source should be in the proportion of three to four,
a square source is easier to construct mechanically and does the same work.

MR. GAGE: The last picture that Mr. Cook showed was a photograph of the
projection objective filled with the filament. That is the way it looks when you
stand at the screen and look at the projection objective through a dark glass,
while the picture is being projected. If you find that the entire surface of the
projection objective appears filled with light, when observing the projection
lens from all points of the screen, the optical system is delivering all the light it is
capable of delivering. If, on the other hand, you find it is not filled with light,
perhaps you can tell by simple inspection where the defect lies. Perhaps the
filament is askew, perhaps the image of the filament in the mirror does not fill
the space between the filament legs with light, or perhaps the filament is not big
enough to fill the aperture. If you find a small image of the filament filling only a
part of the area, a larger filament is required, or perhaps a shorter focus condenser,
to magnify the filament image to a greater extent.

If, now, we go through the back-focusing process, setting up the whole pro-
jection system with the aperture, the objective and so on, and put a light in front
of the objective, with a card at the focus of the condenser, it can readily be seen
that there is no use in having a filament any larger than the spot of light received
on the card.

One thing Mr. Cook did not explain: if the condenser is brought close to the

April, 1932] OPTICS FOR 16 MM. FILM 469

filament, while, at the same time, the surface of the condenser is bent, as can be
done theoretically, a larger amount of the light will impinge on the first surface of
the condenser brought into the optical system by simply bringing the same diam-
eter condenser close to the filament. The acceptance angle becomes greater, and
the filament image becomes larger, until the surface of the condenser comes into
contact with the glass bulb surrounding the filament.

With the present sized filaments, the filament image is sprawled over a larger
area of the objective than can be used.

The greatest possibility of improving the system is to increase the intrinsic
brilliancy of the filament. By using a more efficient condenser it is possible to
use a smaller filament area. Nothing is gained unless the intrinsic brilliancy
of the light source is increased. At the same time its area can be reduced by a
more efficient condenser.

MR. RAYTON: There is one point that Mr. Cook and Mr. Gage have not
touched on, that might be worth mentioning since so much attention has been
paid to the appearance of the front of the objective. If we are to judge whether
the relative aperture of the optical system is completely filled with light, we will
have to do something more than look at the front of the objective under normal
condistions : namely, we will have to insert a pinhole aperture at the center of the
film gate. We may find, by so doing — and we probably will find — that the front
of the objective is filled with light. We will most certainly find, if we move
the pinhole to the corner of the film gate, that the objective is no longer filled
with the image of the filament. We may also find cases in which, with the full
film gate exposed, the front of the objective appears to be filled with the image;
but that when we introduce the pinhole aperture at the center of the film gate
the lens aperture is no longer so filled. If it is not, then under those circumstances
the condenser design is not the most efficient that could be used.

This point ought not to be overlooked, and I want to emphasize it, that we
do not get full information about how an optical system is working without an
exploration carried out with a pinhole aperture at the film gate.

MR. KURLANDER: I should like to ask if it is not true that the objective lens
is seldom filled by each individual point of the aperture.

MR. RAYTON: It is generally true.

MR. KURLANDER: Then I wonder why so much light is spilled over the
aperture plate to get uniform screen illumination when it would be cheaper to
use a cheap lens and a diffusing element in front of the condenser lens.

MR. RAYTON: Usually because the condenser is not large enough. The
relative aperture of the projection lens required in order to get center brightness
is one thing Mr. Cook mentioned the fact that a decrease of brightness of
fifteen per cent or more. at the margin will pass unnoticed. To get uniform
quality of illumination all over the screen, we have to use condenser lenses possibly
somewhat larger than are ordinarily used.

MR. KURLANDER: Do you have to go to such extremes to get evenness?

MR. RAYTON: You do with the set-up for motion picture illumination.

MR. KURLANDER: I have obtained uniform screen illumination of equal
intensity by focusing the filament at the aperture plane, and then smoothing
out the light by placing a diffusing element in front of the condenser lens.

MR. RAYTON: It is quite unreasonable that you should.

470 A. A. COOK

MR. KURLANDER: It seems unreasonable, but I hope some time to be able to
show it.

MR. COOK: Will back-testing according to Mr. Gage's method in this way
show that the filament should be round rather than square?

MR. GAGE: When I tried back- testing the condenser system with the aperture
in place, I obtained, at the position of the filament, not an exact image of the
aperture, but approximately that. It is wider than it is high and is rectangular,
with rounded corners.

MR. FARNHAM: There is an eternal demand for more and more light from
projection optical systems. There are three ways of obtaining it: greater
source brightness, greater efficiency in controlling the light through the optical
system, and utilizing greater source area without reduction of efficiency. I do
not see how we can expect an increase of source brightness of a very high order,
that is, two- or three-fold, as we are now operating the tungsten filaments at
3400 °K., and the melting point of tungsten is 3650°, the highest of any metal we
know. The wire has been so disposed in making a concentrated source as to
secure an optimum effect of black body radiation and a high order of average
source brightness to maximum brightness. Further increases will be a few per
cent at a time. It would appear that the greatest development lies in the direc-
tion of improved optical efficiency and of utilizing greater source areas. This is
particularly emphasized when it is realized that the over-all efficiency of the best
projection optical systems is approximately five per cent.

SILICA GEL AIR CONDITIONING FOR FILM PROCESSING*

E. C. HOLDEN**

Summary. — The need of properly conditioning air in motion picture film process-
ing plants is pointed out and the values of temperature and relative humidity in current
use in such plants are indicated. In particular, the process of humidifying, in which
pure silica which has passed through the sol and gel stages of manufacture, commonly
called silica gel, is used, is described. The principles involved in the process and the
efficiency of the method in controlling the condition of air are explained, and curves
are given showing the efficiency of adsorption.

The most obvious application of air conditioning in the motion
picture industry, aside from the comfort conditioning of theaters, is
in film processing. This seems a relatively simple operation, more
or less satisfactorily performed at a large number of places ; neverthe-
less, a reactionary attitude of secrecy still prevails, even as to this
detail, resulting in wide variations in local practice.

There is probably an ideal set of conditions for film processing, the
determination of which would be hastened, to the benefit of all, if a
more enlightened policy of exchange of experience were practiced,
such as is fostered by this and other technical societies. Cases where
secrecy in the arts is justified and desirable are the exception rather
than the rule.

AIR CONDITIONING REQUIREMENTS

The usual requirements to be met by air conditioning for health and
comfort purposes are that :

(1) The air must have the approximate chemical composition of fresh air.

(2) It must be free from odors.

(3) It must be clean.

(4) It must have an effective temperature within the comfort zone.

In the case of film processing, the important factors are as given
under (3) and (4). Processed air should be super-cleaned, as nearly
free from suspended solids as is possible, and its effective temperature

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Consulting Engineer, Baltimore, Md.

471

472 E. C. HOLDEN [J. S. M. P. E.

must be within the "comfort zone" for films instead of for people. It
must be remembered that the effective temperature with relation to a
moist surface is determined by the dry-bulb temperature, the relative
humidity, and the velocity of the contacting air; and that the requisite
"comfort zone for films" is such that they may be dried rapidly with-
out suffering distortion or becoming brittle.

All are agreed as to the desirability of having the air perfectly clean,
a requirement which has become all the more important with sound
recording. The standard oil-surfaced baffle and the felt filter types of
air cleaners are not adequate for this purpose, as the former puts oil
into the air and the latter lint, both of which may adhere to the film
and produce highly objectionable effects.

There is considerable difference of opinion as to the optimum film
speed, the volume, temperature, and the relative humidity of the air
to be circulated through the drying cabinets. In practice, tempera-
tures from 50° to 110°F., relative humidities varying from 20 to 80
per cent, and film speeds ranging from 15 to 130 feet per minute are
variously used. Even allowing for the difference between positive
and negative film requirements, such extremes cannot all be right.

The ideal conditions can be determined only by making systematic
tests. For this purpose adsorption conditioning units can supply
clean air at any desired temperature and humidity.

SILICA GEL

Silica gel is chemically pure silica, which has passed through
the sol and gel stages in manufacture, and which is therefore amor-
phous and highly porous in structure. The granular silica gel used in
air conditioning units is equivalent in texture to 6 to 14 mesh, and has
the appearance of colorless, semi-transparent sand, although its
specific gravity is less than that of crystalline quartz because of its
porosity. The pores represent the water of hydration which was re-
moved when the material was converted from a gel to a solid. They
are smaller than the wavelength of light, and are invisible under the
ultra-microscope .

This structure gives silica gel remarkable properties. The intense
force of the resultant capillarity enables the granules to adsorb vapors
within the gel granules, thus making it possible to separate vapors and
imperfect gases from air and other perfect gases. The granules will
take up from 30 to 50 per cent of their own weight of water vapor
depending upon the conditions, without swelling or becoming exter-

April, 1932]

SILICA GEL AIR CONDITIONING

473

nally moist. One must think in molecular dimensions to realize that
one cubic inch of the material has an internal surface of over one acre,
and that when ground to the fineness of flour, only two per cent of its
internal structure is destroyed, absorption tests proving that it retains
98 per cent of its original adsorptive capacity.

When the gel has taken up its useful load of vapor, it is readily re-
activated by heat, which drives off the adsorbed vapors; after
reactivation it is ready to be used again. The action in both cases is
the purely mechanical action of capillarity working against vapor

50 €0

IN GEL

O fO £0 30 40

70 CONCENTRATION OF
FIG. 1. Curves showing performance of silica gel in dehydrating air of viiri-
ous initial saturations at 25 °C. and at atmospheric pressure. A, air at 100
per cent saturation; B, 60 per cent; C, 40 per cent; D, 20 per cent.

tension. No chemical reaction is involved, so that there is no deterio-
ration of the gel; and the cycle of operations can be indefinitely
repeated.

The capacity and efficiency of silica gel as an adsorbent is dependent
on a number of factors : the composition, the temperature, and the
pressure of the gas-vapor mixture, the partial pressure of the vapor to
be treated, the rate of flow per unit weight of gel, and whether the
treatment is single or multi-stage, approaching adiabetic or isothermal
operation.

It is impossible in the limits of this paper to give exhaustive data for

474 E. C. HOLDEN [j. S. M. P. E.

all conditions. Fig. 1 shows its performance when dehydrating air of
various initial saturations at 25°C. and at atmospheric pressure. To
determine the useful gel saturation in operation, the residual satura-
tion of the gel of from 4 to 7 per cent should be deducted from the total
saturations shown.

In practice the efficiency of adsorption may be made to exceed 99
per cent, depending on the type and size of the unit used, by keeping
the operating cycle within the "break-point" limit. Some installa-
tions are guaranteed to deliver air at — 30°F. dewpoint. By treating
a regulated fraction of a total flow of gas-vapor mixture, any desired
saturation can be produced; and the high efficiency of adsorption
makes it possible to treat a minimum of the total circulation.

It is because of the power of silica gel to maintain a vacuum greater
than 29 inches of mercury in a vapor system that it finds its applica-
tion in refrigeration, the adsorber taking the place of the compressor.

Silica gel shows a similar selectivity for liquids, due to the character
of the internal gel surfaces and the differences in surface tension of
various liquids, thus making possible the separation and purification
of hydrocarbons and other liquids; however, this class of applications
is not of direct interest in the present paper.

CONDITIONING AIR FOR FILM DRYING

There are two practical stages in drying film, first: the removal of
the excess surface water by the compressed air "squeegee," and,
second, the removal of the water contained in the swollen films down
to approximately 15 per cent residual hydration, required to keep film
flexible and durable. These requirements are quite distinct and
should be considered separately.

In the preliminary stage, blowing off the moisture on the wet film by
compressed air at the squeegee, the air should be clean, oil-free, and
anhydrous, but the treatment actually used to condition the com-
pressed air, so far as the writer knows, is to pass it through the usual
compressed air receiver followed by some form of strainer; or at best,
a simple type of air filter, as described by Crabtree and Ives,1 for re-
moving the condensate of compression and entrained compressor
cylinder oil. No practical mechanical filter is 100 per cent efficient,
and as it cannot remove vapor, a decrease of temperature between
the separators and the squeegee causes further condensation of water
and oil vapor; and finally, as the air expands at the nozzle and thus
becomes chilled, more vapor will condense.

April, 1932]

SILICA GEL AIR CONDITIONING

CONDITIONING COMPRESSED AIR

475

When air is compressed, some of the compressor lubricant^ is
mechanically entrained in the air flow as a fine mist, and some of
it, even though the highest test oil be used, is partially cracked and
vaporized by the heat of compression. If efficient separating re-
ceivers and mechanical filters be used after the compression, a large
part, but not all, of the liquid oil-mist and water-condensate of
compression settles out, although none of the true vapor of the oil or
water is removed, these vapors passing on''and condensing later in the
line, especially at the discharge, due to cooling on re-expansion.

STEAM & WATER
CONNECTIONS

FIG. 2. Compressed gas dehydration unit.

This can be entirely prevented and the air can be dried to a dew-
point below any possible expansion temperature, and all oil vapor as
well as oil-mist will be removed by inserting a silica gel pressure type
adsorbing unit anywhere in the compressed air line following the
receiver. The air passes through a bed of silica gel which adsorbs
both the oil and water vapors and returns practically anhydrous,
clean air to the line. Such air delivered at an effective pressure
through the nozzle at the "squeegee" should do more than merely
blow off the excess water; it should deliver uniformly clean film and
appreciably reduce the duty required of the drying cabinets.

476

E. C. HOLDEN

[J. S. M. P. E.

Fig. 2 shows one type of small compressed air or gas silica gel
dehydrator.

CONDITIONING AIR FOR FILM DRYING CABINETS

Inasmuch as films are made of permeable organic material, they
will distort and lose their durability just as timber warps when
improperly seasoned, and drying requirements cannot be figured a

ATMOSPHERIC VENT
Z70 CFTM.

WET

— »

FILM

DRYING

CABINET

730C.RM.

1000 CF.M.
65-50 y.-

tlOC.FM.

SILICA
DEHYDRATOR

FIG. 3. Silica gel film drying; schematic diagram — 1000 cu. ft.
per minute circulated; 10 pounds of water per hour evaporated;
volumes not corrected for temperature.

priori as can evaporation from metal surfaces, but must be determined
by experience.

It is not, therefore, in the province of this paper to decide, or even to
offer, an opinion as to what are the ideal conditions for treating either
positive or negative films, or how much an anhydrous squeegee that
has not heretofore been available to the industry, may hasten and
simplify the subsequent drying. This can so easily be done, however,
that it would seem worth proving.

April, 1932]

SILICA GEL AIR CONDITIONING

477

The air conditioning system, now in common use in processing film,
of spray cooling or refrigerating to remove some of the water vapor,
or in winter, of spray humidifying followed by reheating, treats the
whole air stream, the used wet air being blown to waste.

The silica gel adsorption system, owing to its ability to deliver
practically anhydrous air, treats only a fraction of the air circulated,
this fraction having an excess moisture capacity corresponding to the
quantity of water being removed from the film. The whole air

FIG. 4. Unit for treating air or other gases continuously at low
pressure.

stream with its dried fraction can then be returned to the cabinets in a
closed circuit. The absorbing operation releases the heat of adsorp-
tion, which varies up to one-third more than the latent heat of the
water removed, so that any additional heat requirement is reduced or
eliminated, and the closed circuit and special filters and the gel bed
assure perfectly clean air and a complete control of temperature and
relative humidity.

478 E. C. HOLDEN

It does not seem logical to have to add water to a drying unit
With the adsorption system, if it be required to increase the humidity,
the hydrometric control automatically slows or stops the adsorber
operation and throttles the waste blow-off, so that the moisture taken
from the film itself quickly builds up the humidity to the desired
point, when the control again automatically regulates the adsorber to
maintain it, and the necessary output of wet air is discharged through
the relief valve.

As an example of how this works quantitatively, the flow-sheet,
Fig. 3, is given, based for convenience on the circulation of 1000 cubic
feet per minute, assuming that 10 pounds of water per hour are to be
removed. It is to be noted that the dehydrator would operate only
when the atmosphere contains more than 90 grains of water vapor
per pound of dry air, if that is the desired entering humidity.

It is apparent from the practice followed in many film laboratories
that the desired absolute humidity of the air entering the cabinets is
higher than the average absolute humidity of the atmosphere, and
that, therefore, the normal pretreatment required for fresh air enter-
ing the drying cabinets is humidification rather than dehumidification.
Whenever the atmospheric humidity exceeds the allowable humidity
of the air entering the cabinet, a drying unit will be useful for main-
taining the drying capacity without increasing the temperature of
operation. This means, however, that predrying is necessary only in
humid summer weather when the drying unit would be a convenient
auxiliary for maintaining production regardless of the weather.

The type of unit required is shown in Fig. 4, and consists of an air
filter with two single-stage bed adsorbers operating alternately,
adsorbing and activating, thus being capable of continuous 24-hour
production.

REFERENCE

1 CRABTREE, J. I., AND IVES, C.E.: "A Pneumatic Film Squeegee." Trans. Soc.
Mot. Pict. Eng., XI (Aug., 1927), No. 30, p. 270.

MEASUREMENTS WITH A REVERBERATION METER*
V. L. CHRISLER AND W. F. SNYDER**

Summary. — A description is given of apparatus with which the rate of decay of
sound energy in a room may be measured. A loud speaker is used as a source of
sound. When the sound reaches a steady state the loud speaker circuit is opened
and at the same time a timer is started. When the sound energy has decayed to some
definite value the timer is automatically stopped. If made in a portable form this
equipment may be used to study the acoustical properties of auditoriums. Attention
is called to the errors which may occur in these measurements.

With the advent of the talking picture, the determination of the
sound absorption values of various materials has become of consider-
able importance. The original method of measuring the coefficients
of these materials is due to W. C. Sabine, and requires the use of a
reverberation room and rather large samples of material. The
inconvenience of this led to attempts to find a method which would
permit the use of smaller samples.

One of these attempts, known as the tube method, is shown in Fig. 1.
A mathematical formula can be derived showing that the sound
absorption coefficient of the sample placed at the end of the tube can
be computed if the relative values of the amplitude at the maximum
and minimum points of the standing wave system in the tube are
measured. Unfortunately, the results obtained in this manner are
not in agreement with those obtained in actual installations. For
this reason the method has been abandoned. At the present time it
is necessary to adhere to the original reverberation method to obtain
results which can be depended upon in actual installations.

Figs. 2 and 3 show a plan and cross-section of the reverberation
room at the Bureau of Standards. To obtain satisfactory measure-
ments it is absolutely essential that all external noise should be
eliminated. The outer walls and roof have therefore been con-

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. Publication
approved by the Director of the Bureau of Standards.

** Bureau of Standards. U. S. Department of Commerce, Washington, D. C.

479

480

V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E.

structed so that they are unconnected with the inner walls and ceiling.
Due to this construction outside noises are seldom heard.

Fig. 4 shows an interior view of the reverberation room with a
sample of material laid on the floor, and Fig. 5 shows the position of
the observer while measuring the absorption of an audience.

To make measurements in this manner requires a trained observer.
The method is very tedious as approximately one thousand observa-
tions are required in order to obtain satisfactory values of the absorp-
tion coefficients of a sample at six frequencies. To eliminate the
personal error of the observer and to make measurements more quickly

Loud Speaker

To Amplifier

g Surface \

FIG. 1. Diagrammatic scheme of the tube method of measuring
sound absorption coefficients.

and more accurately, considerable work has been done at the Bureau
of Standards, as well as at other laboratories, to develop a method in
which all measurements are made with some instrument.

The first attempt was by use of an oscillograph.1 As the sound
waves decay in a very irregular manner in most cases, it is desirable to
take the average of a number of measurements in computing the
results. Figs. 6 and 7 show the irregular way in which the sound may
decay after the source has been cut off. Fig. 6 is for a frequency of
128 cycles and Fig. 7 for a frequency of 512 cycles. If enough records
are taken at each frequency and the measurements averaged, satis-
factory results can be obtained, but this requires too much work.

April, 1932]

REVERBERATION METER

481

O I 2 S 4 S Ft.

FIG. 2. Plan view of the reverberation room at the Bureau of Standards.

SECTION A-A

FIG. 3. Cross-section of the reverberation room at the Bureau of

Standards.

482 V. L. CHRISLER AND W. F. SNYDER [J. S. M. P. E.

The most satisfactory arrangement2 that has been tried is repre-
sented schematically in Fig. 8. The source of sound is a loud speaker
supplied with an alternating current of the desired frequency from a
suitable oscillator and amplifier. The sound is picked up by a
condenser microphone, and is then amplified. The purpose of the
attenuator will appear from the following text. It is desired to call
attention to the section of the circuit following the amplifier marked

FIG. 4. Interior view of the reverberation room at the Bureau of Standards,
showing a sample of acoustical material on the floor.

"tube relay," which consists of a rectifier tube followed by a stage of
direct current amplification. The circuit is shown in Fig. 9.

These tubes are connected in such a manner that, after the alternat-
ing potential applied to the first tube decreases to a definite value, a
very small additional decrease causes a relatively large increase of the
plate current of the last tube. This has been accomplished by util-
izing a "freak" characteristic of the first tube. Fig. 10 shows the

April, 1932]

REVERBERATION METER

483

static characteristic of this tube and gives the variation of the screen-
grid current and the plate current with the grid potential when a

FIG. 5. View showing the position of the observer while measuring the acous-
tical absorption of an audience.

definite screen-grid potential is used. To produce such a characteristic
only a limited range of screen-grid potentials can be used. If the tube

FIG. 6. Oscillogram showing the decay of sound after the source has
been cut off; 128 cycles.

is biased so as to obtain rectification at the upper end of the curve,
advantage can be taken of the abrupt change in plate current with a
very small change in grid voltage.

484

V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E.

Fig. 11 shows the modified plate current when an alternating
voltage is applied to the grid of this tube, and the corresponding
change in plate current in the last tube.

The sudden increase of the plate current of the last tube operates a
relay which stops the timer. As the timer is started automatically
when the loud speaker current is broken, this device gives the time
required for the sound to decay to some level determined by the
amount of amplification used.

FIG. 7. Oscillogram showing the decay of sound after the source has
been cut off; 512 cycles.

By using an attenuator in the amplifier circuit the time required
for the sound to decay to different levels can be determined. In this
way a decay curve can be obtained similar to that obtained in calibrat-
ing a room by the ear method, which uses different intensities, the
ratios of which are known.

There is one marked difference between these two methods. In

CONPCNSER
MICROPHONE

3 STAGE

RESISTANCE -COUPLED

AMPLIFIER

95 db gain

FIG. 8.

Schematic arrangement used at the Bureau of Standards for
making acoustical measurements.

the ear method we start at different intensity levels and end always at
the same lower level, which is the threshold for the ear of the observer.
In the instrumental method we always start at the same intensity
level and end at arbitrary thresholds whose ratios are known. Fig.
12 shows a curve thus obtained. It will be observed that the points
fit a straight line very closely. The points on the curve are not the
results of single measurements but are each the average of ten mea-

April, 1932]

REVERBERATION METER

485

surements. With a timer which adds, several measurements can be
taken rapidly and the average obtained, thus eliminating the un-
certainties of a single measurement.

UY224 .I.I.I.I.M.I.I.I UY224

FIG. 9. Circuit diagram of the vacuum tube relay.

To timer

The slope of this line gives the rate of decay of the sound energy.
From this slope the reverberation time may be computed, as reverbera-
tion time has been defined as the time required for sound to decay

51.0

1.9

I-8
5.7

(B .6

£ 5

Radiotron UY-234

Plate Voltage- 90 v.

Screen-Grid Voltage -34>5v.

Filament Vo/taae-Z^v.
Plate Resistance -^megohm
Grid Bias to Cathode
and neqatt^ Fi lament.

Grid Bias -volts

FIG. 10. Static characteristic of the recti-
fier tube of Fig. 9.

sixty decibels. Knowing the reverberation time, the total absorption
of a room can be computed either by Sabine's formula or Eyring's
general reverberation equation, as may be desired.

486

V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E.

This arrangement gives a satisfactory instrumental method of
measuring sound absorption, and also a method of determining the
reverberation time of any room.

Satisfactory equipment for making these measurements does not
solve all the difficulties of making such measurements. If an ac-
curacy of not greater than ten per cent is desired in the total absorp-
tion, most of the difficulties vanish; but when greater accuracy is
desired several precautions must be taken to obtain a uniform distribu-
tion of sound energy in the room where the measurements are made.

To obtain such a distribution a band of frequencies was used at

.I/
.13

(/*

'224 R
P/ate
^creen-G
niame
Vafe Res
Grid h
rid Has
neyat

ectifier ~i
voltage •
rid • votta
nt vo/fa
Distance
las -1.4-v
^o Ccrfhot
Ve FIla^

rrigger D
•90 v.
ge -34.5
ge-2.Sv.
- .5 mega!
neg.

evice

\ [

V.
int

ie and
nent

\ —

Relay Co

rrent

— *

3.0

<u
QC

1.0

FIG. 11.

25 2. t.5 I 3 Q

A.C. Potential -applied to Grid
Response characteristic of the rectifier tube of Fig. 9.

first ; but later work has shown that this is undesirable, as beat notes
may occur, which appreciably alter the result. The source of sound
is in constant motion, this motion aiding materially in producing a
uniform distribution. At the higher frequencies it was thought that
this motion would be unnecessary, but it was found that the apparent
rate of decay of a sound might vary ten per cent when both the source
of sound and the microphone were stationary. This random varia-
tion rarely exceeds two per cent when the source is in motion.

When making measurements in a reverberation room it is possible
to take these precautions, but in studying the rate of decay in a
theater or auditorium, it becomes more difficult.

April, 1932]

REVERBERATION METER

487

To make an intelligent application of acoustical material in a
theater it is believed that equipment such as described here, or the
reverberation meter as developed by the Bell Telephone Laboratories,
should be used to study typical auditoriums and to learn more about
sound distribution and rates of decay in different portions of the room.

a/;

-§28

c
c

12
4

Q 4Vx2.3(logK)E2-logloE<)

v(t2-t()

/ . \ —

Absorpti(

5n=

__^-_

cycles per sec.

1 1 1 1

^1^

1 1

5 10 15

Time in seconds

FIG. 12. Decay curve obtained by making measurements of a
room, starting at the same intensity level and ending at arbitrary
thresholds whose ratios are known. Each point represents the
average of 10 measurements.

This study should be made at all frequencies so as to aid in determin-
ing the most desirable characteristics of a sound absorbing material
and the locations in which such material should be applied.

REFERENCES

1 MEYER, E., AND JUST, PAUL: "Zur Messung von Nachhalldauer und Schall-
absorption," Elek. Nach. Tech., 5 (1928), p. 293. CHRISLER, V. L., AND SNYDER,
W. F.: "The Measurement of Sound Absorption," Bureau of Standards Jour, of
Research, 5 (Oct., 1930), p. 957.

2 MEYER, E. : "Automatic Reverberation Measurement," Zeit.f. Tech. Physik.,
II (1930), No. 7, p. 253. STRUTT, M. J. O.: "Automatic Reverberation Measur-
ing Instrument," Elek. Nach. Tech., 7 (July, 1930), p. 280. WENTE, E. C., AND
BEDELL, E. H.: "A Chronographic Method of Measuring Reverberation Time,"
/. Acoust. Soc. ofAmer., I (April, 1930), No. 3, p. 422.

16 MM. SOUND FILM DIMENSIONS*
RUSSELL P. MAY**

Summary. — A method is set forth for the derivation of dimensions and locations
of the final projection print and all camera, printer, and recording apertures, con-
sideration being made for film weave, shrinkage, and mechanical tolerances in the
apparatus involved in producing and projecting the film.

Two methods of producing films are considered: (a) Projection positive print
made from a 16 mm. dupe negativz by continuous contact printing, where the dupe
negative is made by optical reduction of the 35 mm. picture from a master positive
and the sound re-recorded from a 35 mm. sound track, and (b) production of the pro-
jection positive print from a 35 mm. picture negative by optical reduction and re-
recording of the sound from the 35 mm. sound film directly to the final 16 mm.
positive. The method provides for modification of these processes so that any com-
bination can be used.

Motion pictures in the home have in the past three or four years
enjoyed a slow but steadily increasing popularity and utility. One
witnesses frequently at the beaches and other resorts, amateur
cinematographers with their cameras making pictorial records of
their children's and friends' animations. Each year these experiences
have become more frequent and now it is not unusual to be enter-
tained, during an evening's visit, with motion pictures whose prin-
cipals are your own friends and acquaintances. At the motion pic-
ture counters of photographic supply houses we see interested people
discussing cameras, projectors, lenses, etc., or leaving cine-films to be
developed. By these activities it is not difficult to conclude that the
public is, in part at least, cinette-minded.

Development of motion picture equipment in all its branches has
advanced in amazing strides since the introduction of sound, and
paralleling this, the home sound movie has likewise been developed.
Numerous devices have already made their appearance on the market.
Thus far, they all employ disk type sound records driven synchro-
nously with the film, but sound projectors utilizing sound on film are
soon to make their debut.

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** RCA Victor Co., Camden, N. J.
488

16 MM. SOUND FILM DIMENSIONS 489

In order to present this subject clearly, it is desirable to review the
difficulties encountered in the early attempts at interchangeability of
films made by the various producers of sound films of the variable
width and variable density types. Innumerable cases of variations
of locations of sound track, recorder, and reproducer optical systems
contributed to endless difficulties in attempts to arrive at universal
operation and satisfactory performance of reproducing equipments.
Augmenting these difficulties another source of trouble arose due to
inherent requirements of the variable width and variable density
type sound records, the former requiring that the scanning slit fully
cover the record at all times, allowance being made for variations
that might be introduced during the production of the projection
print or in the projector. Should the end of the scanning slit fall
within the boundaries of the record, the peaks would be "chopped
off," thereby introducing distortion in the reproduced sound. It is
therefore evident that the sound track width should be somewhat
less than the length of the scanning slit.

In the case of the variable density type record, the opposite re-
quirement, that the scanning slit should at no time fall outside
the boundaries of the sound track, applies. Should this occur,
noise might be introduced by either the sprocket holes or the
picture.

Thus we see that if a universal scanning slit is to be used in pro-
jectors, the first-mentioned record must be narrower than the latter
and the locations of the records and the scanning slit must be held to
close limits.

It is needless to dwell on the desirability of preventing a recurrence
of past difficulties with the advent of home sound motion pictures.
Surely no word of explanation is needed to point out the importance
of standardized film, recording and reproducing slit dimensions,
when considering the potential home and industrial fields for this
class of equipment and the production of apparatus and films, the
success of which depends wholly upon interchangeability of films in
the projection equipment.

With this point in view, this paper sets forth a method for the
determination of film dimensions taking into account the variations
experienced by the films from the making of the original 16 mm.
negative to the projection of the positive print. This method has
been followed in arriving at the projection print dimensions, as well
as the projector sound and picture aperture dimensions.

490

RUSSELL P. MAY

[J. S. M. P. E.

The diagram in Fig. 1 shows all the probable steps in the production
of the positive film from a "dupe" sound and picture negative and
the extreme variations considered. It may be advantageous for

Jo ^f

o ^

§1

•a*a
a D

i-s
'.I

</) C|_|

a, o

3 8

fitf

.2
. VJ3

'•".S
d £5

various reasons to produce projection prints by direct reduction of
the picture from a 35 mm. negative and re-recording of the sound
from the 35 mm. film direct to the projection positive. Fig. 2 shows
the various steps involved in this procedure. It will be noted that

April, 1932] 16 MM. SOUND FlLM DIMENSIONS

Jo »fp» «piCiq |,

491

492

RUSSELL P. MAY

[J. S. M. P. E.

the sound and picture apertures in both diagrams are the same. This
was done in order that the processing equipment might be uni-
versally adaptable to both methods. The latter method involves fewer
steps, which will increase the margin of safety.

The variations used in the original calculations were twice those
shown in the diagrams and were based on the practical limits of the
machines to which they apply. A film 0.660 inch wide resulted,
which did not meet the requirement that the sound and picture be
adapted to a 16 mm. film. Examination of the various steps shown
in Fig. 1 discloses that each factor varies, within certain limits, in-

FIG. 3. Light slit dimensions and loca-
tion for making dupe negatives on 16 mm.
sound recorder. The diagram shows the
film in the recording position with the
emulsion facing the recording light.

dependency of any other. By applying the principle of probability
to the distribution between these limits, the likelihood that all the
variations will simultaneously occur in the same direction is so remote
that it falls within the bounds of safety to reduce the original limits
to one-half those shown in the diagram. By so doing, we are en-
abled to use standard 16 mm. film to carry the sound track in addition
to facilitating the projection of present amateur films.

The projection print will be a 16 mm. film having standard perfora-
tions along one edge only. Eliminating the perforations along the op-
posite edge provides space for the sound track.

April, 1932]

16 MM. SOUND FILM DIMENSIONS

493

The requirements to be fulfilled are as follows:

(1) 16 Mm. film having one row of sprocket holes.

(2) Standard projector picture aperture.

(3) Projector picture aperture within the film image at all times.

(4) Clear space between picture and sound for a film supporting shoe in

recording, printing, and reproducing equipment.

.1186"

3JTIT

P .3981

.5175 -

Emalsion Docun

FIG. 4. Printing aperture dimen-
sions and locations for making direct
reductions to 16 mm. dupe negatives
on optical reduction printer. Diagram
shows film in printing position with the
emulsion facing the printing light.

(5) 60-Mil recording slit.

(6) Sound reproducing slit to cover track at all times with ends riding

in opaque stripes adjacent to the track.

(7) Transparent space at sound edge of film to prevent peeling off of

emulsion.

Starting at the top of the diagram, Fig. 1, the solid areas represent
aperture plates and the clear spaces the apertures through which
the sound recording and picture printing lights pass as indicated at

494

RUSSELL P. MAY

[J. S. M. P. E.

1, 2, 3, and 4. The lines thus numbered will be referred to as the
"principal lines." The departure of latent or developed images from
these starting points is shown by the parallel lines, diverging at each
step, for the reasons indicated. Lines diverging to the left have
been designated a and those to the right b, in each instance. These
lines denote extreme limits only, and when considering picture or

.6269"

Em ul si on Up

.00 55"

FIG. 5. Sound track and printer dimensions and loca-
tions; 16 mm. dupe negative (developed, 0 to 0.5 per cent
shrunk). The sound record leads the picture by twenty-
one frames or 6.3 inches.

sound track widths, the total distance between lines having similar
designations is determined by addition.

The divergence continues in a regular manner in both directions
down through steps A , B, and C. Variations shown in step D deal
with dupe negative shrinkage and therefore are in one direction only,
and since shrinkage of the dupe negative can result only in displace-
ment to the right, the a lines continue down from C to D without
change. The dupe negative is guided in the continuous printer
by the sprocket holes, which in the diagram are represented by the

April, 1932]

16 MM. SOUND FILM DIMENSIONS

495

vertical line at the right-hand edge, and causes the shifting of the
film image, due to the shrinkage, to occur in the direction shown.
The line, 4b, in this step does not show a shrinkage offset This was
purposely omitted as it only amounts to 0.00001 inch and can be
neglected.

Step E shows the film image positions after printing the dupe nega-
tive in the continuous printer, whose variations are as indicated.
The solid areas indicate light-shields built in the printer. The solid
area near the center is a light-shield and film support to hold the dupe
negative and positive films in contact. The space between this and
the shield on the left represents the opening of the sound printing
light aperture. It will be noted that a space exists between the

Dupe Neq>+ive cui4)i
emulsion up.

q sprocket
fo serve &s film

Sound prm-finq
liqhl &per4ure^

.0060

ide.

Positive film cui4h
emulsion docan.

Picfur»e

U-.005L_Nofe=-Heiq)rf of pnVHnq
1 . *per+ ur»c -UTsu^ liqh+ *
film speed. °

Ucjh+ 5Vneld frv-nd' -*j J-^.o7O

Film 5uppor»f

FIG. 6. Printing apertures, dimensions and locations; 16 mm. continuous
contact printer. Diagram shows negative and positive films in printing po-
sition, emulsion facing printing lights.

maximum sound track limits and the ends of this aperture. It is
through these clear spaces that the light of the sound printer passes
to print the black stripes at either side of the sound track. The
space occupied by the light-shields will receive no light, and therefore
will result in transparent stripes on the developed positive film.
The right-hand edge of the center light-shield coincides with the
projected line 3a, and represents the left end of the picture printing
light aperture and, as shown, coincides with the picture negative in
its extreme left position. When the picture negative assumes an
extreme right position a clear space of 0.006 inch will occur between

496

RUSSELL P. MAY

[J. S. M. P. E.

it and the end of the aperture, and will show up as a black stripe on
the positive. The opposite end of this aperture is located at the
projection of line 4b at step E. Addition of the dimensions involved
results in an aperture 0.4027 inch long, the end of which is located
0.002 inch from the sprocket holes. Under these conditions we may
be sure that the picture will never be cut off in the printer, and the
sound track will at all times be bordered with black stripes.

FIG. 7. Sound track and picture dimensions
and locations; 16 mm. positive (developed, 0 to 1.5
per cent shrunk) made from dupe negative.

Step F deals with shrinkage of the positive film, and shows that
the image shifts to the left. This is due to the fact that during pro-
jection the film is guided by the edge of the film adjacent to the sound
track and, therefore, any motion of the film image due to shrinkage
will be in this direction.

The opaque light-shields shown in step E introduce a new set of
secondary lines, 5, 6, and 7, which must be considered in all subse-
quent^ steps. It will be noted that these lines undergo the same

April, 1932]

16 MM. SOUND FILM DIMENSIONS

497

divergence as the principal lines. Line 5a is shown dotted, as it
falls without the boundaries of the reproducer aperture and is un-
important; the same holds true for line 7 a. Line 5b, however, con-
verges toward principal line la, and 6a toward principal line 2b, at
each succeeding step. The divergence of the principal lines continues
through steps G and H for the reason noted, finally meeting the second-
ary converging lines. Attention is directed to the fact that lines
2b and 6a do not actually intersect at step H. This would ordinarily
occur, but in this particular case a modification was necessary to

•^

t-y

> ° .0119 H

, »

• .0066

' .0599

•*•

^ .Ol*ffc

-. .0745

;

— — .

: 0

Emulsion Up

FIG. 8. Light slit dimensions and location
for making direct re-recordings on 16 mm.
sound recorder; diagram shows film in record-
ing position with emulsion facing the record-
ing light.

adapt the reproducing aperture to film made either by printing or
by direct re-recording. These points of intersection define the limits
of the projector sound and picture apertures. The shoe shown be-
tween the two apertures serves as a support for the film and rides
in the clear space indicated by the lines converging from the center
printer shield. The dimension 0.0228 inch defines the limits to which
the edges of the clear space can move toward each other and not its
actual width, thus assuring that the shoe in the position shown will at
all times ride on an emulsionless surface. By using the intersections
of the progressive variations as locating points for the aperture, we
are assured of proper registration with the film images.

498

RUSSELL P. MAY

[J. S. M. P. E,

In the construction of the diagram, the width of the film from the
inside edge of the sprocket holes to the opposite edge of the film,
0.5215 inch, is used as the working basis, and satisfies requirement
No. 1.

Having predetermined all the factors which should be satisfied,

ao-

1I1FE

•* — .380'^ -- H
.5048

Emalsion Up

-*|.l085h—

Vieca •fVcinq pr»o»ecfion
light ^nd. shocjainq film
in projection position.

.O711
.OIO'

FIG. 9. Picture aperture and scanning line dimensions and locations;
16 mm. projector. Diagram shows film in projection position facing
projection light.

the diagram was laid out completely as shown, minus the dimen-
sions. Requirement No. 2 was met by locating the 0.380 inch pro-
jector picture aperture at the center of the film, having the 0.0167 inch
distance to the sprocket holes. By constructing lines 4a and 4b
so that they are located between the picture apertures and sprocket
holes, requirement No. 3 is fulfilled. The dimensions shown are
derived from the 0.0167 inch spacing and the predetermined varia-

April, 1932J

16 MM. SOUND FILM DIMENSIONS

499

tions indicated by the verticaj figures. The point of coincidence at 4
determines the right-hand edge of the printer aperture. Point 3 is
located by similar treatment.

A shoe width of not less than 0.020 inch was considered desirable,
and after an initial calculation 0.0228 inch was chosen as it facilitated
the use of standard gauge sheet metal for the light-shield. Require-
ment No. 4 is thereby satisfied.

FIG. 10. Sound track and picture dimensions
and locations for 16 mm. positive (developed, 0 to
1.5 per cent shrunk) made by direct picture reduc-
tion and sound re-recording.

The position of 2b at step H locates the right-hand end of the sound
reproducing aperture or slit. Requirement No. 5 is met by making
the distance between lines 1 and 2 at step A 0.060 inch and the posi-
tion of la at step H satisfies requirement No. 6; working up from
this point to secondary line 5 results in a distance from the edge of
the film of 0.008 inch, which satisfies requirement No. 7. Fig. 2
shows the method as applied to 16 mm. sound films made by
direct re-recording of sound and reduction of picture from 35 mm.

500 RUSSELL P. MAY [j. s. M. P. E.

dupe inegative. The aperture dimensions and location are identical
to those shown in Fig. 1.

The two methods shown in the diagrams provide a flexible arrange-
ment whereby 16 mm. projection prints may be made using either
method or any combination thereof. In other words, in addition
to the methods shown in the diagrams the sound may be re-recorded
and the picture printed from a 16 mm. dupe negative, or the sound
may be printed from a 16 mm. dupe negative and the picture printed
by optical reduction from a 35 mm. dupe negative. In any case, we
may be assured, however they are made, that they will always
register properly in the projector.

With regard to sound tracks of the variable density type, it will be
observed that adequate provision has been made to permit the
scanning slit to fall within the track area if the methods set forth
are followed, the additional width being provided by utilizing the
space occupied by the black stripes in the variable width type of
record.

The accompanying drawings show in detail all aperture dimensions
and locations, and maximum and minimum film dimensions as de-
rived by the foregoing methods.

DISCUSSION

MR. MITCHELL: Judging from our experience on 35 mm. film, we ought to
consider maximum shrinkage. The fact that 16 mm. film is made of safety stock,
which shrinks more than standard stock, should also be considered.

The elimination of one row of perforations will tend to obsolete a lot of 16 mm.
film already in use both here and abroad.

MR. MAY: We have considered the matter of obsolescence of the existing
16 mm. equipment, and feel that it is not as important as it may seem, due to
the fact that a picture made for sound requires sound with it in order to afford
a satisfactory performance.

We do not think it important to be able to show sound films on existing 16 mm.
projectors, but we do feel that the projection of existing 16 mm. films made with
amateur cameras should be capable of being projected in our new sound equip-
ment. That is possible. As to shrinkages, we have allowed somewhat more than
what we encounter in practice. One and one-half per cent is ample for the
positive.

MR. VICTOR: Is this paper intended as a proposal for a standard?

MR. MAY: Yes.

MR. VICTOR: Have you experimented with contact printing from 16 mm.
negatives to positives, with the sound track?

MR. MAY: We have, and are satisfied that it can be done commercially.

MR. VICTOR: Is the emulsion fine enough?

MR. MAY: I might add that it is not as good as we would like it to be. But

April, 1932] 16 MM. SOUND FlLM DIMENSIONS 501

it is our opinion that the projected picture and the sound quality are commercially
satisfactory.

MR. SPONABLE: I note that the distance between the sound records of the
corresponding pictures is 6.3 inches. That is not a straight reduction from thirty-
five to sixteen, is it?

MR. MAY: No. The two distances that are used are for theater use. The
distance used by RCA Photophone, Inc., is somewhat different from that used
in the Movietone. If I am not mistaken, it is on the opposite side, too, is it not?

MR. SPONABLE: It is my impression that the separation is fifteen and a half
inches, for theater use. It seems to me that if the 16 mm. machine were properly
designed, we could use a straight reduction from 35 mm. film.

MR. MAY: That would be true if we were making it from a sound negative of
which the dupe is already made. It might be more practicable to print first the
sound and then the picture, in which case the dupe negative could be run through
the two printers one after the other, in which case the sound lead would not be
important.

MR. SPONABLE: I wondered whether it was a case of not being able to design
a projector that would give a displacement of four and three-quarter inches?

MR. MAY: There would result little less than the present length of six inches.
This is as short as we can make it and still get the various mechanisms in place.
We have been using a single row of perforations for probably two and a half years
in our development work, and have found that the wearing qualities of the film
show no appreciable difference, whether one or two rows of perforations are used.
In fact, we have run films to destruction, and in most cases cannot keep the splices
in long enough to finish a film. I might add that twelve or fifteen hundred trips
to the gate are not uncommon. And usually the emulsion and the surface of the
film are damaged, rather than the sprocket holes.

MR. RICHARDSON: How is it proposed to adapt the sound to the relatively
tiny figures in the 16 mm. screen image?

Using the relatively narrow 16 mm. sound track, is it possible to obtain the
same quality of the sound that may be had from the wider track?

MR. MAY : The picture can be made about as wide as five feet. When sitting
in a small room, the angle at which the eye subtends the screen is not greatly
different from the similar angle in the theater. In other words, sitting in a living
room and watching a picture as large as can be projected from a 16 mm. film,
one obtains the same illusion of size as he does in the theater. As to the sound,
the sound track is only ten mils narrower than that used on 35 mm. film. The
latter has a seventy-mil track, and we have used a sixty-mil track in our small
film. The difference in size of track is only about sixteen per cent. The sound
output is about the same as that obtained from a radio set — comfortable room
volume.

MR. HICKMAN: This is not an attempt to form standards at any immature
stage in the development of the art. The complaint has been made, in previous
developments of motion picture engineering, that dimensional cooperation is
absent at the start; after the job is done the manufacturers try to get together
and find out how they can simplify matters. Now we have the advantage here of
someone's thinking out clearly and carefully, at the very beginning, the factors
that underly the situation; presenting them quite openly, so that what is being

502 RUSSELL P. MAY

thought by one powerful group will be known by all. I do not think that there
is any question of establishing these standards at this time.

Mr. Hardy asked me to announce that this paper was presented under full
cognizance of the Standards Committee and is being considered by them.

MR. EVANS: One of the reasons why the Standards Committee wanted
to have this paper presented before the Society at this time was to discover
what, if any, objections to it might be raised, so that the Committee would
have as many facts before it as possible when it attempts to standardize 16 mm.
pictures.

To the Committee were presented several communications indicating quite
different viewpoints on the subject. One communication advised that although
we were going to have a 16 mm. sound track, it ought to be 20 mm. Another one
stated that if we were going to have a 16 mm. sound track, two rows of perfora-
tion should be used. Those were important differences. The Committee wel-
comes free discussion of all such important factors, so that it will have all the
various viewpoints before it.

MR. COOK: After an experience of eight years in the Kodascope Libraries,
I can assure you that a single row of perforations is ample to secure a much
greater projection life of the film than is likely to be consistent with the obsoles-
cence of the subject. The Bell and Howell machine has only one claw on one
side; test strips have been run in this machine many hundreds of times without
apparent deterioration of the edges of the perforations.

MR. RICHARDSON: With a single line of perforation, would it not be necessary
to adjust the tension carefully?

MR. COOK: In the 16 mm. film a more accurate registration is possible with
a single row of perforations than is likely in the 35 mm. film with two rows of
perforations. There is no difference that the unaided eye is able to discover
in the image projected from a film with a single row of perforations than from
one with two rows.

PROPOSED CHANGE IN THE PRESENT STANDARDS
OF 35 MM. FILM PERFORATIONS*

A. S. HO WELL AND J. A. DUBRAY**

Summary. — There are at the present time two standards of 3 5 mm. film perfora-
tion, one known as the Bell & Howell perforation for negative films and the other the
rectangular perforation for positive films, both of which have been approved and
adopted by the Society of Motion Picture Engineers. Unfortunately, the use of these
two standards introduces complications found detrimental in certain types of work,
which indicate the advisability of having a single standard.

It is felt that the rectangular style of perforation has advantages that it is de-
sirable to retain. An alternative standard is proposed that will combine the ad-
vantages of both the present styles, and which, at the same time, can be used on prac-
tically all existing equipment without alteration of that equipment. Means are
also suggested for shortening the transitional period of such a change-over.

There are at the present time two standards of 35 mm. film perfora-
tion, one known as the Bell & Howell perforation for negative films
and the other the rectangular perforation for positive films. Both
standards have been approved and adopted by the Society of Motion
Picture Engineers, for reasons which are well known.

It appears that while the decision to adopt two perforation stand-
ards was originally taken with a view of reconciling commercial
requirements and economic dictates, the adoption of a double stand-
ard has now created an undesirable condition, especially in contact
printing, process work, etc., involving exact superimposed regis-
tration.

While it is true that until now the motion picture industry has been
able to get along with the two standards of perforation, it is equally
true that this double standard creates a serious barrier to further
technical advances of motion pictures. This barrier is perhaps only
important at the present time to a limited number of motion picture
technicians, but it is a barrier which will rapidly and seriously hamper
the efforts of the industry toward further achievements.

In this connection, it is interesting to review the discussion that

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Bell & Howell Co., Chicago, 111.

503

504

A. S. HOWELL AND J. A. DUBRAY [J. S. M. P. E.

occurred on the rectangular perforation at the time it was proposed
by the Standards Committee.1 At that time, the difficulties arising
from the use of two standards of perforations were pointed out,
and the situation today, with respect to the impossibility of securing
satisfactory registration when the two sizes of perforations had to
be used together, was forecast.

In present practice where superimposed registration is required,
as in composite photography; dupe negatives for lap dissolves;
step printing with pilot control; color photography, in which two
negatives are exposed simultaneously in contact or in superimposed
relation, or two frames exposed in accurate relation to one another,

I
FIG. 1.

Splice of two films, one with negative and one
with rectangular perforations.

and many other conditions where it may be desirable to use positive
and negative stock together, the existence of two dissimilar perfora-
tion standards is already showing ill effects and will introduce even
more serious obstacles in the near future. Furthermore, the dual
perforation standard may eventually become a cause of trouble in
theater projection, with the practice of indiscriminately using both
positive and negative perforations in release prints. When both
types of film perforation are used indiscriminately in this manner,
it is impracticable, if not impossible, to maintain good splice registra-
tion even with the best facilities available.

For instance, Fig. 1 shows a splice of two films, one with negative
and one with rectangular perforations. In the shaded outlines, two
conditions of the positioning of the splicing machine pilot pins, with

April, 1932] CHANGE IN 35 MM. FlLM PERFORATIONS 505

respect to the perforations, are shown, the pilot being of the same
size and shape for both conditions. In the upper part of the drawing,
the pilot pins are assumed to fit the negative perforation perfectly.
These same pilot pins cannot, however, fill the rectangular perfora-
tion, and the drawing plainly shows what the maximum error in
registration may be.

When two sizes of perforation are used, as it is quite impracticable
to provide two sizes of registering pins in the same splicing machine,
the only alternative is, of course, to use the smaller or negative size
pins, with the resultant probable error as illustrated. The dotted
line portion shows the correct alignment when the splice is made in
correct registration. The possibilities of errors occurring in splicing
under present conditions have been presented here as being perhaps
the most tangible and most easily illustrated.

Such errors of registration, however, assume a much greater
importance in special process cinematography and in general studio
and laboratory processing practice. For instance, in sprocket
control printing machines in which the sprocket design is correct
for obtaining both longitudinal and side control for the negative
form of perforation, there is an essential lack of side control of the
positive film having the rectangular perforation. This amounts to
approximately 0.0045 inch in excess of that existing between sprocket
tooth and negative perforation.

An additional tolerance allowance has to be made on the sound
track to take care of this. In order fully to satisfy this control
condition, it would be necessary to use the negative form of perfora-
tion for both positive and negative film.

We are therefore faced with two alternatives: either to eliminate
one of the two standards now in use or to adopt a perforation that
will combine the advantages of the two present types in such a manner
as to provide the facilities mandatory to good registration in proc-
essing as well as adequate control in projection.

The rectangular form of positive perforation has some advantages;
for instance, it provides for equal longitudinal compensation, while
providing better means for the transverse control with sprockets
as well as with pilot-pin means of registration. It would not be
practicable to eliminate the present negative perforation in favor
of the present positive rectangular perforation; but it would be
entirely possible to eliminate the present positive perforation without
affecting any of the present processing equipment. However, it

506

A. S. HOWELL AND J. A. DUBRAY [J. S. M. P. E.

may be considered desirable to retain the advantages of the rectangu-
lar form of the positive perforation.

It is therefore here proposed, that a dimensional change be made

.99ST ,

(25.375%)

^ ~jfcr

-€

;Cp

-e

70-

.37

795" (35»%0

r4.74«6"AO

(8.79*%)

57598" (34.95'%,)

FIG. 2. Bell & Howell negative perforation; standard 35 mm.
negative film.

in the present positive standard of perforation which will permit
using positive or negative films interchangeably or indiscriminately,
and with equal facility in nearly all existing processing equipment.

.0 195" RAD.

.HO"

.076

0.98--X,)

ALTERNATIVE

FIG. 3. Standard 35 mm. positive perforation and alternative
approved by the S. M. P. E.

The dimensional characteristics of the two standards now in use
will be reviewed and the advantages of the proposed change ex-
plained.

Fig. 2 gives the dimensions of the Bell & Howell negative perfora-

April, 1932] CHANGE IN 35 MM. FILM PERFORATIONS

507

tion, which is bounded by two straight parallel faces 0.073 inch
apart and by two curved faces corresponding to two arcs of a circle
having a diameter of 0.110 inch, the height of the chord of the radial

.743

-&v

•

D
D
D
ID

o ,' ">

°i 4

O JEMULSIOM

IDE UP

~~t j~V

.95

(25.3
1.10

75%)

(28.169%)
1.37795" 1 (35%)

.0133" RAD.

(2.79«%)

1.37598" (34.95%)
4 OF FILM

FIG. 4. Perforation dimensions proposed for 35 mm. positive film.

portion being 0.0139 inch. The pitch of this perforation is 0.187
inch.

Fig. 3 shows the dimension of the standard 35 mm. positive
perforation and an alternative approved by the Society of Motion

-E-J-

FIG. 5. Punches for re-perforating old negatives.

Picture Engineers. It will be necessary to consider only the rec-
tangular perforation with rounded corners of the shorter radius,
since it is fortunately the only one in common use; if both were in
use, the situation would be still more complicated.

508

A. S. HOWELL AND J. A. DUBRAY [J. S. M. P. E.

As shown in the figure, the width of the perforation is 0.110 inch,
as in the negative perforation, while its height is 0.078 instead of
0.073 inch. The radius of the rounded corners of the rectangular
perforation is 0.0195 inch.

Fig. 4 shows the perforation dimensions which are suggested for
consideration as a proposed standard for both negative and positive
film, with certain restrictions imposed with respect to sprocket
controlled processing mechanisms. These restrictions consist in
retaining the present shape of sprocket tooth until the obsolescence
of the present form of negative perforation, which could be safely
anticipated to require a period of about 15 to 20 years. This

FIG. 6. Composite drawing showing present standard
negative film perforations, proposed rectangular form of
perforation, and present positive film perforation, super-
imposed.

period could be shortened if, say, after a definite lapse of time,
whatever old negatives were to be printed could be re-perforated.
This re-perforating has been done in the past with perfect satis-
faction, and would be entirely practical should this proposal be
acceptable.

Fig. 5 shows how this is accomplished. A rectangular punch
is made with the lower end ground to fit the present negative perfora-
tion. The end is pointed to facilitate entry and positioning.

Only one pair of holes is perforated at a time. The sprocket hole
on the control side is perforated to cut out the corners only, while
the opposite perforation would be cut more on one side than the
other, depending on the shrinkage. This assures perfect control
and yet takes care of shrinkage.

April, 1932] CHANGE IN 35 MM. FILM PERFORATIONS

509

The dimensions suggested for the modified perforation are as
follows:

Perforation width 0.110 inch

Perforation height. 0. 073 inch

Rounded corners of a radius of 0. 0139 inch

Pitch 0. 187 inch

It will be noticed that the shape of the proposed perforation is
rectangular with rounded corners. Its width remains 0.110 inch,
as in both of the present perforations.

Its height is that of the present "negative" perforation and the
radius dimension of corners is changed from 0.0195 to 0.0139 inch,

l<2> TEETH
FILM

I'BASEDIA

>7SM/«) ROUND CORNERS

'2ifc75'7«

.p-75'gAD
X(L9I*%0

FIG. 7. Intermittent and feed sprockets.

in order to coincide with the chord height of the radial portions of
the present 35 mm. negative standard perforation, as shown in Fig. 6.

Fig. 6 is an enlarged composite outline drawing in which the
lined portion represents the present standard negative film perfora-
tion, with its major and minor dimensional extremities shown in
coincidence with those of the proposed rectangular form of perfora-
tion.

The dotted line portion of this figure represents the outline of
the present positive film perforation with relation to the proposed
standard perforation showing the differences in the minor dimensions,
amounting to 0.005 inch.

Since the pitch and height of the proposed perforation are identical

510

A. S. HOWELL AND J. A. DUBRAY [J. S. M. P. E.

to those of the present negative perforation, no alteration is necessary
in camera or laboratory apparatus as used today for which the size,
shape, and pitch of the sprocket teeth have been carefully calculated
with regard to the relation of the arc of contact of the film with the
periphery of the sprocket, and to the extent of film shrinkage that is
to be accommodated.

For projection machines, which must accommodate a greater film
shrinkage than cameras and laboratory apparatus, it may or may
not be found advisable to modify the height of the sprocket tooth
and its width at the base, as illustrated in Figs. 7 and 8. These

FIG. 8. Take-up sprocket.

illustrations show: first, the dimensions of an intermittent and feed
sprocket, and second, those of a take-up sprocket. The diameter of
the take-up sprocket is less than that of the others, resulting in a re-
duction of effective tooth pitch for the former, the size and shape of
the teeth being the same for all sprockets. Here again, it is interesting
to refer to the original discussion on the rectangular perforation,1
when it was seriously debated whether it would not be advisable to
standardize the sprocket rather than to change the film.

The figures show that the proposed alteration consists in a reduction
of the width of the tooth base from the present standard of 0.050 to
0.045 inch.

After considering the improved means now available to exchanges

April, 1932] CHANGE IN 35 MM. FILM PERFORATIONS 511

and theaters for handling positive films, and the decreased possibility
of using film that is excessively shrunken, the writers do not feel that
the slight difference between existing sprockets and the proposed
alteration would be of any serious consequence, even though the
present standard tooth width of 0.050 inch is retained.

Incidentally, if such a change in the sprocket should be considered
desirable, it might be pointed out that most of the sprockets in use on
projection machines are probably so worn that they would not have
to be touched; it would be a comparatively simple matter to have
new projector sprockets made to the new proposed dimensions.
If the Society of Motion Picture Engineers, however, considers it
advisable to maintain the same amount of relative clearance as is
now provided in the present positive standard perforation, the sug-
gested alteration of projector sprockets would not present any serious
difficulties.

REFERENCE

1 Report of Film Perforations Committee, Trans. Soc. Mot. Pict. Eng., 16
(May, 1922-23), p. 303.

DISCUSSION

MR. Cox: Are the conditions of the raw stock, or use of it in the machine,
such that the objections originally raised, in 1923, have been obviated in the
present system ?

MR. DUBRAY: I believe that they have. Although little official data are
available on the subject, reports from consumers of film indicate that these ob-
jections are overruled sufficiently at least to make possible the proposed change
in perforation size and shape.

MR. Cox: Have you found any difference in the negative shrinkage or
positive shrinkage in the present process over what it was five years ago?

MR. DUBRAY: With regard to negative films, we would say yes. Films
manufactured now, experimentation shows, shrink less than the films of five
years ago, due, probably, to two main factors: improvement in the manu-
facture of the film base, and more careful and adequate treatment of the film
before and after processing. With regard to positive film, shrinkage accommoda-
tion was extended to as much as 2.5 to 3 per cent five years ago, while today film
manufacturers seem to be in accord in considering 1.5 per cent the maximum
shrinkage that must be accommodated. Again, no official data is available,
but practice seems to be in accordance with these data.

PAST- PRESIDENT CRABTREE: For what percentage shrinkage are these teeth
calculated?

MR. DUBRAY: If I remember correctly, projector sprockets are designed to
accommodate a 2.5 per cent shrinkage, while the sprockets of the Bell and Howell
continuous printer accommodate a shrinkage of 1.5 per cent.

THE ANIMATOPHONE
A NEW TYPE 16 MM. SYNCHRONOUS DISK REPRODUCER*

A. F. VICTOR**

Summary. — A new reproducer employing a vertical turntable and floating pick-up
is described. The reasons for the radical departure from conventional design and the
advantages resulting from this departure are briefly referred to. The reproducer
is of the portable type, being housed in a compact lightweight carrying case.

Two years ago the Victor Animatograph Corporation made a de-
cidedly original contribution in reproducer design by introducing the
vertical turntable. For the first time in the history of reproduction
of sound from disks the conventional horizontal turntable had been
abandoned, the vertical turntable making possible a more compact
design than is practicable with the horizontal turntable, and also
makes possible a rigid direct drive that is absolutely positive.

The close coupling of the turntable and the projection mechanism
in turn made imperative a positive speed control. No mechanical
control could supply the necessary smoothness and uniformity of
drive for perfect sound reproduction. This necessity was supplied
by a new electro-pneumatic control. A current of air originating
in a rotary blower is made to fall upon a thin membrane which
carries an electrical contact. The counteracting force is gravity,
hence the governor is not dependent upon the action of springs.

When the blast is increased to a certain intensity, the contact opens,
causing the current supply of the motor to flow through a resistance,
immediately causing the motor to run more slowly. When the pro-
jector is operating at the correct speed the contact is made and
broken at a rapid rate, thus always maintaining the proper speed
within close limits.

The vertical turntable design, revolutionary as it was, proved to
be a decided advantage. The quality of the sound produced was
distinctly better than had before been obtained from synchronous

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Victor Animatograph Corp., New York, N. Y.
512

THE ANIMATOPHONE

513

disk equipment. But the mechanical ground noise, so common
in all such equipment was present in slight, although appreciable,
degree. Experiment showed that ground noise is inevitable in every
case where the support of the pick-up is rigidly secured to the base
of the mechanism proper.

As insulation and damping did not properly overcome this diffi-
culty, the floating pick-up was designed to do so.

This consists of a support standing beside the reproducer but
mechanically isolated from it. The support carries a rocking arm,
to the upper end of which the pick-up arm is pivoted. Again the

FIG. 1. View of the disk, floating pick-up, and micro-
phone attachment of the Animatopnone.

force of gravity is called into play to keep the pick-up arm in a
vertical position, regardless of the lateral position of the rocker arm.
As a result the reproducing needle is maintained at all times in per-
fectly tangential relation to the groove, a condition common in
recording and the only position which can provide the satisfactory
kind of reproduction that attends the utter absence of needle drag
against the side of the groove.

The floating pick-up may be used with any needle pressure desired,
as merely increasing the distance between the pick-up base and the
projector increases the pressure of the needle upon the record.
Pressures suitable for both lateral cut and hill-and-dale records may
be easily secured.

514 A. F. VICTOR [j. s. M. p. E.

These three features, the vertical turntable, the electro-pneumatic
speed control, and the floating pick-up, have made possible a quality
of reproduction by portable apparatus comparable with the best
theatrical reproduction.

The projector has incorporated in it a safety throw-off which makes
breakage of film impossible. This is a convenience in silent pro-
jection, but is vitally essential in film-disk synchronization for ob-
vious reasons. By thus eliminating film breakage, the principal
objection to the use of disks is overcome. The compact design of

FIG. 2. Interior view of the case showing the
projector and the pick-up and film reels in po-
sition.

the vertical turntable eliminates the sole remaining objection to the
use of disks, so that in the new Animatophone, the disk system is
placed upon a par with film recording, while retaining the low cost
and simplicity of the disk.

The latest model of the Animatophone is enclosed in a carrying
case which also serves as a "blimp," reducing the projector noise to
an unobjectionable level. A small disk in the upper front corner of
the case is pulled out to reveal the pilot lamp for the turntable.
This lamp is connected in the circuit of the speed control so that it
permits direct visual observation of the oscillations of the control

April, 1932]

THE ANIMATOPHONE

515

circuit, a constant check of speed that cannot be overlooked but
which is apparent only to the operator.

The Animatophone may be used with standard theatrical disks,
or with the home type of disk designed to rotate at approximately
eighty revolutions per minute. The change may be effected in a
minute, making it possible to project all synchronous films which
have been placed upon the market.

FIG. 3. Close-up of the turntable and pendulum
type pick-up.

The light used is the 250-watt, 20-volt lamp, energized by a trans-
former concealed within the base of the projector.

The projector may be transported threaded, and ready for imme-
diate operation, and four extra films may be carried within the case.

When in operation the projector requires a table space eight by
eighteen inches, with nine square inches additional for supporting
the pick-up base.

During transportation all items of the equipment, such as the
pick-up, the control, and the microphone, are carried in racks inside

516 A. F. VICTOR

the case where they are instantly available, yet held so securely that
they cannot move when the case is closed. Packed for transporta-
tion, the case measures eight by eighteen inches and weighs thirty-six
pounds.

DISCUSSION

MR. PALMER: Has Mr. Victor any data as to the mechanical accuracy of
the speed control? As in sound recording work it is constantly becoming more
necessary to control the speed of the recorder very accurately, I should like to
know whether this control offers any possibilities in that direction.

MR. VICTOR: I rather think it does. The trouble with the Animatophone,
in the attempt to use it as a recorder, is that the turntable is too light in weight.
There is no stabilizing effect as the turntable weighs only four pounds. But I
rather think this type of speed control would be very serviceable for more accu-
rate work.

THE ACOUSTICS OF LARGE AUDITORIUMS

S. K. WOLF**

Summary. — Extremely large auditoriums present acoustical difficulties which
do not readily yield to the customary methods of analysis and correction. This is
illustrated by measurements of the time of reverberation, made in the Madison Square
Garden, New York, N. Y., which revealed a considerable discrepancy between theo-
retical expectations and the times actually measured throughout the frequency range.
At 500 cycles, for example, analysis of the auditorium indicated a decay period of
35.5 seconds, whereas the time actually measured by the spark chronograph rever-
beration meter was only 7.6 seconds. On the basis of the measured time, 47,000
square feet of one-inch rock wool were installed. This material was distributed in
a manner calculated to suppress undesirable discrete reflections as well as to reduce
the general reverberation time. The result was a reduction in the measured time to
3.5 seconds and the complete elimination of acoustic difficulties. Present reverbera-
tion formulas do not possess sufficient generality to justify application to enclosures
which are extremely atypical in size or shape. Until such formulas are developed,
reliance must be placed on actual measurements.

As has been pointed out so often, the acoustical properties of a
theater should satisfy the acoustical requirements of the performance
to be held in it. Since types of performance vary, unless variable
and controllable conditions are feasible, the selected acoustical
condition should conform to a compromise which will most nearly
fulfill all demands.

By properly considering the use to which the theater is to be put,
prior to constructing it, specifications may be prepared which will
assure adequate acoustical control in the new theater. This is
naturally more economical than recommending treatment or altera-
tions in existing structures. However, the economies thus effected
are not the only reasons which make such scientific acoustical planning
desirable. Of equal importance is the assurance that the best condi-
tions will exist after completion.

It is not always possible to determine all acoustic deficiencies by
theoretical analyses. Certain types of design and construction may

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Electrical Research Products, Inc., New York, N. Y.

517

518 S. K. WOLF [J. s. M. P. E.

contribute to unusual conditions which are not amenable to exact
theoretical analysis. This is particularly true in large auditoriums,
since the discrepancies between the actual and the theoretical condi-
tions increase rapidly with size, and in those of very great volume,
the theoretical analysis is totally inadequate. In modern scientific
design these possible pitfalls may be carefully avoided; in existing
theaters their effects can be determined only by instrumental means.
We hope that research now in progress will soon yield theoretical
methods and formulas capable of diagnosing all acoustic defects.

In the meantime, we must depend on instruments for thoroughly
analyzing the acoustical qualities of many auditoriums which have
been designed without adequate acoustical study. Instruments are
now available which will accurately measure reverberation time,
energy distribution, and associated characteristics. Acoustic
measurements, although not always necessary, are particularly
advisable where peculiar shapes or unusual finishing materials are
employed. Unfortunately, this is true of quite a large number of
theaters. The guiding factors in the past have been the comfort of
the audience, the number of seats in the theater, its beauty, and other
practical considerations. The acoustical quality of the theater was
often neglected. Xhis neglect was not due primarily to oversight, but
rather to a lack of knowledge of acoustical phenomena. As a con-
sequence, the resulting design frequently was acoustically unsatis-
factory. We are considering in this paper a rather unusual problem in
auditorium acoustics, which, however, will serve to demonstrate the
inadequacy of the usually accepted theoretical analysis under certain
conditions. We refer to the acoustical problem that existed in New
York's Madison Square Garden.

Madison Square Garden was designed primarily as an enclosed
sports amphitheater. Acoustics were not seriously considered, since
noise and reverberation are expected in an auditorium of this kind.
Nevertheless, the size, the shape, and the finishing materials that were
used, introduced some very undesirable acoustical effects. When a
band selection was played, the tune was scarcely recognizable.
When an announcer spoke, his voice could not be heard at distant
points. To obviate these deficiencies, a sound reenforcement system
was installed. A standard type of public address system was ob-
tained; yet sound in the enclosure was not intelligible. A new system
was then built, improving the quality of sound emanating from the
system. Still the results were most unsatisfactory. Thus, time and

April, 1932]

ACOUSTICS OF LARGE AUDITORIUMS

519

money were wasted before the real cause of the poor grade of intelligi-
bility was fully recognized. The difficulty was caused by the
acoustics of the auditorium rather than by any equipment de-
ficiency.

Acoustical correction was not easily achieved, since the condition
could not be correctly evaluated solely by theoretical analysis. Such
an analysis showed the requisite correction to be unfeasible. Every
available surface would have had to be treated and the cost would have
been prohibitive. Acoustical measurements of the auditorium, how-
ever, showed that correction was not only feasible, but that it could be
effected at a relatively small cost. Furthermore, the degree of
control would be adequate for both speech and music. This meant
that concerts and operatic selections would be rendered under highly
acceptable circumstances. The auditorium needs no longer depend

FIG. 1. Schematic diagram of chronograph reverberation meter.

entirely on sports for its economic existence. The method employed
in effecting this instrumental analysis, and a comparison of this
analysis with theoretical investigations, will be discussed.

The instrumental analysis consisted largely of measurements of
reverberation time with the chronograph reverberation meter. This
meter was developed by Wente and Bedell of the Bell Telephone
Laboratories, and is here shown in a schematic diagram. (Fig. 1.)
It consists of a variable gain amplifier, a full-wave rectifier, and a
polarized relay with an adjustable bias. This relay is adjusted so
that when the rectified current exceeds the biasing current, the relay
operates, allowing the condenser to charge. When the current falls
below the biasing value, the relay releases, discharging the condenser
through the spark coil, causing a spark to pass to the revolving drum.
If a special waxed paper is placed on the drum, the spark will produce
a spot on its surface. The drum is rotated at a known constant speed.

520 S. K. WOLF [J. S. M. P. E.

The operation of this meter is quite simple. A tone of a given
frequency is amplified, carried through a switch in the meter, and is
then fed to a loud speaker in the room to be measured. With the
drum revolving, the sound is switched on, and the amplification is
adjusted so that the relay will just operate. At a given position in
the rotation of the drum, a trigger arrangement automatically inter-
rupts the source of sound. When this sound has decayed to the level
for which the relay has been set, the spark jumps to the recording
paper as previously explained. The amplification of the meter is
increased 3 decibels, the spark arm is moved a corresponding amount,

FIG. 2. Reverberation time curve at 1000 cycles, Madison Square Garden.

and the same procedure is repeated. In this way a series of points is
obtained, providing an analytical pattern of the decay of the sound.
Since the curve is plotted in decibels, assuming a perfectly diffuse
sound pattern, a rectilinear graph will be obtained. It is not neces-
sary to measure a complete 60-db. decay, since the line can be extrapo-
lated to any required degree. Knowing the speed of the drum, the
time for a 60-db. decay, or the reverberation time, can readily be ob-
tained.

Fig. 2 is a reproduction of an actual curve obtained at a frequency of
1000 cycles in Madison Square Garden. The vertical axis represents
the intensity level in 3-db. steps, and the horizontal axis represents
time. Note that the points weave slightly about the straight line.
This is due to the residual interference pattern in the room.

April, 1932]

ACOUSTICS OF LARGE AUDITORIUMS

521

Fig. 3 shows the reverberation meter.

In making the measurements in Madison Square Garden, specially
recorded warble frequency disks were used as a source of sound. The
warbling of the frequency helps to break up standing waves. The
Madison Square Garden public address system was used to amplify
the pick-up from the records, and to reproduce it in the enclosure.
Measurements made at intervals of approximately one octave indi-
cated a characteristic as shown in Fig. 4. Attention is here called to
the measured time at 500 cycles, 7.65 seconds.

FIG. 3. Photograph of reverberation meter.

Let us now see what theoretical analysis by the classical method
would have yielded. (Fig. 5.) The volume of the Madison Square
Garden was found to be 6,200,000 cu. ft. The ceiling, walls, and
floor were constructed of concrete with a small amount of glass,
metal, and wood. These surfaces comprised 342,300 sq. ft. which,
for the commonly accepted coefficient of absorption of 0.015, provided
5134 sound absorbing units. The 18,000 wooden seats, which are
usually assigned a coefficient of 0.2, provided an additional 3600 units,

522

S. K. WOLF

[J. S. M.P. E.

or a total of 8734 units in the empty auditorium. On this basis
Sabine's formula gives 35.5 seconds, and correction to optimum would
require the installation of 94,600 units, an obvious impracticability.
The use of Eyring's formula offers a refinement which is hardly
within the accuracy of the computations, resulting in a period of 35.0
seconds. Comparing these values with the measured time of 7.65
seconds, the inadequacy of the theoretical analysis is readily ap-
parent.

The reason for this marked discrepancy is not fully explained by

FREQUENCY (CYCLES PER SECOND)

FIG. 4. Measured reverberation characteristics,
Madison Square Garden.

existing data. It is fairly certain that it is partially due to the fact
that the commonly accepted coefficients are too low. Also, there
appears to be a size effect which should be considered in reverberation
computations. A contributing factor may be found in an attenua-
tion by the air or in some other acoustical phenomenon not explicitly
included in current reverberation formulas. Investigations are being
pursued, however, to determine, if possible, the exact reasons for such
a large discrepancy.

The acoustical recommendations for treating Madison Square
Garden were based on the measured values. Using these true values

April, 1932] ACOUSTICS OF LARGE AUDITORIUMS 523

of reverberation time, and correcting them for the presence of the
audience, it was found that with a full attendance of 18,000 people an
optimum time of reverberation existed. Under such conditions,
however, the acoustics were yet remarkably poor. This poor acoustical
condition was obviously due to the distribution of sound energy.
Discrete reflections from the high ceiling eventually reached the seat-
ing area as echoes, or with sufficient time lag seriously to impair
intelligibility. The problem was, therefore, one of properly locating
the absorbing material so as to eliminate these reflections.

By Sabinc's formula: T = -

V = 6,200,000 cu. ft.

0.05 V = 310,000

342,300 X 0.015 = 5134 units

18,000 X 0.2 = 3600 "
Total units A = 8734
310,000

8734

= 35.5 seconds

By Eyring's formula
0.05V

-Sloge(l -a)

35.0 seconds

FIG 5. Computation of reverberation time at 500 cycles,
Madison Square Garden.

The optimum time selected for a volume of 6,200,000 cu. ft. was 3.0
seconds for a frequency of 500 cycles. Since existing optimum curves
do not include data for volumes as large as Madison Square Garden,
this figure represents the projection of the optimums accepted for
smaller volumes, tempered by our experience with the acoustical
conditions of large auditoriums and our knowledge of the uses to
which the Garden would be put after correction. Adjustment to this
optimum value would require 63,800 additional units. Since this
would necessitate the installation of a very large amount of material,
another plan was considered.

The alternative provided for the construction of a false ceiling made
of acoustical material. The ceiling of Madison Square Garden is high

524 S. K. WOLF [j. S. M. P. E.

and arched, supported by trusses. A false ceiling at the lower line of
trusses would reduce the volume by approximately 1,700,000 cu. ft.,
requiring a correspondingly smaller optimum — 2.6 seconds. This
volume reduction would, in itself, effect a large improvement in the
reverberation time so that a great deal less material would be required.

FIG. 6. Interior of Madison Square Garden after
acoustical treatment.

Also the shape of the false ceiling would be greatly superior from an
acoustical standpoint. It was therefore decided to proceed according
to this plan.

The material selected for the installation was a one-inch rock wool
blanket, since this material was found to have nearly the desired
characteristic over the frequency range, and satisfied other physical
requirements. The trusses were spaced 8 feet apart. Since a stand-
ard rock wool blanket is 8 feet long, a rather novel plan was devised.

April, 1932] ACOUSTICS OF LARGE AUDITORIUMS 525

Angle irons were fastened to the upper faces of the blankets to increase
their rigidity, the latter being then laid on the inside of the bottom
trusses as these were of the inverted T-beam construction. This
method of installation was simple, and reduced the cost of the project.
Any section of the ceiling could very conveniently be removed for the
purpose of rearranging the drop lights or for fastening various items of
circus equipment to the steel work. Roughly, 47,000 sq. ft. of blanket
were installed in this manner. (Fig". 6.)

Measurements and observations made since the treatment, indicated
a greatly improved condition. The reflections from the ceiling
surfaces, which had destroyed intelligibility were eliminated. In
addition, measurements show that the reverberation time of the
auditorium when empty, has been corrected to 3.5 seconds at a fre-
quency of 500 cycles with a general characteristic as shown. (Fig. 4.)
A high degree of intelligibility is now obtained from announcements
made on the public address system, and musical selections are heard
with remarkable realism and an agreeable blending of tone.

In a situation such as the one that has just been described, there
is a considerable discrepancy between the usual theoretical expecta-
tions and the actual conditions obtaining. Not all cases are so
pronounced. With a certain amount of justification, we can use the
theoretical analysis as a sort of projected measurement. However, in
acoustics, as in other fields, certain elements are bound to appear
which, for various reasons, are difficult to analyze and evaluate.
Usually this indicates that the empirical law is being applied to situa-
tions beyond the range for which it was intended. It therefore
comes about that considerably more weight must be accorded to
actual measured results when there is any marked disagreement in
the results obtained by the two methods It is hoped, however, that
in the near future, a method of computation may be developed which
will yield results in close agreement with the true values. In the
meantime, the best results are obtained, when measurements are
impossible, by a computation, tempered by the experience gained
from measurements of similar cases.

COMMITTEE ACTIVITIES

REPORT OF THE SOUND COMMITTEE*

At the Hollywood Convention, a paper was presented by Mr. Ben
Schlanger1 entitled "Reversing the Form and Inclination of the Mo-
tion Picture Theater Floor for Improving Vision." As a result of
the discussion which followed, it was suggested that the Sound Com-
mittee consider what difficulties this type of structure might present
from an acoustical standpoint.

The Committee believes that this new design does not present any
radically new problems. From the point of view of reverberation,
the new design should be slightly better than its predecessor because
with a smaller volume of space and equal number of seats, there
should be less reverberation. The considerable reduction in the
amount of curved surfaces, as shown by the architect's statements
and by the proposed plans, will assist in the elimination of undesirable
concentrations of sound and is therefore a praiseworthy feature. It
seems evident that greater care will be needed from the point of view
of distribution of sound from the loud speakers used in a motion pic-
ture house. In the new plan, the orchestra space subtends a smaller
angle at the center of the screen, which means that the equal distribu-
tion of sound from front to back will be somewhat more difficult to
obtain. On the other hand, the opening of the space directly under-
neath the front of the balcony appears to be greater than in the stand-
ard designs, so that there would tend to be less diminution of sound in-
tensity in the region back of this location toward the rear of the main
floor.

It therefore is evident, with the possible exception that greater care
may be required in providing the equal distribution of sound, that the
proposed theater design does not present any problem different from
that of the present type of theater layout. However, the selection
of the proper kind of seat and of the correct type of material for the
wall and ceiling surfaces will remain just as important as it is now.

One of the most important problems now confronting the industry
is the determination of the proper frequency range for the recording

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
526

SOUND COMMITTEE 527

and reproducing systems. It seems to be generally agreed that a
frequency range greater than any yet obtained commercially is es-
sential for an absolutely satisfactory reproduction of sound, but the
exact limits of this range have so far not been agreed upon by the in-
dustry. There are so many unknown or partially known factors
involved that to obtain this information economically requires the
complete coordination of all phases of the industry.

At the present time it is the general practice, of at least the larger
manufacturers of recording and reproducing equipment, to make each
individual piece of apparatus as good as is commercially practicable
so that it may be a fairly permanent investment and not rapidly be-
come obsolete. For example, practically all the amplifiers used in
recording and reproducing are capable of transmitting a much wider
frequency band than they have yet been called upon to transmit.
Therefore, amplifiers as a rule will not have to be changed when other
portions of the circuit, which have heretofore restricted the width of
this frequency band, have been improved. This case simply demon-
strates the desirability and economy of making each link of the chain
of apparatus as good as is commercially practicable so that as the
weaker links are made stronger, the effectiveness of the whole system
likewise increases.

As a part of this problem, the question of flatness of characteristic
within the desirable frequency range must be considered. This Com-
mittee believes that while characteristics other than flat must be used
at times to counteract certain undesirable conditions, nevertheless,
for general application a flat response of each unit, and therefore of
the whole system, is the ultimate to be striven for. At the present
time there are many restrictions placed upon our obtaining such a
characteristic over a wide range, but they should be and are being
gradually removed. When temporary expedients have to be applied
to overcome such conditions, these expedients should be eliminated as
rapidly as possible by alleviating the restricting conditions.

In considering this problem it was fully realized that frequently an
improvement in one part of the circuit will make prominent in other
portions, undesirable features which were previously unimportant.
For example, the introduction of the noiseless recording method has
made evident in theater reproducing systems noisy conditions which
were previously veiled and therefore unnoticed. Considerable
effort and money have, therefore, been spent in improving the con-
dition of the theater equipment.

528 SOUND COMMITTEE [J. S. M. P. E.

Taking all these factors into consideration it is the Committee's
recommendation that the industry attempt to arrive at a conclusion
as to what the ultimate frequency response range should be, and then
to attempt to improve the individual pieces of equipment and the
technic involved so that the circuits will be competent to transmit
this range.

It must not be forgotten that the acoustics of the theater is one of
the most important factors involved and might well be placed at the
head of the list of individual matters to be improved. No matter
how good a recording has been made or how effective the reproducing
equipment may be, if the sound is projected into a house having in-
tolerable acoustical properties, the results will likewise be intolerable.
Knowledge is available to guide this work and materials for correcting
acoustical conditions are at hand. This phase then is not restricted
by lack of knowledge or equipment, but rather by the indifference of
the theater owners. Other important subjects demanding immediate
investigation are loud speaking apparatus in the theaters, film process-
ing methods, and the limitations of the film.

In regard to the last subject, sensitometry presents itself as a matter
of first importance. Individual studios are able to maintain certain
standards and can meet certain requirements in handling both pic-
ture and sound in the negative and also in those positive release prints
made in their own laboratories. On the other hand, every studio has
its own ideas of film processing, which means that additional prints
made by commercial laboratories, either here or in foreign countries,
are apt to differ from the original prints made in the studio labora-
tories. It would seem that the commercial laboratories might well
concern themselves with this problem and, together with the pro-
ducers, insist that an agreement be reached on the proper photographic
measuring instruments to be used, as well as to become thoroughly
familiar with the correct application of the data obtained from these
instruments.

Two steps are necessary: first, standardization of sensitometric
measurements so that data obtained in one place may readily be com-
pared with that obtained in another. At the present time such a
translation of results from one studio or laboratory into terms which
another may understand is almost impossible. Secondly, a better
compromise must be found between the requirements proposed by
purchasers of processed film and the commercial requirements of the
laboratory. A great deal of work in promoting and understanding

April, 1932] SOUND COMMITTEE 529

these needs has been done during the past two or three years; it is
important, however, to take immediate steps to bring about a com-
mon basis of measurement and understanding as well as to prepare a
standard set of specifications to which studios and laboratories may
adhere.

While the domestic situation alone is sufficient to warrant action,
the added confusion in the foreign situation makes this work all the
more imperative. When the release prints are made in a foreign lab-
oratory from a negative produced in an American one, the coordina-
tion between the two laboratories involved is practically non-existent.

The Academy of Motion Picture Arts and Sciences is undertaking
the problem of correlating information to improve this situation. A
duplication of their work is of course unnecessary, but it is recom-
mended that the Society give every assistance possible in this work;
first, it should recognize the need for standardization; second, it
should cooperate with the Academy should the occasion arise. If the
findings of the Academy are acceptable to the Society, the Society
should assist in standardizing them.

H. B. SANTEE, Chairman

M. C. BATSEL N. M. LA PORTE H. C. SILENT

P. H. EVANS W. C. MILLER R. V. TERRY

R. C. HUBBARD S. K. WOLF

REFERENCE

1 SCHLANGER, B.: "Reversing the Form and Inclination of the Motion Picture
Theater Floor for Improving Vision," /. Soc. Mot. Pict. Eng., XVII (Aug., 1931),
No. 2, p. 161.

ABSTRACTS

The views of the readers of the JOURNAL relative to the usefulness to them af the
Abstracts regularly published in the JOURNAL will be appreciated. Favorable views
are of particular interest. In the absence of a substantial body of opinion to the
effect that these Abstracts are desired by the membership, their discontinuance may be
considered.

The Treatment of Auditoriums for Sound Projection. M. SOULIER. Tech-
nique Cinemat., 2, Dec., 1931, p. 35. The steps involved in the preparation of the
auditorium for use in sound reproduction are listed, with critical comments on
standards of acoustic quality, methods, and materials used. C. E. I.

Architectural Acoustics. Study of a Complex Room. G. LYON. Technique
Cinemat., 2, Dec., 1931, p. 5. Maximum efficiency is obtained in an auditorium
when all possible reflected sound reaches the auditor within Via second after the
direct wave. It is asserted that sound impulses arriving within this interval are
integrated in the process of hearing. Certain applications of this principle are
treated geometrically. C. E. I.

Ozaphane Sound Film and Sprocketless Projector. Bull. soc. fran$. phot., 73,
Aug., 1931, p. 168. Sound prints on Ozaphane film used in conjunction with a
scanning device similar to that ordinarily used, have been demonstrated success-
fully in France. A sprocketless projector was employed. The only new feature
of the projector was the take-up mechanism which consisted of a gripping device
with an eccentric rubber covered disk which catches a loop of film and carries it
away with a uniform speed. Automatic framing is accomplished by means of a
photoelectric cell arrangement actuated by the frame lines of the picture.

C. H. S.

Sensitometry at the International Congress for Photography in Dresden.
E. LEHMANN. Kinotechnik, 13, Sept. 20, 1931, p. 346. A report is given of the
action of the Congress with respect to the proposals of the Optical Society of
America for international standards for sensitometry. The Germans proposed to
accept the American specifications regarding the color temperature of the light
source and the filters for altering the spectral distribution to that of sunlight.
They proposed, however, the use of a neutral wedge for determining speeds in-
stead of a time scale, to simplify the procedure for small factories. The matter
was finally referred to the sensitometry committees of the different countries for
consideration until February 15, 1932, with the recommendation that a final
decision be reached in a small commission within a further six months.

M. W. S.

New Lens Eliminates Crane Shots. Film Daily, 58, Jan. 24, 1932, p. 7. This
lens, called the "Varo" is set normally to focus on a definite position, whereupon
various elements in the lens are moved in synchronism, the focal length changing
in smooth progression. Critical definition is claimed to be maintained at all
points. The iris diaphragm is operated by a cam at the same time as the lens
elements. The focal length can be varied from 40 feet to 120 feet. The normal
530

ABSTRACTS 531

focus setting is 150 feet to infinity. Supplementary lenses may be screwed into
the front for changing the focus to other distances. G. E. M.

Automatic Light Change Key Developed for Printing. Film Daily, 58, Jan. 24,
1932, p. 7. The light change is fitted with 28 keys, of which 24 may be set
according to the degree of light intensity. Any single key may be changed
quickly without changing the setting of the others. The change from one light
intensity to another is instantaneous, and the entire board may be cleared by
pressing the clearing key and turning a knob which cuts all exposure values out of
register. G. E. M.

New Portable 16 Mm. Sound-on-Film Projector. Mot. Pict. Herald, 106,
Jan. 23, 1932, p. 20. The machine consists of a projector amplifier unit and a
small loud speaker unit, each operated from any 110- volt, 50- or 60-cycle a-c.
circuit. The film used has sprocket holes along one side only, the sound track
occupying the other border area. A picture size, 52 inches wide by 39 inches
high, is recommended with a projection distance of 23 feet, though a larger picture
may be shown if desired. The exciter lamp is a 4- volt, 0.75-ampere lamp and the
amplifier contains one UX-868 photoelectric cell and the following tubes: one
UY-224, one UY-227, three UX-345, and one UX-280. A dynamic speaker of the
flat baffle type is used, having a volume capacity sufficient for a room of 10,000
cubic feet content. The projector amplifier unit is 14V2 inches long, 13 Y4 inches
high, 8x/4 inches wide, and weighs 43 pounds. The entire unit remains in the
case during use. G. E. M.

Make Talkers for School Use, British Educators Tell Studio. W. H. MOORING.
Mot. Pict. Herald, 106, Jan. 9, 1932, p. 22. A series of tests made on the value of
sound films for educational use in British schools is reported favorably com-
pared with silent film. Over 3500 students were included, of ages varying from
8 to 18 years, and more than 22,000 examinations were conducted by nearly
200 teachers. Fifteen Middlesex schools, including secondary, junior, and
senior institutions, were used for the test, which was under the supervision of the
National Union of School Teachers in collaboration with Western Electric, Ltd.
A future source of films is not known, however, and the likelihood of one is
considered rather uncertain. The detailed report is to appear later.

G. E. M.

Regulating the Acoustics of Large Rooms. E. PETZOLD. /. Acoust. Soc. of
Amer., Ill, No. 2, Part I, Oct., 1931, p. 288. A method of control is proposed,
whereby the acoustic characteristics of a room may be changed readily. The
"controllers" are triangular columns placed in front of the wall (or ceiling).
One side of each column is an absorbing surface, another a reflecting surface, and
the third a resonating surface, such as wood. Various arrangements of the columns
produce the desired effects. W. A. M.

Some Physical Characteristics of Speech and Music. H. FLETCHER. /.
Acoust. Soc. of Amer., Ill, No. 2, Part II, Oct. 1931, p. 1. Kinematic and
statistical descriptions of the physical aspects of speech and music are given in
this paper. As the speech or music proceeds, the kinematic description consists
in giving the principal melodic stream, namely, the pitch variation and also the
intensity and the quality variations. For speech and song, the quality changes
are principally described by giving, besides the main melodic stream, two second-
ary melodic streams corresponding, respectively, to the resonant pitches of the

532 ABSTRACTS

throat and mouth cavities. To this must also be added the positions of the
stops and the high pitched components of the fricative consonant sounds as
functions of the time. The statistical description consists in giving the average,
the peak, and the probable variations of the power involved as the various
kinds of speech and music proceed. These general ideas are illustrated by nu-
merous experimental data taken by various instrumental devices which have
been evolved in the Bell Laboratories during the past fifteen years. AUTHOR.

Vitaphone Develops Monitor Used on Set. Film Daily, 58, Jan. 10, 1932, p. 5.
An announcement of a new type of monitor desk which can be used directly on
the stage near the cameraman and the director. Previously, the monitor desk
has been located either in a separate room off the stage, or in a portable sound-
proof booth, on the stage. Greater flexibility is permitted with the new type of
desk. Technical details are not included. G. E. M.

BOARD OF ABSTRACTORS

BROWNELL, C. E. MACNAIR, W. A.

COOK, A. A. MATTHEWS, G. E.

CRABTREE, J. I. McNicoL, D.

HAAK, A. H. MEULENDYKE, C. E.

HARDY, A. C. MUEHLER, L. E.

HERRIOT, W. PARKER, H.

IRBY, F. S. SANDVICK, O.

IVES, C. E. SCHWINGEL, C. H.

LOVELAND, R. P. SEYMOUR, M. W.

MACFARI.ANE, J. W. WEYERTS, W.

ABSTRACTS OF RECENT U. S. PATENTS

The views of the readers of the JOURNAL relative to the usefulness to them of the
Patent Abstracts regularly published in the JOURNAL will be appreciated. Favorable
views are of particular interest. In the absence of a substantial body of opinion to
the effect that these Patent Abstracts are desired by the membership, their early dis-
continuance may be considered. If, after two weeks from the date of mailing the
April issue of the JOURNAL, no letters concerning the continuance of the depart-
ment will have been received, the Patent Abstracts will be discontinued.

1,830,173. Radio Television System. E. L. NELSON. Assigned to Bell Tele-
phone Laboratories, Inc. Nov. 3, 1931. Circuits for television transmission
and reception having a high degree of selectivity with minimum distortion. At
the transmitting station the weak photoelectric currents are greatly amplified
and used to modulate a carrier current of such high frequency that the distortion
of the side band frequencies in the antenna circuit becomes negligible. This fre-
quency may, for example, be 1500 kilocycles. At the receiver initial selectivity is
obtained by coupling a local circuit containing resistance to the antenna thereby
securing a widened resonance characteristic. The carrier frequency is then com-
bined with a current from a local source of frequency very much higher, so that
the resulting "difference frequency" is much higher than the received carrier.
This reduces the percentage width of the side band to such an extent that selec-
tivity may be obtained in tuned circuits, without undue distortion and at the
same time eliminates interference from the harmonics of the local source, which
are of such high frequency as to be harmless. The frequency of the local source
may be 6500 kilocycles, for example, in which case the difference frequency is
5000 kilocycles. After passage through highly selective circuits this latter current
is combined with current from another local source for producing an intermediate
difference frequency, which, after being selectively amplified, is detected to
produce the image currents. The frequency of the second local source may be
5120 kilocycles, giving an intermediate frequency of 120 kilocycles.

1,830,231. Mirror Disk for Television Systems. A. KAROLUS. Assigned to
Radio Corporation of America. Nov. 3, 1931. Scanning system comprising a
rotary mirror supporting element, a series of wedge-shaped support members of
graduated inclination rigidly secured upon the supporting element, and a reflect-
ing scanning surface rigidly secured to each of said wedge-shaped members.
Centrifugal and compressive forces set up upon rotation of the wheel do not effect
displacement of the members constituting the mirror.

1,830,239. Camera. F. H. OWENS. Assigned to Owens Development Corp.
Nov. 3, 1931. Lens turret for cameras, of either the ordinary "view" or "motion
picture" type. A fixed lens and a turret carrying a plurality of lenses of different
focal lengths are mounted so as to be capable of being selectively brought into
operative relation with the camera. The lenses of different focal lengths can be
independently brought into picture taking position.

1,830,537. Lamp Support for Projection Machines. L. S. FRAPPIER AND E.

533

534 PATENT ABSTRACTS [J. S. M. P. E.

BOECKING. Assigned to International Projector Corp. Nov. 3, 1931. An
auxiliary light source is mounted in the projector upon a rotatable support which
may be turned to bring the auxiliary light source into operative position upon
failure of the other light source. The structure is particularly applicable to a
mounting for the light source of a sound telescope which is used to pass continu-
ous rays of light through the sound record of a projection film.

1,830,538. Support for Light Sources. L. S. FRAPPIER AND E. BOECKING.
Assigned to International Projector Corp. Nov. 3, 1931. A plurality of light
sources spaced peripherally of a rotatable sleeve which sleeve is adapted to be
shifted angularly to bring either light source into position for directing light
through the moving film and sound telescope.

1,830,546. Synchronizing System. J. HERRMANN. Assigned to Siemens &
Halske Aktiengesellschaft. Nov. 3, 1931. A method of synchronizing which
utilizes in simple manner the frequency given by the ripple of the armature cur-
rent as carrier frequency for the transmission of, for instance, a control frequency
to a remote station. As the control frequencies in question in this case are fre-
quencies of the order of 50 to 150 cycles, they can no longer be transmitted over
telephone lines with intermediate repeaters. The invention is directed to the
method for transmitting synchronizing signals which comprises driving a picture
telegraph apparatus by a driving motor and producing from the driving motor a
slot frequency for use as a carrier frequency for synchronizing signals.

1,830,567. Safety Shutter for Cinematographs. A. SHAPIRO. Assigned to
Universal Stamping & Mfg. Co. Nov. 3, 1931. The shutter is adapted to move
automatically into the path of light to interrupt some of the light rays of the lamp
immediately upon the stopping of the light interceptor so as to permit the showing
of a "still" picture without injury to the film. A mechanism is provided operable
by the light interceptor for quickly moving the safety shutter out of the path of
light when the interceptor commences to rotate. The shutter is constructed for
dissipating much of the heat in the path of light. A rotatable clutch element is
connected to the shutter, and there is a pair of radially movable governor bodies
carried directly on the gear which connects to the light interceptor and is movable
into frictional engagement with the clutch element for rotating the shutter out of
the light path upon the rotation of the interceptor.

1,830,586. Transmission of Pictures. E. F. W. ALEXANDERSON. Assigned
to General Electric Co. Nov. 3, 1931. The picture receiver has a plurality of
channels tuned to respond to a different wavelength with a recorder connected
thereto and adapted to respond to all of the transmitted wavelengths. The re-
corder comprises a vibratory member having means for directing a beam of light
on a light-sensitive member, a screen having an opening therein arranged in the
path of the beam of light, a source of alternating current connected to the vibra-
tory member, and means for varying the current actuating said vibratory member
in accordance with the wavelength of the received signal. The different portions
of pictures are transmitted over the different channels and integrated at the
receiver.

1,830,596. Adjustable Mounting for Picture Projection Apparatus. A. DINA.
Assigned to International Projector Corp. Nov. 3, 1931. The plate for support-
ing the projection head is pivoted to the top of the pedestal. The plate for the
lamp house is mounted behind the projection head plate and secured thereto by a

April, 1932] PATENT ABSTRACTS 535

pantograph arrangement to insure continuous parallel relation between the axis
of the projection head and of the lamp house for both still and motion picture pro-
jection. An adjustable bracing device is provided to connect the base of the
pedestal and the lamp house plate, which allows lateral shifting of the lamp
house as well as adjustment of the angle of projection and also aids in imparting
extreme rigidity to the entire mounting structure during operation of the machine.

1.830.601. Sound Telescope. L. S. FRAPPIER AND E. BOECKING. Assigned
to International Projector Corp. Nov. 3, 1931. The sound telescope is rotatably
mounted in a suitable framework and provided with positive means for varying
the angular position of the telescope therein. The framework is mounted for
movement in a horizontal direction transverse to the axis of the telescope by
means of a suitable sliding bracket. The bracket itself may be moved horizon-
tally in a direction parallel to the axis of the telescope. In order to exclude ex-
ternal light from the photographic record, a pair of telescoping members are in-
cluded between the end of the telescope itself and the sound record and are pro-
vided with means for maintaining a positive engagement with both the telescope
and the film guide. A special light source is also provided which includes a pair
of lights and means for alternatingly bringing said lights into operative position.
A second light is accordingly always held in reserve and may be substituted in the
system without material interruption of service.

1.830.602. Distance Releasing Device for Moving Picture Cameras Driven
by a Spring Mechanism. E. GOLDBERG. Nov. 3, 1931. A camera wherein a
film driving or feeding mechanism is employed for moving the film strip in the
path of the camera lens, means being inserted for controlling the operation of the
mechanism as may be required by a user merely by simple adjustment of con-
veniently arranged control devices and, if desirable, allowing such mechanism to
be manually operated both for the photographing of motion and still pictures.
The camera is equipped for remote control of the film feeding mechanism whereby
the same may be started or stopped by a user in a manner which will permit said
user to position himself as a subject to be photographed and when so positioned,
effect elective operation of the camera.

1,830,637. Selector Filter. P. BROSSE. Assignor, by mesne assignments, to
Kislyn Corp. Nov. 3, 1931. A selecting filter for projecting goffered films in
colors, having a set of differently colored selector zones occupying its central por-
tion, and differently colored compensator zones at its opposite end portions. The
colors of the compensator zones are complements of those of the selector zones
which they touch so as to eliminate the noxious colors prevailing.

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.)

BOOK REVIEWS

Geschichte der Kinematographie. WILHELM DOST. W. Knapp, Halle, a.S.,
1925, 51 pages.

This history of motion pictures comprises the following chapters: a brief his-
tory of instantaneous photography; the beginning of "living pictures" from
series photographs to the projection of "living pictures;" further historical de-
velopment of motion pictures (1891-1895, approx.); the Kinematograph of the
Lumiere Bros. (1895-1897); and newer developments and technological improve-
ments (1897-1907). The treatise gives 116 literature references and mentions
151 authors and inventors. Of interest is the reference to the apparent motion of
"series" pictures as described by Titus Lucretius Carus, a Latin poet and philoso-
pher who lived 99-55 B.C. The literature references are principally to the
European literature, and only brief details of American and English work are
given. L. E. MUEHLER

Artificial Sunlight. M. LUCKIESH. D. Van Nostrand & Co., New York, N. Y.,
1930, 264 pp.

The biological effects of radiant energy have been coordinated with the physical
principles underlying the study of radiation in an interesting and concise form.
Much data, some new and some older, have been assembled in this volume in a
manner useful to the engineer, chemist, biologist, or physician.

Application of these data has led to the construction of the G. E. tungsten-
mercury arc lamp Sunlamp, the characteristics of which are fully described.

C. TUTTLE

536

SOCIETY OF MOTION PICTURE
ENGINEERS

OFFICERS
1931-1932

President
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE, Fox Film Corp., New York. N. Y.

Secretary
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Kodak Co., Rochester, N. Y.

Board of Governors

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada

H. T. COWLING, Eastman Kodak Co., 343 State St., Rochester, N. Y.

J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y.

P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y.

O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y.

A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New

York, N. Y.

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLB, Fox Film Corp., 850 Tenth Ave., New York, N. Y.

537

SOCIETY ANNOUNCEMENTS

STANDARDS COMMITTEE

At a meeting of the Standards Committee, held at the General
Office of the Society in New York, N. Y., on March loth, the
question of establishing the dimensional standards for projector
apertures was again considered, this time in respect to recommenda-
tions made recently by the Academy of Motion Picture Arts and
Sciences which resulted from the simultaneous study of the problem
by that organization and the S. M. P. E. The Standards Committee
unanimously agreed to recommend for adoption by the Society the
dimensions 0.600 by 0.825 inch as standard dimensions for 35 mm.
projector apertures, and the dimensions 0.631 by 0.868 inch for the
corresponding camera apertures.

The conclusions arrived at by the subcommittee on 16 mm.
sound-on-film, announced briefly in the March issue of the JOURNAL,
were accepted and approved, with some modifications. These plans
provide two layouts for the film, one involving a single row of per-
forations, the other, two rows. The former is to be recommended
for adoption as a dimensional standard by the Society, with the
suggestion that the latter layout, involving two rows of perforations,
be also published, but as a non-recommended specification. Both
layouts are now being detailed for presentation to the Society at the
Washington Convention.

The Committee also agreed to recommend as standard speed for
16 mm. sound-film equipment a speed of 24 frames per second, and
for the lead of the sound gate an interval of 25 frames.

PROJECTION SCREENS COMMITTEE

At a meeting of the Committee held on February 18th, definite
work was initiated in the various items on the program of the Com-
mittee's work for the year, and preliminary reports on some of these
items were presented. These items include the following:

(A) New developments in screens: (1) metal screens; (2)
screens with embossed surfaces; (3) other types.
538

SOCIETY ANNOUNCEMENTS 539

(B) Matters for standardization, in collaboration with the Stand-
ards Committee: (1) screen sizes; (2) illumination and methods
of measuring it; (3) definitions of brightness; (4) acoustic ratings
of screens; (5) optimum sizes for theater installation.

(D) Tolerable variation of brightness from point to point of the
screen; variation of brightness as a function of the location of the
viewer.

(E) Sixteen millimeter projection screens.

Another meeting of the Projection Screens Committee is to be
held prior to the Washington Convention for the purpose of drafting
the report to be presented at that time.

PROJECTION PRACTICE COMMITTEE

At a meeting held in New York on March 8th, further study was
made of the various tolerances, clearances, and tensions, as en-
countered both in new projectors and in those that have been in
service for some time, bringing nearly to its completion the com-
pilation of the table of tolerances, clearances, and tensions that the
Committee plans to present at the Washington Convention of the
Society. A preliminary draft of a report to be presented by the
chairman, Mr. Harry Rubin, at the Washington Convention,
dealing with the problems of the release print as they effect the
theater, formed the subject of considerable study and discussion.
Some of the items involved in this study are: (1) methods of "pro-
cessing" film; (2) buckling of film in projectors; (3) variations in
the density of prints; (4) film cutting for change-overs; (5) in-
accuracies in punching release prints.

The Society regrets to announce the death of one of its honorary
members,

GEORGE EASTMAN

on March 14, 1932. By action of the Board of Governors, Mr.
Eastman's name is hereby added to the

HONOR ROLL

of the Society of Motion Picture Engineers.

ARRANGEMENTS PROGRAM

SPRING MEETING OF THE SOCIETY, WARDMAN PARK HOTEL
WASHINGTON, D. C.

MAY 9-12, 1932, INCLUSIVE
COMMITTEES IN CHARGE OF ARRANGEMENTS

C. J. NORTH
N. C. HAEFELE
C. N. NICHOLS

C. FRANCIS JENKINS
RAYMOND EVANS
JAMES T. CORRIGAN
W. C. HUBBARD

WASHINGTON LOCAL COMMITTEE

N. D. GOLDEN, Chairman
C. FRANCIS JENKINS

RECEPTION

N. D. GOLDEN
NAT GLASSER
F. J. STORTY

W. C. KUNZMANN

RAYMOND EVANS
NAT GLASSER
JAMES T. CORRIGAN

C. J. NORTH
N. C. HAEFELE
J. C. BROWN
M. W. PALMER

CONVENTION REGISTRARS

H. T. COWLING
E. R. GEIB

W. C. KUNZMANN
S. RENWICK

HOSTESS TO CONVENTION

MRS. N. D. GOLDEN

assisted by

MRS. C. FRANCIS JENKINS Mrs. C. J. NORTH

MRS. RAYMOND EVANS MRS. NELLIE B. CORRIGAN

Miss EVELYN GLASSER

C. FRANCIS JENKINS
NAT GLASSER
C. J. NORTH

W. C. KUNZMANN

ENTERTAINMENT AND AMUSEMENTS

JAMES T. CORRIGAN
N. D. GOLDEN

BANQUET ARRANGEMENTS

W. C. HUBBARD, Chairman
F. C. BADGLEY

RAYMOND EVANS
N. C. HAEFELE
F. J. STORTY

N. D. GOLDEN

540

BANQUET— MASTER OF CEREMONIES

HON. W. P. CONNERY, JR.
Congressman, 7th District, Massachusetts

ARRANGEMENTS PROGRAM 541

SUPERVISORS OF PROJECTION EQUIPMENT, INSTALLATION, AND OPERATION

H. GRIFFIN, Chairman

JAMES FRANK, JR. NAT GLASSER

N. C. HAEFELE F. J. STORTY

Officers and Members of Projectionists Local No. 224, I. A. T. S. E., Washington,

D. C.

PRESS AND PUBLICITY

W. WHITMORE, Chairman

MEMBERSHIP

H. T. COWLING, Chairman

TRANSPORTATION, BULLETINS, AND RESERVATIONS

W. C. KUNZMANN N. D. GOLDEN RAYMOND EVANS

P. A. McGuiRE T. E. SHEA

NEW APPARATUS EXHIBIT

H. GRIFFIN, Chairman
JAMES FRANK, JR. SYLVAN HARRIS

Note: Manufacturers desiring to exhibit new equipment developed within the
past year should communicate with the Editor-Manager of the Society, 33 West
42nd Street, New York, N. Y. The exhibit will be held in the West Lobby of the
hotel, near the entrance to the Little Theater, where all technical sessions will
be held.

CONVENTION SESSIONS

All technical sessions and film exhibitions will be held in the Little
Theater, West Lobby of the Wardman Park Hotel.

BANQUET AND DANCE

The S. M. P. E. semi-annual banquet and dance will be held in
the Gold Room of the Wardman Park Hotel, at 7 : 30 P.M. on Thursday
evening, May 12, 1932.

Note: Banquet tickets and table reservations should be procured at the
registration desk up to noon of the day of the banquet. Tables will be arranged
for six or eight persons.

HOTEL ACCOMMODATIONS

The following special rates have been provided for members of the
Society by the Wardman Park Hotel.

Single room with bath $ 4.00 daily per person

Double room with bath 6 . 00 daily per person

Parlor and bedroom connecting with bath 10.00 daily and up

Note : Room reservation cards should be returned immediately to the Ward-
man Park Hotel in order to assure satisfactory reservations.

542 ARRANGEMENTS PROGRAM [j. S. M. P. E.

A modern fire-proof garage is located on the hotel property and a
special $1.00 rate per day (24 hour parking) has been arranged.

RECREATION

The Wardman Park Hotel management has arranged for golfing
privileges for our members at the Congressional and Indian Springs
Country Clubs. The usual course fee will be charged. The
S. M. P. E. identification card will entitle you to play at either of the
above country clubs during our Convention dates. The weather
permitting, the hotel outdoor swimming pool will be available to
the members. Regulation tennis courts are located on the hotel
property, and riding stables are within a short distance from the
hotel. Transportation can be arranged for sight-seeing tours over the
new Mt. Vernon highway to various points of interest about Wash-
ington. Arrangements for the trip should be made at the registra-
tion desk not later than the afternoon of May 10th.

TENTATIVE PROGRAM
WARDMAN PARK HOTEL

MONDAY, MAY 9th

The morning will be devoted to registration,

Committee meetings, etc.

11:00 P.M. Little Theater: Convention called to order.

Address of Welcome.
Response by the President.
12 : 00 to 2 : 00 P.M. Luncheon.

Reports of the Convention Committee, the

Secretary, and the Treasurer.
Committee Reports.
Consideration of Proposed Amendments of

Constitution and By-Laws.
Technical Papers, if time permits.

7 : 30 P.M. Little Theater: Social gathering and film pro-

gram of especial interest.

TUESDAY, MAY 10th

9: 30 A.M. Little Theater: Papers Program.

12 : 30 to 2 : 00 P.M. Luncheon.

April, 1932] ARRANGEMENTS PROGRAM 543

2 : 00 P.M. Little Theater: Papers Program.

7:30 P.M. Little Theater: Lecture and Film Program.

WEDNESDAY, MAY llth

9: 30 A.M. Little Theater : Papers Program.

11:30 A.M. Department of Commerce Building: Addresses by

by heads of government departments.
1 : 30 to 2 : 00 P.M. Luncheon at Department of Commerce Building.

2 : 00 P.M. Recreation and Sight-Seeing Trips.

7:30 P.M. Little Theater: Film Program.

THURSDAY, MAY 12th

9: 30 A.M. Little Theater: Papers Program.

12 : 30 to 2 : 00 P.M. Luncheon.

2: 00 P.M. Little Theater: Papers Program, Open Forum.

7: 30 P.M. Gold Room, Wardman Park Hotel : Semi-Annual

Banquet and Dance; an evening of frolic.
Adjournment of Convention.

Mr. O. M. Glunt, Chairman of the Papers Committee, promises a
most interesting program of technical papers, which will be listed in
the final issued programs. A reminder for your calendar: S. M.
P. E. Spring Meeting dates, May 9th-12th, inclusive; Wardman
Park Hotel, Washington, D. C.

Respectfully submitted,
Convention Committee

W. C. KUNZMANN, Chairman

W. C. HUBBARD

M. W. PALMER

Papers Committee

O. M. GLUNT, Chairman

544

SOCIETV SUPPLIES

JOURNAL BINDERS

[J. S. M. P. E.

April, 1932] SOCIETY SUPPLIES 545

MEMBERSHIP CERTIFICATE

Society )I(riion Picture Engineers

INCOBPOOATED

Society of Motion Picture Engineers

LAPEL BUTTONS

SUSTAINING MEMBERS

Agfa Ansco Corp.
Bausch & Lomb Optical Co.

Bell & Howell Co.

Bell Telephone Laboratories, Inc.

Carrier Engineering Corp.

Case Research Laboratory

DuPont Film Manufacturing Co.

Eastman Kodak Co.

Electrical Research Products, Inc.

Mole-Richardson, Inc.

National Carbon Co.

RCA Photophone, Inc.

Technicolor Motion Picture Corp.

BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS

No.

Price

1917 { I

$0.25
0.25

1918 7

0.25

1920 <

10
11

1.00
1.00

1921

12
13

1.00
1.00

1922

14
15

1.00
1.00

1923

16
17

2.00
2.00

1924

1925

1926

No.

Price

$2.00

1.25

1927

1928

1929

No.

29
30
31
32
33
34
35
36
37

Price

$1.25

1.25

25
50
50
50
50

3.00

\38 3.00

Beginning with the January, 1930, issue, the JOURNAL of the Society has been
issued monthly, in two volumes per year, of six issues each. Back numbers of all
issues are available at the price of $1.50 each, a complete yearly issue totalling
$18.00. Single copies of the current issue may be obtained for $1.50 each.
Orders for back numbers of Transactions and JOURNALS should be placed through
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and
should be accompanied by check or money-order.
546

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Volume XVIII MAY, 1932 Number 5

CONTENTS

Page

Western Electric Noiseless Recording H. C. SILENT AND J. G. FRAYNE 551

Recent Developments in Theater Loud Speakers of the Directional Baffle

Type H. F. OLSON 571

Gamma by Least Squares D. R. WHITE 584

Mechanical Advantages of the Optical Intermittent Projector . . J. L. SPENCE 593

The Mechanism of Hypersensitization . . .B. H. CARROLL AND D. HUBBARD 600
Advantages of Using 16 Mm. Supersensitive Panchromatic Film in Making

Medical Motion Pictures H. B. TUTTLE AND R. P. SCHWARTZ 609

Some Color Problems G. GEOGHEGAN 619

The Selenophon Sound Recording System P. SCHROTT 622

The Motion Picture Industry in Japan M. RUOT 628

A Machine for Printing Picture and Sound Simultaneously and Automati-
cally O. B. DEPUE 643

Time-and-Temperature vs. the Test System for Development of Motion

Picture Negatives W. LEAHY 649

Studio Projection and Reproduction Practice J. O. AALBERG 652

Size of Image as a Guide to Depth of Focus in Cinematography

J. F. WESTERBERG 655

Sound Recording for Independent Productions L. E. CLARK 659

Special Process Technic V. WALKER 662

Committee Activities:

Report of Studio Lighting Committee 666

Abstracts 676

Officers 680

Committees 681

Obituary — George Eastman 685

Society Announcements 687

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Board of Editors

J. I. CRABTREE, Chairman

L. DE FOREST A. C. HARDY F. F. RENWICK

O. M. GLUNT E. LEHMANN P. E. SABINE

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers.

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930, at the Post Office at Easton,
Pa., under the Act of March 3, 1879.

WESTERN ELECTRIC NOISELESS RECORDING*
H. C. SILENT AND J. G. FRAYNE**

Summary. — The Western Electric method of noiseless recording with the light
valve is described. The general principles are discussed, the circuit diagram is
explained, and the method of adjusting the device for service described. The photo-
graphic characteristics of film are considered, and their application in noiseless
recording is shown in some detail.

The realism of the talking picture is materially enhanced if the
showing in the theater is free from extraneous sounds that are not a
part of the scene shown. The steady grind of surface noise from disk
and film records in the past has given a mechanistic feeling to the
sound accompanying the pictures. The practical elimination of this
noise from film recording has probably contributed more to the con-
vincingness of illusion than any one step of progress that has been
made during the past two years. The result is that the finer shadings
of sound, whispers, and faint noises, once lost in a background
of mechanism, are now elements of reality for facilitating dramatic
presentation. The audience listens without effort; the medium by
which the sound is brought to them is all but forgotten ; the screen is a
stage whose illusion of reality finds its chief limitations in those of
photography.

Before going into a detailed description of the operation of noiseless
recording, the basic principles, according to which the method oper-
ates, will be outlined. It is well known that the noise output from a
light print is higher than that from a dark print when played on the
same fader step. Thus, if an unmodulated sound track be run through
a sound projector, the density of the sound track varying from, say,
that of clear film to extreme opacity, the greatest amount of noise
will be heard when the clear portion of the track is in the sound gate,
while the noise will gradually decrease as the dark portions come be-
fore the sound gate. However, by merely printing a sound track
dark, both the ground noise and the wanted sound are reduced in

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Electrical Research Products, Inc., Los Angeles, Calif.

551

552

H. C. SILENT AND J. G. FRAYNE [j. s. M. P. E.

approximately the same ratio, so that no improvement in the signal-
to-noise ratio results. Under former methods of recording, therefore,
there appeared to be no way out of the dilemma of effectively increasing
the range of sound output above the level of the noise inherent in the
film itself. This apparent difficulty has been successfully overcome
in the Western Electric system of noiseless recording.

In this system of recording it will be remembered that the exposure
on the negative film is made through a light valve whose ribbons are
normally spaced 0.001 inch, giving a certain fixed average density of
unmodulated track on the negative and, in turn, on the print. It is
apparent that this ribbon spacing is necessarily sufficient to permit
the movement required by the loudest sounds, and is therefore con-
siderably greater than necessary for the weaker sound. It is entirely
permissible to reduce considerably the spacing of the light valve

FIG. 1.

Schematic diagram of circuit used for varying the spacing of the
light valve ribbons.

ribbons during periods of no sound or of weak sounds, if in the presence
of louder sound the spacing is in some manner increased sufficiently
that the ribbons do not clash. Since this in no way interferes with
the amount in which the ribbons of the light valve move during the
presence of sound currents, it does not alter the change in light which
these sound currents cause at the film surface. If the film be properly
processed, the change in light which falls on the photoelectric cell
during reproduction is an exact picture of the change in light at the
film during recording. Therefore, the sound output from the photo-
electric cell will be an exact copy, without volume distortion, of the
sound input to the light valve, regardless of progressive changes of
ribbon spacing.

Thus in the new system of recording, the mean spacing of the ribbon
is not constant but is reduced to some predetermined value. This
reduces the density of the negative unmodulated track, and conse-

May, 1932]

NOISELESS RECORDING

553

quently increases the density of the positive unmodulated track, de-
creasing the ground noise. As modulation is impressed on the valve,
the mean spacing of the ribbon increases sufficiently to accommo-
date the further increasing input.

In the ideal case of noiseless recording, the light valve ribbons
should open just to, but never beyond, the spacing required to prevent
their clashing. However, due to more or less inherent circuit limita-
tions, the building up of the current that controls the opening of the
ribbons requires a certain length of time. Under these conditions,
clash might result at the beginning of sudden impulses if it were not
for the fact that an excess of current is supplied to the ribbons, causing

C NOISELESS RCDG.

B NORMAL RCDG.

A SPEECH INPUT

FIG. 2. Action of ribbons in recording.

them to open fast enough to prevent clash, even though they may
eventually open beyond the required amount in the presence of sus-
tained or steady sound. The excess of this opening over the required
amount we will call "margin," and if the ribbons be opened twice the
required spacing, the margin is considered as 6 decibels. The exis-
tence of this margin implies that when sound currents are present of
such a magnitude that the light valve would be loaded up to its normal
carrying capacity, sufficient current would be supplied to cause the
ribbons to exceed their normal spacing. This would result in an
excess of light passing through the ribbons, resulting in a negative
darker than normal and a print lighter than normal, both of which
would exceed the straight-line portions of the emulsion characteristic,
and photographic overload would result. The ribbons of the light

554

H. C. SILENT AND J. G. FRAYNE [j. s. M. p. E.

valve are prevented from exceeding the spacing that they would
ordinarily have in normal recording as described later.

Were the device capable of instantaneous operation, it would, of
course, be possible to reduce the margin theoretically to zero. Practi-
cally, of course, a small amount of margin is always essential. With
extremely fast operating systems, it has been found possible to reduce
this margin to as low as 2 decibels, although at the present time it is
recommended that a margin of 6 decibels be ordinarily employed.

As explained in the second part of this paper, there is a direct

VOLTS INPUT .
A. WITHOUT NOISE REDUCTION

VOLTS INPUT E CLASH

B IODB NOISE REDUCTION WITH ZERO MARGIN

FIG. 3. Ideal

.I56E .5C VOLTS INPUT ECUASH

C IODB NOISE REDUCTION WITH 6DB MARGIN

cteristics of light valve ribbon movement.

relation between the ndibe reproduced from the sound track and the
spacing of the light valve ribbons. Thus, for a given light valve
spacing, if this spacing ibe reduced to one-half, the noise from the
reproduced track will be reduced by 6 decibels. Since the normal
spacing of the light vawe ribbons is 1 mil (0.001 inch) the noise will
be reduced by 10 decibels if the spacing of the ribbons is reduced to
0.316 mil. At the present time this is the most generally used value
of noise reduction, although greater values of noise reduction can be
used. Thus, we see that the light valve is not entirely closed during
periods of very weak sounds, or in the absence of any sound at all.

May, 1932]

NOISELESS RECORDING

555

This failure entirely to close the light valve assists in the recording of
sudden sounds in that the ribbons, never being entirely closed, are
effectively provided with a certain margin against clashing until the
control system can function to open them. It is conceivable that
the ribbons might be brought to full closure if, between the point at
which energy is taken to operate the variable spacing device and the
light valve itself, there could be introduced a delay circuit which would
prevent sound currents from reaching the light valve ribbons until
these sound currents had first acted upon the rectifier, which, in turn,
had acted upon the ribbons to open them. Such a delay circuit
has been employed in certain other forms of voice-operated switching
devices employed in telephony. l Delay circuits are, however, expen-
sive, and, in view of the fact that it is not necessary to close the light

g 10

VOLTS IN 500— OCT.
10 OB N.R AND ZERO MARGIN

VOLTS IN 500" CCT
10 OB N R AND 6 DB MARGIN

FIG. 4. Measured characteristics of light valve ribbon movement.

valve ribbons completely, appear not to be justified in this type of
service at the present time.

In Fig. 1 is shown a schematic circuit of a device for varying the
spacing of the light valve ribbons so that they are always capable of
free vibration without clashing but, except on the very weakest of
sounds, are always kept at the minimum possible spacing. The ac-
tion of the circuit shown in Fig. 1 is essentially as follows: speech
currents are applied directly to the light valve in the normal manner
through the transformer and condenser placed between the light valve
and the main amplifier. A high impedance amplifier with adjustable
gain supplies the power necessary to operate the rectifier B, which in
turn controls the spacing of the light valve. The light valve LV is
strung, spaced, and tuned just as though it were being used for normal
recording, i. e., with a 1-mil spacing.

556

H. C. SILENT AND J. G. FRAYNE [j. S. M. P. E.

Current from the battery S is applied to the light valve through the
rheostat R and causes the light valve to be partially closed to the
required extent. When speech is supplied from the speech line, the
output of the rectifier B opposes the voltage of the battery 5 and
reduces the current in the light valve, allowing the ribbons to open.
Because of the action of the rectifier, which would normally transmit
to the light valve impulses of each half-wave of the speech currents, it
is necessary to interpose between the rectifier and the light valve a
smoothing circuit which will remove the minor variations of the
current and cause the light valve to follow the true envelope of the

--1

-16

-12

-8-4 04

DB INPUT

FIG. 5. Theoretical and measured curves similar to those in Fig. 3; A,
10-db. noise reduction with zero margin; B, 10 db., with 6-db. margin.

speech currents. Condenser C\, resistance F, and inductance L
provide the necessary filtering action on the output of this rectifier.
The resulting current will be of a very low frequency pulsating nature,
the peak value of its pulsations being proportional, of course, to the
strength of the sound currents received from the speech line. Since
the spacing of the light valve ribbons varies directly with the amount
of current in the ribbons, then the spacing of these ribbons will be
increased conformably to the variations in current received from the
rectifier. By properly regulating this action, we may make the spac-
ing of the ribbons of the light valve always just sufficient to permit
the movement that the sound currents received from the speech

May, 1932] NOISELESS RECORDING 557

line will require in these ribbons. Condenser Cz prevents the
transformer J\ from short-circuiting the biasing current supplied to
the light valve. The meter M indicates the biasing current at all
times.

As previously mentioned, in cases where an appreciable margin is
used in the set-up, when the speech currents exceed the value that
reduces the bias in the light valve to zero, there would ordinarily be a
reversal of current in the light valve, and it would overshoot or open
beyond the value corresponding to the normal spacing. In order to
prevent this, the anti-reversing rectifier at A is inserted, which
prevents reversal of the current through the light valve ribbons
and prevents their overshooting.

Fig. 2 illustrates the behavior of the ribbons in a normal light valve
and under the action of the circuit shown in Fig. 1. It will be seen
that in the normal method of recording, the ribbons have a constant
average spacing and their movement is essentially simple, correspond-
ing to the variations of the voice current only (Fig. 2B) . However, in
the method of noiseless recording, the ribbons may be regarded as
having two motions: first, the motion due to the voice currents only,
exactly as in the normal method of recording; and second, a super-
imposed slower movement which follows the envelope of these voice
currents. This is plainly illustrated in Fig. 2C.

A graphical analysis of the movement of the light valve ribbons
under steady-state conditions, with and without noise reduction, is
shown in Fig. 3. Fig. 3A illustrates the extreme limits to which the
ribbons move during normal recording without noise reduction. It
will be seen here that as the input to the ribbons is increased, the
extreme limits to which they travel is proportional to this input, and
that the clash of the ribbons occurs when the instantaneous minimum
becomes zero, i. e., the ribbons strike together. Simultaneously, of
course, their instantaneous maximum is double the mean. The mean
spacing, it will be observed, has remained constant.

Referring to Fig. 3-8, where noise reduction is applied, at low inputs
the mean spacing is reduced from the normal according to the reduction
of noise desired. Thus, if a noise reduction of 10 decibels is desired,
the spacing is reduced to 0.316 of the normal. In the ideal case, as
the input to the ribbons is increased, the minimum spacing of the
ribbons, due to their amplitude, decreases, while the mean spacing
remains constant until the ribbons are almost ready to clash, just as in
Fig. 3A. However, just before clash occurs, when the input is in-

558

H. C. SILENT AND J. G. FRAYNE [j. s. M. p. E.

creased, the mean spacing also increases just sufficiently to prevent
this clash until the mean spacing reaches the normal spacing,
at which time no further action takes place. It will be seen here that
the mean spacing of the ribbons in the absence of sound currents is
never reduced to zero, but is reduced by a ratio corresponding to the
reduction of the noise.

10 20

NEGATIVE EXPOSURE

FIG. 6. Curves showing relation between projected transmission and
negative exposure for two different printer light settings; gamma = 1.

If the system be set up to operate with margin, the ideal conditions
are shown in Fig. 3C. It will be noted that the mean spacing begins
to increase well before the ribbons are quite ready to clash, and since
this figure has been drawn for a 6-db. margin, the minimum spacing
to which the ribbons travel is never less than one-half the mean spac-
ing. When the normal spacing of the ribbon has been reached, no
further increase occurs. Actual measured steadv-state characteristics

May, 1932] NOISELESS RECORDING 559

under the above conditions with the existing noiseless recording equip-
ment are shown in Fig. 4, and are seen to agree closely with the corre-
sponding theoretical curves of Fig. 3.

In Fig. 5 have been plotted theoretical and measured curves similar
to those shown in Fig. 3, except that the mean spacing has been
plotted as the carrying capacity of the valve. Decibel scales instead
of current or voltage scales have been employed.

It has already been pointed out that in this type of recording, the
exposure through the light valve on the negative film is reduced during
periods of silence, while at the same time provision is made for in-
creasing the exposure automatically with increasing modulation of the
light beam by the light valve ribbons. It follows that the density
of the resulting negative sound track will be a minimum during silent
intervals, and will rise to a maximum value with increasing input,
while the density of the print made from this negative will be a
maximum during silent intervals, and will decrease to a fixed minimum
with increasing output from the film.

The question may be raised as to whether any volume or wave-
shape distortion is introduced into the sound reproduced from a
print made in this manner. To clarify these points we shall refer to
the curves in Fig. 6, showing the relation between projected trans-
mission and negative exposure for two different printer light settings,
the effective over-all gamma of the developing process being unity. In
this paper, over-all gamma is defined as the slope of the straight-line
portion of the curve obtained by plotting densities of a series of un-
modulated tracks, as measured by a photoelectric cell in the sound
reproducer, against the logarithm of the light valve openings through
which the exposures were made on the negative.

We shall consider only the straight-line portion of the curves in
Fig. 6, as the range of negative exposure must be confined to this
region if we are to have linearity between projected print transmission
and negative exposure, with a resulting undistorted wave-shape for
the print. In recording with a normal light valve, the mean exposure
is adjusted to the value En, which is the average of the upper and lower
exposures, EI and E2, of the straight-line portion of the over-all trans-
mission-exposure curve. In noiseless recording, the exposure of the
negative during intervals of silence is reduced to some predetermined
fraction of EM of value Eb, while the transmission of the resulting
point on the print will be reduced to some corresponding value Tb.
The carrying capacity, which may be defined as the maximum

560 H. C. SILENT AND J. G. FRAYNE [j. s. M. p. E.

permissible modulation of the print, is limited at this point to a
transmission modulation of amplitude Tb — 7\, but is automatically
raised to its normal maximum by the increase of exposure which
results from increasing the input to the light valve.

Referring to Fig. 6, it will be observed that the normal exposure En
of the negative will give mean transmission values of the print of 15
and 10 per cent, respectively, for the two curves A and B. If a sine
wave of exposure L is now made on the negative, a corresponding sine
wave of transmission will result on the print, the amplitude of the
latter being lower for curve B, which represents the darker print.
If the mean exposure is now reduced to Eb, the same sine wave of
exposure on the negative will give for curve A a sine wave of trans-
mission Mr of unchanged amplitude, even though the transmission of
the carrier gray has now been reduced to the same value as that
previously given for the darker print. It is apparent, therefore, that
for a constant printing light, so long as the negative exposure is at no
time reduced below the value EI, the amplitude of a transmission wave
resulting from a negative exposure wave will be independent of the
mean exposure on the negative and the resulting transmission of the
carrier gray of the print. This allows the signal volume to be main-
tained, while it permits the reduction in ground noise by decreasing
the transmission of the carrier gray of the print.

Since we have experimental evidence that the output of ground noise
from an unmodulated sound track falls off linearly with transmission
over the usual range of transmission used in sound reproduction, we
secure a reduction in noise output by this process similar to what would
be introduced into the signal output if the transmission of the carrier
of the print were reduced by printing, as shown in Fig. 6. This explains
why a movement up and down a definite over-all transmission-
exposure curve, as in Western Electric noiseless recording, results in
noise reduction without distortion of volume. This process should
not be confused in any way with that of the common practice of
controlling output by varying the transmission of the carrier gray
in the printing process. In this case, by alteration of the transmission
of the carrier gray, both signal and ground noise will be reduced in
the same ratio, and hence no effective reduction of noise is obtained.
It is accordingly apparent that any scheme, that starts from a negative
recorded in the normal manner and varies the transmission of the
print by controlling the printer light, will result in volume distortion
of the original sound and will not increase the signal-noise ratio.

May, 1932]

NOISELESS RECORDING

561

The equation of the straight-line portion of curve A in Fig. 6, may
be expressed as follows:

T = CE + K

Where C is the slope of the straight-line portion

.-. Ar = CAE
or

B A

AE =.|£N

5 10 15
PER CENT TRANS.

FIG. 7. ATVr plotted against T, for the
two curves A and B, of Fig. 6.

Now — is proportional to the percentage modulation of trans-
mission of the print, and we may calculate its value for any value of
T when AE is assigned a definite value. The curves of Fig. 7 show

— plotted against T for curves A and B of Fig. 6. AE in this case

has been assigned value corresponding to 10 per cent of the normal
exposure En in Fig. 6. While both these curves show that percentage
modulation of the print varies inversely with decreasing transmission
of the carrier gray for a given input to the valve, the two straight

562

H. C. SILENT AND J. G. FRAYNE [j. s. M. P. E.

AT
lines C and D show that the product— X T is constant throughout

the range of transmission allowed the carrier gray. This is consistent
with the usual requirement that the output be equal to the product of
the percentage modulation and the amplitude of the carrier, familiar in
radio electrical phenomena. Furthermore, since the output of the
photoelectric cell depends only on AT, the fact that this is indepen-
dent of the transmission of the carrier gray, and consequently of the
mean exposure of the negative, serves further to show there is no
volume distortion in the process in which the mean transmission of
the print is allowed to vary to accommodate increasing output.

7=50

1.2

1.0

1.6 1.8 30 2.2 2.4

.8 1.0 1.2 1.4
RELATIVE LOG E.
FIG. 8. Typical negative H & D curve.

The characteristic of the emulsion used in recording sound may be a
factor in limiting the amount of noise reduction permissible with this
system of recording. For example, for an over-all characteristic such
as that in Fig. 6, the theoretical limit is determined by the ratio of
the exposure En to the exposure Eit at which the straight-line portion
begins. Since the curvature at this point is due to the toe region of the
negative H & D curve, we may neglect the positive film characteristic
as a factor in determining the limits of noise reduction.

Fig. 8 shows a typical negative H & D curve in which visual diffuse
densities are plotted against log exposure, the exposures having been
made in a time-scale sensitometer equipped with a tungsten lamp.
We shall assume that this curve simulates the manner in which

May, 1932] NOISELESS RECORDING 563

exposures are made through the light valve on film passing through
the recording machine. The scale of this particular H & D curve is
approximately 20. If we follow the usual practice of making the
normal exposure 10 times the toe exposure, then we have a permissible
light valve modulation of 90 per cent without operating at the toe.
The shoulder exposure will not be reached before 100 per cent modula-
tion (double opening) of the valve is attained. For this particular
characteristic the normal exposure may be reduced to one-tenth its
normal value before entering the toe. This corresponds to a maxi-
mum noise reduction of 20 decibels without any change of exciting
lamp current.

The amount of noise reduction realized from a print will equal the
reduction made in the negative exposure only when the over-all gamma
is unity. Let us assume that a noise reduction of n decibels is desired.
This necessitates, for the ideal case, an equal reduction in exposure of
the negative. We have, therefore:

n = 20 log ^
or

JQ = log En - log Eb

where Eb is the reduced value of negative exposure. Referring to
Fig. 8:

Dn - Db = r (log En - log Eb) = T X ^
or

Dn = Db - T X jfr

If we call the corresponding projection densities of the print made from
this strip Dnl and Dbl we have the relation:

Dnl = zy + r x JQ

where T is the over-all gamma or slope of the line obtained by plotting
projection print densities versus log negative exposure.

The amount of noise reduction realized is given by the relation:

N = 20(ZV - Dbl) = T X n
.'. when r = 1, N = n

In general, therefore, it may be stated that the amount of realized
noise reduction, expressed in decibels, is directly proportional to the

564

H. C. SILENT AND J. G. FRAYN£ [J. S. M. P. E.

value of the over-all gamma of the processed print. It is highly desir-
able, of course, that the amount of noise reduction realized should agree
approximately with the amount expected, for otherwise the processing
will tend to impair the quality of the reproduced sound.

The fact that the negative film characteristic, rather than the
positive, limits the maximum attainable noise reduction is graphically
illustrated in Fig. 9. It is possible, with this combination of negative

POSITIVE H«
7= 2.0

LOG. EXPOSURE

PRINTER LIGHT

NEGATIVE H4.D CURVE
T».55

8 1.2

LOG. REL EXPOSURE

FIG. 9. Illustrating the manner in which the character-
istic of the negative, rather than of the positive, limits
the maximum attainable reduction of noise.

and positive H & D curves, to use printer light settings ranging from 1 1
to 21 in a typical printer without transferring any part of the straight-
line portion of the negative H & D curve into either the toe or shoulder
of the positive H & D curve. In ordinary processing, the upper
printer lights are seldom utilized, and projected densities seldom
exceed 2.4 in order to obtain the maximum output from a print
without resorting to the maximum electrical amplifications available

May, 1932] NOISELESS RECORDING 565

in the reproducing systems. This indicates that the density of the
biased unmodulated positive track is well below the initial shoulder
density of the positive H & D curve, and proves definitely that the
positive characteristic is not the deciding factor in limiting the
amount of noise reduction attainable.

While it is evident that in a properly processed print, no wave-shape
distortion is introduced into the sound output, the question may be
raised as to whether there is any relative loss of output at the higher
frequencies in sound reproduced from the darker portions of the
print. In order to test this, a 1000- and a 5000-cycle frequency record-
ing were made for various openings of the light valve, the input to
the valve being held sufficiently low so as to eliminate any possibility
of ribbon clash for the minimum spacing of the ribbons. These test
recordings were all printed at the same printer light, giving a print
with a wide range of densities. The results of this test are shown in
Fig. 10, where the relative difference in decibels between 1000 and
5000 cycles is plotted against the density of the unmodulated track
corresponding to each setting of the light valve. This curve shows
that the difference in output for 1000 and 5000 cycles remains
essentially constant over a range of print transmission corresponding
to 10 decibels of noise reduction. The relative loss at 5000 cycles
amounts to less than 3 decibels for a 14-db. noise reduction. These
facts indicate that noiseless recording does not produce any serious
loss of high frequencies.

The processing of noiseless recordings offers no peculiarly new
problems. The lamp current of the exciting lamp in the film recorder
is adjusted to give an exposure on the negative that will allow modula-
tions of the light valve of the order of 90 per cent, without operating
into the toe of the negative H & D curve. This is identical to the
method employed in setting the lamp current in ordinary methods of
recording, and is, in fact, made without regard to the fact that the
noiseless method of recording is being used. The development of the
negative film is carried out according to standard practice, the gamma
of the development being chosen so as to permit the attainment of an
over-all gamma of unity, for the particular contrast that is used in the
combined sound and picture print. In order to facilitate printing of
the negative and to act as a guide in setting the printer light, it is
customary to shut off the biasing current and to record a strip of
unmodulated track at the beginning of each roll of film or at the begin-
ning of each scene. The density or transmission of this track may

566

H. C. SILENT AND J. G. FRAYNE

[J.S.M.P.E.

then be used in the usual manner for determining the correct printer
setting required to secure a given transmission of the print.

It is desirable to keep a record of the printer light settings used in
making the daily prints, so that the assembled inter-cut negatives
for release printing, from which the guide densities have been cut
away, may be printed on the proper lights. It is desirable to insert a
strip of negative unmodulated track or other form of standard density
in the leader of each roll of assembled negative. This will give a strip
of normal carrier gray on the print, and a check on the transmission of
this strip will serve as a partial indication of the development of the

10 DB N.R.
I2DB N.R.

I4DB N.R.

5 10 15 20

PER CENT PROJ. PRINT TRANS.

FIG. 10. Results of tests, showing ratio of loss at 5000 cycles to
that at 1000 cycles, for various values of light valve openings.

positive. Even though a simple theoretical relation exists between
the densities of unbiased and biased unmodulated sound tracks, it has
not been found desirable as yet to rely upon the latter for setting
printer lights, as the difference in these densities is sensitive to
fluctuations that might be misleading.

We have assumed in this paper that the classical doctrine
of straight-line H & D recording has been adhered to. Since the
considerations of picture processing often make it desirable to have an
over-all gamma greater than unity, it is desirable to examine what
limitations this condition imposes upon noiseless recording.

D. MacKenzie2 has shown that for an over- all gamma as high as 1.4,

May, 1932]

NOISELESS RECORDING

56?

a relation may be obtained between the projected print transmission
and the negative exposure, which is essentially linear over a limited
range of transmission. The curvature that might be produced by
high over-all gamma in this print is partially offset by extending the
operations into the toe of the positive H & D curve. This introduces
a symmetrical curvature about the mean point, and introduces
similar distortion into both halves of the wave-shape of the projected
sound. Dr. MacKenzie concludes that with an over-all gamma as
high as 1.4, a noise reduction of 10 decibels may be safely attained, as
compared with a noise reduction of 14 decibels, which he considers

OVERALL «T = 1.7

1.0 1.5 2.0

LIGHT VALVE SPACING

5 10 15

PER CENT TRANSMISSION

FIG. 11. Illustrating how considerable curvature is introduced into the
low transmission region of the over-all curve, when gamma is considerably
greater than unity.

safe for classical recording. However, if the over-all gamma is raised
considerably above unity, as, for example, in Fig. 11, where it has a
value of 1.7, considerable curvature is introduced in the low trans-
mission region of the over-all transmission-exposure curve. A curve
of this sort introduces volume and wave-shape distortion, since the
output fails to increase proportionally with increasing modulation of
the negative.

A7 AT
Fig. 11 shows the percentage print modulation — and — X T,

or AT", plotted against T, for such a case of high gamma, A£ being
chosen as 0.05 En. It will be noticed that AT is no longer constant,

568 H. C. SILENT AND J. G. FRAYNE [J. S. M. P. E.

as was shown in Fig. 8. This shows that a film processed in this manner
will give decided volume distortion, the sound output decreasing with
a decrease in transmission of the print. This effect, combined with
the introduction of harmonics due to wave-shape distortion, will give
very poor quality of projected sound. In order to avoid entirely
volume distortion in this print, the valve spacing should not have
been reduced by the biasing current below 0.7 of its normal setting.
While this would mean setting for a 3-db. noise reduction, it would
give an apparent actual noise reduction of 3 X 1.7 or 5.1 decibels,
since the over-all gamma is 1.7 in this case.

While it is possible to obtain an approximately linear relation
between the print transmission and the negative exposure when the
over-all gamma is greater than 1, and obtain undistorted output for
ranges of negative modulation that are confined to the straight-line
portion of the over-all curve, it is desirable in practice to adhere as
closely as possible to the classical straight-line recording methods
with the over-all gamma equal to unity. This is especially to be
recommended for noiseless recording, as it permits the attainment
of a maximum of noise reduction without introducing distortion of
volume or of wave-shape, and makes it posible to obtain the full bene-
fit to be expected from this type of recording.

We have seen that in the Western Electric method of noiseless
recording, the exposure through the light valve is varied: first, accord-
ing to the voice currents in the usual manner and, second, according
to the envelope of these voice currents. These variations reduce the
transmission of the positive for low inputs and allow the transmission
to increase as the sound currents increase ; thus, when a film recorded
by this method is passed through a projector, the ground noise that
results from the film itself is low during intervals of silence of small
sound currents. As the transmission increases with increasing sound
currents, the ground noise will also increase ; but since the sound out-
put increases at the same time the signal-to-noise ratio remains
essentially constant, and the increase in ground noise is obscured by
the increase in signal volume. The net effect of this is to give an
apparent reduction of ground noise that is very real during intervals
of silence, when the ground noise is most objectionable.

We have seen that the processing of noiseless recordings is not
essentially different from that of normal recordings, as both kinds
require a linear relation between the projected transmission of the
print and the exposure on the negative through the light valve. It

May, 1932] NOISELESS RECORDING 569

has been seen that this condition exists when the effective over-all
gamma of the developing process is unity and, to a lesser degree, when
the over-all gamma is somewhat greater than unity. It has also been
pointed out that if considerable curvature exists between the print
transmission and the negative exposure, volume as well as wave-shape
distortion will be introduced, thus distorting the range of sound out-
put as well as introducing harmonics.

We have seen that the amount of noise reduction that can be
realized by this process is limited by the characteristic of the film
emulsion used in recording the negative, rather than in the positive
from which the sound is reproduced. It has been shown that in a
print made by the noiseless recording method, the loss at 5000 cycles
relative to 1000 cycles is of the order of 1 decibel for a noise reduction
setting of 10 decibels, a negligible loss of high frequencies, resulting in
no loss of brilliance in a print made with noiseless recording. From the
photographic standpoint, therefore, it may be stated that sound
recorded in this manner should be equally of as good quality as that
recorded in the normal manner and, in addition, will appear more
natural due to the virtual suppression of all spurious film noise.

REFERENCES

1 WRIGHT, S. B., AND SILENT, H. C.: "New York-London Telephone Circuit,"
Bell System Tech. Jour., VI (Oct., 1927), No. 4, p. 736.

2 MACKENZIE, D.: "Straight- Line and Toe Records with the Light Valve,"
/. Soc. Mot. Pict. Eng., XVII (Aug., 1931), No. 2, p. 172.

DISCUSSION

MR. JENKINS: When the noise reduction method is applied to glow-lamp
recording, is a similar procedure carried out, of superimposing a control current
on the speech current, thus varying the brilliancy?

MR. SILENT: Our experience with glow-lamp recording has been limited
to a few laboratory experiments. From our experience, the answer to that is
yes. There may be other methods.

MR. PALMER: We have often been told that the reason why we could not
obtain good reproduction in the theater was not because the sound was not
recorded on the film, but that we could not reproduce it. Mr. Frederick's
demonstration, reproducing sound from special hill and dale disk records, shows
that it is possible to obtain high quality reproduction from the speakers that we
have. The only two elements in the system by which film records are recorded,
that are not present in the methods used for recording on the disk, are that we
use a light valve to modulate the light and photographic emulsion to record the
sound. Now, why is it that the photographic emulsion cannot do the job as
well as the material of which Mr. Frederick's record was made?

570 H. C. SILENT AND J. G. FRAYNE

MR. SILENT: I may safely say that there is no loud speaker available to the
theaters at the present moment that is quite the equal of the one to which you
refer. The loud speaker generally used in the theaters cuts off approximately at
five thousand cycles. The film is, at the present time and with the present
improved technics, capable of accommodating a frequency range appreciably
greater than is at present being reproduced in the theaters, so that the
limitation of the range cannot be said to be due entirely to the materials used in
recording. The condenser microphones that have been used in the past for
recording constituted, themselves, a limitation in recording.

The moving coil microphone that is now used for recording will provide an
opportunity for the film to prove its recording capabilities. The film is, however,
subject to the limitation that when processed under the commercial conditions
existing in the studios and laboratories, a certain amount of high frequency
loss occurs. It is not possible to state definitely to what this loss should
be ascribed. We believe that the characteristic of the film may be made essen-
tially flat, to a frequency perhaps half an octave higher than is now usual in the
theater, and there is no question but that the film can be made to respond to
frequencies an octave above what was reproduced by the special disk to which
Mr. Palmer referred.

On the other hand, noise constitutes a limitation for the film at the higher
frequencies, and the response at frequencies in the neighborhood of twenty
thousand cycles or higher is considerably attenuated by printer losses and
diffusion in the film; and probably by other causes that enter into the processing
system.

RECENT DEVELOPMENTS IN THEATER LOUD SPEAKERS
OF THE DIRECTIONAL BAFFLE TYPE*

HARRY F. OLSON**

Summary. — This paper describes a group of directional baffle loud speakers that
are designed to combine high efficiency of transformation of electrical into acoustical
energy, with directional characteristics that are particularly adapted to large-scale
reproduction of sound with good fidelity. Three types of directional baffle loud
speakers have been designed, each satisfying a certain set of requirements. The
60-inch directional baffle covers an extremely large frequency range and is designed
for theaters with good acoustic characteristics. The 37-inch directional baffle is
designed to compensate for the acoustics of theaters with high reverberation char-
acteristics. The 25-inch directional baffle is designed for theaters in which the space
behind the screen is extremely limited. Response measurements show that the output
of these loud speakers is uniform over a wide frequency band. The uniform di-
rectional characteristics of these loud speakers eliminate the possibility of frequency
discrimination for points removed from the axis. These loud speakers, due to the
high efficiency and rugged construction, are capable of delivering large acoustic out-
puts without distortion.

INTRODUCTION

The transformation of electrical into acoustical energy may be
accomplished in a multitude of ways. At the present time, while
practically all loud speakers may be classed as of the diaphragm
type, the essential distinguishing characteristic lies in the coupling
between the diaphragm and the medium into which sound is to be
radiated, and in the method of driving the diaphragm. In general,
loss of coupling between the diaphragm and the medium occurs at the
lower frequencies. Among the common methods employed to
increase low frequency radiation from diaphragms are: (1) the use of
large diaphragms, (2) groups of diaphragms, and (3) various shapes of
baffles and horns.

To obtain the maximum of efficiency and minimum of interference
from reflecting surfaces, directional sound radiators have been almost

* Received December 14, 1931.
** Research Division, RCA Photophone, Inc., Camden, N. J.

571

572 HARRY F. OLSON [j. S. M. p. E.

universally employed for large-scale* reproduction of sound. The
directional characteristics of any acoustic radiating system are func-
tions of the dimensions, configuration, and phase of the elements.

The combination of high efficiency and directional characteristics
that is adapted to large-scale reproduction of sound precludes the
use of many possible types of sound radiating systems. An examina-
tion of the inherent characteristics of the directional baffle** type shows
that this loud speaker is particularly suited to satisfy these stringent
requirements. It is the purpose of this paper to describe a group of
directional baffle loud speakers that are designed to combine high
efficiency of transformation of electrical into acoustical energy with
directional characteristics particularly adapted to large-scale repro-
duction of sound with good fidelity.

GENERAL CONSIDERATIONS

A brief discussion of the functions of the essential parts of a direc-
tional baffle type of loud speaker will now be presented. A diaphragm
vibrating with constant velocity, coupled to an infinite tube, generates

pfTftFC^—WSW-

FIG. 1. Equivalent electro-acoustical diagram of dynamic cone and
acoustic impedance.

the same acoustic power at all frequencies. Assume that the dia-
phragm is a dynamic cone of mass m coupled to a tube of acoustic
impedance R (Fig. 1). If the mass m of the cone is chosen so that the
acoustic reactance of the cone is negligible compared with R for the
range in which we are interested, we will obtain a system that dissi-
pates the same power in the acoustic resistance at any frequency
within the range.

* The term "large-scale" reproduction of sound is used to designate the acoustic
powers necessary for reproduction of sound with good fidelity in auditoriums,
theaters, and open-air stadiums and theaters.

** The term "directional baffle" loud speaker has been used to designate a large
throat horn coupled to a cone driving unit.

May, 1932] DEVELOPMENTS IN LOUD SPEAKERS 573

For the infinite tube of constant cross-section we will substitute an
infinite tube of exponentially increasing cross-section. It has been
shown by Webster1 that the acoustic resistance at the small end of this
tube will be a constant at all frequencies above the cut-off frequency.
The cut-off frequency2 is determined by the rate of flare, and may be
located below the lowest frequency to be produced. If we now cut
this tube at some point along its length and terminate the open end in
air, the action will be altered, depending upon the cross-section of the
resulting mouth. If this cross-section is sufficiently large, very slight
reflection will occur at the transition from the mouth to the medium
(air), and the impedance presented to the cone by the tube will be
practically constant above the cut-off frequency. The system, as
before, will dissipate the same power into the tube for the frequency
range we have chosen; and consequently, neglecting slight reflection
at the mouth, will dissipate constant power into the medium for
this range. This system, consisting of a finite flaring tube of expo-
nentially increasing cross-section, coupled to a dynamic cone, essen-
tially constitutes the directional baffle type of loud speaker.

In the wave equation3 for the axial motion in an exponential horn
it is assumed that the phase is the same over a plane normal to the
axis of the horn. This condition is practically satisfied provided
that the cross-section is not greater than a wavelength. It has been
found experimentally that, for any particular frequency within the
transmission band, additional length of horn beyond a certain point
(the radius of ultimate impedance) does not affect the performance of
the horn. That is, the working portion of the horn decreases with
increase of frequency. Therefore, in a horn in which the axis is a
straight line, the condition of the same phase over a plane normal to the
axis is automatically satisfied.

To maintain the same phase over a plane normal to the axis in a
folded or curled horn is exceedingly difficult. The condition is
practically satisfied provided that the diameter at any bend is less
than the wavelength of the highest frequency reproduced. This places
a limitation upon the amount of folding or curling that may be
accomplished without impairing the horn action. If these conditions
are not satisfied, destructive interference will result, and, in addition,
certain portions of the horn will act as reflectors at the higher fre-
quencies. These conditions ultimately result in a non-uniform
response characteristic.

To obviate the occurrence of a non-uniform frequency character-

574

HARRY F. OLSON

[J. S. M. P. E.

istic due to folding, we have employed exclusively a horn with a
straight-line axis. In a horn with a large throat this objective may be
accomplished without making the horn excessively long.

The low frequency cut-off of a finite exponential horn is determined
by the rate of flare and the mouth opening. When the cut-off fre-
quency has been set, the mouth opening and the rate of flare are
fixed. There now remains one factor that determines the length of
the horn, namely, the throat area.

At this point we will digress to point out the limitations imposed
upon the size of the theater loud speaker. In motion picture theaters
in many instances the space behind the screen is limited; and in
theaters having a stage presentation in addition to the motion
picture, portability is a great factor. In view of the fact that the
space occupied by the loud speaker is an important factor, it is
essential that the loud speaker be as short as possible. To accomplish
this objective it is necessary that the throat be made as large as

Ma Me

CIRCU/T

FIG. 2. Complete equivalent circuit of dynamic cone.

practicable. The question then arises as to the proper driving unit
that will properly match the acoustic impedance at the small end of a
large throat horn. It can be shown theoretically, and has been
substantiated experimentally, that a cone type of unit can be designed
for a large throat exponential type of horn to yield high efficiency and
good fidelity of reproduction over a wide frequency range. This type
of loud speaker4 is now supplied with RCA Photophone equipment.
A specific discussion of the essential parts of this loud speaker follows.

SPECIFIC CONSIDERATIONS

The directional baffle type of loud speaker comprises a number of
essential elements, each of which has certain acoustical constants as
shown in Fig. 2.

May, 1932] DEVELOPMENTS IN LOUD SPEAKERS 575

(1) A large throat horn, Zi,

(2) A paper cone and voice coil, Mc,

(3) A box having a felt back, CB> RF,

(4) An air chamber between the cone and horn, C\.

By means of the equivalent circuit, Fig. 2, the magnitude of the
component parts may be adjusted theoretically to yield the greatest
efficiency and best frequency characteristic. It is beyond the scope
of this paper to give a detailed account5 of how this was carried
out. However, a brief discussion of the functions of the important
component parts will now be given.

1. The Plorn. — The horn used in this loud speaker is of the ex-
ponential type. The horn is a kind of acoustic transformer which
matches the acoustic impedance of the relatively heavy diaphragm to a
relatively light sound medium. Two factors influence the radiation
characteristics of an exponential horn, namely, the rate of flare and
the mouth opening. These two factors determine the low frequency
response of the loud speaker. Due to the characteristics of certain
auditoriums, it is desirable to reduce the low frequency response of the
reproducing apparatus. Therefore the average characteristics of
auditoriums will determine the rate of flare and mouth opening.
With the rate of flare and mouth opening fixed, the length of the horn
is determined by the dimensions of the throat. The size of the
throat, that will present a tolerable acoustic impedance to the cone
and, at the same time, will not impair the high frequency response
due to absorption along the walls or cause destructive interference in
the air chamber, has been found to be 4 X 4 inches. The impedance
of the horn at the throat is indicated by Zi (Fig. 2) .

2. The Cone Unit. The unit of this system consists of a paper
cone fitted with an aluminum wire voice coil. An air chamber
couples the area of the cone to the area of the throat of the horn.
The back of the cone is enclosed by a box with a felt back. The
response and dynamic characteristics of a six-inch cone were found
best adapted to this type of loud speaker. As will be seen from the
equivalent circuit (Fig. 2), in order to maintain uniform dissipation
in Zi, it is important that the mass of the cone and voice coil, repre-
sented by Mc, shall be small. This was accomplished by employing
an aluminum wire voice coil and an extremely light, rigid cone. The
non-uniform frequency response at the higher frequencies, commonly
encountered when light paper of great stiffness is employed, was
obviated by suitable corrugation of the cone. The air load upon the
back of the cone is represented by RB and MB.

576

HARRY F. OLSON

[J. S. M. P. E.

3. Cone Box. — The impedance presented behind the cone must be
considered. The air chamber behind the cone is enclosed by a box
having a felt back. The purpose of the felt is to absorb sound striking
it and thus prevent standing wave systems that would cause abrupt
changes with frequency in the impedance presented to the cone. At
the higher frequencies the absorption of the felt is unity and the sound
wave flows from the cone into the felt. At the lower frequencies
the absorption of the felt is not unity, and a stiffness due to the volume
of the cone box is presented to the cone. Therefore, the cone box is

PLAN

ELEVATION

PLAN

BAFFLE

0//?£C T/OWL

FIG. 3. Configurations and dimensions of the three types of baffle loud

speakers.

made large enough so that this stiffness will not reduce the response
of the loud speaker at the lower frequencies. The capacitance of the
cone box is represented by CB and the resistance . of the felt by RF
(Fig. 2). The actual size of the cone box and the felt cover is deter-
mined from this circuit.

4. The Air Chamber. — The purpose of the air chamber is to act as
a transformer between the area of the cone and the smaller areas of
the throat of the horn. To allow freedom of motion of the cone it is
necessary to space the cone from the face of the air chamber. The
volume of this air chamber results in a capacitance indicated by C\ in
the equivalent circuit (Fig. 2). This capacitance is in shunt with

May, 1932] DEVELOPMENTS IN LOUD SPEAKERS 577

the horn impedance. Therefore, to avoid impairing the frequency
characteristic of the loud speaker, the impedance of the air chamber
must be made large as compared with the horn impedance Z\. This
indicates that the spacing must be small. At the same time it is
necessary that the spacing between the cone and air chamber shall
be sufficiently large to allow full power output at the lower frequencies
where the excursion of the cone is large. It has been found that a
spacing of 1/s inch allows full power output of the cone at the lower
frequencies without impairing the response at the higher frequencies
due to the capacitive reactance of the resulting air chamber.

5. The Assembly. — Three types of directional baffle loud speakers
have been designed, each one satisfying a certain set of requirements
encountered in large-scale reproduction of sound. The essential
distinguishing characteristic in these three loud speakers lies in the
design of the horn. The general configuration and dimensions of
the three units are shown in Fig. 3. A discussion of the performance
and the application of these loud speakers will now be given.

RESPONSE AND DIRECTIONAL CHARACTERISTICS

At the present time, response and directional characteristics are
the best criteria of the performance of a loud speaker. The response*

* The response measurements shown in this paper were made using a micro-
phone calibrated with a Rayleigh disk. This gives the sound pressure in the
undisturbed sound field. For the diaphragm type of microphone, the pressure at
the face is twice that in free space at the higher frequencies. In addition, most
condenser microphones exhibit a resonance due to the cavity in front of the
diaphragm, which results in a further increase in response at the resonance fre-
quency. In general, in sound motion picture recording, the practice is to ignore
the greater response exhibited at the higher frequencies by the diaphragm type
of microphones, and to equalize the system for constant sound pressure at the
diaphragm. The argument often advanced in favor of this procedure is that it
overcomes transfer and other losses that occur at the higher frequencies. How-
ever, this is a rather weak argument, because the greater response exhibited by
the diaphragm type does not occur at the proper frequency to compensate for
these losses, and the net result is frequency distortion. If the response char-
acteristics were made with a microphone of the diaphragm type without correcting
for these effects, the loud speaker would show greater response at the higher
frequencies than actually exists. From the standpoint of the performance of the
loud speaker, the logical procedure is to measure the actual sound pressure in
free space and not the pressure at the diaphragm of a microphone, which, ob-
viously, will depend upon the size and geometrical configuration of the instru-
ment employed. One way in which this may be accomplished is to calibrate
the microphone by means of a Rayleigh disk.

578

HARRY F. OLSON

[J. S. M. P. E.

of this loud speaker was taken on the axis at a distance of 20 feet
from the mouth in an unobstructed6 medium, air. It is perhaps
needless to say that response curves made on loud speakers in rooms
have an extremely limited significance unless many curves are taken
and a careful analysis made to determine the influence of the room.

1. The 60-Inch Loud Speaker. — The response characteristic
(Fig. 4) and associated directional characteristics (Fig. 5) indicate
that the acoustic power delivered by this loud speaker does not show
any abrupt change with frequency. This is partially accomplished
by presenting to the cone an acoustic impedance that does not exhibit
abrupt changes with frequency. The uneven response sometimes

10*

10'

3 ^5*7

FIG. 4. Response-frequency characteristic of the 60-inch directional baffle

loud speaker.

encountered in cone type loud speakers has been eliminated by the
reduction in the mass of the cone and moving coil system, by suitable
processing of the paper cone, and by the load imposed by the horn.
As will be seen from Fig. 5, the directional characteristics are un4§
form over the range from 130 to 4000 cycles. Therefore, this loud
speaker will not produce frequency discrimination at points not on the
axis, an inevitable result in loud speakers exhibiting non-uniform
directional characteristics. The response characteristic shows that
the output is uniform from 100 to 7000 cycles, the maximum devia-
tion being 2.5 decibels.

2. The 37-Inch Loud Speaker. For theaters that exhibit high

May, 1932]

DEVELOPMENTS IN LOUD SPEAKERS

579

reverberation characteristics and other acoustic difficulties, it is
necessary to attenuate the low frequency response of the loud speaker
to obtain the most satisfactory results. The response characteristic
of the 37-inch loud speaker is shown in Fig. 6. As will be seen, the
response is attenuated below 300 cycles, falling to 13 decibels below
the 1000-cycle response at 100 cycles. In general, theaters that
exhibit acoustic difficulties show excess reverberation at the lower
frequencies. By using a loud speaker with a response as shown in
Fig. 6, the acoustic characteristics of the theater are compensated
for by the loud speaker, providing a reasonably uniform over-all
acoustic characteristic. The directional characteristics are quite
similar to the 60-inch directional baffle above 250 cycles.

AXIS SHORT AXIS

FIG. 5. Directional characteristics of 60-inch directional baffle loud speaker.

3 . The 2 5 -Inch Loud Speaker. — In many theaters the space behind
the screen is extremely limited, and in theaters having a stage presen-
tation, portability is an important factor. In these instances it is
desirable to fasten the loud speaker to the screen framework, so that
the entire assembly can be hoisted out of the way in a single opera-
tion. A loud speaker suitable for these conditions must be extremely
short. Again the inherent characteristics of the directional baffle
type make it possible to design a loud speaker that will satisfy these
conditions and at the same time retain high efficiency and good direc-
tional characteristics.

There are two general defects of speakers occupying small space;
namely, a deficiency in low response and non-uniform directional
characteristics. A brief discussion will now be given to show how

5SO

HARRY F. OLSON

[J. S. M. p. E.

these defects have been overcome in this loud speaker. This loud
speaker is not designed to operate as a single unit, but two or more
units must be used. Two units are termed a doublet, shown in Fig. 3.
A single unit would exhibit a deficiency in low frequency response.
However, by placing two or more units with their mouths close to-
gether, the impedance at the mouth of each speaker is increased due to
the greater pressure into which each speaker operates. If the distance
between the units is adjusted to match the acoustic impedance of the
individual directional baffle, the response can be maintained at low
frequencies. The response characteristic of the doublet is shown in
Fig. 7. It will be seen that response of this loud speaker is uniform

f\ r

V-

V/^

V-

10' /0J

FREQUENC?

FIG. 6. Response-frequency characteristic of the 37-inch directional baffle

loud speaker.

over a large frequency range. If greater low frequency response is
desired, it can be obtained by using more than two units. The
response of four units is indicated by the dotted curve (Fig. 7) .

The directional characteristics of a single unit of the dimensions
required would show considerable variation of the frequency range.
The directional characteristic of two or more units will be a function
of the directional characteristics of the individual units, the distance
between the units, and the angle between the axis of the various
units. A theoretical analysis was made to find the values of
these various factors that would make the result and directional
characteristic of the multiple units speaker practically independent
of the frequency. The results are shown in Fig. 8. It will be seen

May, 1932]

DEVELOPMENTS IN LOUD SPEAKERS

581

that the directional characteristics of the doublet are uniform in the
horizontal plane. Therefore, the doublet loud speaker could be used
in a single floor house without introducing frequency discrimination
resulting from non-uniform directional characteristics. In a theater
with a balcony four units are used, in which case the directional
characteristics are uniform for all planes.

EFFICIENCY

The efficiency of a loud speaker is the ratio of the sound power out-
put to the electrical power input. In this type of reproducer the
efficiency was determined by three methods: namely, theoretically;

<*5
c[ i r

•4VM/TS —

~*>'7

> V

•^-^

•s

NV-x>

^ in

' L-

2 I/A

({

•j

k—

1 ,

r<?3

7 ^

FIG. 7.

Response-frequency characteristic of the 25-inch doublet direc-
tional baffle loud speaker.

experimentally, by motional impedance measurements; and by the
total sound output determined from a calibrated microphone.

Using the equivalent circuit in Fig. 2, the ratio of the sound power
output to the electrical power input was computed from the constants
of the component parts. The results for the 60-inch directional baffle
are shown in Fig. 8. The theoretical efficiency cannot be predicted
by simple analysis above 2000 cycles because the mode of vibration
of the cone above this frequency is not that of a simple piston. Above
2000 cycles, the inherent stiffness of the cone reduces the effective
mass of the cone. For this reason the output of the cone is greater
than that of a simple piston. This is a desirable characteristic in
view of the fact that the acoustic output is increased.

582

HARRY F. OLSON

[J. S. M. P. E.

The motional resistance7 was determined experimentally, and the
efficiency computed in the customary manner. The results are
shown in Fig. 9.

The efficiency of the reproducer was also determined by measuring
the total acoustic output by means of a calibrated microphone,
comparing this to the electrical input. Pressure measurements
were made on the surface of a sphere with the loud speaker at the
center. The surface of the sphere was divided into elements and the
energy traversing each element determined. The summation of
the increments of the energy gives the total energy emitted by the
loud speaker.

FIG. 8.

Directional characteristics of 25-inch doublet directional baffle
loud speaker.

It will be seen from Fig. 9 that the results obtained from the three
methods are in close agreement.

The decrease in efficiency with frequency (Fig. 9) is not serious
when cognizance is taken of the fact that efficiency is proportional
to the square of the delivered pressure. For this reason, efficiency
expressed in per cent is an extremely sensitive measure of the per-
formance of a loud speaker. Expressed in terms that are more
descriptive from the standpoint of sound reproduction, the maximum
deviation is 3 decibels. As will be seen from the response and direc-
tional characteristics, the slight difference in directional characteristics
between the high and low frequencies, together with the above
efficiency characteristic, leads to a uniform response characteristic.

The high efficiency exhibited by this loud speaker, as compared

Vlay, 1932]

DEVELOPMENTS IN LOUD SPEAKERS

583

with cone loud speakers operating in a flat baffle, is due to the
iction of the large acoustic load of the horn upon an extremely
ight vibrating system. The uneven response, commonly encountered
vhen a cone of light paper of great stiffness is employed, is obviated
)y suitable corrugations. This is an extremely important factor
n that it reduces the effective mass of the cone at the higher fre-
quencies, accounting for the large output at those frequencies. The
arge load imposed upon the cone further assures uniform response
md high efficiency.

|00

^^<

*%^

; /

•-*

:^ ^ j

"^•^

"^ ;

^ 3
2

—

<?/r

1^.

...

/w;

7£?/t

/4^.

/tf

-,)/>AC^ *t£A

S69YKM

r/v/3

XXX

/»^/c

tOf

V£

iA3V/?£M£A

r 6 7

S9 * 3 -f -

to*

~ 6 7 £

J 3^5

b / ^

0"*

FIG. 9. Efficiency of directional baffle loud speaker.

This loud speaker, due to its high efficiency and rugged con-
struction, is capable of delivering large acoustic outputs (from one
to two watts of sound energy) without distortion.

REFERENCES

1 WEBSTER, A. G.: "Acoustical Impedance and the Theory of Horns and the
Phonograph," Jour. Natl. Academy of Sciences (1919), p. 275.

2 HANNA, C. R., AND SLEPIAN, J.:" "Function and Design of Horns for Loud
Speakers," Jour. A.I. E. E. (1924), p. 384.

3 WEBSTER, A.G.: Loc. cit.

4 Photophone Loud Speaker, P. L. 35, P. L. 43, and P. L. 49.

5 OLSON, H. F.: "A New High Efficiency Theater Loud Speaker of the Direc-
tional Baffle Type," /. Acoustical Soc. of Amer., Ill (1931), p. 485.

6 M ALTER, L.: "Loud Speakers and Theater Sound Reproduction," /. Soc.
Mot. Pict. Eng., XIV (June, 1930), No. 6, p. 611.

7 KENNELLY, A. E., AND PIERCE, G. W., "The Impedance of Telephone Re-
ceivers as Affected by the Motion of Their Diaphragms," Proc. Amer. Acad. Arts
and Sciences, 48 (1912), No. 6, p. 111.

GAMMA BY LEAST SQUARES*
D. R. WHITE**

Summary. — It has been found convenient and time-saving to compute gamma
by the method of least squares from data obtained by printing exposures through a
developed sensitometer strip as negative. The computation by this method reduces
to the addition of a set of numbers obtained from tables, one for each density involved.
The paper describes the method followed.

With the H & D printer1 that was developed for making tests of
positive film that would accurately represent printing conditions,
exposures of the positive are made through a negative, and consist of a
developed time-scale sensi tome trie exposure. A typical character-
istic curve of such a negative is shown in Fig. 1 . The densities of the
strips can be regulated by varying the exposure to obtain equal incre-
ments of density between steps, but this involves considerable trial
and error testing that has not been attempted. Instead, strips are
used which are exposed increasingly by factor two steps. The result-
ing negative densities do not, therefore, increase by uniform amounts
so that, when printed, the points of the positive characteristic curve
that correspond to these negative densities are correspondingly non-
uniformly placed on the log E axis of the curve. Plotting such data
proves laborious, time-consuming, and not as accurate as might be
desired, since two workers cannot always draw the same line through
the same group of experimentally determined points, as these points
do not, in general, lie in a mathematically straight line (Fig. 2.)

It was found that by applying the method of least squares the
problem was simplified to such a degree that gamma could be com-
puted very simply on an adding machine. The value of gamma thus
obtained required no plotting, and each person operating the adding
machine would obtain the same value of gamma from a given set of
points. On the experimental side, no requirement was introduced
except that in all tests the printing light intensity level must be ad-
justed to such a value that the densities, produced on the positive

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Du Pont Film Mfg. Co., Parlin, N. J.
584

GAMMA BY LEAST SQUARES

585

behind a chosen group of densities of the negative, are all on
the straight-line portion of the positive characteristic curve. This
group, and only this group, of densities enter into the calculation of
gamma. In practice, this condition is not difficult to realize, because
within the relatively narrow speed range of positive materials, a suit-
able exposure, when once obtained, will serve directly for all other
normal positive materials similarly processed.

The equation of the straight-line portion of the characteristic
curve, relating density, D, with the logarithm of the exposure, log E,
may be written

D = DO + 7 log E (1)

where D0 is the density corresponding to log E = 0, i. e., to E0 =
unity, and 7 is the slope of the straight line (Fig. 3). In general, any

1.5

1.0-

Effective
Density.

Printing
Characteristic
of Negative.

ft el. Log E of Meg. Exposure.

O J .6 .<j 1.2 l.r /.8 2.1 24

FIG 1. Typical characteristic curve of negative film.

observed density, Di} in the straight-line portion of the curve, ob-
tained from exposure log Ei will differ from the value of D calculated
from (1) above, due to experimental errors, by an amount

- D =

- (D0

7 log

(2)

Values of 7 and Do may be chosen from a series of observed pairs of
values (Di, log £,-) such that the sum of the squares of the differences
of the type indicated is a minimum. The expression takes the form
for pairs of points (Di, log EI) ; (D2, log £2) (DN, log EN):

(Dl - (Do + 7 log E<)]2 + [D9 - (D0 + 7 log E2)]2 +

+ [DN — (Di + 7 log EN)]2 = a minimum (3)

586 D. R. WHITE [J. S. M. P. E.

Written more briefly:

X [A — (Do + j log Et-)]2 = a minimum (4)

The minimum value of this summation is attained when, of all the
possible values of 7 and D0, those values are chosen for which the rate
of change of the value of the summation with small changes of 7
or Do is zero. Mathematically, this is stated by:

j- X) [A' - (D0 + 7 log Ei)]2 = 0 (5)

and

5^- Z I A - (A + 7 log Et)32 = 0 (6)

Performing the partial differentiations indicated, and simplifying
somewhat, there result

i = N

X [A - (D0 + 7 log £>)] log Et = 0 (7)

* = i

i-N

J2 [A - (D9 + 7 log Ei)] = 0 (8)

as two equations which maybe solved for 7 and DQ. Solving these
equations and writing

„ N

_____ _ _

(9)

/i = N \2

-( ^logE,)
\i=i /

N dog

(10)

7 = (K log Ei - M} A + (# log E2 - M ) A + (K log E3 - M ) A +

+ (tflogEjvr - Jlf) Z>AT (11)

where the expressions in parentheses can be calculated as soon as
the values of relative log E are known. Gamma is thus obtained as
the sum of a series of terms, one term for each exposure used on the
straight-line portion of the curve.

The nature of equation (11) is such that an error in determining the

May, 1932]

GAMMA BY LEAST SQUARES

587

density has greater and greater effect as it departs more and more from
the average of all densities observed.

In practice the values of the terms (K log EI — M) are calculated for
a given negative as soon as the effective printing densities are deter-
mined, since it can be shown that these expressions depend only upon
the relative exposures, not upon their absolute values. As an aid in
routine work, tables are prepared showing values of the product
(K log Ei — M) DI for the range of densities likely to be encountered in

2.0

/.S

1.0

0.5

Density

Problem-

From Ihest
Tb/nfs.

, LoQi E

1.2

1,5

FIG. 2.

Characteristic curve of positive, showing non-
uniform spacing of points.

the testing. After preparing these tables, the computation of gamma
becomes merely the operation of using the tables in order to find the
contribution toward gamma of any density occurring experimentally,
and adding the tabular values thus found. The term adding is here
used in the algebraic sense, as one or more of the factors (K log Ei — M)
will always be negative. Density is always positive, hence some
of the terms are negative and will be subtracted when finding the sum,
gamma, in the adding machine.

Values taken from tables calculated for the negative film of Fig. 1

588 D. R. WHITE [j. s. M. p. E.

TABLE I

Values of (K log Ei - M)D>

The body of the table gives the values of the terms of equation (11) calculated
for the positive densities indicated at the left and top of the table. The compu-
tations are based upon the negative of Fig. 1, of which, taking the exposure
through the fog area as represented by log EI = 1.50 (E is in arbitrary units),
the relative log E values are

1.50; log£2 = 1.41; log E3 = 1.23; log £4 = 0.98.

Positive
Density 0

.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

+D1 (Add)

1.8

2.52

2.54

2.55

2.56

2.58

2.59

2.61

2.62

2.64

1.9

2.66

2.68

2.69

2.71

2.72

2.73

2.75

2.76

2.77

2.0

2.80

2.82

2.83

2.84

2.86

2.87

2.89

2.90

2.91

2.1 2.93 2.94 2.96 2.97 2.98 3.00 3.01 3.02 3.04 3.05

+£>2 (Add)

1.7

1.40

1.41

1.42

.42

1.43

1.44

1.45

1.46

1.47

1.48

1.8

1.48

1.49

1.50

.51

1.52

1.53

1.54

1.55

1.56

a. 9

1.56

1.57

1.58

.59

1.60

1.61

1.62

1.63

1.64

2.0

1.65

1.66

.67

1.68

1.69

1.70

1.71

1.72

-A

(Subtract)

1.3

0.41

0.42

.42

0.42

0.43

0.44

1.4

0.44

0.45

.45

0.46

0.47

1.5

0.48

.48

0.49

0.50

1.6

0.51

0.52

0.53

0.54

-A

(Subtract)

0.7

1.33

1.35

1.37

1.41

1.43

1.44

1.46

1.48

1.50

0.8

1.52

1.54

1.56

1.60

1.62

1.64

1.65

1.67

1.69

0.9

1.71

1.73

1.75

1.79

1.81

1.82

1.84

1.86

1.88

1.0

1.90

1.92

1.94

1.98

2.00

2.02

2.03

2.05

2.07

For the set of positive densities of Figs. 2 and 3;

A = 1.87; A = 1.75; A = 1.39: A = 0.84;
the gamma computation is

Corresponding Term
D from Table

Di = 1.87 +2.61

D2 = 1.75 +1.44

A = 1.39 -0.44

A = 0.84 -1.60

Therefore 7 = 2.01

May, 1932]

GAMMA BY LEAST SQUARES

589

are given in Table I. These are arranged for use when the fog area
and the three lowest densities of the negative are printed on the
straight-line portion of the positive. Using the densities plotted in
Fig. 2, the gamma is obtained by adding the four tabular values as
shown just below the table.

This application of the method of least squares has been an
outstanding aid, and a time-saver in obtaining gamma from the prints
through a negative sensitometric strip. The saving of time is not as

2ft

1.0

I ,\+

.9 1.2

FIG. 3. Construction used for
deriving equation (1).

important in cases when the values of log E representing the exposures
are separated by equal intervals on the log E axis. For such cases,
where 6 is the uniform interval of separation for the log E's, the equa-
tion for 7 reduces to

for 2 exposures ( 12)

for 3 exposures (13)

- * i £ _ * _ 4
7 ~ 108 + 105 105 105

for 4 exposures

(14)

590 D. R. WHITE [j. S. M. P. E.

* w* m t

All of these equations are written to make gamma positive when DI
is the highest density of the series, DI ..... DN being successively
decreasing densities. Equation (12) is merely the algebraic solution
for the slope of a line passing through two points separated by the
amount d on the log E axis and by the amount (Di — D2) on the
density axis. The other equations give gamma as the sum of a series
of terms.

The calculation of the term Do of equation (1) would follow similar
lines, but this is of relatively little interest by itself since it is not a
constant customarily used in the discussion of H & D curves. The
point of intersection of the straight line and the log E axis (call it log
Eo) is frequently of interest, as it is associated with the H & D speed
of the emulsion. This point of intersection may be found from equa-
tion (8) above by making the substitution (Fig. 3) :

D0 = -7 log E0 (16)

which follows as a consequence of the fact that DQ is the intercept of
the straight line on the density axis and log JEo its intercept on the
log E axis. There results from this substitution:

i = N

£ (Di - 7 (log Et - log Eo)] = 0 (17)

» = i

from which results:

log Eo = av. of the log E»'s — (av. of the ZVs)/7

this is not quite as simple to solve, though it offers no great difficulties
because the first term may be calculated for all tests of the negative
as soon as the log E's are determined, and the second term is obtained
by adding the densities occurring in any test and dividing by 77V,
leaving only the subtraction of the two to give log EQ. In closely
related tests in which only relative values of log E0 are desired, only
the last term needs be calculated, since the first is constant when the
source of light and the negative are maintained the same.

May, 1932] GAMMA BY LEAST SQUARES 591

Carrying out the computation of equation (18) for the positive
test shown in Fig. 3, we have:

N = 4; 2 log E's = 1.50 + 1.41 + 1.23 + 0.98 = 5.12;
2 D's = 1.87 + 1.75 + 1.39 + 0.84 = 5.85; and
7 = 2.01 (as previously computed), yielding

5.12 5.85

log E0 = — - 4-- = 0.55

Although this locates the point of intersection of the straight line
with the log E axis, it does not appear in practice to give an exact
measure of the speed of the emulsion as it is judged in picture work,
and hence is of less importance than gamma.

REFERENCE

1 WHITE, D. R.: "Two Special Sensitometers," J. Soc. Mot. Pict. Eng.,
XVIH (Mar., 1932), No. 3, p. 279.

DISCUSSION

MR. CARLTON : Quite often a point occurs considerably out of line, perhaps
due to a temporary thickness of the emulsion, or some idiosyncrasy of develop-
ment. And while a method of this kind may be applicable for a man who does
not understand the technology of films, I think he is liable to err considerably
by using a formula of this type.

All that is required to determine the gamma or contrast of a set of plots is
to line up a piece of transparent film, and to read the gamma directly. Using
this method it is easy to see whether something is out of order at this point or
that. But if merely the reading is taken for the density, the fact that that
point is out of place will not be noticed. And as Mr. White has derived his
formula from the departures of these points from the mean, the one point in
question might be large enough to cause a considerable variation in the results.
Has it occurred that after determining the gamma by this method, one point
was later found to be greatly off?

MR. WHITE: That occurs, of course. There are two answers to the question:
First, that a person using this method can, after some little experience, very
readily determine by inspection whether the densities are reasonably correct
and whether there are large errors.

Second, as will be noticed by the form of the equation, if one point (indicating
equation) is off, it would not appreciably affect the value of gamma, even for
large errors.

Small errors do, of course, occur; but if a plain blunder occurs in making a
reading, or some unknown things disturb the reading by one- or two-tenths,
as sometimes inexplicably occurs, such extreme cases usually can be detected
by the person doing the work; by glancing over the data and by having a "feel"
for it, as it were. When the differences are small, it is an open question as to what
may be the best method of making the determination ; whether to take the three
points that lie on the line and ignore the fourth, which does not; or to draw the
line between the points, throwing some errors one way and some the other.

592 D. R. WHITE

The mathematical method assumes the line drawn between the points, giving
value to all of them.

It is my belief it is best to assign equal weights to all the points; in other
words, to let the deviations fall where they may in computation, so long at
large blunders have been avoided.

MR. MACNAIR: A method is known, of making the least square adjustment
of a number of points in what might be called a purely mechanical way. All
that is needed is a large drawing board, on which the line is drawn on a large
scale, a rubber band being used for each point. To imitate the line a bar is used.
According to a well-known method of laying out the points and using the rubber
bands, the bar can be suspended in such a way that it will automatically assume
the least square relation with all the points. If a large amount of such work is
to be done, it is easy to train a young laboratory man to do it by this means in
five minutes, for one set of points.

MR. WHITE: There are several such ways of obtaining the same results,
but the computation of gamma by the method described here requires much
less than five minutes. However, I am glad that you called attention to the
mechanical methods of doing such work.

MR. CARLTON: I assume that this method is being used for evaluating
gamma on a positive development machine. Why can we not make our exposed
strip in the same way and have an automatic milliammeter on the end of the
machine? As the exposed strip comes out, either before or after it is dry, a stand-
ard curve could be drawn, in terms of transmissions, of course, rather than in
terms of density. A person used to the technic of reading these curves would
soon be able to tell what the degree of development was, in terms of transmission,
instead of in terms of density, because, after all, gamma is only a superficial
quantity that has been established for the purpose of interpreting what is going on.
With a very little development work, it ought to be possible to design a recorder
that would automatically record the degree of development of each roll, so that a
permanent record would be produced that could be inspected the instant the
end of the roll was reached. No calculations would be required afterward, so
that the technician could immediately determine whether or not the developer
was exhausted and needed replenishing.

MECHANICAL ADVANTAGES OF THE OPTICAL
INTERMITTENT PROJECTOR*

J. L. SPENCE**

Summary. — The mechanical problems of optical intermittent projectors, related
to threading of film, gate structure, running time per reel, operating characteristics,
and difficulties more intimately connected with sound film itself, are briefly described.
In addition, some attention is given to the economic problems involved in film damage
and wear, in the wearing of machine parts, in the use of duplex film on thinner film
base. The paper refers to these problems specifically in relation to 16 mm. machines.

No continuous projector can hope to improve upon the screen
picture produced by an intermittent machine. In the intermittent
projector, the ideal arrangement of a stationary image and lens is
found. With a machine of the non-intermittent type providing
definition the equal of that obtained with present-day projectors, the
inventor could forget his old arguments (correct or otherwise) of
more light on the screen and less eye-strain in view of the many
mechanical advantages.

The mechanical and economic advantages are arguments which
will enable the continuous type of projector to take its place as
an equal in the professional field and a superior one in the amateur
16 mm. field. In the latter case, the addition of sound to the 16 mm.
film will emphasize the inherent advantages of this type of machine.
The questions of cost and simplicity of operation will undoubtedly
determine the extent of its use in this field. It is not the initial
expenditure made for the projection equipment, but rather the
expense involved in obtaining various films for home projection,
that causes the amateur machine to remain upon the shelf.

The intermittent mechanism which complicates film threading
and which requires pressure at the gate, is a major factor responsible
for film damage. This film damage is one of the important causes of

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Akeley Camera, Inc., New York, N. Y.

593

594

J. L. SPENCE

[J. S. M. P. E.

the high rentals which must be charged the user in order to assure
the film exchange a proper return on its investment. With a projector
of the optical intermittent type, film damage would be greatly de-
creased, permitting a considerable reduction in the rental fee, which,
in turn, would conduce to more extensive circulation of films, to
the advantage of all concerned.

Experimentation now being carried on with 16 mm. sound-on-film
is proving the desirability of a machine of the non-intermittent type.
Indications are that satisfactory results are obtainable from a sound
track, the speed of which is reduced 2x/2 times, when care is taken in
designing the film propelling and guiding mechanism. When this

FIG. 1. Comparison diagrams illustrating the rela-
tive complexity of threading the mechanical and the
optical intermittent projectors.

sound reproducing mechanism is incorporated in a projector of the
present intermittent type, certain undesirable features are found to
exist.

The fact that the film must move uninterruptedly for sound re-
production but intermittently for the picture, means that there
would always be necessary two film feeding systems. In an optical
intermittent, the sprocket is already propelling the film correctly
for the sound in drawing the film for the picture, and therefore one
sprocket will perfectly fit both requirements.

The intermittent motion and the "slapping" of the film loops pro-
duces vibration, which is amplified in light, portable machines and is

May, 1932]

INTERMITTENT PROJECTOR

595

apt to cause interference in properly reproducing sound from film
or disk. In an optical machine, no loops are required and, of course,
no intermittent exists.

The intermittent motion and the film loops cause noise. In most
16 mm. installations, the projector is unenclosed, and the noise may
be very objectionable. In an optical machine, the gearing can be
completely encased, eliminating by this means all noise other than
the slight hiss of the film in its smooth journey from one reel to the
other.

A speed of 24 frames per second or greater causes strains in the
present-day light-weight 16 mm. mechanism. As the running speed

FIG. 2.

Sprocket for driving either single- or double-
width film.

is increased, the gate pressure on an intermittent machine must
necessarily be increased also. In the continuous machine, the speed
is practically unlimited and can be obtained by merely increasing
the motor speed without other adjustment.

Pressure is required at the gate of an intermittent machine to stop
and hold the film. The resulting film friction is one of the principal
detriments to a long film life, as is universally recognized. No
pressure is required at the gate in an optical machine, flatness being
secured by virtue of an arcuate form. With the film running on the
celluloid surface, damaging of the emulsion can never occur.

Complexity of threading is exemplified in Fig. 1. In an inter-

596

J. L. SPENCE

[J. S. M. P. E.

mitten t machine, it is necessary to form a loop above and below the
pull-down mechanism. To obtain this loop the film must pass over
a sprocket before and after passing the gate and the intermittent
mechanism, resulting in an average of nine operations for threading.
The addition of another sprocket, which would be necessary for sound-
on-film, adds at least four more operations. Improper threading by
the operator in any of these operations could result in severe film
damage. In some projectors, sprockets A and B are combined into
a single sprocket. The sprocket H, acting as a hold-back sprocket,
might also be eliminated in practice, but at present it does not seem

FIG. 3. Method of feeding film having standard or
doubled perforations.

advisable to do so. In a non-intermittent machine the rollers
C and D could be eliminated, and the film could pass directly
over the sound and picture gates. No doors or pressure shoes
would be necessary, as the film would remain true owing to the
arcuate form. The film then passes around the sprocket and over
the idler roller, which may be stationary and separated a sufficient
distance from the sprocket teeth to permit fitting the film on the
sprocket. It is not necessary in threading this type of machine to
see that the holes of the film engage the teeth of the sprocket. The
starting of the machine will cause proper engagement immediately.
It should be noted that in Fig. 1, horizontal film reels are shown.

May, 1932]

INTERMITTENT PROJECTOR

597

(See projection lamp position.) This is not, of course, imperative;
nor is the design solely applicable to optical intermittent machines.
Experimentation will prove that the horizontal reel has many ad-
vantages over the vertical type. The operator will find it easier to
thread a reel lying in a horizontal plane, and the reels may be single-
flanged, greatly facilitating threading. A uniform take-up tension can
readily be designed, depending for its action on the increased weight

FIG. 4. Double-width film having doubled
and staggered perforations, for increasing the
perforation frequency.

of the film roll. With this type of take-up governed by weight, a
substantially constant drag is maintained from start to finish. A
Porro prism will, of course, be necessary to erect the image for the
screen, but the loss of light due to the prisms is a small price to pay
for the many advantages derived from the horizontal reel.

In using a single sprocket in a non-intermittent machine, it will
be found desirable to have the pull on the film in the direction of
the supply reel greater than the pull in the direction of the take-up

598

J. L. SPENCE

[J. S. M. P. E.

reel. This will result in the sprocket teeth always being in contact
with one side of the film perforations, eliminating the tendency to hunt.
Another desirable, and perhaps necessary, feature will be an in-
crease in the running time per reel. At a speed of 24 pictures per
second, the present 400-ft. roll gives a showing of only eleven minutes.
A considerable increase in capacity is possible in three ways:

(1) Increased roll diameter (easily practicable when the hori-
zontal film reel and weight-actuated take-up are used).

(2) A double-width film carrying two rows of images, as shown in
Fig. 2, one row to be projected first, the reel reversed, and then the
second row projected. This provides the added advantage of having
the rewinding done automatically.

(3) A decreased thickness of base. Not entirely practicable with a
claw-type intermittent, due to the shock encountered at the per-

TABLE I

Comparison of Intermittent with Non-Intermittent Projectors

Items

Sprockets or sprocket con-
tacts

Rollers or equivalent

Vibrating film loops

Intermittent claw or sprock-
ets

Pressure gate or shoe

Noise

Vibration

Threading
Wear on film
Wear on machine
Sliding contact of film emul-
sion
Oiling

Speed

Take-up tension on film
Possibility of threading error
Possibility of film damage
Possibility of using thinner

film base
Application of duplex or

double- width film

Intermittent Projector
3 Minimum

2
1

Disturbing in portable
projector

Apt to affect sound re-
production

Intricate

Considerable

Maximum

At gate

Necessary at a number of

points
Limited

Maximum to minimum
9

Great
Not very good

Inconvenient

Non-Intermittent
Projector

None
None

Insignificant

Simple
Minimum
Minimum
None

None

Unlimited

Constant minimum

Small

Excellent

Simple

May, 1932] INTERMITTENT PROJECTOR 599

f oration edges. There is reason to believe that a film 3l/2 mils in
thickness will be found desirable and practical in the optical inter-
mittent machine, which will give an increase of more than 50 per cent
in the running time per given roll diameter. It appears doubtful
that the thin 16 mm. film can be handled by the laboratories in
processing, owing to its tendency to curl. However, the same film
could be used if the width were increased to 32 mm., accommodating
two rows of pictures. The use of double-width film would impose
greater demands on the present type of 16 mm. intermittent machine,
which is already heavily taxed at a speed of 24 pictures per second.

An important difference between 16 mm. sound-on-film and
35 mm. is the difference between the perforation frequencies, the
conditions being much more favorable in the latter case. It may be
found that the frequency must be increased in the case of 16 mm. film,
thus necessitating additional perforations per picture. In Fig. 3 is
shown a film having twice the number of perforations. Obviously, a
special sprocket with a greater number of teeth would be required,
and in the case of the continuous projector, such a sprocket could be
incorporated as shown, without interfering with the use of standard
perforated film. Fig. 4 is an elaboration of this idea, providing for
the same perforation frequency as the present 35 mm. film.

Table I clearly points out, in a comparative manner, the character-
istic features of the two types of machine.

In conclusion, it is believed that recent developments, which have
taken place, point to the appearance, in the near future, of a 16 mm.
projector of the optical intermittent type.

DISCUSSION

MR. COOK: Will Mr. Spence tell us something about the possibilities of the
non-intermittent machine which he mentioned last?

MR. SPENCE: It is a machine that realizes the possibilities of the non-inter-
mittent machine, as outlined. I hope and feel confident that in the near future we
shall have a machine that will project a picture on the screen the equal of those
that are now projected by intermittent machines.

MR. RICHARDSON: What do you mean by "optical intermittent"?

MR. SPENCE: An optical intermittent projector is one in which the film moves
continuously, the successive pictures being rendered stationary on the screen
by an optical mechanism. The difference between an optical intermittent
projector and a continuous projector would be: In continuous projection the pic-
tures fade one into the other, whereas, in the other case, dark spots occur
between pictures, equivalent to the effect produced by a shutter but not of as long
duration as in intermittent projection.

THE MECHANISM OF HYPERSENSITIZATION*
BURT H. CARROLL AND DONALD HUBBARD**

Summary. — Ammonia treatment of photographic materials used for hyper-
sensitizing leaves them with excess silver over bromine. Working with known dyes
in experimental emulsions, the Bureau of Standards has shown that this excess
silver accounts for the photographic effects of the ammonia.

As hypersensitization has been a controversial subject, some
authorities denying even its existence, it will be well to begin with
our definition. We mean by hypersensitization an increase in the
effect of sensitizing dyes on an emulsion, produced by the addition
of some other material. We are, therefore, including materials added
to a dye bath, as well as those used for treating finished panchromatic
emulsions; the mechanism is the same in both cases. Hypersensitiza-
tion always involves some increase in speed to white light, but the
characteristic feature is the improvement of the relative color-sensi-
tivity, for example, the ratio of yellow to blue sensitivity.

We propose to show that the treatments used in hypersensitization
leave the emulsion with an excess of silver over halogen, and that the
resulting increase in silver-ion concentration, plus the alkalinity of
most of the baths, can account for the observed results.

There are apparent exceptions to this in the literature. A general
discussion of these is unnecessary, although it seems worth while to
point out the possibilities of error resulting from inadequate controls.
The first is the assumption that because treatment with a solution of
some material improves the color-sensitivity of a panchromatic
emulsion, the dissolved substance is necessarily responsible; it may
have been merely the water. We have had cases of emulsions whose
sensitivity to red light was tripled simply by washing out the soluble
bromide. Lack of controls is apparently responsible for claims that
have been made for a mixture of hydrogen peroxide and difficultly
soluble silver salts, ammonia being added to dissolve the latter. As

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Publication
approved by the Director of the Bureau of Standards.

** Bureau of Standards, U. S. Department of Commerce, Washington, D. C.
600

HYPERSENSITIZATION

601

the effect was attributed to the formation of new silver nuclei by the
peroxide, a theory inconsistent with our results, the experiments were
carefully repeated, using the same plates as the original investigator.
The mixture undoubtedly hypersensitizes, but we found that the best
results were obtained with the ammonia alone. The peroxide was
only a handicap.

FIG. 1.

.25 .50

Ammonia normality
Bromide removed from emulsion by ammonia treatment.

.75

Another source of conflicting reports is the dependence of hyper-
sensitization on the dye, and to a lesser extent on the emulsion, which
we hope to show in this paper. We found, for example, that different
brands of panchromatic plates and films from a single maker varied
widely in their capacity for hypersensitizing.

For hypersensitization, in practice and for the purpose of discussion
as well, we need consider only the use of ammonia and of ammonia

602 B. H. CARROLL AND D. HUBBARD [J. S. M. P. E.

plus small amounts of dissolved silver salts. Let us take first the
results of bathing a photographic plate or film in ammonia, as in
hypersensitizing. The ammonia, of course, extracts silver bromide
in amounts increasing with the concentration. But, after this effect
has been eliminated by evaporating or neutralizing the ammonia,
the solution is found still to contain soluble bromide, in amounts
increasing with the original ammonia concentration. The curve in
Fig. 1 is a plot of this soluble bromide, expressed as molecular per cent
of the silver bromide in the emulsion, against the normality of the
ammonia solution used in bathing. This was for Seed's 23 plates; the
results with other emulsions are very similar. Practically all emulsions
contain small amounts of soluble bromide as a preservative, but ex-
traction with water only will remove this. Analysis by other methods
established that there was no such amount of soluble bromide present
in the original emulsion as was found in the ammonia extracts. It,
therefore, must have come from the silver bromide in some way, and,
if there is excess bromide in the extract, there should be excess silver
in the emulsion. This last fact was readily established; extraction
of the ammonia-treated plates with very dilute acid (0.01 N acetic, or
0.001 N sulphuric) removed silver equivalent to the bromide. Hyper-
sensitized emulsions, therefore, contain as much as one per cent of
their silver in some form other than the bromide. Other experiments
have demonstrated that it is in combination with the gelatin.

An explanation of these results requires more physico-chemical
theory than we wish to present here ; it will be given in full in an early
issue of the Bureau of Standards Journal of Research. We changed
our theories repeatedly in the course of the investigation, but the
excess of the silver in the emulsion is an experimental fact and was
therefore unaffected by the reliability of our explanation. The reason
for the nearly unique effectiveness of ammonia is that it is alkaline,
volatile, and capable of dissolving silver bromide by forming a posi-
tively charged complex ion containing the silver; all these properties
are essential if the emulsion is to be left with excess silver after treat-
ment.

Before going on to the experimental demonstration that the
behavior of dyes is the same whether ammonia treatment or addition
of soluble silver salts is used to introduce excess silver into the emulsion,
it will be necessary to review some of the physical chemistry of the
emulsion. There is always a considerable amount of water even in
air-dry gelatin. In the emulsion, this water will be saturated with

May, 1932]

HYPERSENSITIZATION

603

silver bromide, which is a very difficultly soluble salt, but has never-
theless a real and definite solubility, so that there are always silver
and bromide ions present. These obey one of the familiar laws of
physical chemistry: the product of their concentrations in a saturated
solution is always equal to a constant at a given temperature. In the
case of silver bromide, this is 10 ~12; that is, if there is no excess of
silver or bromide ions, each is one one-millionth normal. Because
the solubility product is so small, small amounts of either ion can
produce enormous changes in the relative concentrations. For
example, if we add to a saturated solution of silver bromide 0.1 gram
of potassium bromide per liter, the bromide-ion concentration is 10 ~3

Erythrosin

-9

-5

FIG. 2.

-8 -7 -6

log silver ion concentration

Change in color-sensitivity of erythrosin-sensitized emul-
sions with silver-ion concentration.

normal, while the silver-ion concentration becomes 10 ~9 — that is, it
is divided by one thousand. The stability of an emulsion decreases
with increasing silver-ion concentration, so that small amounts of
soluble bromide can greatly improve the keeping qualities, as is well
known to emulsion makers. Conversely, it is obvious that hyper-
sensitized emulsions, with excess silver, must necessarily be relatively
unstable. The increase in silver-ion concentration with increasing
excess silver is fortunately much slower than the corresponding
changes produced by excess bromide. This is caused by the strong
affinity of gelatin for the silver ion, with a consequent "buffer" action;
most of the excess silver is bound by the gelatin, and only part is

604

B. H. CARROLL AND D. HUBBARD

[J. S. M. p. E.

available to raise the concentration of free silver ions. A somewhat
similar principle is used in most of the hypersensitizing baths contain-
ing silver salts dissolved in the ammonia ; silver chloride or some other
difficultly soluble salt is used so that the final silver-ion concentration
in the plate, after bathing and drying, is determined by the solubility
of the salt. Saturated silver chloride gives a silver-ion concentration
of 1.5 X 10 ~5 N; this is very much greater than that of a bromide
emulsion with excess bromide, but silver chloride in excess of the
amount required for saturation remains as solid so that this value is
the maximum that can be produced. The advantage, if any, of these
baths, is that a higher silver-ion concentration can be obtained with

Pinaflavol

-9

-5

-8 -7 -6

log silver ion concentration

FIG. 3. Change in color-sensitivity of pinaflavol-sensitized emul-
sions with silver-ion concentration.

dilute ammonia and correspondingly low alkalinity, but we have so
far found no marked advantage over plain ammonia.

Increase in silver-ion concentration causes little increase in photo-
graphic sensitivity in the absence of dyes, but in many cases it pro-
duces a marked improvement in sensitivity of a given dye-emulsion
combination to the longer wavelengths. Using known dyes in experi-
mental emulsions, we have compared the effect of excess silver added
to the emulsion, and of ammonia treatment of finished plates, and have
found it to be qualitatively the same in both cases. These experi-
ments have also brought out very clearly the dependence of hyper-
sensitization on the dye. We report here on three sensitizers rep-

May, 1932]

HYPERSENSITIZATION

605

resenting different classes: the acid dye, erythrosin; and the basic
dyes, pinacyanol and pinaflavol, which in water are probably in
colloidal and true solution, respectively.

Figs. 2, 3, and 4 show the change in color-sensitivity of experimental
emulsions with varying silver-ion concentration. Batches of emulsion
were sensitized with one of the dyes, then subdivided into smaller
portions to which were added appropriate amounts of soluble bromide
or silver salts.

We have plotted the speed to light screened by an appropriate
Iter (usually the Wratten Minus Blue) against the logarithm of the

Pinacyenol

-9

-5

FIG. 4

-8 -7 -6

log silver ion concentration

Change in color-sensitivity of pinacyanol-sensitized emul-
sions with silver-ion concentration.

silver-ion concentration; 7 was nearly constant so that the speed
number was a fair measure of sensitivity. Each illustration gives the
results for a single dye. The results for separate batches of emulsion,
representing both neutral and ammonia types, have been plotted on
different scales, so that it was possible to make a graphical average
into a single curve indicating the general trend.

Taking first the erythrosin (Fig. 2), there is a steady increase in
speed with increasing silver-ion concentration. This is to be expected
on theoretical grounds. Erythrosin ionizes into the colorless sodium
ion and a negatively charged dye ion. Since the dye ions are nega-
tively charged, they should be attached to the positive silver ions of sil-
ver bromide crystal surfaces. If bromide ions are present, they com-

606 B. H. CARROLL AND D. HUBBARD [J. S. M. P. E.

pete with the dye ions for the surface of the silver bromide grains. The
resulting change in adsorption of the dye was readily detected in these
experimental emulsions.

Pinaflavol and pinacyanol are both basic dyes, the iodides or
bromides of complex nitrogen bases. Since the dye ions are positively
charged, the theory just applied to erythrosin predicts that their
adsorption to silver bromide should decrease with increasing silver-
ion concentration, in contrast to the acid dye, erythrosin. Experi-
mentally, we found that the sensitivity of the emulsions containing
pinaflavol fell off at the higher silver-ion concentrations, as indicated in
Fig. 3. The results were difficult to reproduce, because the dye has
a strong tendency to cause fog, and are difficult to measure because
the region of sensitization lies so close to the natural sensitivity of
silver bromide. We cannot say positively whether pinaflavol also
has the decrease of sensitivity with increasing excess bromide which
is characteristic of most dyes. Pinacyanol-sensitized emulsions
increased in sensitivity continuously with increasing silver concentra-
tion, as shown in Fig. 4. We have no evidence to account for this
failure of the photographic effect to correspond with the adsorption
theory. We can only point out two things. In the first place,
pinacyanol is almost certainly a colloid in water, and the ionic theory
may not apply. In the second place, it is a reasonable assumption
that the photochemical sensitivity of the dye-silver bromide system in-
creases with increasing silver-ion concentration, so that this indepen-
dent factor opposes the adsorption effect for basic dyes.

In comparing these data with the following data on the effect of
ammonia, again remember that the experimental results can be con-
sidered, independent of our explanations of them.

Comparative data on the effect of ammonia on these dyes are

TABLE I

Erythrosin

Comparison of color -sensitivity of erythrosin-sensitized emulsions with and. without

ammonia treatment.

Speed through Filter
Dye in Emulsion With

Ammonia Without

Bath B 43 22

C 13.2 3.0

44 20.3 13.9

Emulsion 87 40 30

92 49 31.5

May, 1932] HYPERSENSITIZATION 607

TABLE II

Pinaflavol

Comparison of color-sensitivity of pinaflavol-sensitized emulsions with and without

ammonia treatment.

Speed through Filter
Dye In Emulsion With

Ammonia Without

Bath A 13.5 15

44 2.6 6.0

Emulsion 87 13.8 15.2

93 43.5 43.0

TABLE III

Pinacyanol

Comparison of color- sensitivity of pinacyanol-sensitized emulsions with and without

ammonia treatment.

Speed through Filter
Dye In Emulsion With

Ammonia Without

Bath A 54 26

Emulsion 93 79.5 57.5

92 35 33

presented in Tables I, II, and III. The commercial emulsions (desig-
nated by letters) were first washed, then bathed in solutions of
the dyes with and without ammonia. The experimental emulsions
were sensitized before coating, and were bathed in pure water and in
ammonia solutions. Silver-ion concentrations could not be measured
in the set emulsions, but working from the amounts of bromide
extracted we may say that after the ammonia treatment the excess
silver was about the same as the maximum used in the experiments
where it was added to the emulsions. The water wash left little
excess of either bromide or silver. The silver-ion concentration must
have been about 10 ~6 to 10 ~7 N. Contrast being nearly constant,
we have again given speed numbers as measures of color-sensitivity
using the minus blue filter as before.

It is evident from these figures that erythrosin -sensitized and
pinacyanol-sensitized emulsions are readily hypersensitized by
ammonia, while those containing pinaflavol are unchanged or actually
depressed in sensitivity. This corresponds exactly with their
behavior on increasing the silver-ion concentration, thus completing
our chain of proof.

The alkalinity of ammonia is a further asset in hypersensitizing,
as most dyes are more effective when the emulsion is slightly alkaline

608 B. H. CARROLL AND D. HUBBARD

(pH 8.0 to 8.5) than when it is strictly neutral or faintly acid. It also
appears that, with a given excess of silver, there is less fog in faintly
alkaline emulsions, because the free silver-ion concentration is lower.
At any rate, while it is possible to hypersensitize by bathing in dilute
solutions of silver salts, the results are distinctly less satisfactory
than with ammonia.

In this case, a better understanding of a previously obscure process
does not lead at once to practical improvement. The instability of
the emulsion with excess silver imposes a limit to the increase of color-
sensitivity in this way. We have, however, demonstrated the correla-
tion between other properties of the dye and its capacity for hyper-
sensitization, and hope to extend this line of investigation further.

DISCUSSION

PAST-PRESIDENT CRABTREE: The speed and temperature of drying has a
great effect on hypersensitization. Have you made tests on the keeping qualities,
growth of fog, and loss of speed?

MR. CARROLL: The keeping qualities were about the same in both cases,
for plates treated with ammonia and for those in which silver was added to the
emulsion.

MR. PALMER: What is the best way in which to preserve the excess sensi-
tivity as long as possible?

MR. CARROLL: The only thing to do is to put it into the ice-box and to hold
the temperature as low as possible; the lower it is the better, provided that it
does not freeze. The effect of moisture should also be considered. It would
be expected that anything of this nature would keep better if kept thoroughly
dry, or as dry as possible without stripping the emulsion from the support.

ADVANTAGES OF USING 16 MM. SUPERSENSITIVE

PANCHROMATIC FILM IN MAKING MEDICAL

MOTION PICTURES*

HARRIS B. TUTTLE** AND R. PLATO SCHWARTZf

Summary. — Due to the great speed of super sensitive film to Mazda lights, it is
possible to make satisfactory medical motion pictures under lighting conditions
impossible with regular panchromatic film. The high red sensitivity of this new
film permits the recording of detail lost in the past. Telephoto lenses can now be
used, giving both the photographer and the surgeon more freedom of movement, with
less danger of interfering with aseptic precautions.

Although the use of motion pictures as a means of revealing opera-
tive procedures for the training of physicians and surgeons is not ex-
pected ever to become of commercial importance to the motion picture
industry, it is believed that this is an instance where values greater
than economic assets must be considered. Human life and comfort
are dependent on the results of any major surgical operation. Not all
branches of surgery lend themselves readily to the use of motion
pictures in the teaching of surgical technic. It follows, therefore,
that care should be taken in selecting the field chosen for our initial
efforts, lest motion pictures be censured, when in reality the fault lay
in the selection of the subject.

Individual patients present wide variations in the same condition
which requires surgical treatment. All surgeons may agree in prin-
ciples governing operative procedures but differ in methods of execu-
tion. The method by which motion pictures are to be made avail-
able for teaching surgery cannot, therefore, be the same as that which
is represented by Hollywood in the field of entertainment. Honest
effort in this direction has failed, at the cost of several hundred thou-
sands of dollars, chiefly because these two facts were not considered of
great importance in relation to the outcome of the undertaking. Be-
cause of these prevailing conditions, our course is, for the present,

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Eastman Kodak Company, Rochester, New York.
t Strong Memorial Hospital, Rochester, New York.

609

610

H. B. TUTTLE AND R. P. SCHWARTZ [J. S. M. P. E

well defined. It should be directed toward the solution of the prob-
lems confronting the individual surgeon interested in the application
of motion pictures to the problems of surgery.

One of the greatest difficulties is the making of motion pictures in
the operating room. The advent of 16 mm. film and the reversal
process considerably lessened the expense to the medical profession
of making surgical motion pictures. The manufacture of compact
and efficient lighting units has helped considerably in the progress of
making surgical pictures.

Even with panchromatic film, which is sensitive to all colors

FIG. 1. Bausch & Lomb oper-
ating room lamp.

FIG. 2. Bausch & Lomb
floor lamp.

(but slightly more sensitive to blue and to green than to red) , and with
modern lighting equipment, surgeons have been unable to record and
reproduce perfectly all the necessary details. However, the results
obtained during the past year or two have been much more satis-
factory than anything previously attained.

In this branch of motion picture photography there have been a
great many difficulties to overcome. Principal among these was
the rendition of the operative field and of the tissues surrounding the
incision, which are always reddish in color. On the old regular film,
which was not sensitive to red light, these reddish areas in and around
the incision would always come out black, showing little or no de-

May, 1932]

MAKING MEDICAL PICTURES

611

tail in the delicate structures where clarity of detail was very impor-
tant. With the use of 16 mm. panchromatic film, this objection
was somewhat overcome. While the red areas were rendered more
accurately than before, there was still considerable difference between
the photographic reproduction on the screen, and the actual subject.
The second important difficulty in making such pictures was the
size and inconvenience of using artificial lighting equipment. Light-
ing units consisting of three 500-watt bulbs were not readily accepted
by the average surgeon, and the necessity of having an //1. 9 lens cam-

FIG. 3.

Macbeth operating room
lamp.

era limited somewhat the field of surgical motion pictures. For those
few who had lighting units available, the heat, together with the
necessity of meeting the demands of asepsis, prevented the close
proximity of lamps, camera, and the operator to the operative field
to make clear, well-defined motion pictures.

Recently a new emulsion for amateur use, called Cine Kodak super-
sensitive panchromatic film, has been placed on the market. While
this film is twice as fast to daylight as panchromatic film, its chief ad-
vantage lies in its great speed when exposed to artificial light, to
which it is three to four times faster than is the regular panchromatic

612

H. B. TUTTLE AND R. P. SCHWARTZ [J. S. M. P. E.

film. Supersensitive film is especially sensitized in the red portion
of the spectrum, so that most of its speed lies in the area in which
artificial light is richest.

While this emulsion was designed for high speed work in motion
picture studios where artificial lights are used almost exclusively, it
was later adapted to the 16 mm^reversal process so as to enable ama-
teurs to make pictures with low intensity lamps. It so happens

FIG. 4. Scialytic operating room lamp.

that its characteristics are ideal for making surgical motion pictures
Probably the most important advantage of this film from an eco-
nomic point of view is the fact that it is no longer necessary to use any
additional artificial light. The lights which are installed over most
operating tables are sufficient for making good motion pictures.
With the light found in most hospitals, it is possible with supersensi-
tive film to expose at diaphragms //3. 5 or//4.0.

May, 1932] MAKING MEDICAL PICTURES 613

It is also possible with fast film and regular operating room lighting
to use telephoto lenses. It can readily be seen that this is a very
important factor, as heretofore, with a one-inch lens, it has been
necessary for the operator to hold the camera from three to four feet
from the operative field in order to get an image size which would
show sufficient detail to be recognizable on the projection screen.
With a three-inch lens placed six feet from the operative field, it is
now possible to get an image the same size as could be obtained with
the one-inch lens at two feet from the subject.

When using the 3- or 41/2-inch telephoto lens, it has been found ad-
visable to use a tripod. It is almost impossible to hold a camera
absolutely steady. While the small amount of body movement
which does exist is not objectionable when a one-inch lens is used, and
the resulting field is fairly large, the long focus lenses magnify this
movement many times so that the smaller field is unsteady.

Camera Distances Approximate Field Size in Inches

1" lens

2 feet 7 X 9*/2

3 lOYa 14

4 14V2 19

5 18 24

6 21V« 283A
3" lens

6 feet 7 9Va

9 10V2 14

12 14V2 19

15 18 24
41// lens

9 feet 7 91/*

12 9 12

15 12 16

18 14Va 19

The data given above for the use of supersensitive film are for
operating rooms which are illuminated only with artificial light, in
other words, pictures made with no daylight present.

A large percentage of all surgical operations are performed in day-
light. Because of the diffusion of daylight in most operating rooms,
the surgeon requires additional illumination of incisions and cavities
by a spotlight such as the Bausch & Lomb or Scialytic, 100-watt, 110-
volt, floor lamp. This combined use of daylight and tungsten

614 H. B. TUTTLE AND R. P. SCHWARTZ [j. S. M. P. E.

illumination provides advantages both for the surgeon and for the
photographic recording of surgical procedures. A diaphragm open-
ing of //4.0 to//5.6 is correct when these two sources of illumination
are combined.

Variations of season, weather, and time of day so alter the intensity
of daylight that it is impossible to give precise data on the proper ex-
posure. That this variation of daylight in the operating room is dis-

FIG. 5. Mayo operating room lamp.

advantageous to the surgeon, has become more generally recognized
with the development of efficient sources of illumination. For some
years past many operations have been regularly performed to ad-
vantage in special rooms where no daylight was present. It has been
suggested by Dr. J. J. Morton, Professor of Surgery, Rochester
University School of Medicine, that in the future design of operating
rooms, all daylight should be excluded. It is apparent that such

May, 1932] MAKING MEDICAL PICTURES 615

conditions would be helpful to the surgeon and would further the
application of motion pictures to the problems of surgery.

On regular panchromatic film, it was sometimes advisable to use
a K-3 or G filter to render the red areas more faithfully. The use of
such filters, however, made it necessary to allow two or three stops
difference in exposure, thus sacrificing some depth of focus and general
definition. Supersensitive panchromatic film is highly self -corrective,
making it unnecessary to use filters in photographing operative work.

The first experiments with supersensitive film were carried out at
the Strong Memorial Hospital in Rochester, New York. Tests were
made with three very different types of operating room lights having
110- volt circuits. All experiments were conducted at night so that
no daylight or room illumination could cause variations in results.

In order to have a cross-section of results obtainable with other
types of 1 10-volt lamps, similar experiments were made at three other
Rochester hospitals.

The summary of these tests showed that in operating rooms having
from 400 to 600 watts available, at approximately 40 inches from the
operative field, satisfactory exposures could be made at diaphragms
//5.6 to //8.0; where 200 to 400 watts were available, diaphragms
//2.8 to //4.0 ; and where only single units of 100 watts were available,
diaphragms //1. 9 to//2.8.

Strong Memorial Hospital. — Bausch & Lomb mirror spot dome
lamp, 500-watt, 1 10-volt lamp, 40 inches from operative field, at
diaphragms: //1. 9 to//5.6, and telephoto//4.5.

//2.8 — satisfactory exposure
//4.0 — satisfactory exposure
3" telephoto //4.5— satisfactory

Scialytic, one 100-watt, 1 10-volt, mirror reflector, 40 inches from
operative field, at diaphragms: //1. 9 to//5.6, and telephoto //4. 5.

//2.8 — satisfactory exposure
//4.0 — satisfactory exposure
3" telephoto //4.5 — under-exposed

A 100-watt, 110- volt auxiliary operating room spotlight should be
used with a 100-watt, 1 10-volt Scialytic.

Macbeth, twin operating room light, two 150-watt, 1 10-volt lamps
in twin dome receptacles diffused 40 inches from operative field, at
diaphragms: //1. 9 to//5.6, and telephoto //4. 5.

616 H. B. TUTTLE AND R. P. SCHWARTZ [j. S. M. p. E

f/2.8 — satisfactory exposure

//4.0 — satisfactory exposure

3" telephoto //4.5 — satisfactory exposure

Highland Hospital. — Regular operating room lights. Double
dome, two 150-watt, 110-volts, diffused: //1.9,//2.8,//4.0,//5.6, and
telephoto //4.5.

//1. 9 — correct exposure

Repeated with Scialytic spot, 100-watt, 110-volts, diffused:
//2.8, //4.0, //5.6, and telephoto //4.5.

//4.0 — correct exposure
3" telephoto //4. 5— satisfactory

Genesee Hospital. — Scialytic dome lamp, one 100-watt lamp in
center, three 50-watt lamps in cluster, //2.8, //4.0, //5.6, and tele-
photo //4.5.

//4.0 — correct exposure
//5.6 — satisfactory
3" telephoto //4.5 — satisfactory

Repeated with one 100-watt vent light spot added, f/2.8, //4.0,
//5.6, and telephoto//4.5.

//4.0 — correct exposure
//5.6— satisfactory
3" telephoto //4.5 — satisfactory

General Hospital. — Mayo lamps, eight single, 60-watt, 120- volt:
//2.8, //4.0, //5.6, and telephoto //4. 5.

//4.0 — correct exposure
3" telephoto //4.5 — satisfactory

Repeated with 100-watt Bausch & Lomb lamp added: //2.8,
//4.0, //5.6, //8.0, and telephoto //4.5, //5.6, //8.O.

//5.6 — correct exposure

//8.0 — satisfactory
3" telephoto //5.6 — correct exposure
3" telephoto //8.0— satisfactory

Bausch & Lomb, 100-watt, 120-volt lamp alone: //2.8,//4.0,//5.6.

//4.0 — correct
//5.6 — satisfactory

May, 1932]

MAKING MEDICAL PICTURES

617

Summary of Tests

Hospital

Strong Memorial

Highland

Genesee

General

Lamp

Bausch & Lomb
Scialytic
Macbeth
Regular
Regular plus Scialytic

Spot
Scialytic

Scialytic plus vent
light

Mayo

Mayo plus Bausch &
Lomb floor spot

Number

of
Lamps

1
1
2

Wattage
(110 v.)

500

100
100
100

100

100
50

100
60

60
100

Distance Best Exposure
(inches) (//number)

40
40
40
60

2. 8 or 4.0
2. 8 or 4.0
2. 8 or 4.0
1.9

40 4.0

40 4.0 or 5.6

72 4.0

72 5. 6 or 8.0
40

Bausch & Lomb floor

spot
In the average

hospital With average lights

100 40 4.0

250-300 40-45 3. 5 to 4.0

DISCUSSION

MR. KURLANDER: I noticed the lack of blood in the last operation. Was
the subject a live one or not?

DR. SCHWARTZ: For the most part, we have considered the possibility of
making for students, better pictures of a cadaverous foot. For definition of
detail of the various structures in this foot, as compared with the colors of the
tissues, we had a rendition of proper value in grays, blacks, and whites, in order
to set apart the different structures. In other words, when the tendons came
out and could be seen, there was not in this operation, or in any other, the possi-
bility of the blood's interfering with the definition recorded on supersensitive
film, as compared with the recording of the same operation on any film which we
have had heretofore.

The cadaver lends itself particularly well to the teaching of orthopedic surgery,
in as much as the parts are easily available. The depth that we have to reach,
either in a knee, hip, or an ankle, is not great enough but what we can get to
that depth and reveal the details at the bottom.

MR. COWLING: Would it not be a great help if these pictures were made in
color?

DR. SCHWARTZ: When we shall have film available that has a sensitivity
sufficiently great to permit us to reduce the great illumination now required, we
shall be better off. We have made several pictures, including this particular
operation, in color, and the detail that is recorded is remarkable. It is better

618 H. B. TUTTLE AND R. P. SCHWARTZ

than anything that can be obtained in black and white. Kodacolor film serves
the purpose beautifully. I should like, however, to have an emulsion that is
comparable with supersensitive film in Kodacolor.

MR. MITCHELL: I happen to have seen several Kodacolor pictures, taken
with the aid of low wattage lamps, preferably automobile head-lights — the idea
being to use several automobile lights in series. The light thus obtained is
almost "cold" light. Plenty of illumination is obtained for Kodacolor, and no
heat; and all one has to do is to mask the excess of red in the light either by
reversing the ratio diaphragm or by using adhesive plaster over the red portion
of the Kodacolor filter, as some surgeons do. The results are very satisfactory.

MR. KELLOGG: If the heat of the lamp is one of the important problems, one
wonders whether there are not possibilities in the same sort of device as is used
in television, in order to make the exposure with a minimum heating of the
subject; namely, flashing the source synchronously with the opening of the
shutter. Much would not be gained in a quick pull-down camera, but many
cameras have a 180-degree shutter opening, or not much more than that, under
which circumstances there might be a considerable decrease of heating.

Another question that comes to my mind is whether Kodacolor is the logical
choice under the circumstances. As the motions are quite slow, I wonder whether
the old cinema color system, coupled with flashing lamps of different colors,
would not be almost as efficient, or fully as efficient, as the present system,
especially if one did not try for extremes of color saturation.

DR. SCHWARTZ: The lack of progress in this field is obviously proportional to
the lack of money available for its development. We have taken the line not
only of least resistance but of least expense, and we have found Kodacolor suffi-
ciently satisfactory that we have been willing to use it for the present, hoping
that in the meantime something would happen that would permit us to work
along other lines.

MR. MAURER: Has the attempt ever been made in this work to run two
cameras slightly separated, later projecting the images through a viewing device
that would give a stereoscopic effect? It would seem that such a way of getting
additional relief would be very valuable to the medical student who would wish
to study these pictures.

DR. SCHWARTZ: Obviously, the same answer I gave previously applies to
this question. Undoubtedly a third dimension given to this shadow would be
of great value in the teaching of orthopedic surgery.

SOME COLOR PROBLEMS

GERALDINE GEOGHEGAN**

Summary. — Some of the difficulties attending the taking and projection of motion
pictures in actual colors are briefly discussed, particular attention being paid to the
need of matching the taking and projecting lights spectrally. The author concludes
that, despite the great amount of research work being done in emulsions, optics, etc.,
little will be accomplished in achieving motion pictures in natural colors unless the
same particular care be taken with the light sources.

Motion pictures in natural color seem to have developed in a
somewhat jerky fashion. It is quite common, among the various
concerns dealing with color, to have an expert on one or two subjects
concerning the many problems arising in natural color photography,
but, generally speaking, these experts are concerned only with their
particular specialized knowledge, and can offer little or no help when
outside problems upset their calculations.

If the color is produced by an optical arrangement, an expert on
optics is employed; if by dyes, a specialized color printer, etc. It
is quite possible that these men can, and do, produce motion pictures
of astounding beauty and fidelity to color under the standardized
conditions of the laboratory; but when used commercially, con-
siderable trouble arises.

When we consider that color is not part or parcel of the article
observed, but merely its capacity to absorb and transmit such part
of the light waves that fall upon it, we are up against our first prob-
lem— the spectral quality of the "taking" light. If the latter wavers
in wavelength in the slightest degree, the object being photographed
changes its hue. At the same time the human eye is an accommodat-
ing organ and has a very short memory, so that if such changes be
gradual, it is impossible to notice them while under the influence of
the altered light; but, if the same observer and object be again placed
under the correct light with an image taken under the deficient light,

* Received December 14, 1931.
** E. S. S. Color Filter Co., London, England.

619

620 GERALDINE GEOGHEGAN [J. S. M. p. E.

the error can be seen at once. It is common knowledge that it is
very difficult to produce artificially a light with a spectral approxi-
mation to daylight of sufficient volume without heat and noise.
Even if this be done, the motion picture producer of color films is
immediately up against another problem, the spectral quality of the
"projecting" light.

After all, a color transparency is merely a collection of light filters
that absorb, and transmit, according to their power, the light that is
projected through them. Therefore we can deduce from this, that
to obtain pictures on the screen in natural color, the "projecting"
light and the "taking" light must be one and the same spectrally.
Compensation may be attempted, but all filters, it must be clearly
understood, lower volume of illumination.

Let us say, for the sake of argument, that we have our taking and
projecting lights spectrally balanced; minor problems now arise,
such as the absorption and transmission of the screen, the influence
of color in the decoration of the theater, and the general quality of
the approach lights.

Some attempt should be made to screen all interior lights so that
they approximate in quality that of daylight where neutral color
films are to be shown, and only subdued schemes of decoration
should be permitted.

The problem of voltage plays a part in projecting that does not
arise with monochrome work. It would be quite possible for a
motion picture in natural colors to be shown in one theater with
exceedingly pleasing and beautiful results, while its exact duplicate
might be shown in another theater with distorted and repulsive
colors owing to a drop in voltage.

One has to consider that on the stage, where living actors and
actresses appear, the colors of the dresses and the lights that play
upon them are under the control of the producer. He views the
effect as the audience sees it, and is certain that no radical change can
occur; but the producer of motion pictures in color is by no means
in that happy position. Monochrome pictures, once passed by
their director and shown under ordinary standardized conditions,
will please the man in the street even if the expert technician will
notice an error or so. But with natural color pictures it appears
that a private view is necessary in every theater in which each motion
picture is shown, to see that no unforeseen spectral change has oc-
curred. It was found that many theaters were not suited for the

May, 1932] SOME COLOR PROBLEMS 621

sound pictures; some even had to be scrapped, and many altered.
Why not then take the same precautions with color?

Undoubtedly color pictures will take the place of monochrome;
but only by a very strict attention to what may be looked upon as
minor details, can success be obtained. A change in gradation of
tones in a monochrome picture can occur without any appreciable
notice on the part of the spectator, but a change in color will be seen
by every two out of three. The normal vision is trained to recognize
objects not only by shape, but by color; it is not really familiar
with these in monochrome, and therefore allows false gradation to
pass unnoticed.

The writer has purposely ignored such problems as fringing, etc.,
as these are inherent in the processes themselves; and has adopted,
merely as a theme, the difficulties that confront the producer, even
though he have a perfect process of motion picture in color. It is
doubtful whether such a process is yet on the market commercially,
whence the path for color cinematography is beset with many thorns
and snags. But at the moment it is felt that too much thought and
research work are being directed to emulsions, optics, etc., which,
although of themselves invaluable, at the same time are useless
unless the same care be taken with light, etc.

It would be a better box-office proposition to have color fantasti-
cally unreal than to show (as has been done in many cases) true color
degraded and falsified by bad technic.

THE SELENOPHON SOUND RECORDING SYSTEM*
PAULSCHROTT**

Summary. — T his brief description of the Selenophon process of recording was
offered as a contribution to the Progress Committee Report of May, 1931. The
process is employed to record sound photographically on film by the variable width
method, employing a string oscillograph for varying the light intensity; a tightly
strung fiber moving in sympathy with sound vibrations, and acting as a variable
shutter in the path of the beam of light. The reproducer is briefly referred to, in
which a selenium cell of the condenser type is used in conjunction with a five-stage
amplifier.

The Selenophon sound recording system is a process for recording
sound photographically by the variable width method. The width
of the sound track is that determined by international agreement,
namely, 3 millimeters, and the film speed is 24 frames per second
(456 mm. per second). The slit width is 12^, making possible the
reproduction of 8000 cycles.

A string oscillograph (Fig. 1) serves to vary the light intensity. A
tightly strung metal fiber moves in sympathy with the sound vibra-
tions and acts as a variable shutter in the light beam. The string is a
wire about 0.1 millimeter in diameter, of aluminum bronze or tungsten
(which has a high ratio of rigidity to mass, or tension to mass, which
determines the natural frequency) so that the natural frequency
(14,000 cycles) is above the highest recorded frequency. The string
carries the amplified sound currents and hence is made to vibrate
in a strong, uniform, magnetic field (Fig. 2).

As a result of the high natural frequency and the small amplitude,
the displacement of the string is proportional to the current. The
maximum current impulse is about 1 ampere and the resistance is 0.5
ohm; hence the string must carry 0.5 to 1.0 watt. The input energy
from the microphone is amplified to 4 or 5 watts to provide sufficient

* A contribution to the Progress Committee Report of May, 1931. Trans-
lated by J. W. McFarlane, Research Laboratories, Eastman Kodak Co., Rochester,
N. Y.

** Research Professor, Technical High School, Vienna, Austria.
622

SELENOPHON RECORDING SYSTEM 623

reserve power. The string shutter action results from the relatively
high current from a 60 to 1 transformer. As mentioned above, the
recording light beam must have a cross-section of 3 by 0.012 milli-
meters. The string, when at rest, intercepts half the beam, and is
mounted at a small angle to the light beam (about 1 degree) so that
for relatively small vibrations of the string a proportionally large
part of the light beam is affected.

The optical system is arranged as follows: The rays from the source
of light are concentrated by a condenser into the entrance pupil of an

FIG. 1. String oscillo-
graph. (Wire with ten-
sion adjustment.)

objective. Directly behind the condenser is a slit diaphragm, imaged
sharply by the objective in the plane of the oscillograph string. A
second lens images the string and the slit image in the plane of the
film, its focal length being such that the image is 0.12ju wide. A
low voltage lamp is used as a source of light. An audible control can
be effected by adding a light-sensitive cell and amplifier.

The mechanical design is such that the light beam is vertical, com-
ing from underneath, and the film path is horizontal (Fig. 3). The

624

PAUL SCHROTT

[J. vS. M. P. E.

main driving shaft is parallel to the housing, turning at 180 rpm., and is
so designed that the driving, feed, and take-up sprockets are driven

FIG. 2 String oscillograph and electromagnets.

FIG. 3. Sound recorder, open, showing position of
oscillograph.

from this shaft (Fig. 4) . A single small motor drives the take-up spool
in a retort for the exposed film. The main driving shaft is driven
through an 8- to 1-gear reduction unit by a 4-pole synchronous

May, 1932] SELENOPHON RECORDING SYSTEM 625

FIG 4. Selenophon driving unit.

FIG. 5. The reproducer.

626

PAUL SCHROTT

[J. S. M. P. E.

The

motor running at 1440 rpm., requiring a frequency of 48 cycles,
motor draws 250 to 300 watts.

The recording apparatus requires a floor space of 1.20 by 0.75
square meters and is 1.05 meters high. It is so designed that several

FIG. 6. (a)

Home sound record outfit for paper records (closed).
(6) Home sound record outfit (open).

FIG. 7.

Sound records on paper. A and Care printed photographi-
cally; B and D are printed mechanically.

sound tracks can be recorded side by side on the width of a standard
film. For this purpose the whole exposing mechanism can be moved
at right angles to the film drive so that the image of the slit can be
placed on any desired part of the film. In this way, eight ordinary

May, 1932] SELENOPHON RECORDING SYSTEM 627

sound records can be arranged so that they will follow each other in
the proper sequence.

The construction of the reproducer (Fig. 5) is similar to that of the
recorder. A selenium cell designed by Prof. Thirring, a condenser
type, is used as a light-sensitive element in conjunction with a five-
stage amplifier. The record on the sound track, highly enlarged, is
projected on to the cell and is condensed by a cylindrical lens.

The Selenophon Company has developed, at the same time, a home
sound record outfit as a, substitute for the gramophone (Fig. 6) . The
sound record is of the variable width type, and is printed on paper
either photographically or mechanically (Fig. 7). This apparatus
can be used with any radio amplifier and is capable of continuous
performance up to four hours.

THE MOTION PICTURE INDUSTRY IN JAPAN*
MARCEL RUOT**

Summary. — The history of the development of the motion picture industry in
Japan is traced, beginning with the year 1895. The economic problems with which
the Japanese producers have to contend are pointed out, as well as the extent to which
motion pictures are used and theaters are patronized. The relation between the per
capita wealth and the patronage of motion picture theaters is discussed, as well as the
special uses to which motion pictures are put, such as for education and propaganda.

EARLY HISTORY

It can be said of the motion picture business in Japan that it is
almost as old as the invention itself. As early as 1895, K. Inabata, of
Osaka, and the Nichiei Co., of Tokyo, imported, practically at the
same time, the first projectors and films ever seen in Japan, and it
is still a point of controversy as to which of the two importers was
really the first in the field.

Inabata sold his Lumiere Cine*matographe and first series of French
films to a young friend of his, E. Yokota, then only 20 years old, but
today president of one of the most important Japanese motion pic-
ture companies. Messrs. Nichiei & Co., on the other hand, were at
the same moment disposing of their newly imported cinema goods
to K. Kawaura, owner of the firm of Yoshizawa & Co., in Tokyo.

Messrs. Yoshizawa soon gave a motion picture show in Tokyo,
and it seems that, notwithstanding the claims since made by other
parties, this was the very first time that a motion picture film had
ever been shown to the public in Japan. The machine used was an
Edison Vitascope projector, the films being of French origin, and, of
course, all of extremely short length as compared with the present-
day productions.

On his side,' Yokota had already visualized the commercial possi-
bilities of the invention, and also had begun to exhibit in public his
newly acquired films. In order to widen his scope of action the Yo-
kota Company was formed as early as 1896. In 1897, he visited

* Edited from a report submitted to the Progress Committee.
**Kodak, Ltd., London, England: formerly located at Osaka, Japan.
628

MOTION PICTURES IN JAPAN 629

Europe to obtain fresh supplies, and while there he entered into a
contract with Pathe Freres, becoming their agent for Japan. By
1904-1905 Yokota controlled no fewer than 11 circuits.

Due mostly to keener competition and public demand, Japanese
exhibitors were soon to turn to the production of local subjects, and
there again we find Messrs. Yoshizawa first in the field, when, in
1898, they commissioned the popular actor, Baiko Onoue, to produce
Ninin Dojoji, a Kabuki dance performed by two young girls in the
grounds of the Dojoji Temple at Gobo in the Province of Kii. The
cameraman was Mr. Shibata, owner of the Shibata Studios in Kyoto.
The camera used was a Gaumont, and the total length of the film did
not exceed 180 feet.

The following year the same cameraman, working always for
Messrs. Yoshizawa & Co., filmed Momiji Gari (maple viewing) in
Tokyo; in this "super production" two leading stars of the legitimate
stage were featured — Danjuro Ichikawa and Kikugoro Onoue the
Fifth, father of the present equally famous actor, Kikugoro the Sixth.

The Yokota Company built their own studios in Kyoto. These
studios were of the open-air type, as then used generally in America
and Europe. In 191 1, the Yokota Studios were transfered to another
section of Kyoto, and a glass-roofed stage was built, which at that
time created quite a sensation, as it was the first of the kind in Japan.

News events were filmed for the first time in 1899, again by Shibata,
with his Gaumont camera. His first scoop was the filming of the
funeral of the noted actor, Kikugoro Onoue the Fifth. Later in the
same year a Sumo contest (Japanese wrestling) was similarly filmed
for its news value.

In those early days, the exhibiting business meant going the rounds
of the larger cities, renting a public hall, a legitimate theater, or any
large building suitable for the screening of a motion picture pro-
gram; but success, nevertheless, steadily rewarded the two pioneer
firms.

From 1909 to 1930, the fortunes of the industry have been varied.
Besides the two original firms of Yokota and Yoshizawa, other firms
entered the field and trusts were established, until today the indus-
try is controlled largely by six competing producers who also are
exhibitors: namely, the Nippon Katsudo Shoshin K. K. ("Nikkatsu"),
registered in 1912; the Teikoku Kinema Engei K. K. ("Teikini") in
1920; the Schochiku Cinematograph Co. in 1920; the Toa Co. in
1923; the S. Makino Co. in 1927; and the T. Kawei Co. During

630 M. Ruox [J. S. M. P. E.

1931, two producing companies were added to the list, Bantsuma
Productions and Takarazuka Eiga K. K.

FILMS AND EMULSIONS

Three firms are now receiving yearly subsidies from the Japanese
government to help them to manufacture sensitized film of good
marketable quality. Roll film, plates, and papers are already being
produced, and now the attention of these firms is centered on motion
picture film: positive film to start with, and later on, negative emul-
sions. Toyo Kanpan K. K., who were the first to attempt to manu-
facture motion picture films, are a subsidiary of the Dai Nippon
Celluloid Company. During 1930 the research work of both the Dai
Nippon Celluloid Company and the Toyo Kanpan was diverted to
the manufacture of safety film, possibly as the result of the several
serious film fires that recently occurred in Japan.

The Oriental Photo Industry Company, Limited, who have earned
a certain reputation in the Far East for their "Oriental" papers, have
recently enlarged their Tokyo factory and imported from Germany
all the machinery needed to make film in 1000-foot lengths. While
the new buildings were going up and the machines were being erected,
the Oriental research laboratories kept busy experimenting with film
supports and emulsions. It is anticipated that their first film will ap-
pear on the market during 1932. Subsidized firms are encouraged
by the government but their subsidies will be renewed only if, after
a few years of research, a marketable article is produced.

The third runner up, in this subsidized raw stock race, is the Roku-
osha factory, owned by R. Konishi & Co., one of the oldest and most
prosperous wholesalers in photographic goods. Rokuosha entered
the field of sensitized products in 1929 with their Japanese-made
Roll Film, "Sakura." They are now bringing out filmpack, and are
planning the more ambitious manufacture of motion picture film.

On February 25, 1931, the Jiji Shimpo, a daily paper appearing in
Tokyo, announced the completion of the researches made by S.
Hiraumi, who for the past ten years has been trying to make motion
picture raw film in Japan. Manufacturing plans are well under way,
and a factory is being built near Tokyo. Production plans call for
three million feet to be made monthly.

By 1935 we should therefore see three, if not four, Japanese firms
manufacturing 35 mm. positive, and perhaps negative film.

May, 1932] MOTION PICTURES IN JAPAN 631

The annual capacity of the Far East market is at present:

Positive 35 Mm. Positive 16 Mm. Negative 35 Mm.

Japan— 55,000,000 ft. 2,500,000 ft. 8,000,000 ft.

China— 15,000,000 ft. 500,000 ft. 1,500,000 ft.

It is difficult to visualize three film factories, or even two, sharing
between them less than 80,000,000 ft. of raw film, assuming all the
while that they can clear the Japanese market of foreign importa-
tion. This prospect is very doubtful, as in Japan, where quality
sometimes gives way to price, the foreign manufacturers might come
down to the lower prices that are bound to be charged by the Japanese
manufacturers.

STUDIOS

During the past eighteen months, the Japanese studios have
changed over from regular negative to panchromatic emulsions, and
from arc lighting to tungsten lamps. The Teikine Company was
the first to alter the studio lights, having used tungsten lamps for
almost two years. From the older types of panchromatic film
studios willingly passed on to the improved type, but they seem to be
reluctant to adopt the new high speed panchromatic films since their
use necessitates time development, a method that is yet practically
unknown in Japan.

It seems that until the laboratories are convinced of the advantages
of time development, and perhaps until automatic developing ma-
chines are installed, the Japanese studios will continue to use
emulsions that can be handled under safelights, thus permitting the
laboratory attendant to watch the appearance of the image.

Except for special features, the length of pictures is being limited
to 5000 feet, in order to reduce production costs by 25 per cent.
During the summer months of 1930-31, some Japanese studios closed
completely for a week or two, this being done more for reasons of
economy than for granting rest to the personnel.

In 1930, the number of stockholders of eight of the leading com-
panies in the Japanese motion picture trade was five times greater
than it was five years ago, amounting almost to 20,000 holders.
Nearly 4000 people are employed by the Japanese studios, including
"still" cameramen and costumers. The largest number is credited to
the Shochiku Studios at Kamata, which has a payroll of 1412; of
these, 180 are actors regularly employed, 125 are actresses, and 15 are
directors. Teikine, however, leads in the number of actors and ac-

632 M. RUOT [J. S. M. P. E.

tresses, having 320 of the former and 120 of the latter. Nikkatsu
claims the greatest number of directors, with 20. To improve the
technic of camera work and laboratory processing in film production,
the Shochiku Company has organized a research laboratory at their
Kamata studio. All cameramen, laboratory, and electrical tech-
nicians are members of this department and cooperate according to
their respective line of work.

PRODUCTION IN RELATION TO EXHIBITION

There are 1488 cinema theaters in the Japanese empire, of which
1413 are in Japan proper, 15 in Karafuto Island, 10 in Formosa, 35 in
Korea, and 15 in Manchuria. These figures show an increase of 20 per
cent in five years, the total for 1926 being only 1100. While such an
increase might be considered as satisfactory proof of progress, yet it
represents only one theater for 50,000 people, and many of these
theaters have very small seating capacities as the following 1927
census tends to show:

Capacity Number of Theaters Percentage

Less than 250 persons 22 2

From 250 to 500 365 35

From 500 to 750 333 32

From 750 to 1000 213 20

Above 1000 112 11

Total 1045 100

Small theaters, seating from 250 to 750, represent more than two-
thirds of all the cinemas of Japan. The larger houses, of more than
1000 seats, are usually old legitimate theaters converted into motion
picture houses and, of course, are to be found mostly in the larger
cities, like Tokyo, Yokohama, Nagoya, Osaka, Kyoto, and Kobe.
Ten of the larger cinemas of Tokyo have a total seating accommoda-
tion of 13,000, the Denkikan heading the list with 1500, and the Hogo-
kuza with about 1000. The Shockikuza in Osaka boasts of being
the largest cinema in Japan with 1600 seats.

Yet, though the number of theaters in Japan is very limited and
their seating accommodation small, they seldom have a full house.
Only 158 million admissions were registered in 1930, slightly more
than 300 admissions per day and per theater; and as the theaters are
showing twice daily it only means an average house of 150 per show,
which is a long way from a full house for most of the theaters.

May, 1932]

MOTION PICTURES IN JAPAN

633

It is apparent that, under present working conditions, there is suf-
ficient seating accommodation throughout Japan to accommodate
almost twice the present attendance, which should satisfy for a good
many years the average increase in attendance. There is, of course,
very keen competition between exhibitors in their efforts to secure
the limited patronage available, and one of the means used has been
to offer three features, instead of one or two as in other countries;
and, furthermore, to issue these features in lengths of 8 or 10 reels,
so that a program will sometimes last as long as 6 hours, even though
projected at top speed.

As a rule the program is changed once a week, although recently
some special features have had runs of two, three, and even four
weeks. The program is shown every day, including Sundays, and,
generally, twice daily. As it is usually made of three features, the
theaters have to find 156 films each year, and it is this excessive de-
mand for new subjects that has brought about in Japan a situation in
the matter of production that has no equal elsewhere.

New subjects must follow new subjects, and in 1930 alone more than
550 features were produced in the Japanese studios to meet the re-
quirements of the various renting organizations.

According to a recent investigation made by the Movie Times,
Japanese producing companies are equipped as follows:

Name of Company Stages

Cameras

Printers

Laboratory
Capacity
Daily

Staff

Feature
Subjects
Produced

Asia Eiga

Chiezo

Kawai

360

Makino

70,000

561

Nikkatsu

65,000

855

101

Shochiku, Kamato

100,000

815

Shochiku, Shimokamo

70,000

264

Teikine

20,000

655

Toa

. .

30,000

441

Utaemon

104

Total

4193 601

Besides these larger companies, there are several smaller ones.
Their productions brought the grand total of films made in Japan
during 1930 to 630, as compared with 500 produced in America during
the same year. In the number of features produced, Shochiku and
Nikkatsu may claim to be the "biggest in the world."

634 M. RUOT [j. S. M. P. E.

The "big five" producing companies of Japan have a total produc-
tion budget of $18,000,000, of which one-third goes to production
of negatives and the balance to making the positive copies. This
budget has to cover more than 400 features, so that not more than
$50,000, on the average, can be spent on each production.

The average income of the Japanese is exceedingly small, when com-
pared with the income per capita of other nationalities; it is only
one-sixth that of the American, according to an index computed
by the Bureau of Commerce and Industry. Furthermore, this
average income is on the down grade, having dropped by 6 per cent
within the past ten years. The average daily earning for manual
workers during last November was one dollar for men and 40 cents
for women.

33 per cent of the male manual workers earn less than $0.80 daily
16 per cent of the male manual workers earn less than $1.00 daily
21 per cent of the male manual workers earn less than $1.30 daily
10 per cent of the male manual workers earn less than $1.50 daily
20 per cent of the male manual workers earn more than $1.50 daily

After taking the necessities of life into consideration, and based
an income of $45 per month, the Bureau of Home Affairs reckon;
that individuals have only $1.75 left to spend on amusements am
education.

With such a minute sum to spare, people can buy only a very fei
books and enjoy the cinema on rare occasions; and we have a some-
what easy explanation as to why Japanese people do not patroni;
motion picture theaters to the extent found in other industrialize
countries; their average income simply does not permit it, most
them earning barely enough to exist.

SOUND PICTURES

Sound pictures were shown for the first time in Japan eight yean
ago at the branch office of Universal Pictures, by Y. Minagawa, an<
afterward at the Shimbashi Embujo theater, but people seemed onl]
mildly interested in this new curiosity.

In 1929, a regular supply of sound films reached Japan, imported
by the various branches of the American firms established there.
The Shochiku Company, desirous of bringing something fresh intc
its chain of theaters, thought of experimenting with sound films, am
placed an initial order for six sound reproducers for their key
theaters in the larger cities of the Empire.

May, 1932] MOTION PICTURES IN JAPAN 635

Their closest competitors in the exhibition field, the Musashino
chain of theaters, immediately felt that they had to do likewise, and
even managed to get ahead of the Shochiku group in showing two
Movietone films, Parade and Songs and Dances of Hawaii, at the
Musashinokan (Theater) in the Shinjuku district of Tokyo and at
the Denkikan in Asakusa district on May 1, 1 929. A few weeks later,
on May 26th, the Shochikuza Theater and the Kogakuza Theater, in
Tokyo, were both presenting Redskin, a Paramount production.

These premiere presentations of sound films to a regular Japanese
cinema audience did not make the hit that was expected, mainly be-
cause they were shown with the loud speaker and the "Benshi"
competing with each other.

In the days of the silent film, the Japanese audience was able to
follow the action, for both foreign and Japanese films, through the
words of a ''Benshi" or announcer-interpreter who kept a running
fire of flowery comments describing every action, word, or thought
of the players, when not improving upon the scenario by "gags" of
his own. When the talking films were introduced, the exhibitors
discovered that the sound made so much noice that the Benshi could
no longer be heard, and the audience could not understand him.
Consequently, the sound was toned down, allowing the announcer's
voice to stand out. This, of course, was not satisfactory to the for-
eign patrons, who found that they could not hear the words of the
characters on the screen. Foreign patronage fell off to a minimum.
Moreover, the large section of the Japanese public consisting of stu-
dents of the English language and Japanese who have lived abroad,
was almost completely alienated.

There was still another difficulty. In the days of the silent film,
there was a considerable market in the smaller cities. The little
theaters could not afford to install sound equipment and many were
lost as clients for foreign films or continued with much dissatisfac-
tion. The market for the foreign film in Japan contracted to a
marked degree, and fears were expressed in many quarters that it
would be necessary for the film exchanges to close and to do all their
business through agents.

It was finally found possible to overcome this obstacle by run-
ning Japanese titles on a small screen at the side of the main screen,
but this has not proved wholly acceptable as the Japanese audiences
find that, while reading what is written at the side, they often miss
the action on the main screen. One major company hit upon the

636 M. Ruox [j. s. M. P. E.

idea of superimposing the titles upon the actual pictures, and several
films handled in this manner have already been released, and proved
to be quite successful.

Notwithstanding the luke-warm reception accorded to sound films,
far-seeing producers decided to try their hand at Japanese talkies.
To the late S. Makino, owner of the Makino Productions Company,
should go the pioneer's laurels. In cooperation with M. Tojo, owner
of the Eastphone Company, he produced Modoribashi, the first Japan-
ese feature recorded with sound. The disk recording was carried out
at the Nitto record factory. This was followed by Kyoen-roku, but
soon after this second production the untimely death of Makino
brought his meritorious efforts to an end. The Nikkatsu were soon
to follow the lead shown by Makino and, also using the East-
phone system of disk recording, started on the production of Hashi-
suka Koroku, a classic feature in 12 reels. The first show of Makino's
Modoribashi took place on the 17th of July, 1931, at the Daimai Hall,
in Kyoto, whereas Nikkatsu 's Hashisuka Koroku was put on at
the Fujikan Theater in Asakusa district, Tokyo, on the 1st of August.
Unfortunately, both these early sound productions were far from per-
fect, due mostly to the absence of recording experts; they, neverthe-
less, indicated what scope there was for the adoption of sound in
Japanese productions and the real public demand there was for this
type of film, even though its adoption may mean the ultimate dis-
missal of 6000 Benshis now employed by the Japanese cinema
theaters.

Mr. Watanabe of the technical staff of Teikine is the inventor of
a system that uses wax disks. One of the distinctive features claimed
by the inventor is that, with his apparatus, the stylus progresses
across the radius of the wax record instead of the record turntable
moving over.

Besides the sound machines of Japanese manufacture already in-
stalled in various theaters, there are two or three newcomers now
appearing on the market. Of these, is Shochiku's sound-on-film
Shochikuphone, which was presented during August for the first
time. While little detail is available concerning this apparatus, it
is said to be built on lines very similar to RCA recorders.

Katsumi Toyoshima, of Osaka, has invented a recording apparatus
for sound-on-film, the main characteristics of which seem very similar
to those of the Movietone. The makers claim that it can be easily
placed on a truck and will prove most suitable for outdoor work.

May, 1932] MOTION PICTURES IN JAPAN 637

EDUCATIONAL AND PROPAGANDA FILMS

For many years the Japanese Government has made wide use of
motion pictures for educational, health, and social propaganda, and
advertising, through its various departments or ministries.

As regards educational films, while they do not yet form part of
the school program, every encouragement is given to widen the use
of motion pictures in the classroom, and there is good reason to be-
lieve that as soon as economic conditions permit, we shall witness the
adoption of visual education by the Japanese Government; which
means all schools throughout Japan, as education here is organized on
a strictly national basis and even private schools have to follow the
main lines of the education program as laid down by the Department
of Education.

National education has been one of the most powerful factors in
building up modern Japan. The Department of Education was estab-
lished in 1870, and the educational law promulgated in 1873. The
whole area of Japan was divided into eight districts, each one of
which was subdivided into smaller middle school districts, which,
in turn, were also divided into many primary school districts.

The educational expenditure on elementary and secondary schools
is borne by the people of the particular localities in which the schools
are situated, but the government gives an annual subsidy of more than
80 million yen, or 60 per cent of the average total expenditure of 135
million yen.

We have for the whole of the Japanese educational system a grand
total of over 45,000 teaching centers with 300,000 teachers to run
them and 12 million pupils attending them regularly; an immense
field, yet almost untouched !

The Ministry some time ago planned a wide educational campaign
by means of the screen, and a vast program of 180 pictures was de-
cided upon. Several of these pictures were not for the schools but
for the masses, having for their purpose the guidance of national
thought on such things as "the necessity for the government to adopt
a retrenchment policy." In one of those "Government Justification"
pictures, three members of the Cabinet, including the former Prime
Minister, Mr. Yamaguchi, appeared in the course of the action, ad-
dressing the public as they would have done on a platform. The
Governors of Prefecture are strongly supporting the efforts of the
government, by advocating the production of more films of educa-
tional value for both children and adults.

638 M. Ruox [J. S. M. P. E.

The Ministry of Education went even further, and announced that,
in order to promote social education through amusements and recrea-
tions, their intention was to enlarge their field of action, which up
to now was more or less limited to making or recommending films
suitable for education in schools.

The Mombusho (Ministry of Education) would now give free
cinema shows in cities and villages. Two touring units are carrying
out the scheme, one operating from Tokyo, the other from Osaka.
Many of the films shown by these units are specially produced to
educate the adults more than the children, to their duties as good
citizens of the Empire, as may be seen by a short list of the subjects
screened:

Promotion of street etiquette and prevention of street accidents.

Prevention of fire.

Duty to pay one's taxes and the need of thrift.

Other departments of the government have also been regular users
of motion pictures but more for special propaganda than for educa-
tion. The Ministry of Railways has had several films produced
for free distribution in Japan and abroad with a view to inducing
tourists to visit Japan and Korea. Several of these films had copies
printed with English titles.

The Ministry of the Navy has subsidized production of historical
films dealing with Japanese naval victories, while the Ministry of
Home Affairs attends to the recording of the Emperor's and the Im-
perial family's activities and movements, as well as such other sub-
jects as emigration.

All such films are circularized free of charge by the various min-
istries or prefectures responsible for their production, and a good
many among them have been well received by the public, even special
propaganda films explaining the causes of depression and retrench-
ment.

The two leading newspapers of Osaka, the Mainichi and the Asahi,
self-appointed educators of the people of Japan, decided to use mo-
tion pictures as one of the better means of fulfilling their national
task. They have released films of educational nature, as well as
others of a more "newsy" style.

During the summer months the All- Japan Motion Picture Educa-
tion Society runs a summer college for education by motion pictures,
and success has rewarded their efforts, many teachers expressing their

May, 1932] MOTION PICTURES IN JAPAN 639

surprise at the speedy progress witnessed, especially in geography and
biology lessons.

The Japanese branches of the League of Nations and of the Red
Cross Society have also subsidized the making of special films, which
they are now using for propaganda purposes; even the Korean
Government has thought it well to use the screen to educate fisher-
men against the evils of casting their nets in forbidden waters, and
the Hongangi Buddhist temple in Kyoto commissioned the produc-
tion of a propaganda film, feeling certain that new adherents could
be obtained with the help of the screen.

The latest gesture of the Ministry of Education has been the offer-
ing of three national trophies for the best pictures of educational
value. There will be three classes, "Strictly Educational," "Recrea-
tional," and "Artistic," one trophy going to each class.

There are, in Japan, therefore, fine prospects for educational and
propaganda films, and the same might be said of the purely adver-
tising films. Everyone is convinced that the screen is the finest
medium to inculcate knowledge or carry a forceful message to the
public. The moment the government's budget and the various
official funds will permit it, there is no doubt that important sums will
be allotted to motion picture production and distribution.

JAPANESE EXPORT TRADE

There is no firmly established export trade in motion pictures.
Most of the films exported until last year were interest films — scenic,
industries or customs of Japan — the majority having been produced
by the Ministry of Railways as tourist propaganda, or as Japanese
productions for exhibition where Japanese communities in foreign
lands are important enough to justify the transportation of such films.

The main problem is that of language and, with the wide adoption
of sound by motion picture theaters throughout the world, there is
the fear that before long there will be still smaller possibilities of ex-
porting Japanese film. Export business will probably be limited to
cities in which the Japanese population is sufficiently large to keep
a Japanese theater going. The present distribution of Japanese
motion picture theaters in foreign lands is as follows:

In Shanghai, China 2

In Manchuria 7

In California (permanent) 12

In California (traveling) 8

In Hawaiian Islands 8

640 M. RUOT [j. S. M. P. E.

Most of the Japanese films shown in the United States are second-
hand films. They are sold outright for about $.07V2 per foot for
regular subjects, and $.15 per foot for feature films.

Of the motion picture theater companies in the United States, there
are nine owned by Japanese: 4 in San Francisco, 4 in Seattle, 1 in
Stockton, and 3 in Hawaii. On the other hand, there are no theaters
in the South Sea Islands owned by Japan, although the number of
immigrants there is continually increasing. On the whole, the export
of Japanese films is steadily increasing.

POLICE REGULATIONS

Japan is a country where the police rule supreme, satisfied as they
are with the "sanctity" of their mission and aware of the importance
of their position as governmental and not mere municipal agents

Censorship of films comes under the control of the Tokyo office o1
the police department. To each film submitted for censorship must
be attached a statement as to the exact length of each reel, a detailec
synopsis of the film in Japanese, a list of all titles and sub-titles in
English, plus a translation in Japanese of each of them. With the
advent of sound films, it became necessary to add to this already long
list of documents, a copy of the spoken text and a translation in
Japanese of all speech emanating from the loud speaker. This means
a tremendous lot of work for the poor importer of foreign films, but
the police want it and that's all there is to it.

Besides this, the Japanese words that the Benshi, or announcer-
interpreter, is to use when "reading" and commenting on the plot of the
film to the audience during the show must also be written down and
submitted to the censor for his approval or correction.

Once a film has been passed by the police, no alteration or revision
of the censored synopsis, censored film, or censored text as spoken by
the Benshi is permitted. Fines are imposed whenever an infringe-
ment of these regulations is discovered, and the showing of the film
is suspended until the matter is settled with the censor's office, which
sometimes takes several days as there is only one censoring office in
Tokyo.

Foreign films dealing with the army and with war are subject to a
special and additional censorship by the press department of the
gendarmerie headquarters. For some time, the gendarmerie was
using the sound-wired projector owned by the police censor at the
Department of Home Affairs, but this year they decided to have their

May, 1932] MOTION PICTURES IN JAPAN 641

own projector at the gendarmerie headquarters so as to be able to
examine the films at greater leisure.

The gendarmerie, in this matter, represents the army or the navy,
and it is as such that they also had to step in and issue new regula-
tions concerning the making of motion picture scenes within the many
strategic zones that one finds scattered all over Japan. Under their
new regulations, the gendarmerie exercises an extremely strict super-
vision of the "shooting" operations as well as a supervision of all the
film used, besides the censoring of the finished production.

Most important of all were the new regulations issued by the
Tokyo police and in force since April 1, 1931. Under these regula-
tions the projector speed must not exceed 28 meters per minute when
running sound film and 24 meters when showing silent pictures; the
duration of a single show must not exceed four hours, during which
not more than 5500 meters of film can be screened.

These new regulations met plenty of opposition from the exhibitors,
more specially those of the Asakusa district in Tokyo, where cinemas
are many and do a flourishing trade on Sundays and holidays. During
such days, in order to get the most people in, the duration of the
show was shortened but without alterating the number of subjects
shown, by simply increasing the speed of projection to its utmost
limit, often at the rate of 115 feet per minute. Meetings were held,
deputations were sent to the police bureau, all sorts of arguments
were brought forward, and tests made to obtain a reversal of the
regulations, but all to no effect. The police stood firmly on their
ground, stating that the new rules made for greater safety and at the
same time protected the public who were not getting "value" when
seeing a film screened at double the normal speed.

SUB-STANDARD FILM

Amateur movies are becoming one of the most popular hobbies of
the Japanese, and the user of amateur cameras and projectors can
now find in Japan all the very latest accessories, many of which are
already being made locally, even reproducing attachments for 16
mm. projectors. We have the "Marvel" attachment for sound on
disk, retailing at $40. We have had, of course, for several years,
Japanese-made 16 mm. cine cameras and projectors offered at prices
substantially lower than the imported article, but due to the poorer
quality of the domestic apparatus, their presence on the market has
never affected the sale of the foreign machines.

642 M. Ruox

There is quite a large quantity of 16 mm. safety positive film being
used in cameras in lieu of negative, simply because of its cheaper price.
It is reversed by outside firms, who usually blame the film manufac-
turer for the unsatisfactory results obtained with the reversed positive
film. This pernicious habit is growing, and unless manufacturers
take steps to prevent it, the reputation of home movies for producing
good quality films will be seriously damaged, as the amateurs are
never told by the unscrupulous dealer that they are being supplied
with a film never manufactured for use in a camera and, still less,
for ultimate reversal.

A MACHINE FOR PRINTING PICTURE AND SOUND
SIMULTANEOUSLY AND AUTOMATICALLY*

O. B. DEPUE**

Summary. — This paper describes a machine designed for printing picture and
sound records simultaneously and automatically. It is so constructed that it can
print full-width pictures, with or without masking the sound track, or the sound track
only. Provision is made for printing news weeklies or single-system negatives, as
well as the double-system or separate sound negative. Two prints can be made
simultaneously from a single negative, or one reel can be printed on one sprocket and
a second and different reel on another. The capacity of the machine is 5100 feet per
hour per finished print.

With the adoption of sound pictures by the motion picture world
came new demands on the laboratory to print both picture and sound
on a single positive, but of necessity not at the same position on the
positive. This has been accomplished in most cases by passing the
positive through the printing machine twice, or by having two
machines, one arranged for printing the picture with the sound track
masked, the second for printing the sound track with the picture
masked. This reduces the capacity of the printer to less than half
its normal capacity when printing the picture only.

A printer that will not only print the picture but the sound track
as well, both at the same time, should more than double the footage
in a day's work. It would have other advantages as well: eliminat-
ing mistakes in handling the positive from machine to machine,
and keeping records as to what has already been done or is still to be
done. Instances have been known in which as many as ten reels
in one lot have been marked as having been completely printed,
only to discover after development that no sound track had been
printed. The one-operation sound printer eliminates such possi-
bilities. Also, in the matter of economy in operation, one has but
to calculate the saving of employers' general overhead, in providing
space for more printers, the insurance on employees, and necessary
additional equipment for employees, such as lockers, washrooms,

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** O. B. Depue, Chicago, 111.

643

644

O. B. DEPUE

[J. S. M. P. E.

recreation, etc. A printing machine that will print both sound and
picture rapidly and automatically, and in a single operation, would
appear to be very desirable. If such a machine could be used for
regular printing as well, and to double the production by simul-
taneously printing two positives from a single negative, or
simultaneously printing two different reels by simply threading the
machine differently, its place in the laboratory should be assured.

PQ$iT/V£
HAND WHEEL

FIG. 1

Showing complete double machine with full automatic light control
for each printing sprocket.

In the multiple sound and silent printer (Fig. 1) described in this
paper all these features are combined. The machine is, in effect,
two printers in one; each printing head is alike in construction
and can print full-width pictures, with or without masking the
sound track, or it can print the sound track only. Its high speed
and automatic operation are two of its distinct features. Blinking,
unsteadiness, and flicker have been overcome by the driving mecha-
nism and by providing for perfect contact at the printing aperture.

May, 1932]

PRINTING PICTURE AND SOUND

645

The driving mechanism, as the rear view (Fig. 3) shows, is positive,
from the directly coupled motor at the bottom to the sprocket at the
top. The speed reducer is directly coupled to the motor, and a posi-

FIG. 2.

Front view showing thread-up when printing two different reels
at same time.

tive four-jawed clutch drives a single vertical shaft which, in turn,
operates all the sprocket wheels by right and left helical gears, hard-
ened and lapped. The one rewind belt is driven positively; and each

646 O. B. DEPUE [j. s. M. p. E.

rewind disk driven by it is individually controlled by an adjustable
pair of knurled thumb-nuts and a coiled spring pressing together two
friction plates separated by a thin pad of oil-soaked felt. Once
adjusted, the pressure remains unchanged for months without
further attention, the gentle tension thus provided allowing the
stopping and starting of a 1000-foot reel at any stage of the printing,
without unduly straining the film at the hold-back sprocket.

Referring to the front view of the printer (Figs. 1 and 2) attention is
directed to the printing sprocket housing, which is also the lamp
housing. It is here that one of the main features of the printer is
to be found. The printing lamps have a capacity of but 40 watts,
a frosted stock mazda lamp being sufficient to furnish ample illumi-
nation for printing at a speed of 85 feet per minute with an aperture
opening 5/i6 inch. Referring to the rear view (Fig. 3) the three-way
adjustment of the lamp base and socket will be seen. The scale
is graduated uniformly in sixteenths of an inch, the forward and
backward movement being l!/2 inches. The two lamps are easily
matched and may also be easily matched to other machines in daily
testing.

The light is controlled by a rheostat, the small filaments of the
40-watt bulbs responding more quickly to changes of voltage than
larger filaments. The light controls provide for 152 changes and 22
densities, or a smaller number if desired. The rheostats are easily
detached, and rheostats of any capacity up to 500 watts can be
substituted. This provision, of adapting the control to other printers
having different lamp capacities, is another economic feature of
the machine. By merely shifting the lever provided on the back
of the light control, one rheostat is substituted for the other. A
250-watt lamp is suitable for the optical printer, and a 40-watt lamp
for the sound printer. A mechanical filter of sufficient latitude is
provided by a Lovejoy flexible coupling having a four-point rubber
element 5/ie inch in diameter, which acts as a shock absorber when
throwing in the positive clutch, and assures smooth operation while
printing.

The multiple, or double, positive printing feature is worthy of
mention. On the left of the front view (Fig. 1) will be seen an
additional sprocket and rollers. The first printed positive is passed
over the top of this roller and is rewound on the upper left disk.
The negative passes down to the lower printing sprocket after passing
under the tension roller and the interrupter roller, and the second

May, 1932]

PRINTING PICTURE AND SOUND

647

unprinted positive film is taken from the lower disk and passes
under the lower side of this feed sprocket. It then passes under
the tension roller and joins the negative at the printing aperture,

CLUTCH

FIG. 3.

Rear view showing motor speed reducer, main vertical driving
shaft by helical gears.

and a second print is made and rewound on the bottom left-hand
disk while the negative is rewound on the bottom right-hand disk.

648 O. B. DEPUE

The left-hand light control is connected to the upper printing sprocket,
and the right-hand light control is connected to the lower printing
sprocket. These light controls are provided with the male half
of a three-circuit plug connector. The printer is provided with a
flexible armoured cable terminating in the female half of the con-
nectors, so that any light control can be connected to any printer
so far furnished. At the top of the printer unit, on the left of Fig. 2,
is located a footage meter, in addition to a hand wheel for enabling
the operator to advance the film readily when threading and syn-
chronizing.

On the lamp housing at the upper right-hand corner (Fig. 2) is a
small plate with numbers 1, 2, and 3, and a small knob and sliding
lever. No. 1 is the sound track printing aperture, No. 2 is the picture
aperture and sound track masked, while No. 3 is the full-sized picture
aperture. On either side of both printing sprockets is a small screw
knob which allows an almost instantaneous opening and closing of
footage number apertures on each side of the sprocket. This simple
device makes footage printing selective and useful in film printing,
or in printing for disk reproduction of sound.

The machine is very flexible in use, being in effect a twin printer.
News weeklies or single-system negatives may be handled as easily
as the double-system or separate sound negative. Two prints
from a single negative can be made; one reel can be printed on the
upper printer sprocket and a second and different reel on the lower
head, the No. 3 setting providing for two distinct printers, each one
having its own light control independently of the other. Of course,
both are operated at the same time. The capacity of the machine is
5100 feet per hour per finished print. In actual use, sound pictures
have been printed at a rate greater than 4000 feet per hour. An
output of 302,000 feet of film was recently made by this machine
in 100 hours of continuous operation. When used as a multiple
unit, printing two positives from a single negative, the capacity is
somewhat greater than 60,000 feet per day. Only one negative has
to be rewound and handled. A VVhp., 1750-rpm., 60-cycle, a-c.
motor is used. The machine is wired for a-c. operation except for the
two printing lamps, which are operated on a separate 110-volt, d-c.
circuit The bearings are of bronze; space is provided at each
sprocket wheel bearing for a felt pad soaked in oil, so as to assure
proper lubrication.

TIME-AND-TEMPERATURE VS. THE TEST SYSTEM FOR
DEVELOPMENT OF MOTION PICTURE NEGATIVES*

WILSON LEAHY**

Summary. — -Two methods are in current use for controlling the developed density
of picture negatives. These are: (1} the test system, and (2) the time-and-temperature
system. In this paper the two systems are compared briefly and their advantages and
disadvantages are given.

Upon the advent of sound as an integral part of the motion picture,
and through the necessity of combining on one positive both the
sound and the picture, standard laboratory practice at that time
underwent an unprecedented change. After a rather hectic period
of controversy with sound engineers it became obvious to most
laboratory men that the heretofore satisfactory rule-of-thumb
methods must be discarded; that the standardization of solution
contrasts in processing sound tracks must be accompanied by a like
standardization of densities and exposures of the picture negative,
in order that the highest quality of picture may result in composite
printing through a positive solution primarily formulated and main-
tained for sound.

The introduction of sensitometric measurements enabled the
laboratory to control the degree of contrast and density in solu-
tions; but opinion was, and still is, divided as to the most exact
means of regulating the developed density of the picture negative.
Two methods are in use at present: namely, (1) the test system,
and (2) the time-and-temperature system. In this paper the two
systems will be compared briefly, the facts on which the comparison
is grounded having been gained from lengthy and intimate experience
with both systems.

In a laboratory where the time-and-temperature method is em-
ployed, the first requisite is, of course, a smooth, evenly balanced,
negative formula, maintained so as to provide a constant trans-

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Universal Pictures Corp., Universal City, Calif.

649

650 WILSON LEAHY [j. s. M. p. E.

mission and gamma. It may be assumed, also, that there exists
the cooperation between the laboratory and the cameramen requisite
to assuring even and uniform exposures and to establish in the camera-
men a feeling of security as to the unvarying development of his
negative from day to day. Admitting this condition, the enumerated
advantages of the method are as follows :

(1) Continuous machine development without alteration of speed or strength
of solution and precluding the possibility of error on the part of the operator.

(2) Eliminating the need of making test shots on the set, which are at once
expensive and in many cases inaccurate.

(3) Eliminating the personal variability of the man reading the tests.

(4) Reproducing on the screen a replica of what was shot without exaggerating
or losing contrast or tone value.

(5) Securing full, deep development, proportionate to exposure.

(6) Eliminating the necessity of rewinding to detach and sort tests.

(7) In the end, producing an even and uniform release negative as a result of
dispensing with the human element as much as is possible.

In a laboratory where the test system is employed, the same
amount of care is necessary in maintaining the solutions. It is
obvious, however, that the same spirit of cooperation between
the cameramen and the laboratory cannot exist here as in the
laboratory employing the time-and-temperature method. The timer,
who is only human and who is therefore liable to err in his judg-
ment, is as a rule the cause of no little friction between the two
departments.

The necessity of obtaining good tests is also a contributing factor
of contention. The system, however, has its advantages; it provides:

(1) Partial protection to cameramen who have inadvertently made mistakes
in exposure.

(2) Partial protection to cameramen ignorant of solution conditions in the
laboratory.

(3) Partial protection to cameramen who have been forced to shoot under
adverse conditions.

It is notable that the test system seems to have been devised to
protect the cameramen regardless of the cost of processing or of the
hazards to which the film is exposed in the laboratory. Experience
has shown that the average cameraman, possessing sufficient "film
sense" to deserve the name, is plastic and ingenious enough to adapt
himself to the time-and-temperature method in a comparatively
short time; whereas, with the test system, and with negative de-

May, 1932] DEVELOPMENT OF NEGATIVES 651

velopment proceeding at varying speeds from night to night, he is
unable to settle down to any standard of exposure.

The principal advantage offered by the test system is its ability
to smooth off rough negative to conform to the laboratory printing
scale regardless of quality. It is an acknowledged fact that the
printing scale has an all-important bearing on the general appearance
of the release print, particularly on the evenness and uniformity of
the blacks. But in obtaining this uniformity of printing scale, it
is often necessary in the test laboratory to under-develop an over-
timed, flat negative, and to over-develop an under-timed, contrasty
negative, procedures which at once sacrifice quality. This is self-
evident and beyond dispute, and can be charged, first, to lack of
cooperation between the cameramen and the laboratory and, second,
to the maintenance of a system that permits a variable element to
intrude between the cameraman and the screen. It is a condition
brought about by under- developing an over-exposed negative that
possesses considerable inherent contrast, and over-developing a com-
paratively flat negative, instilling in the cameraman a false confidence
which in the end proves his undoing. For, without much thought
or hesitation, he will over-time a flat subject and under-time a con-
trasty one.

In the time-and-temperature laboratory, this constant see-saw
is avoided. The effect of exposure, filter values, lighting arrange-
ments, etc., are all illustrated clearly to the cameraman by the
unjuggled negative and print. He assumes the mental attitude of a
student, with his own work as a text, the result being almost immedi-
ately evident on the screen.

It has often been argued that this standardization of exposures
stifles a cameraman's individuality, but such is obviously untrue.
No restraint is exercised over the use of gauze, filters, or composition ;
the cameraman is free to use any means at hand to enhance the
artistic beauty of his product; and an additional advantage is
placed at his disposal through the consistently uniform development
of his negative.

STUDIO PROJECTION AND REPRODUCTION PRACTICE*
JOHN O. AALBERG**

Summary. — The number of projection rooms in Hollywood studios varies between
one and fifteen, depending on the production capacity of the studio. Projection
distances average about 60 feet. In general, reproducing equipment is furnished
by the company' whose recording apparatus is used. During shooting and editing
of a picture, the sound track and picture are on separate films, practically doubling
the amount of equipment needed and calling for special synchronizing devices.

Daily and weekly routine checks covering frequency characteristics, power levels,
and screen brightness are described, as well as small projectors and reproducers used
for inspecting release prints in film processing laboratories. The paper also covers
special applications, such as reproducers on stages (play-backs] used for furnishing
music or for special work, as in split mat photography, special uses in scoring,
trick work, etc.

The final link in the technicians' daily work of making motion
pictures is projecting and reproducing them. Here on the screen,
the art department, responsible for sets and costumes, sees the
results of its work; the make-up artists see that their skill in beauti-
fying the stars has registered properly; and the cinematographers
check their photography and the processing of the film by the lab-
oratory. From the screen, the sound technicians hear their recording
and ascertain whether or not it matches the photographic action
for sound perspective and for other points of recording finesse.

The studio projection rooms are small, ranging in seating capacity
from ten to two hundred, and varying in appointments from a bare
acoustic plastered room, where the film editor checks his picture
cutting, to the elaborate rooms of the executives. A large studio
may have fifteen such theaters and the smaller ones possibly one or
two.

Care is taken in these theaters that the equipment produces only
average theater quality so that the technicians do not become too
optimistic about their results. To insure uniform work from day
to day, routine tests are made on the equipment characteristics.

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** RKO Studios, Inc., Hollywood, Calif.
652

PROJECTION AND REPRODUCTION 653

The latter is essential because scenes are made from day to day
that must match sequences made weeks or perhaps months before
in both picture and sound quality. Screen brightness is determined
with a Macbeth illuminometer and is kept at approximately twelve
foot candles. Daily power level tests are made on all reproducers
by running a variable frequency film. During the day the sound
level of the reproducers is verified frequently by reading the level
obtained by running a piece of 1000-cycle sound track, the optimum
level for each theater having been previously determined by the
sound technicians. At one time it was practice in a studio here to
attach a piece of 1000-cycle film to the end of each reel of "dailies"
so that the level could be verified after every reel. It is essential
that the equipment be maintained with systematic care and kept
as nearly perfect as possible so as to disclose any imperfections in
either picture or sound; otherwise costly retakes might be ordered
when the imperfection noted was actually the result of faulty re-
production in either picture or sound.

Studio reproducing equipment is furnished by the company
whose recording machinery is in use in the studio. It is usually
maintained under the supervision of the sound department and is
operated by skilled projectionists. During the filming and editing
of a picture in the majority of studios, the sound track is kept on a
film separate from that of the picture so as to simplify editing.
This procedure practically doubles the reproducing equipment
otherwise required, and necessitates special synchronizing devices,
such as electrical interlocks for synchronizing the projector with its
sound reproducer. Such an arrangement may also interlock two
projectors so that the sound track and the picture may be viewed
simultaneously by using the proper masks in the picture apertures.
Projecting the sound track affords an easy method of inspecting it
for dirt, scratches, placement, modulation depth, etc.

Every studio, in addition to the regular projection rooms, has a
scoring room where appropriate music or special sound effects may
be .synchronized with the picture. These rooms are of such di-
mensions and acoustic design as to accommodate large orchestras
for scoring. Projection rooms and recording channels are attached
to them, and in many cases a re-recording channel. With the latter
arrangement it is possible to add music and other effects to the sound
that was recorded when the picture was photographed.

One laboratory has developed a small projector and reproducer

654 JOHN O. AALBERG

for inspecting every reel of release print that they make. The
picture projected by it is approximately two feet square and the
sound is heard by the inspector through ear-phones. Special features
of the machine are quick starting, stopping, ease of threading, and
the absence of rollers or gates that might damage the film. In one
of the color laboratories use is made of the Selsyn motor remote-
control principle to allow focusing of the picture from the theater.

Unique uses are sometimes made of projection and reproduction
apparatus. In one case the projected picture is utilized by the trick
photographer to secure a scene usually made by special photographic
processes. In this instance a picture background is projected on a
treated glass screen by rear projection and rephotographed with
foreground action. This process has been simplified by the electrical
interlock which maintains synchronization between the camera
shutter and the projector pull-down mechanism and may, therefore,
operate with the projector shutter removed, insuring maximum
light on the screen.

Another special use of reproduction is immediate play-back of a
scene recorded on wax. Play-backs, whether from film or disk,
are useful in instances where an actor is playing a dual role and must
speak to his counterpart when it is necessary for both to appear on
the screen simultaneously in the final result. He is photographed by
split mat photography as he speaks the lines of one role. These are
then played back for timing as he is photographed in his counterpart,
allowing him to talk to himself.

The foregoing is not the practice of any one company, but a de-
scription of the general practices in Hollywood studios.

SIZE OF IMAGE AS A GUIDE TO DEPTH OF FOCUS IN
CINEM ATO GRAPH Y *

J. F. WESTERBERG**

Summary. — The questions discussed in this paper are: (1) The depth of focus
vs. depth of field. (2) The size of the permissible circle of confusion. Should it
be a constant value or should it vary as in still photography? It is concluded that
it should remain fixed. (3) Magnification as an index to depth. The simple
rule that depth varies inversely as the square of the magnification may prove to be a
very practical yardstick in the appraisal of depth in the photographing of near-by
objects. A table is shown in which the magnification is estimated from the size of
the figures on the ground glass. The corresponding depth is given for various stops.

In A. C. Hardy's paper on "The Depth of Field of Camera Lenses,"1
several questions were brought up that merit discussion from the
point of view of practical cinematography.

The common misunderstanding in regard to the terms "depth of
focus" and "depth of field" was mentioned, and also the question of
whether in motion picture photography the size of the permissible
circle of confusion should vary or remain constant. The most
significant point that was brought up, however, was the relation of
magnification to depth. The simple rule, that depth varies inversely
as the square of the magnification, may prove to be a very practical
yardstick in the appraisal of depth in photographing near-by ob-
jects. In the past, too many factors have been involved to warrant
any other method than direct visual examination of the image on
the ground glass.

DEPTH OF FOCUS VS. DEPTH OF FIELD

It is common practice among photographers to refer to all problems
of depth by the expression "depth of focus." Strictly speaking,
depth of focus should only be used in referring to the leeway that one
has in focusing upon an object at a fixed distance. Thus, under
certain conditions, if it is desired to focus at 15 feet, a satisfactory

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Hollywood, Calif.

655

656 J. F. WESTERBERG [J. S. M. P. E.

image may be obtained of an object at that distance if the point of
sharp focus falls anywhere between 13 and 18 feet.

Depth of field, on the other hand, assumes that the lens has
been focused correctly at the desired distance, in this case 15 feet,
and indicates that all objects between 13 and 18 feet would be in
practical focus.

It is doubtful if in most cases cameramen actually adjust to the
exact focus, especially when following focus. Depth of focus is,
therefore, an ever-present life-saver. When attempting to focus
so that objects both near and far shall seem sharp, it would be tech-
nically exact to say that depth of field is under consideration, although
practically it is much simpler to retain the term depth of focus in an
all-inclusive sense.

FIXED CIRCLE OF CONFUSION

It is well known that in still photography a larger circle of con-
fusion can be tolerated when the picture is big and the lens of long
focal length than when the picture is small and the lens of short
focal length. Big pictures are usually looked at from a distance,
while small pictures have to be examined close at hand. It is assumed,
therefore, that in still photography the size of the permissible circle
of confusion may vary directly as the focal length. A value of
Vioooth of the focal length is the accepted figure. This works out,
for example, as Viooth an inch for a ten-inch lens and V25oth an inch
for a four-inch lens.

Can this line of reasoning be followed in the case of a motion
picture? Apparently not, because the size of the picture and the
viewing distance are constant for any one spectator. The fact that
the faces on the screen vary in size is evidently immaterial in this
case.

The result of allowing a sliding scale is quite apparent to the
eye when looking through a camera. The longer focal lengths, on
the one hand, do not live up to their higher rating and the wide angle
lenses, on the other hand, have depth to spare. A fixed circle of
confusion of Vsooth an inch seems to be about correct in practice.
This is Vioooth of the focal length of a two-inch lens.

SIZE OF IMAGE A MEASURE OF DEPTH CAPACITY

Hardy's proposal to consider depth as a function of the magnifica-
tion has practical possibilities that should not be ignored. Nearly

May, 1932]

IMAGE A GUIDE TO Focus

657

all the scenes in a motion picture are made at close range, some-
where between a full-length figure and a close-up. The subject of
depth becomes greatly simplified when we consider, for instance,
that the depth in photographing a waist figure is always practically
the same for any given stop, regardless of the focal length of the lens
used. With this in mind it becomes possible to construct a table
in which any reference to focal length or distance of object becomes
superfluous. This simplifies matters considerably. All that re-
mains to consider is the stop and the magnification. The stops
are easily read, of course, and the magnification can be readily
estimated with sufficient accuracy by reference to the ground glass.

Relation of Depth to Magnification in Motion Picture Lenses
Image Data Total Depth

At least one-half of total depth available
Based on aperture 0.6 X 0.8 of an inch beyond plane of critical focus

F/2 F/2.8 F/4: F/5.6 F/8

1.0 in. 1.4 in. 2.0 in. 2.8 in. 4.0 in.
2.0 in. 2.8 in. 4.0 in. 5.6 in. 8.0 in.

Magnifi
cation

Height of Subject
Included at Point
of Focus

Character
of Scene

1/11.2

6.7 in.

Insert of

hands

1/15.6

9.4 in.

Action

insert

1/22.4

13.4 in.

Large

head

1/38.7

23 .2 in.

Close-up

1/46

27.6 in.

Bust

1/55

2 ft. 9 in.

Waist

4.0 in. 5.6 in. 8.0 in. 11.0 in. 16.0 in.
Close-up 1.0ft. 1.4ft. 2.0ft. 2.8ft. 4.0ft.
1.4ft. 2.0ft. 2.8ft. 4.0ft. 5.6ft.

figure 2.0ft. 2.8ft. 4.0ft. 5.6ft. 8.0ft.
Cutting

at hips 2.8ft. 4.0ft. 5.6ft. 8.0ft. lift.
Hands

showing 4.0ft. 5.6ft. 8.0ft. lift. 16ft.
Cutting

at knees 5.6ft. 8.0ft. lift. 16ft. 22ft.
Cutting

at ankles 8ft. lift. 16ft. 22ft. 32ft.
Full

length lift. 16ft. 22ft. 32ft. 45ft.
Medium

long shot 16ft. 22ft. 32ft. 45ft. 64ft.

The above table illustrates the simplicity of this method when
used for motion picture work. Purely minor variations such as the
effect of distance on the // value have been completely ignored.

1/65.5

3 ft. 3 in.

1/77.5

3 ft. 10 in.

1/90.7

4 ft. 6 in.

1/110

5 ft. 6 in.

1/130

6 ft. 6 in.

1/155

7 ft. 9 in.

658 J. F. WESTERBERG

It is hoped in this way to make the table simple enough to be of some
practical use in production. A table of this sort should prove
useful in many ways.

(1) It indicates at a glance the capacity in regard to depth of any particular
set-up.

(2) It indicates to what extent stopping down the lens will improve depth.

(3) It indicates to what extent a larger stop is justified under any given
circumstances.

A table like this, based on magnification of the image, should make
it possible for any one to obtain an accurate yet simple grasp of the
depth situation in photographing near-by objects with a motion
picture camera, and to know, without difficulty, how much depth
can be relied upon and utilized in any given case.

REFERENCE

1 HARDY, A. C.: "The Depth of Field of Camera Lenses with Special Reference
to Wide Film," J. Soc. Mot. Pict. Eng., XVI (March, 1931), No. 3, p. 286.

SOUND RECORDING FOR INDEPENDENT PRODUCTION S:

L. E. CLARK**

Summary. — The conditions of independent production are contrasted with major
studio activities, personnel, equipment, etc. The economics, time, and quality
requirements for sound recording in the independent field are discussed, and the
relations between independent producers and recording equipment manufacturers
are briefly referred to, as well as the technical and business problems to be met, and
the probable future developments.

Practically every technical phase of sound recording in the motion
picture field has been thoroughly discussed and reported in the great
number of invaluable articles that have been prepared by the engi-
neers in charge of this work. Several authors have discussed the
major problems of operation and have touched upon the economic
angles of studio recording; but all these papers have been written
from the view-point of the large studio that maintains a complete
sound department to operate its several sound channels. Inde-
pendent production has been at such a low ebb during the past two
years that the special problems of that field have not assumed suffi-
cient weight to warrant reporting.

In the past six months the trend has turned toward this type of
production, with the result that many new picture companies have
been organized, while several of the older independents have pro-
ceeded with renewed vigor. And with this increase in production
there is being developed a new method of sound operation, peculiarly
adapted to suit the needs of this type of work. In the first place,
independent production can be defined as production that must be
sold to another organization for release. Sometimes the release
agreement is completed before production begins, although many
pictures are begun before definite sales arrangements are made.
In any event, the producer is only certain of an outlet for relatively
few pictures and consequently is not in a position to make heavy
commitments over any considerable period in the future. He

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Clarco, Inc., Hollywood, Calif.

659

660 L. E. CLARK [j. s. M. P. E.

must write off his entire expense on one or two, or at the most,
six pictures.

Because of the huge development charges and patent expense,
the cost of owning and operating licensed sound equipment has been
prohibitive to the independent producer. Frequently he is not in
a position to make five-year, or longer, contracts; even in cases
where his outlet is assured, the royalties owed to the sound equip-
ment manufacturer accruing from a year's product fall far below the
minimum that he must guarantee, while the initial cost and operating
expense of even a single channel is an expensive item in itself. These
reasons are the principal factors that tended to discourage inde-
pendent production during 1929 and 1930.

An analysis of sound recording expense shows that fully 90 per
cent of the cost of owning a channel of equipment appears as fixed
charges — royalty, guarantees, amortization of the original capital
investment, sinking funds for properly caring for changes and
improvements in methods and equipment, and insurance on ex-
pensive equipment, being only a few of the major items under this
classification. These fixed charges constitute a five- to ten-year
contingent liability that is enough to make the owning of sound
equipment practically an impossibility for the producer of small vol-
ume. It is essential to keep the sound equipment busy most of the time
in order to reduce this enormous fixed charge to an operating point.

To this end, more and more of the independent companies are
adopting the system of renting sound equipment and facilities.
The company that rents out the sound equipment assumes the
contingent liability, and depends upon a steady volume of work
from many producers to keep its equipment busy. The individual
producer pays for only his own expense — the royalty accruing
against his particular picture and the operating expense of the
sound rental company. In this way it becomes possible for any
independent producer to secure the highest type of recording equip-
ment and personnel, assuring him a job of sound recording equal
to those turned out by the major studios, at a price that permits
him to operate without the risk of long-term contracts and guarantees.

Experience with these rental companies discloses a considerable
difference between recording methods in this field and in the major
studios. The sound engineer in the latter case is held to a constant
standard of speed and quality in every picture. He knows by
experience on previous productions just how much time spent in

May, 1932] SOUND RECORDING 661

rehearsals is economically justified for the particular type of product
turned out. He works for the most part with relatively few actors
and learns to correct for their individual voice characteristics. The
sound engineer with the rental organization, on the other hand,
must be prepared to be as careful and precise as the best when
operating with a company that takes six months to complete a single
production, and a few weeks later must be able to shoot as fast as any
newsreel cameraman when shooting a serial for another company.

It is next to impossible to find a single mixer who can adapt
himself to do both these types of work. If he is exactly the man for
the first organization he will be too slow and precise for the serial
company. In fact, the biggest single operating factor is the human
equation. The personality of the mixer and his assistants must be
suited to the particular type of work in hand: precise, patient,
thorough-going mixers for the picture that takes months to com-
plete, and quick-thinking, alert, well-founded men who know where
time can be saved without materially affecting quality for the pro-
ductions with ten-day schedules.

Equipment problems accordingly are reduced to a few paramount
considerations; high quality, dependable, and easily operated
equipment must be furnished always, the particular method of using
this equipment depending upon the type of production under way.
The equipment must be complete within itself. One channel with
spares mounted in a panel-body truck, together with complete power
supply for operating recorder and cameras, as well as the amplifiers,
is the preferred layout. With this equipment, high-quality re-
cording can be produced both on the sound stage and on location,
as long as one of the licensed types of equipment is used. With
equipment of either Electrical Research Products, Inc., or of RCA
Photophone, Inc., assembled in such a fashion, the independent opera-
tor can rent sound recording and be assured of the best possible job,
every bit as good as he could obtain by owning and operating equip-
ment himself.

Summarizing, the rental system of providing sound recording
facilities is greatly aiding the growth of independent production,
while at the same time this growth permits more and more invest-
ment by the sound rental companies in equipment. What the
future holds in store for this method of operation is of course un-
known, but the great economies that are effected by it speak well
for its permanence and growth.

SPECIAL PROCESS TECHNIC

VERN WALKER**

Summary. — The reasons for the increasing importance of special process photog-
raphy in motion picture production are briefly discussed. The influence of sound
in enlarging the scope of application of special process photography, the avoidance of
difficulties of recording sound in many natural locations, the use of special process
photography in avoiding hazardous stunts, the economic aspects, and various other
phases of the subject are briefly treated of.

The various methods in commercial use are outlined, and the technical difficulties
encountered in special process and trick work, the mistakes to be avoided, and the
solution of special problems are pointed out.

The improvements that have occurred in recent years in making
"trick work" have brought about a complete change of method, the
original single "stupendous trick shot" of the picture being sup-
planted by a number of shots, and, in many instances, entire sequences
being made by utilizing trick photographic processes.

Process photography has been widely used for some time, although
its general application was at first limited to the photographing of
"stunts" and spectacular scenes that were impossible to photograph
in the usual manner. The addition of sound to the picture brought
about a situation that required special photographic processes,
not only for stunts and spectacular scenes, but for the photographing
and recording of sound in the regular production scenes in the studio
under the requisite conditions that could not be obtained on actual
location. For instance, a scene to be taken within a trolley car
would be practically impossible from the standpoint of recording
sound. The extraneous noises, the transporting of sound equip-
ment, the tying up of traffic in the desired location, and the expense
involved make a suitable case for the use of process photography.

To complete such a scene, using the methods of process photog-
raphy, the specialist in this type of work obtains from the director
instructions as to the angles required and the approximate footage

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** RKO Studios, Inc., Hollywood, Calif.
662

SPECIAL PROCESS TECHNIC 663

of the scenes to be processed. A camera crew is sent out to "shoot"
the backgrounds for the scenes, under conditions exactly as they
would be in the actual circumstances of life. The section of the
trolley car is duplicated on the sound-proof stage, the action and the
sound are recorded, the background being put in either at this time
or after the foreground scene is completed, depending on the method
used. After the processed scene is completed, the sound depart-
ment puts in the desired background noises without impairing the
dialog.

The economic value of utilizing process photography is quite
apparent in the above situation, although this is a relatively un-
important example of what is being done. The expense of sending
a troupe to distant locations is to be compared with the cost of sending
only one or two men to the location to obtain the backgrounds,
the action to be included in the picture at the studio. Other ex-
amples involve unusual locations that cannot readily be found, or
imagined locations that do not exist in nature. Such are built in
miniature, the action, including full-sized actors and sets, being
processed in. Hazardous scenes that require the actors to ride fast
cars, scenes of runaway wagons, aeroplane stunts, etc., that in any
way endanger life or property, are usually processed into the picture.

The different process methods commonly used are briefly de-
scribed as follows :

The Williams processes, one of which employs the traveling matte
for blocking out the foreground action while the background is being
superimposed. This method allows the action to be shot first and
the background to be put in later.

The HandsMegl process uses two films, when photographing the
foreground action, and employs the principle of color separation
for withholding the exposure of the portion of negative required
for the background. The background is put in later.

The Dunning process is known as a transparency process; two
films are employed when shooting the foreground action. The
film in front of the unexposed negative is a transparent positive of
the background, its color being complementary to the color of a
backing placed behind the foreground set, the action being photo-
graphed with a light not complementary to the transparent positive
of the background. This method puts in the background at the
same time the foreground scene is photographed.

The Projection process, which, at this time, is being used with con-

664 VERN WALKER [j. s. M. P. E.

siderable success, consists in projecting on a transparent screen
the background desired, placing the actors and set in front of the
screen, and synchronizing the camera and projector at the time of
photographing. The manner of illuminating the screen for most
large shots has been quite a problem, but this difficulty will, no doubt,
be overcome in the near future, whereupon this method should then
be used extensively.

Many difficulties are encountered in making process shots, as two
scenes very seldom call for the same kind of treatment. Not only
does the photographic technic of superimposing the scenes become
a problem, but there are added the actual mechanics involved in
adjusting the action, the speed, and the lighting of the background
and the foreground, so that the finished picture may appear as one
in the same plane and atmosphere. Often these mechanical adjust-
ments are not painstakingly and properly made and, as a result,
the picture on the screen shows obvious signs of having been ''faked."
When properly done and cut into the picture, the average processed
shot is not noticeable to the uninitiated.

The technical hazards in process work are numerous. Photo-
graphicallyj the worst is known as "phantom," or "ghost." This
is the case when the foreground appears transparent, the back-
ground showing through. Another very common fault is known
as a "fringe," in which an obvious line, black or white, appears
around the foreground objects, giving an undesirable matte-like
effect.

Mechanically, in the photographic phase of processing, about the
only real trouble encountered is that of registration, necessitating
micrometric adjustments of the cameras and printing apparatus.
This trouble appears most often when a stationary foreground is
processed over a stationary background. If the two scenes do not
register exceedingly accurately, the projected picture appears
unsteady, the background and foreground wavering slightly in
opposite directions. The chance that this might be due to poor
perforations in the film has been found negligible. The greater
number of cases of poor registration are found to be due to the
shrinkage of the film or to faulty matching of the registering pins
of the printers and cameras.

The laying out of the process scenes is an engineering problem.
Angles, heights, and speed are of very great importance. At the
time the background is photographed, the height of the camera,

May, 1932] SPECIAL PROCESS TECHNIC 665

the angle it is tilted, the lens used, and the speed, if traveling, are
recorded. When the foreground is superimposed, the corresponding
conditions of the camera must conform to these measurements or
the results will appear obviously bad in perspective.

The things to be avoided in process work are so numerous, and so
seldom appear twice, that no two men experienced in this kind of
work will attack the problems in the same way, or even arrive at
identical solutions of a given problem, so that it is impossible to
prepare a blanket statement of what not to do.

A group of highly trained men specializing in nothing but process
photography has been developed to handle this work for the ma-
jority of the Hollywood studios. Every day a new commercial
use is being found for process technic, not only to overcome photo-
graphic and sound recording obstacles, but to save a considerable
amount of time and money.

ERRATUM

The following corrections should be made in the paper, Resume of the Pro-
ceedings of the Dresden International Photographic Congress, by S. E. Sheppard
beginning on page 232 of the February, 1932, issue of the JOURNAL. The di-
mensions given on p. 234 should read:

(I) Perforation Pitch 475 mm.

-0

(II) Width of Take- Up (also Feed) Sprocket

between Centers of Sprocket Teeth 28.15 mm.

-0.05
+0

(III) Over- All Width of Take- Up (and Feed) Sprocket 35.00 mm.

-0.20

COMMITTEE ACTIVITIES

REPORT OF STUDIO LIGHTING COMMITTEE*

EQUIPMENT

The report of the Studio Lighting Committee presented at the
Hollywood meeting dealt with the various illuminants that could be
employed for motion picture photography. This report supplements
the preceding one, and discusses the various kinds of lighting equip-
ment, power supply, and distribution systems and wiring practice,
in order to make available the information on lighting equipment and
practices employed in producing professional motion pictures.

An analysis of the characteristics of studio lighting equipment is
facilitated by grouping them into two general classes: (a) those
employed for general illumination, and (b) equipment particularly
adapted for modeling lighting. Lighting units of the first group are
characterized by a broad light distribution, 60 degrees or more, and are
used to produce a relatively uniform illumination over a considerable
area. Into this class fall the Broadside, the Rifle, the Dome, the
Scoop, Strip Lights, Backing Lights, Floodlights, and various other
devices giving a wide distribution of light.

Modeling lighting equipment gives a relatively narrow beam spread,
2 to 30 degrees, producing high intensities over limited areas. Typical
units of this class are the reflector spot (also called sun spot), the
lens spot, and the soft spot.

This grouping of lighting equipment is based on their more general
usage. However, studio lighting requirements frequently necessitate
the use of modeling lights for general illumination, and vice versa.

GENERAL LIGHTING DEVICES

Broadside unit. — The broadside unit (Fig. 1), available with both
incandescent and arc lamps, is provided with one, but more often two,
light sources. The lamp housing has a porcelain enameled steel
reflector for redirecting light, that would otherwise be wasted, back to
the area illuminated. The housing is equipped with holders so that
glass or silk diffusing screens may be used for creating, in effect, a

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
666

STUDIO LIGHTING COMMITTEE

667

larger area source, thus softening the light. The light distribution is
quite uniform over a vertical and horizontal angle of 130 to 140
degrees. The incandescent lamp broadside uses 1000- or 1500- watt
pear-shaped bulb lamps, and the arc type uses two 35-ampere auto-
matic arcs operating in series. Broadsides are mounted on a three-
legged adjustable stand, which permits the lamp to be raised from
about 4*/2 to 8 feet, and tilted. Means are also provided for attaching
the lamp house to the base when light is required near the floor.

FIG. 1. M-R Type 20.
Double side lamp.

M-R Type
Rifle lamp.

211,

The twin arc broadside is still the conventional general illumination
unit where arc lighting is employed. The incandescent broadside
is being largely superseded by the more efficient "rifle" (Fig. 2) units
and floodlights. The broadside is most generally employed as a floor
unit for the general lighting of small and medium sized sets. More
detailed information relative to its use is given in the section on
lighting practice.

668

STUDIO LIGHTING COMMITTEE

[J. S. M. P. E.

Rifle Unit. — The rifle unit is a product of incandescent lighting. It
consists of a deep circular reflector about 18 inches in diameter, and a
1000- or 1500-watt, PS-52 bulb lamp is generally used. The reflector

FIG. 3. Fifty-amp, arc
dome with silent working
mechanism and built-in re-
sistance for 110 volts.

FIG. 5. M-R Type 125.
Bowl lamp.

is made either of silvered glass or of chromium plated metal. The
reflecting surface possesses spiral flutes which break up striations and
irregularities in the illumination; hence the name "rifle."

FIG. 4. M-R Type 30. Overhead strip lamp.

This lighting device is employed and mounted in a manner similar
to the broadside, the greater part of the light distribution is confined

May, 1932]

STUDIO LIGHTING COMMITTEE

to an angle of about 60 degrees. The efficiency of the unit is very high
and from 50 to 70 per cent of the light output of the lamp is available
at the area to be illuminated.

Scoop. — The scoop is similar in general design to the broadside
except that the reflector is shaped so as to direct the greater part of the
light through a vertical angle extending downward from the horizontal.
Since the scoop is designed primarily to be mounted overhead it is not
provided with a floor stand. It is available either with the arc lamp

FIG. 6 (a). One hundred and
fifty-amp, arc illuminator.
(Available with facetted mirror
and ground glass parabolic
mirror.)

(b). Two- or three-kilowatt incandes-
cent spot (available with ground glass,
facetted, and stippled parabolic mirrors)
on spot rail fitting. (Illustration with
stippled mirror.)

or with 1000- or 1500-watt incandescent lamps. It is used relatively
little in incandescent lighting since the rifle unit can be readily sub-
stituted. The greater efficiency of the rifle unit gives it a decided
advantage over the older forms of scoop.

Dome Light. — This unit (Fig. 3) is designed primarily to be mounted
above motion picture sets and to give a general uniform flood of light
throughout the set. In arc lighting practice, domes are available hav-
ing from one to four lamps. In incandescent practice, domes are used

670 STUDIO LIGHTING COMMITTEE [j. S. M. p. E.

to a limited extent, and possess usually ten or twelve 1000- or 1500-
watt PS-52 bulb lamps.

In incandescent lighting practice, the dome has largely been super-
seded by an overhead lighting unit consisting of 4 to 12 rifle reflectors
mounted on a single suspension device. Since individual reflectors
around each lamp are far more efficient in directing light where it is
desired than a single large reflector for a group of light sources, this
latter device gives far greater illumination intensities for the same
wattage than the dome.

Strip Lights. — These (Fig. 4) are an outgrowth of incandescent
lighting practice, and they usually consist of a long, porcelain enamel,
trough-shaped reflector about 18 inches wide and 60 inches long.
Five 1000-watt PS-52 bulb lamps are mounted in a row. These units
are available with floor stands, and without stands but with a number
of suspension rings. The strip light finds its greatest use as a substitute
for the dome unit, and when used in this manner several strips are
hung side by side. The strip light is also used to direct light through a
doorway, behind columns, etc. The greater compactness of this unit
over broadsides giving the same light output makes it desirable where
it is necessary to use the doorway.

Backing Lights. — The backing of a motion picture set is the large
curtain that often surrounds three sides of the set and is used to give
the effect of sky or to produce a background. It is necessary to light
this backing very uniformly and to a high intensity.

In incandescent lighting practice there are available large shallow
chromium plated metal reflectors that use the 5000- and 10,000-watt
incandescent lamps (Fig. 5). These reflectors are designed to give
a very wide uniform distribution so that they can be used quite close
to the backing.

Floodlights. — In arc lighting practice high intensity arc lamps either
in their housings but without reflectors, or the bare lamps themselves,
are commonly used. Where space is available large numbers of
broadsides or floodlights are used, especially to give a high intensity
of illumination near the floor.

There are also available a number of small miscellaneous lighting
devices, some consisting of only a socket and a semi-cylindrical metal
reflector. These devices usually employ the 1000-watt, 1 15- volt tubu-
lar bulb projection lamp. They are primarily used to secure a higl
intensity of illumination over a limited space, such as behind statues,
clocks, vases, etc.

May, 1932]

STUDIO LIGHTING COMMITTEE

671

FIG. 8. Twenty-five amp. arc
spot on spot rail fitting.

FIG. 7. M-R Type 324.
Twenty-four inch "Integral
Inkie"sun spot.

FIG. 9. M-R Type 26. Two thousand-
watt "Integral Inkie" studio spot.

672

STUDIO LIGHTING COMMITTEE

MODELING LIGHTING DEVICES

[J. S. M. P. E.

Reflector Spots. — The most generally used modeling lighting devices
are the reflector spot lamps. For arc lighting practice they are avail-
able with 18-(Fig. 6), 24-(Fig. 7), 36-, and even 60-inch reflectors, and
in incandescent practice 18-, 24-, and 36-inch reflectors are used. The
reflectors are generally mirrored glass of a parabolic contour.

Mirrors employed for incandescent service usually have a shorter
focal length than those of the same diameter used with arc lamps.
The incandescent reflector spots operate with beam spots having
spreads varying from 7 to 30 degrees. The arc spots vary from 2 to

FIG. 10 (a). Fifty-amp, choke. (6) Resistance and choke combined for
150-amp. sun arc. (c) Three hundred amp. choke for 300-amp. arc lens, or
two 150-amp. suns, or four 75-amp. arcs.

30 degrees. Practically all the arc lamp reflector spots use the 150-
ampere, high intensity arc lamp with rotating electrode. For in-
candescent lighting, the 2000-watt, 11 5- volt G-48 bulb monoplane
filament lamp is used with the 18-inch reflector; the 5000-watt, G-64
bulb with the 24-inch spot; and the 10,000- watt, 115-volt, G-96 bulb
lamp is used with a 36-inch mirror.

Control of the beam spread is obtained by moving the light source
from the focal point where maximum concentration is obtained toward
the mirror. The great advantage of this type of lamp is that the

May, 1932] STUDIO LIGHTING COMMITTEE 673

mirror intercepts light through an angle of 120 to 140 degrees, thus
utilizing a relatively large proportion of the available light. These
reflector spot lamps are usually provided with mountings so that
diffusing glass doors or prismatic lens doors, giving a horizontal beam
spread, can be attached. There have recently been made available a
number of metal mirrors, usually chromium plated, designed to give
some diffusion so that the illuminated spot produced has a high inten-
sity center, and the illumination gradually falls off toward the edges.

Lens Spots. — The lens spot lamp employs a plano-convex lens 6, 8,
10, and 12 inches in diameter (Figs. 8 and 9). The particular ad-
vantage of the lens spot is that all the light emitted is contained within
the beam and there is no spill light; also, the beam spread can be
controlled with great uniformity, through a wide range at all times.
Its particular disadvantage is that the light is intercepted at the lens
in a small angle, 30 to 45 degrees, and hence the volume of light con-
tained within the beam of a lens spot is much less than that of a
reflector spot of equal wattage. In arc lighting practice, lens spots
are available using both open and the high intensity arc with 70-, 80-,
100-, 120-, and 150-ampere ratings. Incandescent lamp lens spots
employ either the 1000- or 2000-watt, monoplane filament, 115-volt
lamps. A spherical mirrored reflector placed behind the lamp is always
employed with incandescent spot lamps for redirecting into the beam
much of the light that would otherwise be wasted.

Soft Spot. — The soft spot is another outgrowth of incandescent
lighting, and consists of a glass reflector of a modified parabolic
contour, in some instances the surface of the reflector being stippled.
The illumination is produced by a fairly well-defined beam having a
high intensity center that tapers off at the edges. Movement of the
lamp in and out of the reflector produces some control of the beam
spread. The soft spot is largely used in close-up work.

CHOKE COILS

Various types of rugged induction coils have been developed for use
in series with d-c. arcs for filtering out commutator hum. Three types
of these are shown in Fig. 10.

DISCUSSION

PAST-PRESIDENT CRABTREE: Has any practical application been made of
photometers for measuring intensity in studios?

MR. PALMER: No. We have tried to use photometers, and have spent a
great deal of time in the effort to do so, but have always encountered the diffi-

674 STUDIO LIGHTING COMMITTEE [J. S. M. P. E.

culty that photoelectric cells are not constant in their reactions, and that a reading
obtained from a certain cell one day does not check with the reading obtained
under the same conditions the next day.

PAST-PRESIDENT CRABTREE: A cameraman in Hollywood suggested that, if a
rheostat or some means of controlling the intensity were attached to each lighting
unit, it would be of great assistance to him in his work.

MR. FARNHAM: In one of the West Coast studios, a number of banks of
semi-portable rheostats have been made up, that can be moved to the set and
into which various lighting units can be plugged, so as to obtain various dimming
effects.

MR. PALMER: Mr. Crabtree's suggestion is to apply a control unit to each
individual lamp. We frequently have occasion to dim a single lamp, and find it
necessary instead to put on another diffuser or, perhaps, two more diffusers, in
order to soften the light. A simple, light, easily worked device for reducing
the voltage of the individual lamps would certainly help in many cases, and
would save a lot of time in the studio.

MR. MOLE: The banks to which Mr. Farnham referred were made only for
effects; for certain sunrise effects or to dim the entire set and the like. The
cameramen always wanted a control at each lamp, instead of having to apply
diffusers. But a great deal of equipment would be required, and, if any more
gadgets are connected to a lamp, difficulties will result. We have found that in
studios the simplest equipment, having the least number of connections, is the
successful equipment. The personnel is not as well trained as that in the pro-
jection room where the equipment remains in one spot and where it is not difficult
to add auxiliary parts to supplement the main equipment.

MR. BARTON: Does not the actinic value of incandescent lamps change
rapidly in the useful range, so that the effect of the resistors may be to decrease
the actinic value considerably without decreasing greatly the apparent bril-
liancy?

MR. PALMER: That is one of the difficulties that would be encountered if we
should dim a lamp by using a resistor. But experience quickly teaches how
much dimming is necessary, and how much difference a slight change will make
in the photographic value of the lamp. The new film is quite sensitive to red
and yellow light, so that the introduction of a resistance into the lamp circuit
would not necessarily render the lamp useless.

PAST- PRESIDENT CRABTREE: Are the studios taking advantage of the in-
creased sensitivity of the film? Are they reducing the intensity of the lamps,
or are they using the same number of lamps and simply adding a few more
diffusers?

MR. MOLE: When the new film first came to Hollywood, many cameramen
used it in tests and found that excellent results could be obtained with about
fifty per cent of the former illumination. It appeared as though half the number
of lamps would be needed, and half the wattage. The studios were very much
encouraged over it, as they felt that it was going to cut down their expenses.

But actually, in a production, the cameramen do not have the time to adjust
each light. They cannot take the time to fuss around with the adjustments,
as the saving achieved in using less wattage or fewer units would not warrant
the additional time required to shoot the picture.

May, 1932] STUDIO LIGHTING COMMITTEE 675

After a few months, it was found that, although the wattage was reduced to
about seventy-five or eighty-five per cent of the former value, the number of
units was about the same. No appreciable reduction in lighting expense was
noticed. However, the new film is being used in many productions; I should
say that seventy per cent of the productions in Hollywood are being made with
the new film.

MR. FARNHAM : I wonder if the cameramen are not stopping down the lenses
more, improving the photographic quality, and thus taking advantage of the
new film in that manner, instead of endeavoring to reduce the wattage?

MR. MOLE: That depends on the cameraman; some feel that sharp photog-
raphy is not artistic photography and would not prefer the sharp pictures to the
so-called artistic pictures that are continually being produced.

MR. MITCHELL: I think the question is not so much that of saving light,
as in having a sufficient number of point sources of light to permit satisfactory
adjustment of the shadows, or to obtain the requisite detail in the shadows;
in many cases, and, in fact, in most cases, these requirements involve quite a
number of sources of light. By using the same number of lamps they can be
controlled by diffusers, reducing the over-all illumination, but keeping the
general illumination unchanged — that is, the balance of the illumination. If
the lens is stopped down, the desired effect is entirely lost.

PAST-PRESIDENT CRABTREE: Could not someone deal with one particular
treatment? In the matter of lighting, the conception of the artist is definite.
He has a picture in his mind of what he wants or at least he should have, before
photographing the set. It is purely a matter of technic in getting the result,
and I wonder if someone could not outline in black and white how to get it.

MR. MOLE: The same result can be obtained using various technics. The
cameraman can obtain about the same results with entirely different forms of
lighting. A paper written on such a subject would describe Mr. Jones' lighting;
another would describe Mr. Smith's lighting; and so on. That is their stock-in-
trade, and the cameraman cannot be expected really to disclose it or publish it.
I dare say you could place every lamp in the same manner that he does, and
shoot the picture; and you would not get the same result that he does. There
is some individual touch that he has, in painting that picture with light, in being
able to obtain certain effects that another cameraman would not obtain with
the same set-up.

PAST-PRESIDENT CRABTREE: I disagree with Mr. Mole. If the lamps were
placed in the same position, with the same intensities and at the same angles,
under the same conditions the results would be identical.

MR. MITCHELL: I agree with Mr. Mole. I have seen a cameraman photo-
graph the same scene that another cameraman had previously photographed,
with lights approximately in the same position, and the results would be entirely
different. They develop, through experience, an uncanny sense of light. The
cameraman may put a diffuser on, or move a light back two feet, and although
the change may not be noticeable to the eyes, it makes a difference in the photog-
raphy of the picture.

ABSTRACTS

The Demand for Stereoscopic X-Ray Motion Pictures in Diagnosis. G.
KOGEL. Kinotechnik 13, Nov. 5, 1931, p. 399. It is maintained that the stereo-
scopic impression of an object obtained in looking at a pair of stereograms with
a suitable optical device depends largely on the observer's previous experience
with similar objects, i. e., on his "memory images." For this reason, in order to
achieve the ability to see x-ray stereograms correctly, the student must familiarize
himself with x-ray photographs. It is believed that the field for stereoscopic
x-ray motion pictures lies in detecting the faulty functioning of organs before the
disease has had time to alter their form, especially in those cases in which long
irradiation of the patient is undesirable. M. W. S.

Thomas A. Edison and His Relations to Motion Pictures. C. FORCH. Kino-
technik, 13, Nov. 5, 1931, p. 397. By autumn of 1891, Edison had constructed
an operable motion picture camera in which Eastman perforated film was moved
intermittently. The film was advanced by a sprocket driven by a friction belt.
Suitable members served to arrest the rotation of the sprocket during the intervals
when the exposure was being made. Edison is reported to have employed a
Maltese cross for securing the intermittent movement but he discarded it for the
mechanism described. Edison's Kinetoskop was a device enabling only one
person to view a motion picture. In it, the film moved continuously; a very
narrow shutter opening gave such a short view of each picture that a sharp
image was obtained. In his American patent no. 493,426, applied for Aug. 24,
1891, he described another viewing device by means of which pictures were pro-
jected to a screen. The system was intended to give stereoscopic relief, but the
principle was wrong, and incapable of giving a true stereoscopic effect. The
apparatus was not designed, however, to project large pictures visible to more than
a few persons at a time. M. W. S.

Vacuum Photoelectric Cells of High Sensitivity. M. C. TEVES. Technique
Cinemat. 2, Dec. 1931, p. 13. Increased sensitivity, especially to light of longer
wavelengths, has been attained in Philip's caesium vacuum photoelectric cells
with the purpose of increasing their usefulness with tungsten light sources.
Caesium is deposited to a depth of 100 molecules on a foundation coating of a salt
or oxide. Sensitivity extends to 12,000 A. with a maximum between 6000 and
8000 A. A response of 20 or 30 X 10 ~8 amperes per lumen for illumination by a
source at 2680° Kelvin is attained regularly. Quantum efficiency is therefore
as high as 1 : 20. After 3 hours' use the sensitivity diminishes 5 per cent, but is
recovered in 20 hours' rest. Forty to fifty volts' potential is recommended. With
such cells the maximum of absorption variation among colored films of a well-
known manufacturer measures only 25 per cent. Two (geometric) types of cell
are made. C. E. I.

The Use of the Color Filter in the Production of Photographic Images That Are
True to Reality. P. LOB. Kinotechnik, 13, Nov. 5, 1931, p. 400. The ab-
676

ABSTRACTS 677

sorptions of five filters — red, yellow, green, bright blue, and deep blue — were
measured at five different positions in the spectrum. For this purpose, a mono-
chroma tor was used, the intensity of the monochromatic light before and after the
insertion of the filter into the beam being measured by means of a thermocell.
The sensitivity of a photographic plate for the same wavelengths was measured
by first adjusting the light source so that it produced the same effect on the
thermocell at each wavelength, and then exposing the plate to the monochromatic
light. The sensitivity of the plate was taken as directly proportional to the
density produced. Then, in order to show the difference between the absorption
of a filter as determined photographically and as measured by its effect on a caesium
cell, as well as to show the necessity of knowing the spectral sensitivity of a photo-
graphic plate in determining the absorption of filters, exposures were made
through each of the five filters by monochromatic light of each of the five spectral
regions, and the densities compared to the density produced without the filter in
the beam. The absorption of each filter was then measured at each wavelength
by means of a caesium cell. In general, the effect produced on the plate fails to
correspond to the effect on the cell. It is concluded that for exact work with
filters, the following three items must be known: (1) the spectral distribution of
the light source, (2) the spectral sensitivity of the photographic emulsion, and (3)
the characteristic absorption curve of the filter. M. W. S.

The Motion Picture in Rockefeller City. G. SCHUTZ. Mot. Pict. Herald,
106, Feb. 13, 1932, Sect. 2, p. 13. Building No. 8 in the huge construction pro-
gram in progress in New York City under the Rockefeller sponsorship is to be a
motion picture theater seating 3509 persons. A topographical sketch is shown of
the entire project and detailed plans of the theater. The auditorium will be
158 feet wide and 128 feet deep, from the rear wall to the curtain, with the
average height of 65 feet. The stage area measures 92 by 46 feet. There will be
three shallow mezzanines, each seating approximately 500 persons. Although
the large overhanging balcony with its objectionable acoustic character is elimi-
nated, the added height of three levels places the two upper levels above the
normal line of the screen, and will require patrons of these sections to lean forward
to see the picture, which means some physical discomfort. An excessive and
objectionable screen angle for projection is also introduced, since the projection
room will be located above the uppermost section. Certain other features of the
theater are •commented on in the light of modern knowledge of theater con-
struction. G. E. M.

A New Type Projection Lamp. F. H. RICHARDSON. Mot. Pict. Herald, 106,
Feb. 13, 1932, Sect. 2, p. 40. A detailed description of an improved lamp for
theater projection. Specially designed, quickly acting clamps have been intro-
duced, both for negative and positive carbons, which permit rapid change of
carbons but insure firm retention when burning. Control of the arc is intended to
be accomplished chiefly by means of a thermostat. A lens projects a side view
of the burning positive crater to a mirror which reflects the image to a thermostat.
As the crater burns away, the image falls nearer and nearer the thermostat
until a set of electrical contacts is brought together which speeds up the motor
and the crater is brought forward to its normal position. Additional features are
mentioned. G. E. M.

New A-C. Amplifier. Film Daily, 58, Feb. 21, 1932, p. 6. This instrument

678 ABSTRACTS [J. S. M. p. E.

has been designed especially for sound-on-film reproduction, and constitutes the
entire electrical apparatus necessary between the photoelectric cells and the stage
horns. The unit is equipped with a new type of transformer which is stated to
supply the current to the exciter lamps without the need of filtering. The de-
vice is designed for use in theaters having about 1200 seats. G. E. M.

Sound Equipped Theaters in U. S. in 1931. Mot. Pict. Herald, 106, Jan. 30,
1932, p. 9. According to figures supplied by the Film Boards of Trade, there
were 13,223 sound equipped theaters in the United States at the close of the
year 1931. Of these, 6434 have sound-on-film equipment; 3609 use disk only;
and 4898 were equipped for both disk and film. One thousand five hundred
eighty-two theaters having sound equipment were not operating. A total of
20,100 theaters, having an approximate seating capacity of 10,767,000, are listed
on the books of national distributors. G. E. M.

Planning Today's Simplified Cinema. B. SCHLANGER. Mot. Pict. Herald,
105, Nov. 21, 1931, Sect. 2, p. 18. Two theater plans are discussed in some detail
for 300-seat and 600-seat structures, respectively, which are designed to be built
within limited spaces. Both theaters are planned to occupy only a portion of a
structure used also for other purposes. The reverse slope floor plan is used in
each design. G. E. M.

Sound Control in Air Conditioning Installations. V. O. KNUDSEN. ' Mot.
Pict. Herald, 105, Nov. 21, 1931, Sect. 2, p. 37. Attention given to sources of ex-
traneous noise within and without the sound picture theater has resulted in con-
siderable study of causes of and means for elimination of noise in the ventilating
system. It is important that all mechanical equipment used in air conditioning
be carefully insulated from the solid structure of the building. Detailed mathe-
matical equations are presented for the determination of suitable insulation, know-
ing certain measurable factors. Absorptive filters are necessary between the
ventilating fan and the outlets to eliminate noises transmitted through the ducts.

G. E. M.

A Radically New Studio Camera. W. STULL. Amer. Cinemat., 12, Feb.,
1932, p. 12; Internal. Phot., 4, Feb., 1932, p. 4. The novel feature of
this new camera, designed by T. L. Tally and T. M. de la Garde, is that the
magazines are placed beneath the camera case, thereby lowering the center of
gravity and providing better balance. The range of tilt is increased. Sprockets
are a part of the magazine, and act as a light trap in this position. The camera
has a four-lens turret, movable as a unit for focusing — a 240-degree shutter, and
a view finder in which the film aperture can be observed directly. A. A. C.

New B. & H. Lens Eliminates Crane Shots in Professional Movies. Amer.
Cinemat., 12, Feb., 1932, p. 31. This objective is a variable focus outfit, with
mechanical shifting of the elements to maintain accurate focus and diaphragm
opening throughout a range of 40 to 120 mm. focal length. It is thus possible to
approach a subject or recede from it without moving either the camera or the
scene. The speed of the unit ranges from //3. 5 for 40 mm. to//5.6 at 120 mm.
focus. It is made on special order only. A. A. C.

RCA Presents 16 Mm. Sound-on-Fihn Projector. Amer. Cinemat., 12, Feb.,
1932, p. 36; Internal. Phot., 4, Feb., 1932, p. 25. This new equipment is said
to show a good 4X6 foot picture, with excellent quality of sound reproduction.
Since it is practically the first of the sound-on-film 16 mm. outfits, its performance

May, 1932] ABSTRACTS 679

will be noted with much interest. The projector amplifier unit weighs 43
pounds, with its case; all the equipment is readily accessible for necessary ad-
justment so that it need not be removed from the case during operation. The
loud speaker fits in a 21 -pound case, 19 X 16 X 9*/2 inches. Space for eight
400-foot reels is also provided. Sound volume is sufficient for a room of 10,000
cu. ft. A. A. C.

Internationalizing Talking Pictures. A. GRADENWITZ. Proj. Eng., 4, -Feb.,
1932, p. 7. A new rhythmic method of recording sound effects enables directors
to add the foreign text after a film has been finished in English. It is based on a
new means of remote control, invented by C. R. Blum, of Berlin, by which
synchronism can be attained between any number of electrical devices. It is
independent of the actual speed of motion. The recording from the film is re-
peated on a band arranged to move in front of the operator on an electrical re-
corder. Text and music are accurately spaced in accordance with the rhythm of
the picture so that actors have only to read or play their parts from the band in
order to be sure of perfect agreement between picture and sound record.

A. A. C.

A New Zoom Lens. Amer. Cinemat., 12, March, 1932, p. 16. Describes a
lens of adjustable focus announced by O. Durholz, of Paterson, New Jersey.
"The lens snaps over the standard Mitchell type cup mount in a few seconds ready
to focus. . . . From long shot to close-up it maintains focus automatically from 40
to 160 mm. (equivalent focal length). The effective aperture is//8 at full range,
//5.6 at 3x, increasing as the range is limited." An outline of the problems
of mechanical construction met by the designer is given in some detail.

A. A. C.

Agfa-Novopan Reversal Film. L. KUTZLEB. Kinotechnik, 13, Sept. 10,
1931, p. 333. A new panchromatic 16 mm. reversal film has been placed on the
market. This is said to have a speed standing in the ratio, 16 :6 : 1, to the speeds
of Agfa Pan and Ortho Reversal films by tungsten light, and in the ratio 4:2:1
to the speeds of these same films by daylight. This increased speed is
stated to be the result of increased color-sensitivity, particularly for the longer
wavelengths. An anti-halation layer is inserted between the emulsion and the
support. The film is recommended especially for use under artificial lighting.

M. W. S.

Agfa Leica-Superpanfilm. L. KUTZLEB. Kinotechnik, 13, Dec. 20, 1931,
p. 466. This film is panchromatic and is said to have three times the speed of
Agfa Leica-Isochromfilm by incandescent lighting, or twice the speed by day-
light. The film is said to make possible the making of snapshots in well lighted
rooms by the aid of high aperture objectives without a yellow filter. A double
emulsion layer and an anti-halation layer are used. A fine grain developer is
recommended for developing small negatives for enlargement. M. W. S.

SOCIETY OF MOTION PICTURE
ENGINEERS

•

OFFICERS
1931-1932

President
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester. N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE, Fox Film Corp., New York. N. Y.

Secretary
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Kodak Co., Rochester, N. Y.

Board of Governors

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada

H. T. COWLING, Eastman Kodak Co., 343 State St., Rochester, N. Y.

J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y.

P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y.

O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y.

A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New

York, N. Y.

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLB, Fox Film Corp., 850 Tenth Ave., New York, N. Y.
680

COMMITTEES

681

COMMITTEES
1931-1932

J. L. CASS

W. T. CRESPINEL

F. E. IVES

Color

P. D. BREWSTER, Chairman
R. M. EVANS, Vice-Chairman

J. F. KlENNINGER

N. M. LA PORTE

G. E. MATTHEWS
H. B. TUTTLE
L. T. TROLAND

W. C. HUBBARD

Convention

W. C. KUNZMANN, Chairman

M. W. PALMER

J. I. CRABTREE

Constitution and By-Laws
P. H. EVANS, Chairman
O. M. GLUNT
D. McNicoL

F. H. RICHARDSON

J. CRABTREE
J. I. CRABTREE
T. FAULKNER

Development and Care of Film

R. F. NICHOLSON, Chairman
R. C. HUBBARD, Vice-Chairman
E. D. LEISHMAN
K. MAC!LVAIN
D. MACKENZIE

J. S. MACLEOD
H. RUBIN
J. H. SPRAY

H. T. COWLING
W. B, COOK

Finance

L. A. JONES, Chairman
J. I. CRABTREE
W. C. HUBBARD

J. H. KURLANDER

L. C. PORTER

E. W. ADAMS
W. CLARK
O. B. DEPUE

Historical

C. L. GREGORY, Chairman
N. D. GOLDEN
C. F. JENKINS

G. E. MATTHEWS
O. NELSON
T. RAMSAYB

J. A. BALL

Journal and Progress Medal Awards
C. E. K. MEES, Chairman

O. M. GLUNT

E. A. WILLIFORD

682

COMMITTEES

[J. S. M. P. E.

D. M. BALTIMORE

E. J. COUR

B. W. DEPUE

C. D. ELMS
R. EVANS

Membership and Subscription
H. T. COWLING, Chairman

W. H. CARSON, Vice-Chairman
E. R. GEIB
J. G. T. GILMOUR*
J. KLENKE
W. C. KUNZMANN

E. E. LAMB
M. L. MISTRY
T. NAGASE
N. F. OAKLEY
E. C. SCHMITZ

B. W. DEPUE
O. B. DEPUE

C. L. GREGORY

Museum

W. E. THEISEN, Chairman
C. F. JENKINS
F. H. RICHARDSON

T. RAMSAYE
A. REEVES
A. F. VICTOR

A. A. COOK
W. B. COOK
E. R. GEIB
H. A. DEBRY

Non-Theatrical Equipment
R. E. FARNHAM, Chairman

N. B. GREEN

H. GRIFFIN

L. A. JONES

J. H. KURLANDER

R. P. MAY
R. F. MITCHELL
A. SHAPIRO
A. F. VICTOR

J. A. BALL
C. DREHER
P. H. EVANS
A. C. HARDY
N. M. LA PORTE

Papers

O. M. GLUNT, Chairman
G. E. MATTHEWS
P. A. McGuiRE
G. A. MITCHELL
D. McNicoL

P. MOLE
K. F. MORGAN
C. N. REIFSTECK
P. H. REIS
T. E. SHEAMAN

H. T. COWLING
J. I. CRABTREE
A. S. DICKINSON

Preservation of Film
W. H. CARSON, Chairman
R. EVANS
C. L. GREGORY

J. M. JOY
T. RAMSAYE
V. B. SEASE

M. ABRIBAT
L. BUSCH

F. CAHILL

G. A. CHAMBERS
C. DREHER

R. E. FARNHAM

Progress

J. G. FRAYNE, Chairman
W. C. HARCUS

F. S. IRBY

G. E. MATTHEWS
J. P. MAXFIELD
G. A. MITCHELL
J. M. NICKOLAUS

M. W. PALMER
G. F. RACKETT

P. SCHROTT

N. TRONOLONE
S. S. A. WATKINS
S. K. WOLF

May, 1932]

COMMITTEES

683

J. O. BAKER
T. BARROWS
W. H. BELTS
G. C. EDWARDS
S. GLAUBER
J H. GOLDBERG

Projection Practice
H. RUBIN, Chairman
C. GREENE
H. GRIFFIN
J. HOPKINS
W. C. KUNZMANN

R. H. McCULLOUGH

P. A. McGuiRE

R. MlEHLING

F. H. RICHARDSON
M. RUBEN
P. T. SHERIDAN
L. M. TOWNSEND

J. L. CASS
E. R. GEIB
H. GRIFFIN

Projection Screens
S. K. WOLF, Chairman

J. H. KURLANDER

W. F. LITTLE
A. L. RAVEN

H. RUBIN

L. T. TROLAND

C. TUTTLE

R. E. FARNHAM
H. P. GAGE

Projection Theory
A. C. HARDY, Chairman
W. F. LITTLE
W. B. RAYTON

L. T. TROLAND
C. TUTTLE

F. C. BADGLEY
B. W. DEPUE

Publicity

W. WHITMORE, Chairman
D. E. HYNDMAN
F. S. IRBY
W. C. KUNZMANN

G. E. MATTHEWS
D. McNicoL

M. C. BATSEL
P. H. EVANS
N. M. LA PORTE

Sound

H. B. SANTEE, Chairman
E. W. KELLOGG
C. L. LOOTENS
W. C. MILLER

H. C. SILENT
R. V. TERRY
S. K. WOLF

L. E. CLARK
L. DE FOREST

J. A. DUBRAY

P. H. EVANS
R. E. FARNHAM
H. GRIFFIN
A. C. HARDY

Standards and Nomenclature
M. C. BATSEL, Chairman

R. C. HUBBARD

L. A. JONES
N. M. LA PORTE
D. MACKENZIE
G. A. MITCHELL
G. F. RACKETT

W. B. RAYTON
C. N. REIFSTECK
V. B. SEASE
T. E. SHEA
J. L. SPENCE
E. I. SPONABLE
L. T. TROLAND

684

COMMITTEES

L. J. BUTTOLPH

R. E. FARNHAM

Studio Lighting
M. W. PALMER, Chairman
C. W. HANDLEY
K. C. D. HICKMAN

J. H. KURLANDER

E. C. RICHARDSON

R. S. BURNAP
W. H. CARSON

Ways and Means
D. McNicoL, Chairman
H. GRIFFIN
F. S. IRBY

R. F. MITCHELL, Chairman
B. W. DEPUE, Sec.-Treas.

Chicago Section

J. H. KURLANDER
J. A. NORLING

R. P. BURNS, Manager
O. B. DEPUE, Manager

New York Section

P. H. EVANS, Chairman

D. E. HYNDMAN, Sec.-Treas.

Pacific Coast Section

D. MACKENZIE, Chairman
W. C. HARCUS, Sec.-Treas.

M. C. BATSEL, Manager
J. L. SPENCE, Manager

C. DREHER, Manager
H. C. SILENT, Manager

GEORGE EASTMAN

JULY 12, 1854— MARCH 14, 1932

The name, George Eastman, will always be linked inseparably
with the growth of photography, particularly amateur and motion
picture photography. Mr. Eastman began his career with an idea —
to make photography available to every one. He lived to see the
growth of a great industry built around this idea, for there are millions

GEORGE EASTMAN

of persons in all parts of the world who now use photography. His
active interest in photography began about 1878. During his
lifetime he was successful in introducing small cameras, roll film,
the folding Kodak (daylight loading and daylight developing),
improved photographic papers, motion pictures for the amateur,
first as black and white and later in natural colors, as well as many

other developments.

685

686 GEORGE EASTMAN

Almost equally significant, however, were his contributions to the
growth of the motion picture industry. Within a few months after
his discovery of a method of making film on a transparent support,
Edison's purchase of some of the new product stimulated its manu-
facture. For many years this "picture ribbon," as it was called,
was made only in two grades, negative and positive; but additional
refinements were added as the industry grew, until in 1914 panchro-
matic film was introduced, making possible more accurate and
pleasing tone reproduction. More than a decade elapsed, however,
before the industry came to use this film extensively and, under the
stimulus of greater use, further improvements were announced in
1931, both as regards speed and color-sensitivity.

Many of these developments were made possible by Mr. East-
man's life-long conviction of the value of research. Besides experi-
menting himself during his earlier years, he employed the services
of others, until a large research organization was built up which
today investigates all branches of photographic endeavor, from
theoretical as well as practical standpoints.

Besides his fame as an industrial leader, he gained public dis-
tinction and satisfied his personal responsibility as a philanthropist
through his gifts to the upbuilding of his native city of Rochester
and other cities. These took the form of endowments for research
and teaching, erection of buildings for education in engineering,
a school of music, a college of medicine, dental clinics, and for other
useful purposes.

Mr. Eastman was elected to honorary membership in the Society
of Motion Picture Engineers on April 13, 1928; and at the banquet
honoring pioneers of the industry, which was held at Swampscott,
Mass., on October 7, 1931, was one of seven honorary members to
whom formal scrolls were presented. On that occasion he designated
J. I. Crabtree to receive his scroll for him. Concrete evidence of Mr.
Eastman's respect for the work of the Society was shown recently by
his donation of a fund for the establishment of a Motion Picture Engi-
neering Fellowship, under the supervision of the Society.

Honored by many nations and international societies, George
Eastman's greatest contribution was undoubtedly the develop-
ment of the medium of film photography, which resulted in a world-
wide hobby for the amateur and exerted an important influence
in the establishment and growth of the motion picture industry.

GLENN E. MATTHEWS

SOCIETY ANNOUNCEMENTS

BOARD OF GOVERNORS

At a meeting held on March 25th at New York, further details of
the Spring Convention to be held at Washington, D. C., were
arranged, the general scheme of which was published in the April
issue of the JOURNAL. Other details concerning the Convention are
given below.

Authorization was given for the formation of a "Constitutional
Committee," the function of which would be to consider recom-
mended amendments of the Constitution and By-Laws of the
Society. Among other amendments proposed at this meeting of
the Board of Governors, the recommendation was made that the
admission fee to the grade of Active membership be reduced to ten
dollars and to the grade of Associate membership, five dollars; and
that the transfer fee from the Associate to the Active grade be the
difference between the two admission fees, or five dollars.

It was also ruled that the Honor Roll of the Society, established
at the Swampscott Convention for the purpose of perpetuating the
names of distinguished pioneers in the motion picture art, who are
now deceased, be published each month in the JOURNAL.

SPRING, 1932, MEETING

MAY 9 TO 12, 1932
WARDMAN PARK HOTEL, WASHINGTON, D. C.

A rather complete schedule of activities for the approaching
Washington Convention was submitted to the Board of Governors
at its recent meeting by Mr. W. C. Kunzmann, Chairman of the
Convention Arrangements Committee, and Mr. O. M. Glunt,
Chairman of the Papers Committee. The final plan adopted by
the Board of Governors included, among other details, the following
features :

The morning of Monday, May 9th, will be devoted to registra-
tion, committee meetings, etc. The Convention will be formally
opened at 11:00 A.M. with a welcoming address by Hon. Con-
gressman Sol Bloom, followed by the response of the President.

687

688 SOCIETY ANNOUNCEMENTS [J. S. M. P. E.

The afternoon of Monday, May 9th, will be devoted to the pre-
sentation of S. M. P. E. committee reports.

On Wednesday, May llth, a session will be held at the Audi-
torium of the Department of Commerce, where addresses will be
delivered by various government departmental heads. Sight-
seeing trips and other means of recreation will be provided for the
afternoon of this day. The semi-annual banquet of the Society will
be held on the evening of Thursday, May 12th, in the Gold Room of
the Wardman Park Hotel.

An interesting papers program has been arranged by the Papers
Committee, separate sessions being devoted to (1) the problems of
theater operations; (2) problems of the release print, in production,
theaters, and exchanges; (3) lectures by members of the staff of the
U. S. Bureau of Standards; (4) motion picture photography, and
various other interesting subjects.

An exhibition of newly developed motion picture apparatus will
be held at the Wardman Park Hotel, the Convention Headquarters.
Manufacturers desiring to exhibit their new apparatus should com-
municate with the General Office of the Society at 33 West 42nd
Street, New York, N. Y.

NEW YORK SECTION

At a meeting of the New York Section held on March 23rd at the
Electrical Institute in New York, N. Y., an interesting address
on the subject of "Animated Cartoons in the Making" was pre-
sented by Mr. Harry Bailey, of Fables Pictures, Inc., illustrated
by hand drawings and a motion picture parody of the subject of the
talk.

The next meeting of the Section is scheduled for April 30th, at the
Electrical Institute, at which time Mr. H. G. Tasker, of the
United Research Corporation, will present a paper dealing with
the problems of recording sound on sixteen millimeter film and of
the corresponding problems of projection and reproduction.

CHICAGO SECTION

The March meeting of the Section was held on March 3rd at the
headquarters of the Electric Association in Chicago. A paper pre-
sented by Mr. H. Shotwell, dealing with portable a-c. amplifiers,

May, 1932] SOCIETY ANNOUNCEMENTS 689

was followed by a general discussion of the problems attending the
use of this type of equipment in reproducing sound from film.

At the following meeting held on April 7th at the Electric Asso-
ciation headquarters in Chicago, Mr. E. Cour demonstrated the
Artreeves recorder and described its operation. Mr. W. A. Holtz
also gave a demonstration of the new sixteen millimeter sound-on-
film projector.

PROJECTION PRACTICE COMMITTEE

At a meeting held at New York, N. Y., on April 4th, a further
study was made of the various problems attending the use of the
release print in the theater, and a preliminary draft of that section
of the Committee's report, to be presented at the Washington Con-
vention, was drawn up. Further consideration, also, was given to
a proposed method of equalizing the sound output of projectors in
theaters, and of the data that are now being accumulated by the
Committee with regard to the illumination of projection screens in
theaters. It is probable that, on account of the magnitude of the
work of collecting and analyzing all the requisite data on projector
tolerances, clearances, and tensions, the description of that part of
the Committee's work will be deferred until the following report, as
it was felt that unless the data were reasonably complete, their
great importance to the motion picture industry might not be fully
appreciated and their utilization might be more limited than is
desirable.

SOUND COMMITTEE

At a meeting held at New York, N. Y., on March 18th, the report
of the subcommittee on frequency characteristics was considered,
particularly with reference to the compensation of frequency char-
acteristics of reproducing and recording apparatus, and as to the
manner in which compensation should be made for the slit losses in
recording. The report also included a recommendation on the
method of adjusting the azimuth of the recorder slit.

A study was also made of the desirable volume range of repro-
duction, the limitations of reproducing equipment, and the over-
load to ground noise ratio in sound records.

Another meeting of the Committee will be held prior to the Wash-
ington Convention for the purpose of drafting the final report to be
presented at that time.

690 SOCIETY ANNOUNCEMENTS [J. S. M. p. E.

STANDARDS COMMITTEE

At a meeting of the Standards Committee, held at the General
Office of the Society on March 15th, further consideration was given
to the establishment of dimensional standards for sixteen milli-
meter sound film, and to various items recommended for standardi-
zation by the Projection Screens Committee. Among these were
the standardization of tolerances and methods of test for determin-
ing the acceptability of projection screens, the method of making
measurements of the reflectivity of screens, the definitive names of
various types of projection screens, and the relation of the size of
screen to the distance of the nearest observer.

After reviewing the circumstances attending the problem of
establishing dimensional standards for the apertures of 35 milli-
meter projectors, the Committee passed for recommendation to the
Society the dimensions 0.600 X 0.825 inch for the projector aper-
ture, and the dimensions 0.631 X 0.868 inch for the corresponding
camera aperture.

Final recommendations were made concerning the layout for
16 mm. sound film to be proposed for standardization, and arrange-
ments were made to have these final layouts ready for submission
and action at the Washington Convention.

JOURNAL AND PROGRESS AWARDS

At a meeting of the Board of Governors held May 24, 1931, it
was decided that the following actions of the Board, relating to the
JOURNAL Award and the Ft ogress Medal, should be published an-
nually in the JOURNAL.

JOURNAL AWARD

The motion was made and passed that "an award of $100.00 shall
be made annually, at the Fall Convention of the Society, for the
most outstanding paper published in the JOURNAL of the Society
during the preceding calendar year. An appropriate certificate
shall accompany the presentation.

"The JOURNAL Award Committee shall consist of not less than six
Active members of the Society, to be appointed by the President
subject to ratification by the Board of Governors. The Chairman
of the Committee shall be named by the President and a two-thirds
vote is necessary for election to the award. (Proxies are permitted.)

May, 1932] SOCIETY ANNOUNCEMENTS 691

PROGRESS MEDAL

"The Board of Governors may consider annually the award of a
Progress Medal in recognition of any invention, research, or de-
velopment, which in the opinion of the Progress Award Committee
shall have resulted in a significant advance in the development of
motion picture technology.

"The Committee shall consist of not less than six Active members
of the Society, to be appointed by the President subject to ratifica-
tion by the Board of Governors. Names of persons deemed worthy
of the award may be proposed and seconded, in writing, by any two
Active members of the Society and shall be considered by the Com-
mittee during the month of June; a written statement of accom-
plishments shall accompany each proposal.

"Notice of the meeting of the Progress Award Committee must
appear in the March and April issues of the JOURNAL. All names
shall reach the Chairman not later than April 20th.

"Recipients of the Progress Medal shall be asked to present their
portraits to the Society, and, at the discretion of the Committee,
the recipients may be asked to prepare a paper for publication in
the JOURNAL of the Society. These regulations, the names of those
who have received the medal, the year of each award, and a state-
ment of the reason for the award shall be published annually in the
JOURNAL of the Society."

692 SOCIETY ANNOUNCEMENTS

A written statement of accomplishments should accompany each
proposal, which should reach the Chairman not later than April
20th.

The two committees have this year been amalgamated into a
single committee known as the "Committee on Journal and Progress
Medal Awards."

SUSTAINING MEMBERS

Agfa Ansco Corp.
Bausch & Lomb Optical Co.

Bell & Howell Co.

Bell Telephone Laboratories, Inc.

Carrier Engineering Corp.

Case Research Laboratory

Eastman Kodak Co.

Electrical Research Products, Inc.

Mole-Richardson, Inc.

National Carbon Co.

RCA Photophone, Inc.

Technicolor Motion Picture Corp.

HONOR ROLL

OF THE

SOCIETY OF MOTION PICTURE ENGINEERS

By action of the Board of Governors, October 4, 1931, this Honor Roll was estab-
lished for the purpose of perpetuating the names of distinguished pioneers who are
now deceased:

Louis AIME AUGUSTIN LE PRINCE

WILLIAM FRIESE-GREENE

THOMAS ALVA EDISON

GEORGE EASTMAN

Statement of the Ownership, Management, Circulation, Etc., Required by the
Act of Congress of August 24, 1912, of Journal of the Society of Motion Picture
Engineers, published monthly at Easton, Pa., for April 1, 1932.
State of New York
County of New York

Before me, a Notary Public in and for the State and County aforesaid, person-
ally appeared Sylvan Harris, who, having been duly sworn according to law,
deposes and says that he is the Editor of the Journal of the Society of Motion
Picture Engineers and that the following is, to the best of his knowledge and
belief, a true statement of the ownership, management (and if a daily paper,
the circulation), etc., of the aforesaid publication for the date shown in the
above caption, required by the Act of August 24, 1912, embodied in section 411,
Postal Laws and Regulations, printed on the reverse of this form, to wit:

1. That the names and addresses of the publisher, editor, managing editor,
and business managers are:

Name of — Post Office Address —

Publisher, Society of Motion Picture Engineers, 33 W. 42nd St., New York, N. Y.
Editor, Sylvan Harris, 33 W. 42nd St., New York, N. Y.
Managing Editor, Sylvan Harris, 33 W. 42nd St., New York, N. Y.
Business Manager, Sylvan Harris, 33 W. 42nd St., New York, N. Y.

2. That the owner is: (If owned by a corporation, its name and address
must be stated and also immediately thereunder the names and addresses of
stockholders owning or holding one per cent or more of total amount of stock.
If not owned by a corporation, the names and addresses of the individual owners
must be given. If owned by a firm, company, or other unincorporated concern,
its name and address, as well as those of each individual member, must be given.)
Society of Motion Picture Engineers, 33 West 42nd St., New York, N. Y.

A. N. Goldsmith, President, 570 Lexington Ave., New York, N. Y.
J. H. Kurlander, Secretary, 2 Clearfield Ave., Bloomfield, N. J.
H. T. Cowling, Treasurer, 343 State St., Rochester, N. Y.

3. That the known bondholders, mortgagees, and other security holders
owning or holding 1 per cent or more of total amount of bonds, mortgages, or
other securities are: (If there are none, so state.)

None.

4. That the two paragraphs next above, giving the names of the owners,
stockholders, and security holders, if any, contain not only the list of stockholders
and security holders as they appear upon the books of the company but also,
in cases where the stockholder or security holder appears upon the books of the
company as trustee or in any other fiduciary relation, the name of the person or
corporation for whom such trustee is acting, is given; also that the said two
paragraphs contain statements embracing affiant's full knowledge and belief
as to the circumstances and conditions under which stockholders and security
holders who do not appear upon the books of the company as trustees, hold stock
and securities in a capacity other than that of a bona fide owner; and this affiant
has no reason to believe that any other person, association, or corporation has
any interest direct or indirect in the said stock, bonds, or other securities than
as so stated by him.

5. That the average number of copies of each issue of this publication sold
or distributed, through the mails or otherwise, to paid subscribers during the
six months preceding the date shown above is: (This information is required
from daily publications only.)

SYLVAN HARRIS, Editor-Manager.
Sworn to and subscribed before me this 14th day of March, 1932.

(Seal) KENNETH L. JEFFERY.
Notary Public, Westchester County,
Certificate filed in New York County,
Clerk's No. 48, Reg. No. 2-J-37.
(My commission expires March 30, 1932)

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR
Volume XVIII JUNE, 1932 Number 6

CONTENTS

Page

The Principles of the Light Valve

T. E. SHEA, W. HERRIOTT, AND W. R. GOEHNER 697

Variation of Photographic Sensitivity with Different Light

Sources R. DAVIS AND G. K. NEELAND 732

Variation of Photographic Sensitivity with Development Time. .

R. DAVIS AND G. K. NEELAND 742

A Reflector Arc Lamp for Portable Projectors . . . . H. H. STRONG 752

Vacuum Tube and Photoelectric Tube Developments for Sound

Picture Systems M. J. KELLY 761

Process Photography G. A. CHAMBERS 782

A Shrinkage-Compensating Sound Printer R. V. WOOD 788

Committee Activities:

Report of the Color Committee 792

Officers 795

Obituary— Thomas A. Edison 796

Society Announcements 799

Author Index, Volume XVIII 803

Classified Index, Volume XVIII. . 806

JOURNAL

OF THE SOCIETY OF

MOTION PICTURE ENGINEERS

SYLVAN HARRIS, EDITOR

Board of Editors

J. I. CRABTREE, Chairman

L. DE FOREST A. C. HARDY F. F. RENWICK

O. M. GLUNT E. LEHMANN P. E. SABINE

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers.

Publication Office, 20th & Northampton Sts., Easton, Pa.
General and Editorial Office, 33 West 42nd St., New York, N. Y.

Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc.

Subscription to non-members, $12.00 per annum; to members, $9.00 per annum,
included in their annual membership dues; single copies, $1.50. A discount
on subscriptions or single copies of 15 per cent is allowed to accredited agencies.
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y.

The Society is not responsible for statements made by authors.

Entered as second class matter January 15, 1930, at the Post Office at Eastt
Pa., under the Act of March 3, 1879.

THE PRINCIPLES OF THE LIGHT VALVE*
T. E. SHEA, W. HERRIOTT, AND W. R. GOEHNER**

Summary. — The light valve has been used very widely as the modulating device
in systems of film sound recording. In this paper the principles of operation of the
light valve are discussed, and those engineering factors which prescribe limitations
on performance and indicate operating advantages are described in detail. The type
of distortion which results when a light valve is overloaded is depicted both for
single-plane and two-plane valves. Finally, a new type of light valve having
advantages from the standpoints of weight, size, and stability of operation is described.

I. Introduction

The light valve, as a sound recording instrument, has seen very
wide use during the past three years. It has undoubtedly been used
for more recording and re-recording in the motion picture industry
during that time than have all other types of light modulating
devices combined. The extensive experience acquired with its use
in studio, location, and newsreel recording has shown it to be a
rugged and efficient instrument capable of making sound records of
excellent quality.

During this time extensive studies have been carried on to perfect
the light valve. For instance, as the producers have become better
acquainted with sound recording systems and have been able to
utilize them more nearly to their full capabilities, the quality of sound
recording has improved, and it has been necessary to improve various
elements of recording systems in order to extend their range of opera-
tion and to reduce to a minimum production delays due to recording
difficulties. How great a change in recording conditions has taken
place during the last three years may be seen by considering, as a
case in point, the early difficulties encountered with outdoor re-
cording on location in contrast with the smoothness and regularity
with which recording equipment is now operated under similar
circumstances.

As often happens with an instrument which has been perfected to a

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Bell Telephone Laboratories, Inc., New York, N. Y.

697

698

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

high degree prior to commercial use, the improvements which have
resulted from the studies mentioned above, though numerous, have
not been fundamental. They represent rather an aggregation of
minor improvements which, taken as a body, constitute an important
advance. A new type of light valve, described later in this paper,
does, however, represent fundamental advances.

Although various earlier attempts to construct light gates of
variable aperture for sound recording had been made, the first

FIG. 1. The receiving end (optical system) of a picture transmis-
ion system ; V, the light source ; D and S, condensing and ob-

FIG. 1.

sion system; v , tne ngnt source
jective lenses; C, the moving film.

FIG. 2. The single-ribbon type of
light valve; R, the ribbon; /, the
aperture jaws.

practical form of light valve was developed by E. C. Wente in 1922.
Subsequently, the light valve has been used in the "single-ribbon"
form since 1924 for the regular commercial transmission of pictures
over telephone circuits.1

In this development, the picture is broken into a series of long,
narrow sections, similar to sound tracks, which when illuminated
are scanned by a slit and photoelectric cell. The electric currents
thus generated are amplified and sent over telephone wires. Proc-

June, 1932] PRINCIPLES OF THE LlGHT VALVE 699

esses of frequency modulation are used which need not be described
here. At the receiving end these currents modulate a light valve
which varies the exposure of a moving film, and provide the proper
latent image for the re-creation of the transmitted picture. Fig. 1
shows the receiving end optical system of a picture transmission
system, V being the light valve, L the light source, D and S the
condensing and objective lenses, and C the moving film. Fig. 2

FIG. 3. Enlarged portion of a transmit-
ted picture of the variable density type.

shows the single-ribbon type of light valve employed, the ribbon R
vibrating in front of the gap between the aperture jaws /. Fig. 3
shows an enlarged portion of a transmitted picture of the variable
density type; the similarity of its horizontal sections to sound
tracks is obvious.

The conditions of use of light valves in sound recording are rather
different from those of picture transmission, and a consideration of
some of them led to the choice of the double-ribbon type of valve
for this field. In general, it was recognized that in the latter field

700

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

the proposed conditions of operation were more severe; and it was
believed that, if the duty of modulating could be carried out by two
ribbons instead of one, the sensitivity of the valve would be in-
creased, the internal temperature rise due to conductor heating
reduced, and tones freer from spurious harmonics obtained. Subse-
quent experience has verified these suppositions.

This is not to say that the single-ribbon valve is not suited to sound
recording, but that under present conditions, at least, it will do so
only at a disadvantage from several fundamental design standpoints.
In what follows, where a comparison between the two types is made,
an endeavor will be made to separate these factors out of any spe-
cific design and consider them on a general basis, so that the facts
involved may not be clouded.

FIG. 4.

An early double-ribbon valve whose moving con-
densers lie in different planes.

It should be kept in mind, however, in order to dissipate a preva-
lent but erroneous belief, that either the single- or the double-ribbi
valve may have two forms: (a) one in which light barriers or gates
adjacent to the aperture are in the same plane, and (b) one in whicl
these barriers are in different planes. Thus, in the general form
either type of valve, excessive modulation does not lead to "light
valve clash" but merely to a cutting off of the peaks of one side of
the signal wave. Indeed, in one of the earliest discussions2 of light
valve operation there is described (Fig. 4) a double-ribbon valve
whose moving conductors lie in different planes.

In the use of the light valve as an optical rectifier, valves of tl
two-plane type are requisite. In the use of a light valve as a simpl
modulator, the choice between the one-plane and two-plane types is t(
be determined by efficiency, quality, and maintenance considerate
These will be discussed later.

June, 1932] PRINCIPLES OF THE LlGHT VALVE 701

The general method of employing the light valve in variable
density sound recording has been described at length by MacKenzie.3
The photographic and recording technic outlined by him is suffi-
ciently representative of present-day procedure to be assumed in
what follows. The light valve itself has undergone changes in form
which will be described. It is important to note in this connection,
however, the specific changes in recording technic involved (1) in
the use of a 1.0-mil normal aperture (instead of 2.0 mils) imaged on
the film (Fig. 5) as an exposure beam 0.5 mil wide, and (2) in the
use of noise reduction equipment such as that described by Silent.4

Either the single- or double-ribbon valve may be used not only
for variable density, but for variable width recording. In the former
case, the direction of the ribbons is transverse to the film; in the

PLANE OF
RIBBON OF
LIGHT

PLANE OF
VALVE
RIBBONS
/O.OOI"X tt2
( SLIT

*")

5l-:i^

PLANE OF
IMAGE ON
FILM „
/0.0005 X ai28 \
^ IMAGE )

___..-——'"'"'"

"~ ———____

.---

CONDENSING OBJECTIVE

LENS SYSTEM LENS SYSTEM

FIG. 5. Illustrating the specific changes in recording technic involved in
the use of a 1-mil normal aperture (instead of 2 mils).

latter case, lengthwise with respect to it. The light valve has been
used for variable width work in picture transmission. Inasmuch
as all commercial light valve sound recording is of the variable
density type, we shall consider only the latter.

The remainder of this paper will be divided into a consideration of
theoretical aspects of light valve modulation (Part II), a discussion of
practical factors important in light valve design and use (Part III),
and a general description of a new type of light valve (Part IV).

77. Theoretical Aspects of the Light Valve

1. VIBRATORY ACTION OF THE LIGHT VALVE

The light valve, considered in its electromechanical aspects, is
similar in operation to the Einthoven string galvanometer. Con-
sidered in a transverse section (Fig. 6), a current 7 flows in a con-
ductor A in a uniform magnetic field. (A second conductor B is

702

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

used in the double-ribbon valve for the return of the current and
moves always in opposite direction to the first, but the individual
conductors may be considered to act separately.)

The conductor (ribbon) A will move at right angles to both the
direction of the magnetic field and of the current, and the force F

acting upon it is

F = klH

where H is the intensity of the magnetic field linked by the con-
ductor, and k is a constant. Thus the force acting upon the ribbon
is proportional to (1) the strength of the magnetic field and (2) the
instantaneous current. If / is an alternating current, such as corre-

FIG. 6. Transverse section of light
valve used in analyzing its action.

spends to speech and music, the force F will be of similar character
and the motion of the ribbon will alternate accordingly.

The extent of motion of the conductor per unit of current at low
frequencies will depend solely on the total tensioning force exerted
on the conductor as it lies stretched between its supports. That is,
the force due to the current will deflect the ribbon until it is offset
by an equal restoring force; the latter is a component of the ten-
sioning force and is proportional to the displacement of the con-
ductor from its line of support.

Under these conditions, the power input to the conductors is
simply that dissipated by their resistance. The conductor, however,
has uniformly distributed mass as well as elasticity, and the former
becomes of increasing importance at high frequencies. The inert ial

June, 1932]

PRINCIPLES OF THE LIGHT VA.LVE

703

force of the moving conductor tends to keep the conductor moving
and to offset the effect of the restoring force. As the influence of
the mass increases, an increase in the motion of the conductor for a
given amount of current occurs, and the valve becomes more sensitive
or responsive. At a particular high frequency, the effect of distrib-
uted mass and elastance will offset each other and light valve
"resonance" will occur. For this condition, the valve is highly
sensitive ; and for frequencies in the vicinity of resonance, distortion
of the signal takes place in that the response is excessive compared
with that at low frequencies. Fig. 7 shows relative response curves

10000

2000O

5000
FREQUENCY-C.P. S.

FIG. 7. Relative response curves of light valves whose resonances
occur at 7000, 10,000, and 15,000 cycles.

of light valves whose resonances take place at 7000, 10,000, and
15,000 cycles, respectively.

2. LIGHT VALVE RESONANCE AND TUNING

This resonance, of course, is controllable and is generally caused
to fall outside the useful range of recording. The resonance or
"tuning" frequency is given5 by

f = ±JI = ± JT

where / = resonance frequency

/ = length of vibrating conductor

T = tension

M = mass per unit length

A = area of conductor in cross-section

p = density of conductor material'

704

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

The resonance or "tuning" frequency may be controlled, therefore,
by changing the tension of the conductor, the length of its vibrating
span, the cross-section of the conductor, or the density of its material.
For a given design of light valve, / is fixed. For a given type of
conductor ribbon, M, A, and p are fixed. Consequently, in practice
the resonance frequency is set by adjusting the tension of the ribbon
to a sufficiently high value.

3. IMPEDANCE OF THE LIGHT VALVE

The impedance characteristic of the light valve is of interest in
considering the efficiency of the light valve. Since the motion of
the conductors depends on the value of current flowing, the valve
should be connected to its supply circuit under the most efficient

A A A A A A A A

A A A A /\A—t

FIG. 8. Electrical equivalent of the light valve.

conditions, i. e., when the impedance of the valve and the circuit are
matched so that for a given emf. in the supply circuit, maximum
current flows in the valve.

At low frequencies, the power delivered to the valve is entirely
used in conductor heating, and yet the valve is most highly respon-
sive when maximum current is delivered because the mechanical
force set up is greatest. At high frequencies, the mass and elasticity
of the conductor must be represented in the electrical impedance,
for, as the conductors move, reaction emfs. are generated, which
create reactive impedance in the valve circuit. At resonance,
the reactances due to mass and elasticity offset each other and the
impedance is controlled largely by the damping resistance (r) , due to
mechanical friction. Fig. 8 shows the electrical equivalent of a
light valve. Fig. 9 shows a typical light valve frequency-impedance
characteristic. The same data, when put in polar form (Fig. 10),

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

705

display the ' 'motional impedance characteristic" typical of the
telephone receiver and other vibrating instruments.

4. SCANNING LOSSES AND HARMONIC DISTORTION

The "Ribbon Velocity" Effect. — The foregoing discussion has dealt
with the movement of the ribbons in response to an alternating
current and, accordingly, with the variations of light flux passed
through the valve. We are interested, however, in the variations

NOTE: REACTANCE CURVE RE-
VERSES AND PASSES THROUGH
ZERO APPROXIMATELY AT RES
ONANCE FREQUENCY.

10000
FREQUENCY- C.RS.

FIG. 9.

The optical light valve frequency-
impedance characteristic.

in point-to-point exposure of the moving film on which the light
valve slit is imaged in a recording machine. The exposure given
to the film, it may be readily seen, is not determined by changes in
intensity of the light flux, but by the time required for any point on
the film sound track to pass through the image of the light valve
slit. This time, and the effective exposure of any point on the film,
is therefore affected by the film velocity.

If the film moves very rapidly, the average exposure of the sound

706

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

track will be low, and vice versa. The brilliancy of the lamp source,
the condensing lens system, and the average opening of the light
valve must be arranged, for any given film speed (e. g., 90 ft. per
min.), to give the proper average film exposure.

If the frequency being recorded is low, so that the velocity of the
ribbons is small compared with the velocity of the film, the varia-
tions in film exposure will represent faithfully a pattern of the light

90'

2 0"

FIG. 10. Same data as in Fig. 9, plotted in polar coordinates.

valve modulation. As the frequency becomes high enough, however,
the velocity of the ribbons increases, so that "the ribbon velocity
effect," as it is called, comes into play. This results (1) in a loss
of effective variation in exposure, which means a loss of recorded
volume, and (2) in a degradation of wave-shape which includes the
production of spurious harmonic frequencies.

The ribbon velocity effect is somewhat different in the cases of
the single- and double-ribbon valves. It may be analyzed6 as follows :

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

707

Double-Ribbon Case. — In Fig. 11, let a transverse line (infinitesimal
striation) P of a film moving with velocity v be, at any time /, at the
center of the exposure image, and let the instantaneous width of
the image be 2w. The half-width of the image is then w. Let the
half-width of the image at that previous time /i, when P just entered
the image, be w\t and let the half-width at that subsequent time /2,
when P will leave the image, be w%. It will be assumed that the
film velocity v always is greater than the rate of change of the half-
image size (dw/dt).

"FILM

.1.

FIG. 11. "Ribbon velocity," effect
diagram.

The total light received by P is proportional to

Wi + W2 = V(t2 — /l)

This must be expressed in terms of t.
Now

Wi = V (t — /i)

and

W2 = V (fe ~ 0

If the image varies sinusoidally, that is,

iv = a + & sin co/
then

and

vti = vt — a — b sin wt\
vtz = vt + a + b sin cof2

708 SHEA, HERRIOTT, AND GOEHNER [j. S. M. P. E.

or, multiplying by co/y for convenience

to/i = co (t — a/v} — (b co/y) sin co/i
and

co/2 = co (t + a/v) + (b ca/v) sin co/2

These equations are of the type

x = y -\- a. sin x

so that x and y are odd functions of each other. Hence x— y can
be expanded into a Fourier series of y, containing only sine terms, i. e.,

x — y = ^ an sin ny

n = l
Hence

2 C*
An = - (x - y)sinnydy

* Jo
Integrating by parts,

An = — — (x — y) cos ny I + I cos ny d(x — y)
W7r L |o Jo

The integrated term vanishes since x = y for both 0 and TT; also

X7T
cos ny dy = 0

Hence, putting

-if

""Jo

x — a sin x = y

cos n (x — a sin x) dx

?/.(*.)

by the Bessel integral.

Thus, we obtain the solutions

and

n = l

Whence

v(h — ti} = 2a + — -

(t + l

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

709

By expanding and regrouping,

- cos - sin

o /2 ( — - ) sin — " cos 2a)t + - 73 ( — - Jcos — " sin 3 ut + . .

2 \ v / v 3 \ v / v J

6
where 2a is the normal image width and~ the fractional modulation.

Single-Ribbon Case. — It may readily be shown that the character
of the alternating exposure is not affected by the direction of motion
of the film relative to the fixed edge of the image. We shall assume
that the film approaches the fixed edge of the image first. Whence,
from the equation for co/2 above,

100

S- SINGLE RIBBON CASE
D- DOUBLE RIBBON CASE

SECOND HARMONIC

THIRD HARMONIC'

FIG. 12.

2000 4000 6000 8000 10000

FREQUENCY~C.PS

"Ribbon velocity," effect for 0.5-mil normal image.

* ( i j\ i t.

W*a — t) = a -\- o r-

From the formulas, typical curves may be drawn which show the
loss of amplitude of the fundamental component of the film exposure
with increasing frequency, and the magnitudes of the various har-
monics.

Such curves are shown for the case of a normal 0.5-mil image for
the single- and double-ribbon valve, in Fig. 12. Here it will be noted

710

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

that for 100 per cent modulation, the fundamental of both types of
valve suffers a loss of several decibels at 10,000 cycles, the double-
ribbon valve suffering about a decibel more than the single-ribbon
valve.

In the matter of harmonic distortion, however, the double-ribbon
valve is markedly superior, and this is especially true for the third
harmonic, which relatively is very weak in the double-ribbon valve.

At lower modulations, the frequency characteristic of the funda-
mental improves, in the case of the single-ribbon valve, more rapidly

100

g 60

u, 40

S- SINGLE RIBBON CASE
D- DOUBLE RIBBON CASE

2000 4000 6000 8000 10000

FREQUENCY~C P S.

FIG. 13. "Ribbon velocity," effect for 0.167-mii normal
image.

than for the double-ribbon valve, but the situation on harmonic
distortion remains relatively much the same.

The diminution of the fundamental at high frequencies is of
minor importance, because the influence of light valve resonance in
present or future practice may be considered to offset it.

The illustration chosen (for 100 per cent modulation) is fairly
typical of most recording situations where "noise reduction" ap-
paratus is employed. It should be pointed out that, for the low
valve spacings obtained on weak sounds with such apparatus, dis-
tortion of fundamental and harmonic production from the causes
mentioned is greatly reduced. Fig. 13 shows curves corresponding

June, 1932] PRINCIPLES OF THE LlGHT VALVE 711

to those of Fig. 12, except that the valve spacing is reduced to 0.3
its former value.

This latter point is especially important in comparing the light
valve with other light modulating devices, such as the flashing lamp,
for it indicates that by reducing the amount of the average exposure,
the light valve distortion may be reduced accordingly.

///. Practical Aspects of the Light Valve

1. FACTORS GOVERNING SENSITIVITY OF VALVE

For any given normal separation of the ribbons, the sensitivity
of the light valve depends on (1) the force on the ribbons per unit
current, and (2) the deflection of the ribbons per unit force.

The force on the ribbons per unit current depends, as we have seen,
on the strength of the field in which the ribbons move. Aside from
the use of (a) magnetic material having high permeability and (b)
an efficient winding, in the case of an electromagnetic field; or the
use of material having high residual permeability in the case of
a permanent magnet field; the principal factor influencing the
magnetic field is the length of the air gap. The air gap must be wide
enough to accommodate the moving ribbon or ribbons and any
additional light barrier placed between the magnetic poles.

In general, then, a valve of the one-plane type is more efficient
magnetically than a valve of the two-plane type, for in the latter
the air gap must, in general, be somewhat longer. Although in
practice the magnetic yoke is brought to a high saturation point,
so long as the reluctance of the air gap forms an appreciable part of
the total reluctance, the magnetic efficiency of the circuit will be
greater with a narrower gap. If this is put on the basis that a definite
magnetic flux is required through the gap, then, with the narrower
gap generally pertaining to the one-plane type of valve, the field
magnetizing current required is smaller.

In considering the sensitivity of the valve for a given strength
of magnetic field the following factors are important:

(a) Low Resistivity of the Cottductor Material. — Since the de-
flecting force of the ribbons depends on the current flowing through
them, the amount of a-c. power which must be supplied to the ribbons
for a given deflection is obviously proportional to the resistivity of
the ribbon material. For frequencies substantially below the
resonance frequency, the impedance of the valve is closely equal

712 SHEA, HERRIOTT, AND GOEHNER [j. S. M. P. E.

to its d-c. resistance. The power required to drive the ribbons is
therefore similar to that dissipated in the ribbon as a conductor.
It is assumed in this discussion that the valve input transformer
matches closely the valve impedance. Under this condition each
doubling of the resistivity means a doubling of the power supplied
to the valve per unit of current in the ribbons and therefore a loss of 3
decibels in sensitivity.

(b) Resonance Frequency of the Valve.— From the formulas, given
in Part II, for the resonance frequency of a light valve, it is seen that
the tension which must be applied to the ribbon is proportional to
the square of the resonance frequency. This means that the higher
the tuning frequency the less sensitive the valve, for the amount
of the tension determines the size of the restoring force which tends
to prevent displacement of the ribbons. With any given ribbon
material, therefore, a doubling of the tuning frequency means a
loss of 12 decibels in valve sensitivity.

(c) Density of the Ribbon Material. — The density, or specific
gravity, of the ribbon material has an influence on the sensitivity
of the valve. If two valves be alike except for the material of
which their ribbons are composed, and if each be tuned to the same
frequency, it is obvious, from the formulas, that the tensioning
force will be greater for the valve having ribbon material of higher
density. The tension required for any given resonance frequency
will be proportional to the density of the material, and, therefore,
the sensitivity of the valve varies inversely as the density of its
conductor material. This means that each doubling of the density
of the ribbon material causes a loss of 6 decibels in sensitivity.

(d) Length of Vibrating Span. — In considering the influence upon
sensitivity of the length of span of the vibrating ribbon, it is necessary
to consider only the resistance of the conductor material. If the
length of span be doubled, the power supplied to the ribbon must
be doubled; that is, there is a loss of 3 decibels in sensitivity. While
it is true that the force created in the conductor by its reaction in
the magnetic field is proportional to the length of the vibrating
span, this increase in force is directly offset by the fact that the force
doubled must move a conductor which, for any given tuning fre-
quency, has a total restoring force proportional to the length of the
vibrating span. That is to say, doubling the length of the span
quadruples the tension for a given tuning frequency, but halves the

June, 1932] PRINCIPLES OF THE LlGHT VALVE 713

angular displacement of the ribbon. The net result, therefore, is
that the sensitivity of the valve varies inversely as the square root
of the length of the vibrating span.

(e) Sensitivity of Single-Ribbon and Double-Ribbon Valves. —
The length of ribbon required in the double-ribbon valve is funda-
mentally twice that required in the single-ribbon valve. Therefore,
for a given current in the vibrating ribbons, twice as much power
must be supplied to the double-ribbon valve. This means an ap-
parent loss of 3 decibels in sensitivity. However, the displacement
obtained from two ribbons in the double-ribbon valve is, of course,
twice that obtained with the same current in the single-ribbon valve.
Therefore, for a given percentage modulation of the recording
illumination, a factor of 6 decibels must be added in favor of the
double- ribbon valve. The net result is that the double-ribbon valve
is inherently 3 decibels more sensitive, for a given percentage of
light modulation and consequent volume of reproduced sound, than
the single-ribbon valve. This figure, of course, assumes valves
which are alike in other design details, such as the nature of the
conductor material employed, the flux density of the air gap, etc.
This estimate of 3 decibels is conservative, for it assumes that ribbon
material of the same cross-section is employed in either type of
valve. Since the ribbon of the single-ribbon valve must be dis-
placed twice as far as either of the ribbons of the double-ribbon
valve, and since the width of the vibrating conductor is determined
primarily by considerations of mechanical tolerances in relation to
the amount of ribbon displacement required, it is more fundamen-
tally correct to assume that in the single-ribbon light valve, for a
fair comparison, the ribbon material should be twice as wide. If
this assumption is made, the wider ribbon is equivalent to two of
the narrower ribbons, vibrating side by side, and a further factor of 3
decibels should be allowed for the additional power required to dis-
place the heavier ribbon. Thus, from a fundamental design stand-
point the single-ribbon valve is 6 decibels lower in efficiency than the
double-ribbon valve.

2. PROPERTIES OF LIGHT VALVE RIBBON

It is of major importance in the successful use of the light valve
that the metal ribbon or tape used to form the vibrating light gate
shall be adequate for the purpose it is to serve. In general, the
ribbon should possess the following properties: (1) low resistivity,

714

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

(2) low specific gravity, (3) high tensile strength, (4) straightness of
ribbon edges, (5) stability under continuous tension, (6) non-corrosive-
ness, and (7) non-magnetic character. The importance of (1),
(2), and (3) have been discussed.

The importance of straight optical edges is apparent when it is
considered that the variations from straightness cause changes
from point to point in the light valve slit width and hence in average
film exposure. In an average slit width of 1 mil an effort is made to
keep edge straightness deviations below 0.1 mil. This represents
a change of 10 per cent in average exposure for the corresponding
portion of the sound track. This does not ordinarily mean a change
in signal volume recorded, because the actual displacement of the
ribbons is unaltered; but it means a slight shifting, from point to
point along the light valve, of the exposure in relation to the straight-
line part of the H & D curve. It can also affect the maximum re-
cordable volume by altering the clash point of the valve.

Among the materials which are suitable for use duralumin has
been found greatly superior and has, in addition, proved to be
fairly workable material. The following table shows the more
important constants of various metals which might be considered.

Constants of Various Metals Used for Light Valve Ribbons

Tensile Figures of Merit

Strength Sensi- Breaking
Density Resistivity Density tivity Frequency

2.7 3.0 10,200 1.29 0.48

Tensile
Strength

27,600

90,000*
65,500
75,000*

8.3

8.93

2.8

2.0
1.8
4.6

10,800

7,350

26,800

0.51
0.50
1.00

0.51
0.35
1.27*

Material

Aluminum
Aluminum (90% Cu)

Bronze (10% Al)
Copper (hard drawn)
Duralumin
Duralumin (light

valve ribbon)
Molybdenum
Molybdenum (0.002

wire)
Silver
Tungsten
Tungsten (ribbon)

* Probably less for ribbon form.

f Material difficult to work smoothly.

The figure of merit for sensitivity indicates directly the compara-
tive sensitivity of valves employing the different materials, and is

59,000

2.8

4.6

21,000

1.00

154,000 f

10.0

5.7

15,400

0.25

0.73

200,000* f

9.0

5.7

22,000

0.28

1.05*

42,600

10.5

1.6

4,050

0.50

0.19

590,000*f

18.8

5.5

31,400

0.14

1.50*

450,000f

18.8

5.5

24,000

0.14

1.14

June, 1932] PRINCIPLES OF THE LlGHT VALVE 715

obtained by multiplying the ratio of the densities by the square
root of the ratio of the resistivities. The figure of merit for breaking
frequency indicates directly the ratio of maximum allowable tuning
frequencies for light valves using the various metals. Where the
figures given are not otherwise noted, they are for the metal in bar
form, and it should be realized that neither the tensile strength of
metal in this form nor that in drawn wire form may, as a general
thing, be realized in the case of metallic ribbon of the dimensions
required for light valves. It is readily seen that duralumin is the
only metal listed which has a high figure of merit for both sensitivity
and breaking frequency. To give specific illustrations, a light
valve of any character employing molybdenum must be inherently
about 12 decibels less sensitive than one employing duralumin, and a
light valve employing tungsten would have a loss of sensitivity, in

61"

.000492'

-J 1— - .0005"
FIG. 14. Cross-section diagram of light valve ribbon.

comparison with one using duralumin, of 17 decibels. The compo-
sition of duralumin varies somewhat; the alloy at present used for
light valves has the following composition :

Aluminum 94

Copper 4

Manganese 0.5

Magnesium 0.5

Silicon, iron, etc. 1

As is well known, heat treatment and aging have an influence on the
tensile properties of duralumin, and much effort has been expended
in recent years to increase the tensile strength. As a result of such
efforts, the ribbon now employed has a tensile strength about 75
per cent greater than that of the earlier light valve ribbon.

There are two general methods which may be employed for the
production of duralumin ribbon. In the first, wire is drawn to the
proper cross-section and flattened into the ribbon form by rolling;

716

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

in the second, ribbons 0.006 inch wide are sheared directly from
sheets of duralumin foil 0.0005 inch thick.

In the rolling method, the primary obstacles are the extreme
accuracy required of the rolls and the uniformity throughout its
length of the material. Fig. 14 represents a cross-section of ribbon
6 mils by 0.5 mil. If manufacturing tolerances hold the width to

FIG. 15. Illustrating a microscope
fixture attached for visual examination
of ribbon edges.

±0.1 mil and the density of the material remains constant, the thick-
ness must be held to 0.008 mil, or eight one-millionths of an inch.
In the shearing process accurate alignment of cutting shears must
be supplemented by a technic for producing foil of uniform thickness,
free from pinholes, embedded impurities, etc. Because of the empiri-
cal nature of alloy processing, it is customary, as a check inspection
after regular manufacturing inspection has been completed, to
test a substantial portion, about 5 per cent, of all supposedly satis-
factory ribbon. Valves are actually strung with this material

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

717

which is then inspected for ribbon edge straightness, and the ribbons
are tuned to destruction to determine their breaking point. During
the past two and one-half years approximately 2500 valves have
been thus strung and inspected. Fig. 15 shows a microscope fixture
attached for visual examination of ribbon edges.

Fig. 16 shows an illustration of rough edges, taken from an early
grade of ribbon. Fig. 17 shows a type of variation in ribbon due to
undulating edges. Fig. 18 shows a sample of ribbon excellent in its
quality of edge straightness.

3. LIGHT VALVE "HYSTERESIS" AND RIBBON SLIPPAGE

In much of the earlier studio recording, a phenomenon was often
noticed called light valve "hysteresis." This was due to the failure

FIG. 16. Light valve ribbon
having rough edges.

FIG. 17. Light valve ribbon
having wavy edges.

FIG. 18. Light valve ribbon
having straight edges.

of the valve ribbons to return, after a large impulsive displacement,
to their original normal position. This gave the d-c. amplitude
characteristic, for wide variations in current, the appearance of a
magnetic hysteresis loop. Fig. 19 shows the hysteresis loops for
an older type of valve for two tunings, at 7000 and 10,000 cycles,
respectively. From these curves, it is seen that higher tuning
reduces the magnitude of the hysteresis effect, if we judge this magni-
tude by the displacement (in mils) of the opposite sides of the loop.
This might be expected, because the vertical component of the ribbon
tension, tending to hold the ribbon in place by means of friction at
the ribbon supports, is increased in proportion to the increased
tension required for the higher tuning frequencies.

718

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

Improvements in recording technic required that the valve tuning
frequency be raised considerably above the earlier value of 7000
cycles. This development, in combination with the availability
of the recently developed stronger type of duralumin ribbon, per-
mitted the general adoption of higher tuning frequencies, which
automatically reduced the hysteresis effect. However, with the
advent of noise reduction equipment to reduce film background

FIG. 19.

VALVE CURRENT

Light valve hysteresis (old valve).

noise, the requirements for the exact biasing of the valve ribbons
to small average slit widths made necessary the elimination of most
of the hysteresis tolerated in the older equipment.

Various methods were tried experimentally to secure greater
stability of average valve spacing, such as pin locating stops, metal
clamps, paper spacers, cement, etc. The cement method was given
a field trial, but did not prove satisfactory for general use.

The valve modification adopted for general field use practically
eliminated the hysteresis effect and was simple enough to enable

Jui

line, 1932] PRINCIPLES OF THE LlGHT VALVE 719

the ready conversion of available light valves; the slight modifica-
tion of the bridge support and spacing pincers, shown in Fig. 20,
has practically eliminated the hysteresis effect. The curves of Fig.
21 demonstrate this for 7500 and 10,500 cycles' tuning.

In interpreting the true importance of light valve "hysteresis,"
we must realize that it is only superficially like magnetic hysteresis.
In the first place, it does not represent a loss in energy. Secondly,
it is generally not present except for current cycles which exceed
a critical (and large) value; thus, for the smaller displacements,
frictional anchoring forces at the ribbon supports are adequate.
Thirdly, if the sides of the hysteresis loop are straight and parallel,
the spurious harmonics produced are small.

FIG. 20. Illustrating the slight modification of the bridge support and
spacing pincers for eliminating hysteresis.

The principal detrimental effects of ribbon slippage were therefore
(1) material variation in average exposure of the film, which inter-
fered with exact sensitometric control, and (2) departures from
normal of the load capacity of the valve, which interfered with
standardization of recording technic and the securing of maximum
recorded volume range. It is for both of these reasons that the
introduction of noise reduction equipment required the reduction
of light valve hysteresis.

4. AZIMUTH AND FOCUSING ERRORS IN RECORDING

The light valve is an electromagnetic shutter, and it translates
the amplified electrical energy from the microphone into mechanical
energy in such a manner that the light passing through the valve is
proportional to the speech waves impinging upon the diaphragm

720

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

of the microphone. It is, therefore, necessary to photograph the
light valve action as accurately as possible. This is made possible
by two important adjustments, the first of which is the azimuth
adjustment of the light valve, and the second, the focal adjustment
of the objective lens.

The azimuth adjustment of the light valve consists in locating
the horizontal plane of the valve perpendicular to the direction in
which the film is traveling. This adjustment also positions the

VALVE CURRENT

FIG 21. Light valve hysteresis (improved
valve).

striations on the film so that they are perpendicular to the direction
of the film travel. An error in the azimuth adjustment of the light
valve produces an azimuth deviation on the recorded film.

The azimuth deviation on the recorded film must be considered
in relation to the azimuth deviation of the scanning image in the
reproducer. Unless these values of azimuth deviation are identical
in degree and direction, the losses at the higher frequencies are
greater than those for optimum conditions of adjustment. If,

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

721

however, the azimuth deviation of the film varies from that present
in the reproducer, we have to deal with the sums and differences in
the deviation of each to obtain the effective value.

Experimental measurements of azimuth deviations in both the
recorded film and the scanning image in the reproducer have indicated
that the effect of an azimuth deviation on the recorded film and no
azimuth deviation in the scanning image is equivalent to a similar
deviation of the scanning image and no deviation in the recorded
film for small values of azimuth deviation. The effect of the azimuth
deviation of the scanning image has been treated both theoretically
and experimentally and presented in a paper7 before this Society.

3 *

0 6

3 8

12
0.0

, — —

-^-^,

+ TOWARD LIGHT VALVE
- AWAY FROM LIGHT VALVE

06 0.004 0.002 0 0.002 0.004 0.0<

DEVIATION OF THE IMAGE PLANE FROM THE FILM PLANE IN INCHES

FIG. 22. Influence of improper focus on reading, etc.

Figs. 6 and 8 in that paper show the loss for various azimuth de-
viations of the 0.0005-inch light valve image.

The adjustment of the //1. 5 objective lens consists in the move-
ment of the lens along the optical axis until the light valve ribbons
are focused on the film emulsion at a reduction of 2 :1.

As the objective lens system is moved along the optical axis, the
plane of the image of the light valve ribbons also moves, but at a
slower rate than that of the objective lens. Fig. 22 illustrates the
influence of improper focus in the recording of a 7000-cycle sound
track. A 2- and a 4-mil deviation of the image plane from the
film plane results in an approximate additional loss of 1 and 3 decibels,
respectively, at 7000 cycles. A more general expression of the effect

722

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

of improper focus is given in Fig. 23, in which the average effective
image width is shown to vary with the deviation of the image plane
from the film plane. With this data, the loss at any frequency,
due to an improper objective lens adjustment, may be computed.
When the lens is improperly focused the average effective image
width is increased and greater losses occur at high frequencies.

As shown in the paper by Stryker,7 when the loss due to both
improper focus or an increase in the average image width and the
azimuth deviation occur simultaneously, as they may in practice,
the total loss of reproduction due to the two of them jointly will be
the sum of the individual losses produced by each separately.

(ft
M

I
1-
O

— — ^

^^•^

£ 0.0006

•^-^

— „

; ^"

+*• —

ill

h- -^

•*

•

4- TOWARD LIGHT VALVE
- AWAY FROM LIGHT VALVE

-f-

0.006 0.004 0.002 0 0.002 0.004 0-0

DEVIATION OF THE IMAGE PLANE FROM THE FILM PLANE IN INCHES

FIG. 23. Effect of improper focus on average image width.

It is, therefore, apparent that either the azimuth adjustment of
the light valve or the focal adjustment of the objective lens, or both,
are important factors that may seriously affect the quality of re-
produced sound records. If the analysis of light valve action given
in Part II is to hold, the average image width assumed must be
consistent with the azimuth and focus conditions.

5. TUNING METHODS

An amplitude resonance curve for a light valve is shown by Fig. 7.
The resonance peak indicates that the valve requires 30 decibels
less power at the resonant frequency than at the low frequencies
for the same modulation.

Several methods of measuring the resonant frequency of light
valves have been employed. A visual method, used to tune the

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

723

earlier telephoto and sound film valves utilized a microscope to
observe the maximum deflection of the ribbons when the frequency
of an oscillator connected to the ribbons was varied.

The visual method proved satisfactory until widespread use of
recording equipment placed exacting limits on the resonant fre-

FIG. 24. Separate valve tuning unit supplied with Western Electric
recording equipment.

quency, especially where it became desirable to have several valves
operate at nearly the same clash point. The sound recorder itself
provided ready means to determine the tuning frequency more
accurately than the visual method. The procedure was to use the
photoelectric cell monitoring system to measure the degree of light
modulation by the valve when known current levels were supplied

724

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

to the valve from a variable oscillator. An output level variation
of the monitoring circuit of =*=0.5 decibel could be measured with the
ordinary volume indicator method and, therefore, from the resonance
curve (Fig. 7) it is seen that it is possible to measure the tuning
frequency to within ± 50 cycles from the resonant frequency.

The use of recording equipment for tuning more than a few valves
would have proved inefficient; and, therefore, separate valve tuning
units were supplied with Western Electric recording equipment.
This equipment is shown in Fig. 24. An oscillator, amplifier rectifier,
light source, light valve field coil, and photoelectric cell are mounted
on a 17- by 19-inch panel. The schematic circuit of Fig. 25 shows the

LIGHT
VALVE

PHOTO

ELECTRIC

CELL

LIGHT VALVE
TUNING PANEL

j^Hfvp

AMPLIFIER

VOLUME
INDICATOR

LATOR

OSCIL

FIG. 25. A schematic circuit of the tuning unit of Fig. 24, showing the
valve operation in the tuning circuit to be similar to that in an actual re-
cording circuit.

valve operation in the tuning circuit to be similar to that in the
actual recording circuit.

Another method of tuning valves, apparently due to R. D. Gibson,
makes use of the motional impedance characteristic of the valve,
shown in Fig. 10. The electrical impedance of the light valve with
the magnetic field applied is equivalent to an anti-resonant electrical
circuit as shown by Fig. 8. Many electrical methods of observing
the tuning point of such circuits are familiar to the electrical art and a
simple device consisting of an oscillator and a thermocouple in series
with the valve has been described by Ceccarini.8 Other arrange-
ments employ volume indicators, thermocouples, or rectifiers to
measure the voltage across the valve terminals when the valve is
connected to an oscillator source supplying approximately constant
current to the valve. It is evident that the series method of observ-

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

725

ing the peak will not indicate the correct peak sharpness unless the
thermocouple resistance is small compared with the d-c. resistance
of the light valve. Similarly, the voltage method will not indicate
the correct sharpness unless the voltmeter impedance is high com-
pared with the resonant impedance of the light valve.

6. LIGHT VALVE OVERLOAD AND CLASH

A phenomenon, concerning which little experimental evidence
has been presented, is that of wave-form distortion due to light
valve overload. In the two-plane type of valve the action which

-2 DB LEVEL

-2DB LEVEL

4-6 DB LEVEL

V J \ /

+ 6 DB LEVEL

+ 13 DB LEVEL

FIG. 26. Overload wave forms of single-
plane valve for 100 cycles.

+ 13 DB LEVEL

FIG. 27. Overload wave forms of
single-plane valve for 250 cycles.

might be expected would be a simple cutting off of the negative
troughs of the wave whenever the ribbons were sufficiently dis-
placed so as to cut off all light. This assumes that, in the main,
modification of the overload distortion, due to photographic con-
siderations and due to non-linearity of valve displacement for ex-
cessive movements of the ribbon, might be neglected.

In the one-plane type of valve, however, it has not been clear what
phenomena took place under similar conditions. Figs. 26, 27, and
28 are illuminating in this respect, as they show the relative wave-

726

SHEA, HERRIOTT, AND GOEHNER

[J. S. M. P. E.

forms at frequencies of 100, 250, and 1000 cycles, respectively, which
are encountered under the following conditions:

(a) modulation 2 db. below overload

(b) modulation 6 db. above overload

It is readily seen in these cases that the type of distortion obtained
is a relatively simple one, and is very much of the type that might be
expected. Figs. 29, 30, and 31 show experimental wave-forms

- 2 DB LEVEL

-2DB LEVEL

+ 6 DB LEVEL

(\ A l\ (\ (\ (\ A

+ 6 DB LEVEL

4- 14 DB LEVEL

4-1 3 DB LEVEL

FIG. 28. Overload wave-forms for FIG. 29. Overload wave-forms of double-

single-plane valve at 1000 cycles.

plane valve for 100 cycles.

obtained from a two-plane type of valve for the same frequencies,
respectively, at the following levels:

(a) 2 db. below overload

(b) 6 db. above overload

It is quite apparent, however, from the overload wave-forms of
both types of valve that distortion of this kind is highly objectionable
from a sound quality standpoint, and is to be avoided in either type
of valve by observing proper recording margins against overload.

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

727

Under average recording circumstances the frequency of occurrence
of overload on speech and music sounds may be expected to follow
the curve given by Sivian9 for the relative distribution of instan-
taneous amplitudes of speech and music throughout the frequency
spectrum. Thus, at very high frequencies sounds of high amplitude
are seldom to be expected.

The relation of light valve overload to conductor heating and
valve sensitivity should also be considered. If, for example, we

+ 6DB LEVEL

+ 6 DB LEVEL

A A A A A A

4-14 DB

LEVEL

FIG. 30. Overload wave-forms of
double-plane valve for 250 cycles.

+ 14DB LEVEL

FIG. 31. Overload wave-forms of
double-plane valve for 1000 cycles.

compare two types of valve differing in sensitivity by 15 decibels, it is
obvious that the normal power required to bring the more efficient
valve to the clash point is 15 decibels less than for the second valve.
This means, of course, a much smaller internal heating of the valve
since less power is dissipated. Consequently, one may expect a
smaller temperature rise in relatively sensitive valves and corre-
spondingly somewhat greater mechanical stability. One may also
expect that the maximum overload level to which a valve may be
subjected, from a temperature standpoint, relative to that level

728 SHEA, HERRIOTT, AND GOEHNER [j. s. M. p. E.

at which overload just takes place, will be largely dependent on the
relative sensitivity of the valve.

IV. A Permanent Magnetic Type of Light Valve

A new type of light valve will now be described which represents
several fundamental advances in light valve design. It is a valve
which uses a minimum total ribbon length in operation so that its
electrical efficiency is high. It is readily tuned to very high fre-
quencies. It is, within experimental error, entirely free from hys-
teresis and ribbon slippage. It contains temperature compensation
features to maintain greater constancy of spacing and tuning and,
in general, is very rugged and stable. These ribbons are clamped
in place in such a manner as to add to the constancy of spacing and

•^p^ii* ^585P^

FIG. 32. The principal parts of a new type of light valve

tuning. Possessing a permanent magnet field, it requires no field
exciting current, yet a higher flux density is secured in its air gap
than was accomplished electromagnetically in the earlier light valve.
Last but not least, it is especially light and compact.

Fig. 32 shows the principal parts of this valve. On the right and
on the left are two permanent magnets which fasten to the central
portion of the valve, the objective and condenser sides of which
unit are shown in the center of the figure. The size of the units of
the valve may be estimated by considering that each of the magnets
is 1 Y2 inches in length and that the base, as shown, of the top pole-
piece is 7/s by 3/4 inch. The ribbon when in place lies under the
glyptol clamps on the condenser side of the valve. When the
condenser side and the objective side of the valve are placed together
the slit in the center of the latter lies opposite the aperture between
the ribbons.

June, 1932]

PRINCIPLES OF THE LIGHT VALVE

729

The valve in an assembled form with associated condensing lens
is shown in Fig. 33. The total length of the unit as assembled there
is 3 5/s inches. Fig. 33 also shows in place the terminal strip to which
connections are made from the string unit of the valve.

The light valve weighs 300 grams, or about 11 ounces. The
permanent magnets account for about 220 grams of this weight, so
that the central unit of the light valve comprising the two compo-
nents shown in the center represents a weight of 80 grams or about
2.8 ounces.

The total amount of ribbon contained in the electrical circuit of
the valve as used in recording is approximately 1 inch. For the

FIG. 33.

The new valve in its assembled form with
the associated condenser lens.

present studio light valve the corresponding length is about 8 inches.
The valve can without difficulty be tuned to frequencies of the order
of 12,500 cycles. Of course, as in any valve, tuning to such a high
frequency can be accomplished only by a corresponding sacrifice in
sensitivity. But the electrical and magnetic efficiency of the valve
permit the use of relatively high tuning with a corresponding sensi-
tivity equal to or greater than the present studio light valve when a
considerably lower tuning frequency is used.

REFERENCES

1 IVES, H. E., HORTON, J. W., PARKER, R. D., AND CLARK, A. B.: "The
Transmission of Pictures over Telephone Circuits," Bell System Tech. J., 4
(April, 1925), No. 2, p. 187.

2 WENTE, E. C.: U. S. Patent 1,686,355.

730 SHEA, HERRIOTT, AND GOEHNER [J. S. M. P. E.

8 MACKENZIE, D.: "Sound Recording with the Light Valve," Trans. Soc.
Mot. Pict. Eng., 12 (Sept., 1928), No. 35, p. 730.

4 SILENT, H. C., AND FRAYNE, J. G.: "Western Electric Noiseless Recording,"
J. Soc. Mot. Pict. Eng., 18 (May, 1932), No. 5, p. 551.

5 RAYLEIGH, "Theory of Sound," Vol. 1, Macmillan & Co., London, England
(1896), p. 175.

6 This analysis follows closely an unpublished memorandum by C. F. SACIA of
the Bell Telephone Laboratories.

7 STRYKER, N. R.: "Scanning Losses in Reproduction," /. Soc. Mot. Pict.
Eng., 15 (Nov., 1930), No. 5, p. 610.

8 CECCARINI, O. O.: "The Measurement of Light Valve Resonance by the
Absorption Method," 7. Soc. Mot. Pict. Eng., 15 (July, 1930), No. 1, p. 60.

9 SIVIAN, L. J.: "Speech Power and Its Measurement," /. Acoust. Soc. of
Amer., 1 (Jan., 1930), No. 2, Part 2, p. 1.

DISCUSSION

MR. KELLOGG: Has it been found necessary or desirable to try to introduce
more damping than what naturally comes from friction at the support?

MR. SHEA: There was described before the Hollywood Convention an oil-
damped type of light valve used at one of the studios on the West Coast. Gener-
ally speaking, we prefer to move the resonance out of the recording range, so
that the resonance does not influence the recording.

MR. PALMER: Is the fact that this light valve is a more efficient and better
piece of apparatus shown in the results obtained in the theater, when the sound is
taken off the film?

MR. SHEA: There are many causes of loss of quality in reproduction, and of
those, the leakage due to the light valve is among the smallest, and it has been
made yet smaller in this valve. I am quite sure that such losses in quality are
negligible compared with other losses encountered commercially.

The improvement in size and weight, the stability of the valve, and its ability
to retain its spacing and tuning, mean a great deal either in portable or studio
work. The advantages of the new valve are chiefly its operating advantages;
our extensive laboratory tests convinced us that the old valve, with proper
care, gave excellent quality. But with the new valve less care will be required.

MR. OLSON: Is there any tendency for ribbons to turn out of their normal
plane?

MR. SHEA: Not if one can judge by the oscillograms of the modulated light.
They do not apparently turn to such an extent that any amount of light gets
through gaps that might be so created.

MR. EVANS: We use both types of valves in our studio, and our experience
to date with the baby light valve is a favorable one. It has one advantage and
one disadvantage — with a compensation available for the latter. The ad-
vantage is that a higher tuning frequency reduces the frequency at which over-
loading occurs. If the valve is tuned to 8600 cycles, as the old ones are, sounds of
that frequency, of which there are many, may, when added to the other fre-
quencies existing, cause the valve to clash. When it is tuned to 11,000 cycles,

June, 1932] PRINCIPLES OF THE LlGHT VALVE 731

as are the baby light valves, we avoid clashing at 8600 cycles, and 11,000 cycle
sounds are less frequent and generally weaker.

On the other hand, the disadvantage is that there are frequency losses at 5000
and 6000 cycles in film processing and other places that cause quite a con-
siderable drop in the reproduction at those frequencies. If the light valve
is tuned to 8600 cycles, the resonance of the light valve tends to compensate
for these losses, so that a loss is incurred at frequencies of 5000 to 7000 cycles by
going to the baby light valve. This loss can be compensated for, however, by
equalization in the electrical circuit. So the net result, I think, is that the baby
light valve is quite a bit better than the other. Certainly it is more stable.

MR. SHEA: I believe that such a comparison should be made, not between
one design of valve and another, but rather on the basis that the higher the
tuning frequency, the greater the sacrifice to be made at the present time in
eliminating the equalizing action of the light valve. With the old valve the
same difference exists as between six or eight thousand cycle tuning, and ten
thousand cycle tuning. We have employed experimentally many of the old
light valves, some with the shorter bridge length, at quite high tuning frequencies,
exceeding those that you mentioned for the small light valve; and I think we
ought to make the comparison rather on the basis that the choice of tuning
frequencies for certain reasons leads to a sacrifice in another direction at the
present time.

MR. KURLANDER: Is there any effect on recording? Do the two legs of the
ribbons cut each other's magnetic field, or is that field neutralized?

MR. SHEA: The ribbons act independently as far as any one can tell. There
must be minor effects, such as that due to skin effect, at the high frequencies, but
they appear to be very minor.

VARIATION OF PHOTOGRAPHIC SENSITIVITY WITH
DIFFERENT LIGHT SOURCES*

RAYMOND DAVIS AND GERALD K. NEELAND**

Summary. — The variation of photographic sensitivity (as measured by the index
10/Em) with different sources of light of equal visual intensity, but having different
distributions of energy, was experimentally obtained. Ordinary, orthochromatic,
and three new fast panchromatic plates were investigated. Distinction was made
between two types of sensitivity comparisons. Thus it was found that: first, the
ratio of the sensitivity of any one of the panchromatic plates to that of the ordinary
plate was greater with incandescent lighting than when "sunlight" was used, and
second, in all cases the panchromatic plates were less sensitive to incandescent lighting
than to "sunlight." Approximate factors are included (for the particular emulsions
studied) by which visual exposure meter readings should be multiplied when certain
types of illuminants are encountered.

I. INTRODUCTION

The sensitivity of photographic emulsions to light from various
sources having different spectral energy distributions has recently
become of considerable interest. This is due, in part, to the dis-
covery of new dyes which make possible an increased sensitivity to
light in the red end of the spectrum.

Artificial illuminants, particularly incandescent lamps, have a
relatively large amount of energy in the longer wavelengths. With
such illumination, therefore, it is natural to expect the increase in red
sensitivity would make the new emulsions more efficient than former
ones, i. e., less exposure would be required to obtain the same photo-
graphic effect.

Comparisons of sensitivity may be made in one of two ways.
First, one may compare the sensitivity of two different emulsions
using the same source of light. Second, the sensitivity of an emulsion
to light from one source may be compared with its sensitivity to light
from another source of equal "intensity" but having a different rela-
tive spectral energy distribution.*

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Publication
approved by the Director, Bureau of Standards, U. S. Department of Commerce.
** U. S. Bureau of Standards, Department of Commerce, Washington, D. C.
732

SENSITIVITY WITH LIGHT SOURCES 733

The question arises as to whether the expression "equal intensity"
should be taken to mean equal energy or equal visual intensity. That
the two would not be necessarily the same is obvious. In certain
problems, such as determining what type of illuminant is most
efficient with a given emulsion, it is desirable to have the sources
under comparison arranged to deliver to the plate the same amount
of energy per unit time. With this arrangement, the source to which
the emulsion has the higher sensitivity would be the most efficient,
i. e., a greater photographic effect is obtained with a given amount of
energy.

However, in carrying out the experimental work described in this
paper, the expression "equal intensity" has been taken to mean "equal
visual intensity." That this condition for comparisons of sensitivity
has practical significance may be seen from consideration of a "visual
photometer" type of exposure meter in common use. With this
instrument, the exposures for two objects are indicated as being the
same when the light reflected from each object appears equally
bright, regardless of its energy distribution.** Therefore, to use
this type of meter properly, it is necessary to know how the sensi-
tivity of the emulsion varies with different types of illuminants.

In this paper are given several examples showing the variation in
sensitivity (of panchromatic, orthochromatic, and ordinary plates)
with the energy distribution of the light where the visual intensity of
the light incident on the emulsion is kept constant. This informa-
tion is of interest, not only with regard to the use of an exposure meter,
but particularly because it furnishes actual data on the performance of
new panchromatic emulsions as compared with ordinary emulsions.

H. PROCEDURE AND RESULTS

The emulsions were exposed to three sources, each having a differ-
ent energy distribution. These were: (1) an incandescent lamp
operating at a color temperature* of 2360° K., (2) an incandescent
lamp operating at a color temperature of 2810°K., and (3) an incan-

* The terms "relative spectral energy distribution" and "energy distribution"
have the same meaning in this paper.

** It is assumed here that the photometer employs no blue, or other "correc-
tion" filter. However, the mere inclusion of such a filter would not, necessarily,
correct for the different energy distributions of various sources. Thus, only
with ordinary plates, could reliable readings be obtained with a blue filter.

734

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

descent lamp operating at 2360 °K. in combination with a Davis-Gib-
son filter such that the resulting light closely approximates the chroma-
ticity and the energy distribution of mean noon sunlight.**

Energy distribution curves for these three sources, together with
the visibility curve V, are given in Fig. 1 .

These sources were chosen because it was desired to approximate,
as far as possible with available equipment, light conditions, both

150

125

fc 100

350

400

450

500

750

550 600 650 700
Wave Length -- millimicrons

FIG. 1. Relative spectral energy distributions of the light sources used. The
circles represent the spectral energy distribution of the lamp and filter combina-
tion producing artificial sunlight. The Eighth International Congress of Pho-
tography adopted this combination as standard for the sensitometry of negative
materials. The dashed line represents the spectral energy distribution of the
2810°K. source and the solid line that of 2360°K., both arbitrarily adjusted to
100 at 560 mn The visibility curve is marked V.

out of doors in sunlight and inside with artificial illuminants. Thus, a
60- watt, gas-filled incandescent lamp approximates a color tempera-
ture of 2800°K., while the old vacuum tungsten lamp approximates a

* By a color temperature of 2360 °K., is meant that the chromaticity of the
light approximates that of a Planckian radiator at an absolute temperature of
2360°. In the case of an incandescent lamp, the relative spectral energy distri-
bution, as well as the chromaticity, is also very close to that of the Planckian
radiator at the specified temperature.

* * The chromaticity of mean noon sunlight is approximately that of a Planckian
radiator at 5400 °K.

June, 1932] SENSITIVITY WITH LlGHT SOURCES 735

color temperature of 2400 °K. The incandescent lamps, such as are
used for studio work in motion pictures, are close to a color tempera-
ture of 3000 °K.

The sources were placed in a non-intermittent sector-wheel sensi-
tometer1 at such distances from the plane of the emulsions that the
visual intensity at this plane was 1 meter candle.

As pointed out in a previous paper,2 we believe that the most
adequate representation of the sensitivity of an emulsion is given by a
curve showing the relation between sensitivity and the time of
development. It was thought best to use the index of sensitivity
10/£w23, where Em is the exposure of that point in the "toe"
region of the characteristic curve of the emulsion where the gradient
is 0.2

In each test, 24 sensitometric strips, backed by a black shellac
mixture for preventing halation, were exposed, 8 to each of the three
sources of light. In several instances, the tests were repeated a
number of times. The 24 exposed strips were placed in eight racks,
each containing 3 strips, one exposed to each source. The racks
were then placed together in a single tray of developer which was
kept at a temperature of 20 °C. by a water bath. They were
removed, one rack at a time, after periods giving a series of times of
development from 2 to 22 minutes. Thus the usual range of times of
development was more than covered. The developer was agitated
by brushing the plates with a camel 's-hair brush.

Pyrogallol developer, compounded according to the following
formula, was employed:

U) Water 1000 cc.

Potassium metabisulphite 12 g.

Pyrogallol 60 g.

(5) Water 1000 cc.

Sodium sulphite (anhydrous) 90 g.

Sodium carbonate (anhydrous) 75 g.
(To develop, one part each of A , B, and C was mixed with seven parts of water.)

The densities of the developed strips were measured in diffuse light
with a Martens polarization photometer. A family of eight charac-
teristic curves (total density vs. log exposure) was obtained for
each source. From each curve a value of the index 10/Em was de-

736

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

rived. Thus, a measure of the variation of sensitivity with the time of
development was obtained for each source. From these data the
sensitivity vs. development- time curves were drawn.

TABLE 1

Photographic Plates Investigated

Description of plate Identification Figure

American "medium speed" ordinary A 2

American "high speed" orthochromatic B 3

American "high speed" panchromatic C 4

English "high speed" panchromatic D 5

German "high speed" panchromatic E 6

15
minutes

FIG. 2.

5 10

Development time

Sensitivity vs. development-time curves for plates A under
three conditions of illumination.

In Table 1 are listed the plates included in the study, the types of
emulsions, and the identifying letters used in the figures and tables.
The selection is intended to be representative of the most sensitive
emulsions now available, and to include an example of each of the
three common types of emulsions — ordinary, orthochromatic, and
panchromatic.

June, 1932 J SENSITIVITY WITH LlGHT SOURCES

737

III. DISCUSSION OF RESULTS

Curves representing the change in sensitivity (as measured by the
index 10/Em) with the time of development for each of the three
sources are given in Figs. 2 to 6, inclusive.

In Fig. 2, the sensitivity vs. development- time curves are given for
plates A. The three curves are similar in shape, as would be ex-
pected, and show the usual3 optimum time of development (in this in-
stance 5 minutes). Since these plates are "ordinary" (not "color-sen-
sitive"), the drop in sensitivity to the light from sources of lower color
temperature is large compared with that of the panchromatic plates.

I5UU

1000
JO

500

FIG. 3

Plates EyX"
Orthochr^mottic

o ~~^~

}

~>-^_

o-/flw/7 noon sun

Q-28/0' 'K

D 5 10 15 20 21
Development time in minutes
. Sensitivity vs. development-time curves for plates B unde

three conditions of illumination.

In Fig. 3, similar curves are given for plates B. Although these
plates are orthochromatic, the drop in sensitivity is nearly as great as
with the ordinary plates.

In Fig. 4, the curves for the panchromatic plates (C) indicate a
drop in sensitivity that is, for practical purposes, negligible. Al-
though these are not as sensitive as plates D, their sensitivity is less
diminished by the incandescent type of illumination.

Fig. 5 presents the curves for plates D, the most sensitive of the five
brands studied. Even though their loss in sensitivity with in-
candescent illumination is greater than with plates C, the plates D* are
more sensitive with all three sources.

* It is interesting to note that these plates, D, advertised to have a higher
H & D speed to "half- watt" light than to daylight, do not appear to justify the
claim.

738

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

In Fig. 6, plates £,* we have an example of a panchromatic
plate in which the maximum sensitivity is reached at a long time of
development.

Table 2 gives the data on the sensitivity (as measured by 10/£OT)

1000

5 10 15

Development time in minutes

FIG. 4. Sensitivity vs. development-time curves for plates C under
three conditions of illumination.

1500

1000

Plates 0
Panchromatic
o /

o- -Mean noon sun

05 10 15 20 25

Development time in minxjtet

FIG. 5. Sensitivity vs. development-time curves for plates D under
three conditions of illumination.

* Although advertised as "special rapid," these plates are considerably slower
than the other two brands of panchromatic plates.

June, 1932]

SENSITIVITY WITH LIGHT SOURCES

739

of the five brands of plates in a more easily visualized form. To ob-
tain these values, each sensitivity vs. development-time curve is
represented by its maximum value.* The resulting number is thus

TABLE 2

Maximum Values of 10 /Em Taken from the Sensitivity vs. Development- Time

Plate

A (Ordinary)
B (Orthochromatic)
C (Panchromatic)
D (Panchromatic)
E (Panchromatic)

Curves

Maximum Value of the Sensitivity Index
2360 °K. 281 0°K. Mean Sun

150
450
600
725
240

225
620
625
975
300

505
1350

755
1450

380

300

100

Plates E
Panchromatic

o-/1ean noon sun
D-25/0" K

15
in minutes

FIG. 6.

5 10

Development time
Sensitivity vs. development-time curves for plates E under
three conditions of illumination.

more accurately representative than any of the experimentally ob-
tained values of the index. The table shows that: (1) the ratio
of the sensitivity of the panchromatic to the ordinary plates is greater
with incandescent lighting than with "sunlight," and (2) with every

* The representation of the panchromatic plates £ by a single number is not
strictly comparable with the others, since the maximum value of the index
occurs at a development time exceeding 22 minutes and was, therefore, unobtain-
able.

740 R. DAVIS AND G. K. NEELAND [J. S. M. P. E.

brand of plate the sensitivity decreases with decreasing color tempera-
ture of the source.

The values of Table 3 were obtained from the data in Table 2.
This table gives the correct exposure for a given plate to a given
source, that of the ordinary plates to sunlight being taken as
unity.

TABLE 3

Factors Indicating the Exposure Necessary to Obtain Similar Results with Each

of the Five Plates under the Three Conditions of Illumination, That for the Ordinary

Plates to Sunlight Being Taken as Unity

Exposure Factor
Indoors
Plate Vacuum Lamps Gas-Filled Lamps Sunlight

A (Ordinary) 3.4 2.2 1.0

B (Orthochromatic) 1.1 0.81 0.37

C (Panchromatic) 0 . 84 0.81 0 . 67

D (Panchromatic) 0.70 0.52 0.35

£ (Panchromatic) 2.1 1.7 1.3

At present the. use of the 'Visual photometer" type of exposure
meter is fairly common. Usually with such a device provision is
made for the difference in sensitivity of the several types of photo-
graphic materials. However, assuming that the correct exposure
to sunlight for a certain brand of plates is known, it is important to
know further how the exposure should be modified when using the
meter under "incandescent" lighting conditions. This information,
for the five brands of plates studied, is supplied by Table 4.

TABLE 4

Factors Indicating the Exposure Necessary to Obtain a Similar Result with the Two
"Incandescent" Illuminants, That with Sunlight Being Taken as Unity

Exposure Factor
Indoors
Plate Vacuum Lamps Gas-Filled Lamps Sunlight*

^(Ordinary) 3.4 2.2 1.0

B (Orthochromatic) 3.0 2.2 1.0

C (Panchromatic) 1.3 1.2 1.0

D (Panchromatic) 2.0 1.5 1.0

£ (Panchromatic) 1.6 1.3 1.0

* These values do not indicate that the plates are equally sensitive to sunlight.

June, 1932] SENSITIVITY WITH LlGHT SOURCES 741

REFERENCES

1 Bureau of Standards Science Paper No. 511.

2 DAVIS, RAYMOND, AND NEELAND, G. K.: "An Experimental Study of Several
Methods of Representing Photographic Sensitivity," Bureau of Standards
Research Paper No. 355.

3 JONES, L. A., AND RUSSEL, M. E.: "The Expression of Plate Speed in Terms
of Minimum Useful Gradient," Proc. 7th International Congress of Photography
(1928), p. 130.

DISCUSSION

MR. HARDY: The specification of the speed of a photographic material to
different light sources may be misleading if it is forgotten that one seldom, in
practice, photographs the light source. In other words, the only time these
data apply directly is when one takes a picture, in one case of the sun, or, in the
other case, of the tungsten lamp. Actually, the scenes photographed are, of
course, colored, and may represent practically any distribution of energy through-
out the spectrum. It is difficult to suggest any better answer than the one the
author has given to this question, but it is nevertheless necessary to recognize the
fact that if the subject happens to be colored, a different factor applies.

PAST-PRESIDENT CRABTREE: I might add to what Mr. Hardy says that the
only way in which to measure speed is to measure it under the actual conditions
that obtain when using photographic material.

VARIATION OF PHOTOGRAPHIC SENSITIVITY WITH
DEVELOPMENT TIME*

RAYMOND DAVIS AND GERALD K. NEELAND**

Summary. — The relation between photographic sensitivity and the time of de-
velopment has been studied. Three modifications of Hurter and Drijfield's method
of measuring plate speeds have been used as indexes of sensitivity. The results are
generally in accord with those of Bullock, who investigated the relation of "threshold
speed" to development time. An optimum time of development is indicated, which
varies with the type of emulsion.

The differences between the three indexes have been brought out. Just which index
is most satisfactory depends somewhat on the nature of the individual problem.
It is concluded that, in many cases, a single value of any index is inadequate to
represent the sensitivity of a given emulsion. The clearest representation may be
had by a curve plotted between a sensitivity index and development time.

I. INTRODUCTION

Several methods of representing the sensitivity of photographic
emulsions have been proposed from time to time. At first, the
sensitivity, or "speed," was taken as inversely proportional to the
exposure required to produce (on development) a just perceptible
density. This has been called threshold speed. The difficulty in
specifying accurately a "just perceptible" density has discouraged
the use of this method, although the method and its modifications
are still employed to some extent.

Hurter and Driffield's system was based on a characteristic curve
representing density as a function of the logarithm of the exposure.
The exposure at the intersection of an extension of the straight-line
portion of this curve with the base line was called the "inertia" of the
plate. The "speed" of the plate was taken as proportional to the
reciprocal of the inertia. It was found that, in the absence of soluble
bromides, the speed of a given plate was constant regardless of the
time of development. Later it was shown that the H & D speed of a

SENSITIVITY WITH DEVELOPMENT TIME

743

plate varies somewhat with development. The nature of this varia-
tion has not, to our knowledge, been systematically investigated.
The discrepancy between the results of Hurter and Driffield and those
of later workers is probably due to changes in the methods of manu-
facturing emulsions.

Bullock1 investigated the relation between threshold speed and
development time (using a Chapman- Jones plate tester), and ob-
tained curves indicating an optimum time of development.

3.0

2.4

1.6

fog I ,- 73 1- to
'oglz'i3d-io
/ogfm;7.?0-f
fog Em;7.75-tO

, -f 2)0

-(2)

FIG. 1. Illustrating the methods of obtaining the three indexes of sensitivity
from each of two characteristic curves.

We have investigated the relation between sensitivity and time
of development, using three modifications of Hurter and Driffield' s
system as indexes of sensitivity. We shall, first, show the nature
of the relation; second, indicate the relative merits of the three
indexes; and third, show that a single value of any index is, in many
cases, inadequate to represent the sensitivity of an emulsion.

The first method substitutes the tangent of the curve at the point
of maximum gradient for the extended straight-line portion. The
value of the exposure at the intersection of this tangent with a hori-

744 R. DAVIS AND G. K. NEELAND [j. S. M. P. E.

zontal line representing the fog density is designated as i. The sen-
sitivity index on this basis is 10/e.

An index of sensitivity, that has frequently been used, is derived
in a manner similar to that adopted by Hurter and Driffield. That is,
the sensitivity is taken as inversely proportional to the value
of the exposure at the intersection of an extension of the straight-
line portion of the characteristic curve with the exposure axis,
where total density is plotted against log exposure. The exposure
at this point of intersection is sometimes erroneously called the
inertia.

In the second method employed in the present work, a quantity /
is defined as the value of the exposure at the intersection of the
tangent of the characteristic curve, at the point of maximum
gradient, with the exposure axis. The sensitivity index on this basis
is 10/7.

Note that, when the characteristic curve has a central straight
portion, i will be equal to the inertia, and / will be equal to the above-
mentioned quantity, erroneously called "inertia."

The third method is that of Jones and Russell,3 who take "speed''
as proportional to the reciprocal of a quantity Em, which is defined
as the exposure corresponding to the point in the "toe" region of the
characteristic curve where the gradient is 0.2. The sensitivity index
on this basis is 10 /Em.

This method is applicable in those cases where the characteristic
curve is not straight over an appreciable length, or where an exposure
range that more than covers the straight-line portion is encountered.
The latter condition often occurs in practical photography. The
expression of sensitivity in terms of inertia is obviously unsatis-
factory under either condition.

II. PROCEDURE

In order to determine for each emulsion the nature of the relation
between the sensitivity and the time of development, sets of 11 test
strips each, backed with a black shellac mixture for preventing
halation, were exposed in a non-intermittent sector wheel sensi-
tometer. The source of light, having a visual intensity of one meter
candle and the quality of sunlight, was produced by an incandescent
lamp operated at a color temperature of 2360°K., in combination
with a Davis-Gibson filter.2 The 11 exposed test strips were de-

June, 1932]

SENSITIVITY WITH DEVELOPMENT TIME

745

veloped together in total darkness in a tray of pyro* at 20 °C. The
strips were removed, one at a time, giving development times ranging
from 1 to 22 minutes. The developer was agitated by brushing the
plates with a camel's-hair brush. The resulting densities were
measured in diffuse light with a Martens polarization photometer.

The characteristic curve corresponding to each test strip was
plotted, and from it values of the indexes 10/t, 10/7, and 10 /Em

5 10 15 20 25

Development Time in minutes

FIG. 2. Sensitivity vs. development- time curves
for plates A. The solid curves show the variation
in each of the indexes with the time of development.
The dashed curve shows the change in the maximum
gradient, G (max.), with the time of development.

were obtained. From this data three curves were drawn for each
emulsion, showing the variation of the indexes with the time of de-

* Pyrogallol developer,
was employed:

compounded according to the following formula,

(C)

Water .................................. 1000 cc.

Potassium metabisulfite ................... 12 g.

Pyrogallol ............................... 60 g.

Water .................................. 1000 cc.

Sodium sulfite (anhydrous) ................ 90 g.

Water ................. ................. 1000 cc.

Sodium carbonate (anhydrous) ............ 75 g.

(To develop, one part each of A, B, and C was mixed with seven parts of water.)

746

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

velopment. The curves are given in Figs. 2 to 6, inclusive. In
each figure a curve is included which shows the relation of the maxi-
mum gradient to the time of development for that particular emul-
sion. In Table 1 are listed the plates included in this study, the
types of emulsions, and the identifying letters used in the figures.

TABLE 1

Photographic Plates Investigated

Description of plate Identification Figure

American "high speed" orthochromatic A 2

American "medium speed" ordinary B 3

German "high speed" orthochromatic C 4

English "high speed" panchromatic D 5

American "process" ordinary E 6

1200

iOOO

5 10 15 20 25

Development Time In minutes

FIG. 3. Sensitivity vs. development- time curves for plates B. The solid
curves show the variation in each of the three indexes with the time of de-
velopment. The dashed curve shows the change in the maximum gradient,
G (max.), with the time of development.

HI. DISCUSSION OF RESULTS

1. Characteristics of the Three Indexes. — Figs. 2 to 6 show the
variation of sensitivity, as measured by the three indexes, with the

June, 1932] SENSITIVITY WITH DEVELOPMENT TlME

747

time of development. For plates A, Fig. 2, all three indexes show a
pronounced maximum in the neighborhood of 3 to 4 minutes' develop-
ment. However, it is to be seen that the curve for 10/7 is distinctly
different in shape from the other two, showing a second upward trend
after 8 minutes' development. Note that, with these plates, the
curve for 10/7 lies above that for 10 /Em.

For plates B, Fig. 3, we again have all indexes a maximum at

?000

1500

"> 1000

500

5 10

Development time

15 20

in minutes

FIG. 4. Sensitivity vs. development- time curves for plates C. The solid
curves show the variation in each of the three indexes with the time of de-
velopment. The dashed curve shows the change in the maximum gradient,
G (max.), with the time of development.

about 3 to 4 minutes' development. As before, the curve for 10/7
differs in shape from the other two.

For plates C, Fig. 4, similar results are obtained. The curve for
10/7 differs less from the other two than in the two preceding cases,
but the beginning of the second upward trend is indicated at about 18
minutes' development.

For plates D, Fig. 5, the curves for both 10/t and 10/7 indicate an
early maximum not shown by the curve for I0/Em. Here the second
upward trend of the 10/7 curve begins at about 14 minutes' develop-
ment.

748

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

For plates E (Fig. 6) the curves for both 10/t and 10/7 again indi-
cate an early maximum sensitivity. Note that the G (max.) curve
does not flatten out here as with the other plates. As with plates D,
the curve for W/Em indicates maximum sensitivity at time of de-
velopment in excess of 22 minutes.

1400

1200

• 000

800

400

200

Plates D

*- to/ i

D - 10/C
« - 6- (max.)

2.4

1.6 -x
o

1.2

0.6

5 10 15

Development time in

20
minutes

FIG. 5. Sensitivity vs. development-time curves for plates D.
The solid curves show the variation in each of the three indexes
with the time of development. The dashed curves show the
change in the maximum gradient, G (max.), with the time of
development.

Figs. 2 to 6, inclusive, show that the second upward trend in the
10/7 curve usually occurs in the region where the rate of increase of
maximum gradient has fallen off.* That is, after this point is
reached, the maximum gradient line tends to shift so as to remain
parallel to itself, resulting in an increase in the values of 10/7. That

* On all except the process plates, the rate of growth of fog was practically
constant over the entire range of development times.

June, 1932]

SENSITIVITY WITH DEVELOPMENT TIME

749

this rise does not necessarily signify an actual increase in sensitivity,
has been shown by the "hypothetical" case discussed in a note4
by the authors. In this "hypothetical" case, we have two character-
istic curves identical in form but one raised above the other. (Total
density is here plotted against log exposure.) The tangents at the
points of maximum gradient are, of course, parallel, and intersect
the zero density line so that the upper curve gives the higher value of

Plates E

A- to/f
a- 'O/t
• — & (max)

6.0

5jO

3.0

2.0

1.0

0 5 1O

Development time in minutes

FIG. 6. Sensitivity vs. development- time curves for plates E. The solid
curves show the variation in each of the three indexes with the time of de-
velopment. The dashed curve shows the change in the maximum gradient,
G (max.), with the time of development.

10/7. It is obvious that the two emulsions represented by these
curves should be considered equally sensitive since, if two negatives
were made under the same conditions, one with each emulsion, they
could be made to yield identical positives by properly adjusting the
printing exposure.

2. Method of Expressing Relative Sensitivity. — In Fig. 7, the
1Q/Em curves are plotted for plates A, B, C, and D. The superiority

750

R. DAVIS AND G. K. NEELAND

[J. S. M. P. E.

of such a curve over a single quantity for expressing sensitivity is
obvious from this figure. The sensitivity of an emulsion having a
comparatively flat curve can be satisfactorily represented by a
single value; an emulsion having a curve with a pronounced maxi-
mum obviously can not. The procedure, followed by some workers,
of giving values at 3 times of development is sufficient in some cases

2000

1500

1000

500

5 10

Development

time in minutes

FIG. 7. Illustrating the manner of expressing the relative sensitivity of
emulsions. All curves are plotted to the same scale. The curve for the
process plate E is not included, as its sensitivity is too low to show to ad-
vantage.

provided that the times are properly chosen. However, an adequate
representation of the sensitivity of emulsions seems to require their
sensitivity vs. development- time curves.

REFERENCES

1 BULLOCK, E. R.: "On Variations in Threshold Speed according to the
Development and Conditions of Development," Communication No. 268, Parts I
and II, Abridged Scientific Publications, Research Laboratories, Eastman Kodak
Company (1926), p. 144.

2 DAVIS, RAYMOND, AND GIBSON, K. S.: "Filters for the Reproduction of
Sunlight and Daylight and the Determination of Color Temperature," Bureau
of Standards, Mis. Pub., No. 114.

June, 1932] SENSITIVITY WITH DEVELOPMENT TlME 751

3 JONES, L. A., AND RUSSELL, M. E.: "The Expression of Plate Speed in
Terms of Minimum Useful Gradient," Proc. 7th Inter national Congress of Photog-
raphy (1928), p. 130.

4 DAVIS, RAYMOND, AND NEELAND, G. K.: "A Note on the Speed of Photo-
graphic Emulsions," /. Opt. Soc. of America, 21 (July, 1931), No. 7, p. 416.

DISCUSSION

PAST-PRESIDENT CRABTREE: The variation in the speed was not more than
twenty-five per cent, which, from a practical standpoint, is not very serious.

MR. NEELAND: In other cases the variation is greater than fifty per cent.
We have plates where the sensitivity is reduced to one-half its maximum value
after eighteen or twenty minutes of development.

MR. SEASE: What fog levels did you reach?

MR. NEELAND: Fog values of 1 to 1.5 might be reached in certain cases,
but usually they are not greater than 1 .

PAST-PRESIDENT CRABTREE: Gammas in practice are much lower than those
limits to which you have been developing. Furthermore, the solutions that you
used were absolutely fresh and new. Do the results apply to solutions that are
partially exhausted?

MR. NEELAND: Solutions that are not fresh would have a maximum, at least
at a higher time of development.

PAST-PRESIDENT CRABTREE: We rarely use a solution that is absolutely
fresh. As soon as a certain number of feet of film are developed the stock solution
in the system of the developing machine is partially contaminated by exhaustion
products that cut down the fog values, and would change that data materially.

MR. CARROLL: With regard to the lower gammas in practice, the change in
sensitivity number was greatest at about seven-tenths gamma, a fact which
indicates the practical importance of the thing.

A REFLECTOR ARC LAMP FOR PORTABLE PROJECTORS
HARRY H. STRONG**

Summary. — A description of a portable reflector arc lamp and current rectifying
device designed especially for use with portable sound projection equipment. The
lamp is small, is adapted to portable projectors, is well proportioned, and of exceptional
power. The rectifier is of the familiar Tungar type and gives full-wave rectification.
It is compact and portable in design, and the whole equipment is of an efficiency
high enough so that ample screen illumination may be secured with current drawn
from the 110-volt lamp socket.

A new field for motion picture projection has been created by the
perfecting of portable sound equipment. This type of equipment is
finding extensive application for educational and advertising pur-
poses, as well as for entertainment in small theaters and in auditoriums
of moderate size.

In all of these uses, however, the audience is composed of
individuals accustomed to the large projected image and brilliant
screen illumination characteristic of the motion picture theater
today. They are no longer satisfied with a picture three or four feet
wide, a low intensity of screen illumination, and sound coming from
a position at one side of the picture.

The attainment of satisfactory results in the use of portable
sound equipment requires a picture eight to twelve feet wide, a
porous screen permitting the sound to come from the screen itself, and
a light source of sufficient power to afford a screen illumination and
brilliancy comparable with that seen in the popular theaters.

The d-c. carbon arc is the only available source of light possessing
sufficient power and concentration to satisfy the requirements of this
newly created condition. Adaptation of the d-c. carbon arc to
portable equipment, however, presents certain problems, the solution
of which has required careful study and extensive experimentation.

Portability places definite restrictions on the weight and bulk of
equipment, and these factors must be given careful consideration in

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Strong Electric Co., Toledo, Ohio.
752

A REFLECTOR ARC LAMP 753

working out details of design. Direct current is no longer available
in most localities, except in central urban districts. This condition
makes it necessary to provide some means of converting the alter-
nating current of the power supply to direct current suitable for use
at the arc.

Simplicity of operation is a requisite of prime importance in port-
able equipment, since the equipment is more likely to be used by an
operator of little experience than by one having extensive knowledge
of motion picture equipment. Finally, results must be attained in a
manner to afford efficient use of electrical power so that connections
may be made to any available light socket.

It is the purpose of this paper to describe a portable, reflector
type, carbon arc lamp and rectifier, developed to meet the difficult
conditions outlined in the preceding paragraphs.

The remarkable success of the reflector type of arc lamp in standard
theater projection has led to the adoption of the reflector principle in
the development of the portable unit. The superiority 'of d-c.
reflector arcs is due basically to the interception of a large angle of
radiated light emanating from the crater of the positive carbon and
the reflection of this light, as a converging beam, to the film
aperture. The adaptation of these principles has resulted in the
development of a compact unit only 18 inches long, 12 inches high, and
10 inches wide, well proportioned and having exceptional power. Fig.
1 shows this lamp adapted to the modern portable projector.

The urgency of reliable performance in the hands of the lay operator
requires the elimination of hand control of the arc. This is accom-
plished by means of a fully enclosed control motor, which is
mounted as an integral part of the lamp house, and which automati-
cally feeds the carbons at exactly the same rate at which they are
consumed, thus maintaining the proper arc length throughout the
entire burning of the carbons.

The automatic arc control system operates upon the principle
that certain electrical characteristics of an arc are changed as the
carbons are consumed. Use is made of these changes to control
directly the speed and direction of rotation of a differentially com-
pounded motor. A wiring diagram of this motor is shown in Fig. 2.

The armature of the control motor is geared to the carbon carriages
in such a manner that the rotation of the armature will cause the car-
bons to move closer together, or farther apart, depending upon the
direction of rotation of the armature. Since the armature is so con-

754

HARRY H. STRONG

[J. S. M. P. E.

nected that it is electrically energized at all times, its speed and direc-
tion of rotation, and hence the movement of the carbon carriages, will
depend upon the direction and strength of the magnetic field passing
through the armature.

The field of the motor consists of two windings which are differen-
tially connected — a shunt or potential winding, and a series or current

FIG. 1.

Portable reflector arc lamp, adapted to Western
Electric portable sound projector.

winding. The shunt winding, comprising many turns of fine wire, is
connected through a rheostat across the arc. This sets up a magnetic
field which tends to rotate the armature in a direction that feeds the
carbons toward each other. The series winding, comprising a few
turns of heavy wire, is connected in series with the arc, and sets up a
magnetic field opposed to that of the shunt winding, thus causing the
armature to rotate in a direction that separates the carbons.

June, 1932]

A REFLECTOR ARC LAMP

755

When the arc is burning properly, and the rheostat which controls
the strength of the shunt winding is adjusted so that both the series
and shunt windings are of equal strength, the two field windings
neutralize each other and the armature does not rotate. As the arc
gap increases, due to the normal burning of the carbons, the potential
across the arc is increased and the current through the arc decreased.
An unbalanced condition is thus created in the field windings, and the
resultant magnetic flux through the armature is equal to the difference
in strength between the shunt and series field windings. Under these

FIG. 2. Wiring diagram of arc control system.

conditions the armature will rotate in a direction to bring the carbons
closer together and compensate for the burning loss.

The rheostat, in series with the potential field windings, permits
adjustment of the strength of the potential field in relation to that of
the series field, so that proper motor speed may be established and the
proper arc length maintained for any given operating current. The
rheostat is provided with an adjusting knob and indicating dial at the
top of the lamp house, as shown in Fig. 3.

The operation of the arc control is entirely automatic and
continuous. Once the arc has been struck and the carbons separated
to the proper arc length, the control motor rotates slowly and
continuously, feeding the carbons toward each other at a rate that

756

HARRY H. STRONG

[J. S. M. P. E.

exactly equals their consumption. A uniform arc gap is thus main-
tained without manual control.

An arc imager is mounted on the side door of the lamp house ad-
jacent to the window. The imager projects an image of the arc and
the incandescent carbon tips to a small screen secured to the side of
the vent stack. This device can be seen in Fig. 4. While the lamp is
in operation, the lines on the imager screen will indicate the proper
position of the positive crater in its relation to the focus of the reflector,
as well as the correct position of the negative carbon in relation to the
positive.

FIG. 3.

Portable arc lamp showing knurled knob for adjusting
focus and motor control rheostat with dial.

The operator may adjust the position of the positive crater to the
exact focus of the mirror by turning the knurled knob at the lower
left-hand corner of the lamp house, as shown in Fig. 3.

"Striking the arc" is accomplished by turning the ball crank at the
rear of the lamp house. This crank is clearly shown in Fig. 4. The
ball crank further permits manual adjustment of the arc length, i. e.,
adjustment of the negative carbon in relation to the positive. Once
this relation has been set, the carbons seldom require further manual

June, 1932] A REFLECTOR ARC LAMP 757

adjustment. The automatic control, under normal conditions, will
maintain the proper arc length and position of the positive crater for
the entire burning period of one complete trim.

The optical system comprises an elliptical mirror 6 5/s inches in
diameter, having a working distance of 4 inches from the* arc crater to
the vertex of the mirror and 19 inches from the mirror to the film
aperture. This gives a working speed to the optical system slightly
faster than //3, which is sufficient for the quarter-size lenses regularly
supplied with portable projectors.

FIG. 4. Portable arc lamp showing ammeter, arc imager, and
manual controls for mirror and carbon feeds.

The mirror is adjusted for horizontal and vertical alignment of the
spot at the film aperture by means of two knurled knobs projecting
from the back of the lamp house, as shown in Fig. 4.

This lamp is designed for use with either standard 35 mm. or with
16 mm. film. The arc current required will vary with the width of
the film used and the size of the projected image.

The diameter of the carbons is determined by the arc current. The
following table indicates the correct carbon trims for different condi-
tions of operation.

758

HARRY H. STRONG

[J. S. M. P. E.

Positive Carbon

Negative Carbon

Arc Current

Film Width

7 mm.

5 mm.

9 amp.

16 mm.

9 "

6.4 "

13 "

35 "

10 "

7 "

15-16 "

35 "

Under the operating conditions indicated in this table, the lamp will
produce a screen brilliancy comparable with that produced by standard
theater equipment at equal current. Ample illumination is provided
for the projection on perforated screens of an 8- to 12-foot picture
from standard 35 mm. film. On a solid screen the intensity of
illumination is ample for a picture 14 feet or more wide.

FIG. 5. -interior view of portable arc lamp.

To avoid the possibility of trimming improperly, the lamp has been
designed to accommodate carbons 4 inches long, when clamped at
their extreme ends. This arrangement eliminates any necessity of
adjusting the carbons in the holders during the burning period of
slightly more than an hour.

In trimming the lamp, the spring clutch is released, the carbon is
placed against a stop at the back of the holder, and the clutch is then
allowed to engage the carbon. By confining the carbon length to 4
inches, a perfect alignment of positive and negative carbons is assured
without any necessity of adjustment.

June, 1932]

A REFLECTOR ARC LAMP

759

The interior mechanism of the lamp is shown in Fig. 5. It should
be noted that the parts are few in number, as well as simple and sturdy
in construction.

To avoid the necessity of purchasing carbons of special length, and
so as to permit the use of standard 8-inch carbons, a cutter is provided
at the rear of the lamp house. By using this cutter, standard 8-inch
carbons may be scored and broken to exactly 4 inches in length.

Should an experienced operator, who understands the readjust-
ment of carbons in the holders, desire to use the full 8-inch carbons,
the stops can easily be removed from the back of the carbon holders,

FIG. 6. Exterior view of full-wave
rectifier for use with portable arc lamp.

and a tubular guard can be attached to the rear of the lamp. This
arrangement will prevent contact of the end of the negative carbon
which extends through the mirror.

An opening is provided in the bottom of the lamp house for connec-
tion to the ventilating system provided with some makes of portable
sound equipment. The exhaust from the lamp house is carried out
through a chimney at the top.

An ammeter is mounted on the rear of the lamp house (Fig. 4).
This is surrounded by a ventilating duct that connects to an annular

760 HARRY H. STRONG

passage around the main ventilating flue. In this manner an induced
draft of cool air is drawn in around the ammeter, to maintain the
instrument at normal temperature and prevent any disturbance of its
calibration.

The rectifier has been chosen as a means of converting alternating
current to direct current because of its light weight for the required
capacity, its freedom from moving parts and intricate mechanism, and
its simplicity and safety of operation.

The rectifier, herein described and shown in Fig. 6, was developed
particularly for use with the lamp described above. Its elements
comprise a special transformer for changing the alternating line
voltage to the correct potential for operation of an arc; a radial switch
for regulating the current to the desired value ; two Tungar tubes for
rectifying the current; a substantial housing; and necessary sockets
and lead wires.

The transformer is of a special design in which the output pos-
sesses constant current characteristics, thus allowing commercial
fluctuations of line voltage without affecting the stability of the arc.
The primary and secondary coils are separate and are effectively
insulated from each other, which construction allows only the low
voltage necessary for operating the arc to enter the lamp house, thus
avoiding any possibility of the operator's sustaining an electric shock.

The rectifier tubes are of the familiar Tungar type, i. e., thermionic
tubes filled with argon at low pressure. These tubes provide a valve
action, permitting the alternating current to pass in one direction only.
Connection is made in such a way that full-wave rectification is
secured without the use of moving parts, relays, or other intricate
devices.

A radial switch, placed within convenient reach, gives eight points
of current adjustment. This permits the arc current to be adjusted to
values ranging from 8 to 16 amperes when the rectifier is connected to
a 115-volt supply.

The electrical efficiency of the rectifier unit is 80 per cent. From
this fact, it is evident that the lamp may be operated at an arc current
of 15 amperes with a line consumption of only 1000 watts.

VACUUM TUBE AND PHOTOELECTRIC TUBE DEVELOP-
MENTS FOR SOUND PICTURE SYSTEMS*

M. J. KELLY**

Summary. — This paper reviews some recent vacuum tube and photoelectric cell
developments which are of interest in sound recording and reproduction systems.
An indirectly heated cathode triode is described, in the output circuit of which the
current components due to the a-c. power supply of the heater have been reduced
approximately 20 decibels below previously obtained levels. This tube makes it
possible to use an a-c. supply in amplifiers having flat frequency characteristics
with over-all gains of the order of 100 decibels. The microphonic disturbances in
vacuum tubes are discussed. A measuring system for evaluating the microphonic
noise currents is described, and the characteristics of a filamentary cathode tube of
low microphonic noise level are given. The characteristics of a double anode, thermi-
onic, gas- filled, rectifier tube for use in a d-c. power supply unit for the sound lamp
and vacuum tube filaments of reproducing systems are given. A photoelectric cell
of high sensitivity for use in sound reproduction work is described.

Since the standardization of sound recording and reproducing
systems many technical developments have been made resulting in
less distortion in reproduction, in a decrease of extraneous (back-
ground) noise, and in systems having improved operation and main-
tenance characteristics. The thermionic vacuum tubes used in the
recording and reproducing systems and the photoelectric cell used
in reproduction are important elements in determining the quality of
reproduction, the level of background noise, and the operation and
maintenance characteristics of the systems. A review of the vacuum
tube and photoelectric cell developments at the Bell Telephone
Laboratories during the past year that have contributed toward such
improvements in sound recording and reproducing systems will be
given, and the characteristics of the new devices described.

1. A LOW "HUM LEVEL" AMPLIFYING TUBE

The advantages of obtaining the filament supply of vacuum tubes
directly from a-c. lighting circuits have long been recognized. Prior
to the time when the indirectly heated cathode vacuum tube became

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Bell Telephone Laboratories, Inc., New York, N. Y.

761

762 M. J. KELLY [J. S. M. P. E.

available, much consideration was given to the use of "raw "alternat-
ing current for heating the filamentary type of cathode. Due to the
magnitude of the 60 and 120 cycle disturbances introduced into the
plate current even under the most favorable cathode design conditions,
it has not been possible to make at all general the application of
alternating current to the heating of filamentary cathodes. In
audio frequency amplifiers the 60 and 120 cycle disturbances generally
restrict the use of alternating current for heating the filamentary
cathodes to the final stage of amplification.

The introduction of the indirectly heated cathode into the vacuum
triode made immediately possible a further extension of the use of
"raw" alternating current as the source of the cathode energy. A
heater and cathode unit consisting of a hairpin of tungsten wire im-
bedded in a cylindrical insulator of magnesia or similar material, with
a tightly fitting nickel sleeve surrounding the insulator upon which
is deposited the active cathode material, has been generally stand-
ardized in triodes for broadcast radio receiver use. The heater
element of such a tube can be lighted by alternating current without
introducing 60 and 120 cycle disturbances in radio frequency stages.
However, its use in audio frequency circuits having flat frequency
characteristics above 60 cycles is, in general, limited to circuits having
gains less than 50 decibels ahead of the first tube heated by alternating
current. If such a tube, employing alternating current for the heater
supply, is used in amplifiers having flat frequency characteristics and
appreciably greater gain, the 60 and 120 cycle disturbances from the
heater supply are too great to be tolerated.

The amplifying units of sound recording and reproducing systems
have over-all gains of the order of 100 decibels. In systems having
such great gain, it is possible to use alternating current for heating
the filaments of all the tubes only by suffering a reduced response at
frequencies lower than 120 cycles, or by tolerating in the output a high
level of extraneous noise arising from the 60 and 120 cycle disturbances
in the tubes of the preceding stages.

The advantages of using alternating current for the filament supply
in high-quality amplifiers used for sound reproduction as well as for
public address systems, radio broadcast pick-up systems, and other
high-gain audio frequency amplifiers, made desirable the study of 60
and 120 cycle disturbance levels in the plate circuits of the indirectly
heated cathode tubes and an investigation of means of making these
disturbances sufficiently small to permit the use of alternating current

June, 1932]

VACUUM AND PHOTOELECTRIC TUBES

763

for heating the cathodes of all the tubes in such systems. In order
that alternating current might be generally used for such purposes, the
disturbances in the plate circuit of the first tube should not be greater,
in order of magnitude, than the resistance and thermionic emission
noises. Alternating current could be then used for heating all the
cathodes in any amplifier whose gain was not limited by these
fundamental causes of noise.

As the first step in these studies, a measuring system was developed
with which one could measure a 60 or 120 cycle current to 120 decibels
below a level of 1 milliampere. This system is shown schematically
in Fig. 1.

TUBE UNDER TEST
CIRCUIT

FIG. 1. Circuit used for measuring disturbance current.

The tube under test is placed in a single-stage amplifier circuit, and
its heater supply is so arranged that there is no 60 and 120 cycle
disturbance in the plate circuit except that produced by pick-up in the
tube itself. The output from this circuit passes through a variable
attenuation network to the input of a resistance coupled amplifier.
The output of the resistance coupled amplifier is fed into a harmonic
analyzer which permits the separation and measurement of the 60 and
120 cycle currents. In order to calibrate the analyzer, an oscillator
is provided whose 60 or 120 cycle output can be fed into the variable
attenuation network, amplifier, and analyzer.

Two variable 100 ohm resistances of the dial box type are connected
in series across the heater terminals of the tube under test. The

764

M. J. KELLY

[J. S. M. P. E.

equipotential cathode is connected to their common point. By keep-
ing the sum of the two resistances equal to 100 ohms, a potentiometer
is provided, by means of which the cathode may be maintained at any
potential varying from that of one end of the heater to that of the other.
It is then possible to determine the value of the disturbance currents
as a function of the potential of the cathode with respect to the heater.

>ECOND

HARMO

NIC,12C

CYCLE

O 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

RATIO 3/b

FIG. 2. Disturbance currents in typical indirectly heated cathode triodes.

Measurements of the disturbance currents in the plate circuits of
typical standard indirectly heated cathode triodes were made with
the measuring equipment described above. The curves of Fig. 2
exhibit representative results. The levels of the disturbance currents
in the output circuit are shown as functions of the position of the

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 765

common point of the heater and cathode. For a/b equal to 0, the
common point is at one end of the heater; for a/b equal to 1, it is at
the opposite end of the heater. The measurements were made at the
rated plate, grid, and heater voltages of the tube, with an output
impedance equal to that of the tube and with an input resistance from
100 to 200 ohms.

The disturbance currents resulting from the a-c. supply used for
the heaters of indirectly heated cathode triodes are introduced into
the output circuits by the :

(1) Electric field of the heater,

(2) Magnetic field of the heater,

(3) Resistance between heater and grid and between heater and plate, and
capacitance between heater and grid and between heater and plate.

The electric field of the heater element in the space between the
cathode and anode will afTect the electron current to the plate in
precisely the same manner as does the electric field of the control
grid. With one point of the heater circuit connected to the cathode,
the electric field of the heater at each point in the cathode-anode space
will be the sum of the fields due to each segment of the heater element.
As the common point of the heater and cathode is shifted along the
heater wire, the value of the field will change. It would be expected
that, when the common point was located at the mid-point of symmetry
of the heater circuit, the electric field would have its minimum value.

The results given in Fig. 2 confirm this expectation for the funda-
mental disturbance current. A definite minimum is shown in most
cases. However, since in the tubes under test the 60 cycle distur-
bance current is due to factors in addition to the electric field, the
expected characteristic variation of the disturbance currrent due to
shift of the common point is masked in varying degrees in the different
tubes.

The second harmonic disturbance current does not vary with the
position of the common point for the tubes of Fig. 2. From this
it might be assumed that there was no second harmonic component
due to the electric field. In general, this is not the case. The
grid action of the heater circuit varies non-linearly as the effective
voltage of the heater system changes with respect to the cathode.
This non-linearity of the grid action would be expected to produce
second harmonic components in precisely the same manner as they
are produced in the familiar case of "mu" modulation with the standard

766

M. J. KELLY

[J. S. M. P. E.

control grid. The second harmonic disturbance current due to the
electric field is, in general, very much smaller than that due to the
magnetic field, and is masked by it. This is the case with the data of
Fig. 2. Experiments have been arranged where the magnetic effects
were eliminated and the presence of second harmonic current due to
the electric field demonstrated.*

The magnetic field of the heater in the space between the anode and
the cathode will affect the electron current to the plate. The electrons

c- L

«— \Vv\- -

— AAA/V-

lllll

— l|l|l

(b) (c)

FIG. 3. Equivalent circuits of tube, used for analyzing disturbance currents:
a, complete circuit; b, input circuit; c, output circuit.

will be deflected by the magnetic field according to force relations of
the magnetic field. The deflection of electrons by this field causes a
double frequency change in the electron space charge which results
in a second harmonic component of the disturbance current in the
anode circuit. Due to asymmetries in the space charge system, the
two changes in space charge per cycle of the heater current are not
equal. The inequality in the two changes will produce a disturbance
current in the plate circuit of the same frequency as that of the heater
current.

* To be described in a forthcoming paper by J. O. McNally on "Disturbance
Currents in the Output Circuit of an Indirectly Heated Cathode Triode," Proc.
I. R. E.

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 707

Heater circuit voltages are introduced into the grid circuit and into
the plate circuit through resistance and capacitance between the
heater and each of these elements. The circuit diagram shown in
Fig. 3 indicates the paths. For simplicity, one side of the heater is
shown connected to the cathode, and the resistance and capacitance
from the heater to the other elements are connected to the opposite
side of the heater.

Experimental tubes of special construction were made in order to
evaluate the contribution to the disturbance current of the factors
described above. The various means of decreasing the disturbance
currents were considered, and experimental models were made in order
to check the relative effectiveness of the different means. From these
data the best tube, from a manufacturing view-point, that would give a
sufficiently low level of disturbance current, was designed.

In order to decrease the electric field effect, the heater circuit was
electrically shielded. The cathode itself acts as a shield over a portion
of the heater circuit. In order to make the shielding more complete, the
upper end of the cathode sheath was completely enclosed and the sheath
was lengthened so as to extend well below the lower ends of the plate
and grid. A drawn metal thimble was then placed around the heater
leads below the end of the cathode, extending to the stem press.

In order to reduce the disturbance currents due to the magnetic
field, the magnetic field of the heater in the space between the cathode
and the anode must be made as small as possible. There are several
ways in which this may be done. The heater circuit may be com-
pletely enclosed by a sheath of material that will act as a magnetic
shield; the heater unit may be so designed that the field outside its
surface is substantially zero; or the heater can be made of high
resistance so that the heater current is small and its voltage drop is
large.

It was found that the most practicable solution lay in combining
the last two methods. The heater current was adjusted to 0.32
ampere and the voltage drop to 10 volts. The heater is a closely
wound spiral of tungsten wire, mounted in the form of a hairpin in a
twin bore magnesia insulator. The geometry of the hairpin is such
that the magnetic field in the cathode-anode space is as small as can
be realized in a commercial mounting; and the reduction of the heater
current to 0.32 ampere, which is approximately one-fifth the value
normally used, gives an adequate reduction of disturbance current
from the magnetic effect. By reducing the heater current to this

768

M. J. KELLY

[J. S. M. P. E.

extent the potential drop across the heater is increased from four to
five times the value normally used. This increase in voltage increases
the electrostatic effect of the heater, increasing the 60 cycle distur-
bance current. However, it is possible in a commercial structure to
shield the structure sufficiently so that even with the increased potential
drop the contribution to the disturbance current by the electric
field is adequately small.

In order to decrease the disturbance currents due to resistance
leakage, the tube elements are held together at the two ends by means

FIG. 4. Type 262A mounting
and completed tube.

of a specially designed insulator. The insulator is so designed that
there is not a continuous path between any two of the tube elements
on the side of the insulator facing the tube elements. This makes
impossible the formation of leakage paths of metal vaporized in the
pumping process or of active material vaporized from the cathode.
With the tube elements at operating temperature, the leakage between
the tube elements, or between the heater and the tube elements, is
maintained at a value greater than 100,000 megohms throughout the
life of the tube.

For general services the normal values of capacity between heater

June, 1932]

VACUUM AND PHOTOELECTRIC TUBES

769

— ru r\) GO Cu
0 o M
ooooofo ^ o> o>

10,000

0.325

0.3 00

0.275

QC<

- 0.250

0.225

6 8 10 12

HEATER VOLTS

.X^ >

Eb-1

80 157.

5 135 1

2.5 9(

)

-20 -18 -16

-4

-14 -12 -10 -8 -6

GRID VOLTAGE
FIG. 5. Electrical characteristics of the 262A tube.

-2

770

M. J. KELLY

[J. S. M. P. E.

and grid and between heater and plate, that are obtained with
standard mechanical designs, are of sufficiently low value that the
disturbance currents introduced through them are not important.
However, in such cases as when a tube works directly from the output
of a photoelectric cell, it is desirable that the heater-grid capacitance
be lower than that obtainable by standard design. In order to reduce
this capacitance to a sufficiently low value, the grid lead for the tube
has been brought out at the top of the tube through a cap of the type
used in screen grid tubes. No grid supports are placed in the stem
press of the tube and all the constructional details of the grid are in
keeping with the minimum capacity requirements.

FUNDAMENTAL, 60 CYCLES

NG DISTURBANCE CURRENT
THE VALUES INDICATED
BSCISSAS

M Jt • •

O O O 0 0

t-v

f=0

•/

*•»>.,

PER CENT OF TUBES HAV
OUTPUTS GREATER THA^
BY THE

o 8 o S S !

*=»/

0 90 100 110 120 I3<

SECOND HARMONIC, 120 CYCLES

4-1*1

FIG. 6.

90 100 110 120 130 80 90 100 110 120 130
DISTURBANCE CURRENT OUTPUTS IN DECIBELS BELOW 1.0 MILLIAMPERE

Distribution of disturbance output currents in a representative
number of 262 A tubes.

This tube has been standardized by the Western Electric Company
and coded 262 A. The completed tube and its mount are shown in
Fig. 4. Its electrical characteristics are given in Fig. 5. The tube is
normally used with a plate potential of 135 volts and a grid bias of
— 4.5 volts. Under these conditions the plate current is 3.0 milli-
amperes, the output impedance 15,000 ohms, and the voltage amplifi-
cation factor is 15. The tube is satisfactory for use with a plate
potential of 180 volts and a plate current of 10.0 milliamperes.

Distribution curves of disturbance currents in the output circuits
of typical tubes, taken under normal conditions of operation with an

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 771

input resistance of less than 1000 ohms, are shown in Fig. 6. These
data indicate that for a/b = 0.5, the level of fundamental disturbance
current for all tubes is less than 95 decibels below 1.0 milliampere, and
the level of the second harmonic disturbance current is less than 105
decibels below 1.0 milliampere.

The level of the output noise derived from sources other than the
a-c. supply used for heating is of interest. The noise level in represen-
tative tubes was measured in a voice frequency amplifier that had a
flat frequency characteristic. The heater was operated on direct
current. With an input resistance of less than 100 ohms, the noise
level of the output circuit varied between 118 and 127 decibels below
1.0 milliampere. This noise is principally due to the Schott effect
from the cathode. With 2 megohms in the grid circuit the noise level
in the output circuit is approximately 105 decibels below 1.0 milli-
ampere. This noise is almost entirely due to the resistance noise of the
grid circuit.

It is necessary to have the disturbance currents due to acoustic
pick-up or mechanical shock sufficiently low that they will not place a
limitation on the fields of application of the tube. The mechanical
structure of the tube has been determined with these requirements in
view. The tube has a sufficiently low response to acoustic or me-
chanical stimulus so that, when mounted in a suitably cushioned and
shielded socket, the disturbance currents from acoustic and me-
chanical sources will not be of greater magnitude than the resistance
noise and heater current disturbance noise.

2. AGITATION NOISE IN AMPLIFIER TUBES

When a vacuum tube in an amplifying circuit is subjected to me-
chanical agitation the resulting motion of the elements of the tube
relative to each other gives rise to small transient changes in the
electrical characteristics of the tube, which produce transient changes
in its plate current. The plate current changes are usually of the
form of complex damped oscillations, corresponding in their general
character to the damped vibration of the tube elements. When these
plate current changes are amplified and reproduced by a loud speaker
they produce the unpleasant, usually discordant, ringing sound
generally designated as microphonic noise.

There is another kind of disturbance whose existence has not been
generally recognized, which also arises from mechanical agitation.
It is often as much of a limiting factor in noiseless reproduction as is

772 M. J. KELLY [J. S. M. P. E.

the microphonic noise. It manifests itself in the loud speaker as an
irregular scratching or sputtering as contrasted with the more or less
sustained ringing sound of the microphonic noise. This sputtering is
caused less directly by the relative motion of the elements in the tube,
in that it depends on the making and breaking of electrical contacts be-
tween metallic parts, which are not otherwise electrically connected,
or by the discontinuous change in a relatively high resistance between
tube elements.

One of the most common causes of the sputtering noise in the
filamentary type of tube lies at the center point of a filament "V,"
which is ordinarily supported by a small hook attached to a spring
imbedded in an insulating support. If this hook is in contact with the
filament, its potential will, of course, be the same as the potential of
the filament at the point of contact. If, however, it is not in electrical
contact with the filament, it will assume some potential, depending
upon its degree of insulation from the other elements which, in
general, will not be the same as that of the contact point on the fila-
ment. If the contact between the filament and hook is alternately
made and broken, as easily happens when the filament is suspended
loosely on the hook and mechanical agitation occurs, the potential of
the hook changes discontinuously ; and, by a grid-like action, produces
corresponding discontinuous changes in the plate current. These
plate current changes when amplified, produce the disturbance
designated as the sputtering noise. This type of noise is also due to the
imperfect welding of the parts constituting the grid or plate structure.
It has also been traced to discontinuous changes in the resistance of
thin films of conducting material covering the insulating materials
between tube elements.

The level of agitation noise currents in the initial stages of amplify-
ing systems having over-all gains of the order of 100 decibels is suffi-
ciently great to produce an objectionable level of background noise in
recording and reproducing systems using such amplifiers.

As the first step toward decreasing these agitation noises, a measur-
ing system was developed in which the microphonic and sputtering
noise currents could be separated and quantitatively measured when
the tube under test was subjected to a reproducible agitation
stimulus.

Although the aural demonstration of microphonic and sputtering
noises requires nothing more than a high-gain amplifier and loud
speaker, the measurement of these quantities presents a number of

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 773

difficulties. Instead of applying an arbitrary thump with the finger
or pencil to the tube under test, it is necessary to provide a constant
and reliable agitating agent. If the test is to form a part of a factory
acceptance test, the agitation should be continuous, since ballistic
readings are slow and unsatisfactory. The agitation and mounting
of the tube should be such that a periodic excitation is applied; other-
wise, mechanical resonances may occur between certain tubes in the
testing apparatus. Such resonances give rise to misleading results,
since a tube with a resonant point at a predominant frequency of
excitation will respond much more strongly than other tubes, which
on the whole may have very similar microphonic characteristics.

The chief problem in measuring the sputtering noise is not the
agitation of the tube under test, although this is important, but
rather the separation of the sputtering noise from the microphonic
noise.

Sputtering is often more disagreeable to the ear than microphonic
noise, and although the intensity of the noise may sometimes be much
higher than the intensity of the microphonic noise, the total energy
of the sputtering noise over an interval of time is usually considerably
less than that of the microphonic noise for the same interval. This is
due to the discontinuous character of the sputtering noise. Since
microphonic noise is always present and varies in magnitude from
tube to tube and from one operating condition to another, it is practi-
cally impossible to measure the sputtering noise by taking differences
between measurements of total noise.

Advantage, therefore, has been taken of the fact that a dis-
continuous impulse may be resolved into a continuous spectrum of
frequencies. The frequency spectrum of the sputtering noise extends
even into the radio frequency band, and has given trouble in radio
frequency amplifiers. The microphonic noise spectrum, on the other
hand, lies largely in the audio frequency band, and no components of
microphonic noise of measurable intensity have been observed above
15,000 cycles per second. If a high-pass filter which cuts off below
15,000 cycles is included in an amplifier having a flat frequency
characteristic, the microphonic disturbance currents will be effectively
suppressed, while the components of the sputtering noise above 15,000
cycles are transmitted with only slight attenuation. The sputtering
noise currents of frequencies greater than 15,000 cycles may then be
measured by ordinary means. If it be assumed that the distribution
of energy over the entire spectrum is the same for all sputtering

M. J. KELLY

[J. S. M. P. E.

noises, then such measurements may be taken as an indication of
sputtering noise intensities. While this assumption is certainly not
exactly true, it has been found to be approximately so, and the
measurements of the components of the sputtering noise at frequencies
greater than 15,000 cycles has proved of much value in conducting
investigations of tube noises.

On the basis of these considerations, a measuring system was
developed which comprised four essential parts: a tube mounting and
agitating system, a flat-frequency amplifier, a high-pass filter, and an
indicator. These units are arranged as shown in the schematic
diagram of Fig. 7.

The agitator consists of a thick slate base on which is rigidly
mounted at one end an uncushioned tube socket, and at the other end

JICROPMONIC

FIG. 7. Circuit for measuring agitation noise.

a vibrating hammer which directs horizontal blows against a steel
block firmly mounted on the base near its center of gravity. The
hammer consists of a good vibrating electric bell, the gong of which
has been replaced by the steel block just mentioned and whose
clapper has been weighted by the steel hammer. This hammer
strikes eight times per second ; and because of the rigidity of the base
mounting, for all practical purposes, it causes shock excitation of the
tube under test. The frequency of the blows is so low that there is
little likelihood of encountering resonance in tubes under test.

The tube under test is mounted in the socket of the agitator, and is
operated under its standard plate, grid, and filament voltages, with a
resistance of 60,000 ohms in the plate circuit. By means of taps

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 775

brought out from this resistance the input to a coupling tube is
controlled. A high-pass filter with a cut-off at 15,000 cycles follows
the coupling tube. The filter is followed by a two-stage, 50 decibel
amplifier, a gain control unit, and, finally, a two-stage, 70 decibel
amplifier. The output of this amplifier is measured by means of a
suitable thermocouple galvanometer. The amplifying system has
substantially a flat frequency characteristic over a range from 50 to
30,000 cycles. It is down 3 decibels at 30,000 cycles and 2 decibels at
100 cycles, and is calibrated by means of a 20,000 cycle oscillator.

When the total agitation noise of the tube is measured, the 50
decibel amplifier, the high-pass filter, and the coupling tube are not
included in the circuit, the output of the tube under test working
directly into the 70 decibel amplifier. For separately examining the
discontinuous noise, which has been designated as sputtering noise,
the 50 decibel amplifier, the high-pass filter, and the coupling tube are
inserted. A measurement is then made of the components of the
sputtering noise having frequencies greater than 15,000 cycles. The
variable potential drop produced across a fixed resistance of 1000
ohms in the output circuit of the tube by the standard agitator is
taken as a measure of the microphonic noise level of the tube. This
potential drop is expressed in terms of decibels below 1.0 volt.

This measuring system has been of great value in studying the
agitation noise levels during the development of sufficiently quiet
tubes for high-gain amplifiers. It has also been of service in making
comparisons of the cushioning action of different types of tube sockets
and mountings. For this purpose the agitation noise characteristics
of a group of tubes of a given type are determined with the rigid
mounting described above. The noise characteristics of the same
tubes are again determined with the tube mounted in the socket under
examination, or, if cushioning material is under investigation, with
the cushioning material inserted between the agitator slate base and
the socket. A comparison of the two sets of readings gives a measure
of the effectiveness of the cushioning material.

A set of this type has been found satisfactory for use in acceptance
tests for agitation noise in vacuum tubes. The manufacturing depart-
ment's sets are kept in calibration with respect to a master set in the
laboratory by means of a group of reference tubes.

The detailed mechanical design of the low hum level tube, the
262A, described in Section 1, has been based on agitation noise level
studies made in this system. Both its microphonic and sputtering

776 M. J. KELLY [J. S. M. P. E.

noise levels are sufficiently low that with standard cushioning and
shielding, the agitation noise currents in its plate circuit, when used as
the first tube of a 100 decibel amplifier, will be no greater than that of
the emission noise currents.

There are many applications in which a cathode that consumes
less energy than that of the indirectly heated cathode of the 262A tube
is desirable. A tube having such a filamentary cathode has been
developed for those services demanding microphonic noise levels
much lower than those of tubes previously standardized for such
systems.

In Western Electric systems the 239A tube has been used in the past
for preliminary stages of high-gain amplifiers. The new filamentary
cathode tube, which has been coded 264A, has been made identical
with the 239A in mechanical dimensions and in electrical character-
istics, except for a slight change in the filament characteristics. With
a plate potential of 100 volts and a grid potential of —8 volts, its
average output impedance is 12,500 ohms. The average amplifica-
tion factor is 7, and the average plate current is 2.0 milliamperes.
The filament current is 0.30 ampere and the nominal filament potential
drop is 1.5 volts.

The microphonic noise level of a 239A tube measured in the equip-
ment described above has an average level of 28 decibels below 1.0
volt. The corresponding value for the 264A tube is 45 decibels
below 1.0 volt, while it is 50 decibels below 1.0 volt for the 262 A tube.
It is again pointed out that these measurements were made in un-
cushioned sockets, and with direct transmission of mechanical
disturbance from a relatively high level source. Significance should,
therefore, be attached only to the relative values of noise levels and
not to absolute values. The sputtering noise level of the 239A tube
under the same conditions of measurement has an average value of
80 decibels below 1.0 volt, while the 264A tube has an average value
of 95 decibels below 1.0 volt, the corresponding value of the emission
noises of both tubes being approximately 95 decibels below 1 .0 volt.

The improvements in microphonic noise level have followed an
analysis of the resonances occurring among the various elements of
the tube structure, and have been made by designing structures that
avoid such resonances. It was found that a rigid structure built as
close to the stem press as possible exhibited very little tendency to
resonate. A structure in which the three elements were bound to-
gether as rigidly as possible and mounted as close to the glass stem as

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 777

practicable was, therefore, adopted. For reasons of interchange-
ability, it was necessary to limit the size of the parts to dimensions
suitable to the over-all dimensions of the 239A tube, which it was
replacing, and also to permit the use of the small push type base.

With these limitations it was possible to obtain a more rigid and
lower mounted structure by using a special means of constructing the
glass stem press. Due to the dimensional limitations, it would not
have been possible to use glass tubing for the stem press of greater
average inside diameter than 0.53 inch. With such tubing, and with
standard methods of stem construction, the maximum distance be-
tween the two plate supports would be approximately 0.53 inch.
With new means of stem construction it was possible to use the same
size of tubing and to make a stem press in which the distance be-
tween the plate supports was 0.64 inch. In order to obtain this plate
support spacing with standard methods of stem construction, it
would have been necessary to have used a stem tubing whose mean
inside diameter was 0.66 inch. It would not have been possible to
have sealed a stem made from such tubing into a bulb that could have
been used with the small push type base.

Fig. 8 shows, in solid lines, an outline of the stem press for the
tube using the 0.53 inch tubing and, in dotted lines, the stem press
that could be made with this size of tubing with standard stem con-
struction practice.

It is evident that the base of the mounting has been increased from
0.53 to 0.64 inch, adding considerably to the rigidity of the structure
and making possible the use of straight plate support wires. By using
the straight plate support wires, the assembly can be mounted closer
to the stem press with greater facility than when the plate support
wires are bent outward. The increased distance between the plate
supports makes possible a greater separation between the leads and
permits the insertion of adequate shields above the stem press to
maintain insulation paths free from thin films of vaporized material.
The thin films of vaporized material on the glass produce variable
resistances which contribute to the sputtering noise as described
above.

The filament and its mounting contribute materially to the agita-
tion noise. If the filament is placed under considerable tension so
that the contact between the filament hook and the filament is
maintained at all times, the production of sputtering noise at this
point is eliminated. However, the degree of tension to which the

778

M. J. KELLY

[J. S. M. P. E.

filament is subjected materially affects the level of the microphonic
noise deriving from the filament unit. In general, the higher the
tension, the greater the microphonic noise level. It is, therefore,
necessary to balance the two requirements. With zero tension a
considerable number of the tubes will give evidence of sputtering
noise originating at the filament hook, whereas a minimum of micro-
phonic noise will result from the filament unit. As the tension is
applied and is gradually increased, the microphonic noise deriving
from the filament unit will also increase. A spring and hook unit
has, therefore, been adopted which^will place the filament under a

STEM OUTLINE WITH SPECIAL CONSTRUCTION

STEM OUTLINE WITH STANDARD CONSTRUCTION

FIG. 8.

Vacuum tube stem of
264A tube.

FIG. 9. Rectifier tube,
type 263 A.

tension of a few tenths of a gram, and will have adequate displace-
ment to keep the hook and filament in contact. In this way both the
microphonic and sputtering noise of the^filament unit are kept at the
lowest practicable level.

All welds in the structure which are a part of the electrical circuit
are made with special care to assure the elimination of variable
resistances. As described above, the stem press is shielded against
the deposition of material from the filament or from the metallic parts
during pumping. This shield prevents throughout life the formation

June, 1932]

VACUUM AND PHOTOELECTRIC TUBES

779

on the stem press of thin film resistances that are variable and produce
the sputtering noise. The glass insulator tying the parts together at
the top of the structure is located at the back of the plate so that no
material from the filament can be deposited across insulating paths
during the life of the tube.

3. A DOUBLE ANODE THERMIONIC GAS-FILLED RECTIFYING TUBE

By operating the vacuum tubes of sound reproducing amplifiers on
alternating current, only a portion of the storage battery equipment
that is necessary in the projection of sound films is eliminated. The
sound lamp is heated by direct current, and because of its high current
rating requires considerable storage battery capacity.

A rectifier tube, which has been coded "Western Electric 263 A,"
has been developed to supply the direct current for the sound lamp
and thus completely to eliminate the storage batteries from reproduc-
ing equipment. It is shown in Fig. 9. The rectifier tube has two
anodes and a filamentary cathode of the oxide-coated type. The
filamentary cathode is mounted between the two anodes, with its
housing so arranged that the necessary peak potential can be obtained
between anodes without voltage breakdown. The tube is filled with
argon at a sufficiently high pressure to give the minimum anode-
cathode potential drop during the conducting half of the cycle. The
characteristics of the rectifier tube are as follows:

Filament potential 2.50 volts

Filament current 16 amperes

Anode-cathode potential 5 to 10 volts

Maximum value of peak space current 6 amperes

Maximum peak potential between anodes 100 volts

With a suitable filter system this tube will supply a direct current of
4.0 amperes. The available d-c. potential will, of course, depend
upon the voltage drop in the filter system. With a filter system
designed to give direct current with a ripple small enough to permit
the output to be used for heating the filaments of vacuum tubes, a
voltage of 15 to 20 volts should be available when the peak voltage
between the anodes does not exceed 100 volts.

4; A PHOTOELECTRIC CELL OF HIGH SENSITIVITY

The photoelectric cells initially used in sound picture reproducing
systems were filled with gas and were of the potassium hydride
cathode type. They were operated with an anode potential of 90

780

M. J. KELLY

[J.S.M.P.E.

volts. The average cathode sensitivity of the potassium hydride
surface was 1.0 microampere per lumen for a light source having a color
temperature of 2710°K. The gas was maintained at such a pressure
that a gas amplification of approximately 4 was obtained at 90 volts,
giving an output current of 4 microamperes per lumen for the light
source described above.

It was recognized from the beginning that a photoelectric cell of
higher sensitivity would be of material assistance in reducing back-
ground noise. Any increase in the level of the photoelectric cell out-

10. Photoelectric cell, type 3A.

put would proportionately decrease the necessary amplification by
vacuum tubes and thus increase the ratio of the signal to the noise
deriving from the amplifier tubes in the output.

The potassium hydride photoelectric cell, in addition to having a
lower sensitivity than was desirable, had other inherent properties
that were not ideal for a commercial device. A potassium hydride
surface, even when made under the most favorable conditions,
exhibits on standing a gradual decay in surface activity. This decay
is due to the covering of the hydride surface by a film of potassium.
This coverage is accelerated by an increase in temperature. In

June, 1932] VACUUM AND PHOTOELECTRIC TUBES 781

commercial parlance, this type of cell has a shelf life which is a func-
tion of the temperature.

Searches were instituted for cathode surfaces of greater sensitivity
and, if possible, free from shelf life characteristics. The work of
Langmuir and Kingdon and of Becker on the effect of oxygen films in
lowering the thermionic work function of thin films of caesium on
tungsten gave a valuable indication of the most fruitful direction of
investigation. The use of oxygen as well as of sulfur in connection
with thin films of sodium and caesium was found to be effective in
lowering the electronic work function of cathode surfaces.

Manufacturing considerations, such as the cost and the control of
quality, led to the standardization of a thin film surface of the caesium
type rather than of other surfaces of substantially the same sensitivity.
The cathode of the cell so standardized is a silver sheet upon which
there is formed during exhaustion, by irreversible chemical processes, a
matrix of caesium oxide, silver oxide, and finely divided silver. Upon
this matrix there is placed by reversible processes a thin film (of
atomic thickness) of caesium.

This cathode surface has a long-wave limit beyond 12,000 A and a
maximum of sensitivity at about 8000 A. This spectral sensitivity
makes the surface unusually suitable for use with light from a tungsten
filament. With a light source having a color temperature of 2710 °K.,
the stabilized sensitivity of this surface is approximately 35 micro-
amperes per lumen. The surface exhibits no shelf life characteristics,
even at temperatures of the order of 65 °C., which is well above normal
storing and operating temperatures.

The cell employing this cathode was filled with argon at a pressure
suitable for operation at 90 volts' plate potential with the light flux
normally employed. At this pressure the gas amplification factor at
90 volts is approximately 3.

Late in 1929, this cell was placed in service trials in equipment
previously using the potassium hydride cell. The trials indicated a
considerable improvement over the hydride cell in all operating
characteristics. It was coded 3 A in the Western Electric series, and
has been made available as a replacement of the hydride cells in
existing equipment. It is shown in Fig. 10. While its substitution for
the hydride cell in existing equipment is fully justified by the im-
proved service and the lowering of operating costs, the full advantage
of its improved characteristics is realized only in equipment designed
to take full advantage of them.

PROCESS PHOTOGRAPHY

GORDON A. CHAMBERS**

Summary. — The several methods used in process photography are briefly de-
scribed, in the beginning, from the historical point of view. The various methods
of applying these processes, and the technics involved in applying them, are discussed.

Marked advances have been made in recent years in the technic of
process photography, particularly in the field of application of the
so-called traveling matte. Many of these advances have come
about as a result of the limitations placed upon the cinematographer
by the addition of sound recording to his medium. It is the purpose
of this paper to bring to the attention of those who are unfamiliar
with process work a general survey of the methods in use, rather than
to attempt to give an intimate description of those methods in the
form of a text.

Process photography has as its objective two kinds of effects, those
that are recognized by the audience as a deception or illusion, and those
which, unrecognized as "trick" shots, are inserted in a picture to lend
production value. These latter effects are obtained by means of
process methods because mechanical or economic reasons make it
impossible or impractical to secure them by ordinary photography.

While the results obtained by high-speed and trick crank photog-
raphy fall properly in the field of special effects, it is intended to
limit this discussion only to those forms of process work that employ
mattes. Two forms of the latter are used, known as still and traveling
mattes.

The earliest and certainly the simplest of still mattes is the so-called
"split" matte, used immediately in front of the focal plane in the
camera to facilitate the making of multiple exposures. Peters1
mentions a ninety-foot production available in 1902 entitled The
Inexhaustible Cab in which ". . .thirty-two persons enter the carriage
built to hold but four, but none are seen to get out." This kind of
effect was extensively employed in the earlier days of photography,

* Presented before the Pacific Coast Section, S. M. P. E.
** West Coast Division, Eastman Kodak Co., Hollywood, Calif.

782

PROCESS PHOTOGRAPHY

783

especially to obtain ludicrous or comical effects such as that mentioned
above.

At the present time the form of still matte commonly referred to as
a "glass" is very often employed. This form of photography is
mentioned by C. L. Gregory2 and by A. B. Hitchins3 who illustrate
the method of use. This process is very versatile in its ability to
introduce into the picture detail that did not exist in the actual set
just so long as the action on the set does not overlap that portion of
the frame to be exposed to the "glass." While glass is often used as
the medium on which the painting is made, many of the paintings so
introduced into a motion picture are made on an opaque support.
This is usually done when the film is subsequently doubly exposed to
the painting on that portion of the frame that was suitably masked in
the camera. The latter method is finding more general use than is
the method of painting on glass and shooting through this glass on the
set. The painting can be prepared after the set action is photographed
without tying up the space or set fojr the length of time necessary to
prepare a painting on glass. One expert worker in this field has found
that the use of chalk in various shades from black to white makes it
easier to prepare the background, as the final matching of tones at the
point where the two exposures join on the frame is more easily
accomplished with chalk than with oil colors. Also, the quality of
the image obtained is different from that obtained from a brushed-on
oil paint, and it is felt by this particular worker that the resulting
composite is more uniform in quality.

While a glass is in every sense a miniature, even though confined to
two dimensions, three-dimensional miniatures in the form of models
are often employed to great advantage. With these also, it is possible
to employ either single or multiple exposure to obtain the composite
negative. For some effects it is desirable to employ both "glasses"
and models. Some very realistic effects have been obtained by using
model airplanes and dirigibles moving before the camera with one or
more glasses interposed, on which the clouds are painted. The result-
ing shot of a dirigible moving through beautiful cloud banks would be
difficult to obtain by natural means without expending a great deal of
time and money.

In using miniatures or models, it is extremely important that not
only the perspective of space be maintained but also what might be
termed the "perspective of time." Attention has been called to the
importance of this by several writers. Both J. A. Ball4 and G. F.

784 GORDON A. CHAMBERS [J. S. M. p. E.

Hutchins5 have discussed this subject from the theoretical standpoint,
and have stressed the relation that must exist between time and linear
magnification for a perfect illusion. Several interesting applications
of the use of models for creating special effects have been described
by F. Waller.6

With increased development in the application of process photog-
raphy has come the creation in the various studios of a "special
effects department," which devotes its time entirely to the creation of
these illusions.

One of the most useful tools of the specialists who comprise these
departments is the optical printer. Such an instrument has been
described by C. L. Gregory.7 Essentially, it consists of a camera
so mounted as to be able to copy one or more negatives moving in
synchronism with the raw stock in the copying camera. One of the
well-known results to be obtained with the optical printer is the
kaleidoscopic effect used to convey to the audience a train of thought
in a character's mind. Such an example is cited in the paper by Waller,
mentioned above. The optical printer is also used extensively for
the routine production of duplicate negatives from master positives,
because of the fact that the duplicate image is an optical one rather
than one secured from a contact printer. Greater sharpness is thus
obtained. The motor-driven cameras of today can not be used
easily for making fades and lap dissolves, and it is common practice
to produce these on an optical printer. Mere fades are often made
chemically, however, because of the ease and rapidity of this process.

Various applications have been found for prisms and other re-
flecting surfaces in the production of effects. These may be used on
the camera itself, or subsequently in making a dupe in the optical
printer. Sequences showing a ballet have been made to appear as
though the dance were performed on a glass floor, whereas the re-
flected image was obtained by using a prism or sheet of optical glass
in front of the taking lens. In the optical printer, several images
from as many different negatives may be superimposed in a single
composite by optical means, involving prisms for reflecting the
respective images into the same plane.

It must not be imagined that these processes are used only singly,
as often two or more are used to obtain the desired effect. The major
problems encountered in this work are those of obtaining accuracy of
registration of the images, and equal accuracy of timing of the action
of multiple exposures, so that the events take place in their proper

Tune, 1932]

PROCESS PHOTOGRAPHY

785

sequence without overlapping. Fig. 1 in the paper by Waller6 is a
chart illustrating the intricate attention to detail requisite to the
proper timing of the events in such a sequence of multiple exposures.

The developing of traveling matte processes into their present
state of perfection has opened a large field for the producer. In the
generic term "traveling matte" is included that form of image-carry-
ing film used before an unexposed negative in the camera. This image
is usually of a dye that has replaced a silver image. This film
has been called a "transparency" and is also sometimes referred to as a
"key." It might be mentioned at this time that a complete picture,
The Subway Express, made by Columbia was produced entirely by
such a transparency process. Two traveling matte processes are
available at the present time. One of these, known as the Williams
Process, has been described by its inventor, F. Williams.8 In this
process, the action is photographed against a black background.
From the negative of this, a duplicate negative is made which is
intensified in order to produce a silhouette of the action. This
silhoutte is then used as a traveling matte in a projection printer
during the printing of the background negative, as it covers and leaves
unexposed the space that is later to be used by the foreground action,
which is doubly printed from the original negative with a print from
this negative as a matte. Several variations of this process are
possible; either black or white backgrounds, or colored ones in
connection with filters, may be used to obtain contrast. Inasmuch
as it has been found that a spreading of the image occurs on the
silhouette because of the full exposure and the subsequent intensifica-
tion, and further, because of the increase in graininess incidental to
the multiplicity of duplicating processes employed, Mr. Williams has
been engaged in a method of simplifying his process to overcome these
difficulties. No information is publicly available at the present time
as to the solution of these in the new process.

Another traveling matte process that has enjoyed a great deal of
favor is the one commonly known as the Dunning Process. This
process is the result of work by C. Dunning and R. Pomeroy. The
original Dunning Process has been described previously.9 The
procedure has been modified somewhat so that at the present time
the methods used by the Dunning Process Company and the special
effects department of the Paramount Publix organization are essen-
tially identical.

From the background negative is made a positive transparency

786 GORDON A. CHAMBERS [J. S. M. P. E.

from which the silver is bleached, a yellow-orange dye being substi-
tuted for the silver. This transparency is run through the camera
together with the panchromatic film on which the foreground action is
being photographed against a blue backdrop. The foreground action
is illuminated by tungsten lamps screened by filters having a spectral
transmission comparable with that of the dye in the transparency.
The blue backdrop is illuminated by white light. This drop is
painted with a special blue complementary to the yellow-orange
dye.

Suppose for the moment that no foreground is present. The
running of the transparency through the camera in contact with the
emulsion of the panchromatic film would result in an exposure due
solely to the blue light reflected by the backdrop. This exposure
would be so regulated that the selective transmission of the dye in its
various depths would result in exposing on the panchromatic film a
duplicate negative of the transparency.

Suppose now that the entire field of view of the camera is occu-
pied by action. This portion would be photographed through the
transparency as though the latter did not exist by virtue of the color
of the light in the foreground. The composite obtained is a combina-
tion of the two exposures, the background being a duplicate negative
and the foreground an original exposed directly through the
transparency.

Just as great care is taken in making shots involving the use of still
mattes in order to have the shadows match, as these indicate the direc-
tion of lighting, so care is taken to simulate the lighting of the original
background in the lighting of the foreground during the making of a
transparency shot. The timing, where the background is moving,
such as in the case of a traveling shot showing a car moving along a
road, is another item that requires great care. The perspectives of
the background and foreground must be the same, and the depth of
focus of lenses ordinarily used must be used when making the back-
ground in order that the composite will have the appearance of
reality. The results that are possible with traveling matte processes
are so diverse and amazing, and at the same time invisible to the
audience, that the extreme care necessary is well worth the trouble.

There has been a recent revival in several of the special effects
departments of one process of making shots of the traveling matte
type which was employed nearly ten years ago. The advent of higher
speed negative emulsions has made the process practicable. The

June, 1932]

PROCESS PHOTOGRAPHY

787

process referred to is the one in which the foreground action is per-
formed in front of a translucent screen on which the background is
projected by a standard projector. It is possible by this method to
obtain effects similar to those obtained by using a traveling matte,
that is, the superposition of action on a moving background.

At the present time a great deal of study is being given to the
desired nature of the screen, the type of projection print to be used as
far as general density and contrast are concerned, and other problems
that have arisen in the course of trial of the process.

The variety of effects possible, and the ability of the workers in this
field to realize the possibilities of the various processes, have con-
tributed to making process photography a very useful and economical
branch of the motion picture industry.

REFERENCES

1 PETERS, T. K.: "A Museum of Motion Picture History," Trans. Soc.
Mot. Pict. Eng. (May, 1925), No. 22, p. 54.

2 GREGORY, C. L.: "Trick Photography," Trans. Soc. Mot. Pict. Eng. (Sept.,
1926), No. 25, p. 99.

3 HITCHINS, A. B.: "Method of Using Miniatures or Models for the Intro-
duction of Extra Detail in Motion Pictures," Trans. Soc. Mot. Pict. Eng. (Oct.,
1922), No. 15, p. 41.

4 BALL, J. A.: "Theory of Mechanical Miniatures in Cinematography,"
Trans. Soc. Mot. Pict. Eng. (May, 1924), No. 18, p. 119.

5 HUTCHINS, G. F.: "Dimensional Analysis as an Aid to Miniature Cine-
matography," Trans. Soc. Mot. Pict. Eng., 14 (Apr., 1930), No. 4, p. 377.

6 WALLER, FRED.: "Illusions in Cinematography," Trans. Soc. Mot. Pict.
Eng., 11 (July, 1927), No. 29, p. 61.

7 GREGORY, C. L.: "An Optical Printer for Trick Work," Trans. Soc. Mot.
Pict. Eng., 12 (Apr., 1928), No. 34, p. 419.

8 WILLIAMS, FRANK: "Trick Photography," Trans. Soc. Mot. Pict. Eng.,
12 (Apr., 1928), No. 34, p. 537.

9 DUNNING, CARROLL: "Composite Photography," Trans. Soc. Mot. Pict.
Eng., 12 (Sept., 1928), No. 36, p. 975.

A SHRINKAGE-COMPENSATING SOUND PRINTER*
R. V. WOOD**

Summary. — The shrinkage-compensating sound printer described is designed on
the principle of bending the shrunken negative film so that its emulsion surface will
temporarily regain the exact length that matches the positive. A means of achieving
this automatically is described and the advantages are noted.

When designing machines for printing sound on film the first
point to consider is the shrinkage of the negative. In general,
some arbitrary value of shrinkage is chosen, say, one-third of one
per cent, the machine being designed on this basis, so that for a
negative of this shrinkage no creeping between negative and positive
will occur. For a negative of a different shrinkage, creeping will
occur, and the attempt is made to make the creeping uniform.
The possibility of creeping necessitates a design that will make the
creeping as uniform as possible. This is generally attempted by
exercising extreme care in the workmanship, by using a very ac-
curately cut sprocket, and by extending the printing area over a
length sufficient to blur out creeping noises. The extension of the
printing area, combined with the creeping, results in a loss of sound
at the higher frequencies. The higher frequencies also suffer a loss
due to the difficulty of establishing contact between the films over an
extended area.

In the printer described here, the problem of design is approached
from a different view-point. No arbitrary figure is chosen for the
shrinkage, but the negative is stretched until it is of the same length
as the positive, thereby entirely eliminating creeping. The negative
is stretched until it exactly fits the positive; more accurately, the
emulsion surface of the negative is stretched by bending the film
until its emulsion surface exactly fits the emulsion surface of the
positive. In previous designs the matching of the emulsion sur-
faces was only approximate; and it has been found that an approxi-
mate match does not eliminate the creeping.

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Rochester, New York.
788

SHRINKAGE-COMPENSATING PRINTER

789

In Fig. 1, S is a sprocket that feeds and takes up the negative and
positive films (negative inside, positive outside, emulsions facing
each other) . R is a freely rotating drum of large diameter D driven
by contact with the positive film through tension at T and 7\.

Ri is a freely rotating roller of small diameter held by its own
weight and the weight of its mounting against the roller R. RL
is rotated by contact with the negative film, which is moved by con-
tact with the positive film.

The negative approaches the point of contact C through an arc of
diameter DI, which is determined approximately in threading the

UMJ*

FIG. 1. A sprocket for feeding and
taking up the negative and positive
films (negative inside, positive outside,
emulsions facing each other).

FIG. 2. Polished plate
made of Allegheny metal,
against which the edges of
the films are brought, in
order to keep them in line.

machine, and accurately by the movement of the films when the
machine is in motion.

It will be seen from the following analysis that DI is adjusted
automatically to its exact value for any condition of shrinkage as the
machine runs.

In feet per minute:

Speed of PPS for unshrunken positive = 90

Speed of PPR = 90 (assuming the pitch line is the middle of the film, which

moves at the speed of the film)
90 ^ + thickness of film

Speed of CPR

Speed of CNP

and the film as 0.0055)
90.108

790 R. V. WOOD [j. S. M. P. E.

Speed of PNR = 90.108 X D ^QQ^

Speed of PNS = 90 — 90s (where 5 is the per cent of shrinkage)

Now if the

Speed of PNR = speed of PNS

then

In other words, there is one value of D\ for any shrinkage.

FIG. 3. View of complete printer.

Now if DI is larger than this value, then the speed of PNR is
greater than the speed of PNS, and the loop PNR to PNS will
shorten until the correct value of DI is reached. Now if DI is smaller
than the correct value, then the speed of PNR is less than the speed
of PNS, and the loop PNR to PNS will increase until the correct
value of D! is reached.

So, at the point or line C, both emulsions move at the same speed

June, 1932]

SHRINKAGE-COMPENSATING PRINTER

791

and there is no tendency to creep. C is the printing point, an area
of approximately 10 by 100 mils.

The films are kept in line laterally by bringing their edges against
the polished Allegheny metal plate P (refer to Fig. 2). This plate
is raised slightly at the point CP.

CP is slightly out of line with SP, causing the negative film to
guide against it. This also assures a downward pressure at the
printing line. The roller RI is slightly out of perpendicular with
the plate P, which forces the film to guide against the plate.

The positive film is brought against the roller R at a point slightly
back of the plate P so that it guides against the plate P which lines it
with the negative and also causes a tendency to buckle up slightly
at the printing line, insuring perfect contact.

The advantages to be expected from this design are: (a) Better
definition; this is particularly important in printing 16 mm. film.
(b) A considerable saving in printer cost, (c) All degrees of shrink-
age within the intention of the design are printed with equally good
definition, (d) Narrow width negative may be printed on larger
stock; this is because at the printing point the films are driven and
guided entirely by their surfaces and one edge, without regard to
the other edge or the perforations.

DISCUSSION

MF. JENKINS: Do you know the difference in shrinkage of the positives and
negatives before you adjust the tension?

MR. WOOD: The tension adjusts itself as the machine runs; the loop takes
its own natural course and adjusts itself.

MR. KELLOGG: The film is rather slippery, and yet the adhesion between
the two films that are being pressed together must be depended upon to determine
the motion of one of the films — the negative, in this case. In view of the slight
tendency of the film to bend more at one point than at another, has not some
difficulty been experienced in this respect?

Also, how much does the length of the loop have to change in order to compen-
sate for a given difference in thickness? It would seem, from the general layout,
that it would be necessary to produce quite a change in the amount of film in
the loop in order to give the necessary change in curvature. That might be
somewhat of a problem in maintaining synchronism.

MR. WOOD: Regarding the amount of friction between the two films, if a
screw-driver is inserted in an attempt to displace the loop one way or the other
while it is running, it seems to resist, unbelievably, any pressure.

As regards the lack of synchronism, it is true that there is a variation of two
sprocket holes, at most. The sound track goes either one way or the other.
But as a variation in synchronism of the sound can not be detected within, say,
two frames, that point is not important.

COMMITTEE REPORTS

COLOR COMMITTEE REPORT*

The producing organizations working on color motion picture
processes in the United States may be grouped conveniently into two
classes, according as their process is of the additive or subtractive
type. In the additive method, the image to be projected is an original
black and white image, the color being obtained by interposing the
proper color filters in the light beam during projection. Kodacolor
(16 mm.) and its "parent," the Kellor-Dorian method, are representa-
tive additive processes.

In the subtractive method, which has enjoyed considerable favor
for several years, the original positive silver images are converted
wholly or in part to colored images composed of inorganic salts or
dyes, so that the final picture may be projected under normal condi-
tions on a standard projector. Typical examples of the subtractive
process are: Technicolor, Multicolor, Sennettcolor, Colorcraft, Photo-
color, Brewster Color, and Kodachrome.

It is not certain that any of the systems mentioned meet the desires
of the producers. Lower print costs are the immediate requirement.
Ability to make prompt deliveries is the second important require-
ment. Producers also resent the presence of strange cameramen on
the lots, and the wait for "rushes." Nothing will satisfy the producer
other than to make his own picture in his own way, on his own lot,
with his own men and equipment. This applies to the Class A pro-
ducers, while the independent will always welcome the independent
color print maker.

The question as to how much the theaters will stand for in making
changes of or addition to projectors in order to accommodate the
"additive processes" is a pertinent question for the Committee to
investigate. The trend in color picture projection appears to be
toward additive systems,** which require some changes in the pro-
jectors. Additive color prints are usually in black and white, making

* Presented at the Fall, 1931, Meeting at Swampscott, Mass.
** Some members of the Committee do not agree with this statement.
792

COLOR COMMITTEE 793

the rental cost of prints nearly equal to what exhibitors are accus-
tomed to pay.

It is not believed, however, that color pictures will ever be produced
and released at the same cost as a similar subject in black and white.
That this is so follows from the fact that by whatever color process
pictures are produced on the screen, there must be present, in the
observer's mind, pictures that are made up of two or three separate
components, each photographed individually by light of different colors
and each group of two or three representing only a single frame in black
and white. Such a multiplication of images must be more expensive
at some part of the process, whether it be in studio technic, negative
or positive materials, in projection, or in all four.

The definite trend toward the additive processes seems to be a step
in the right direction, since it appears to be directed toward the type
of process that will give the best color, and so be the most likely to
make colored pictures a necessity to the theater manager. It brings
with it, however, the problem of somehow increasing the amount of
light available for projection, since all pictures using the additive
process in projection must of necessity use as many times more light
as there are picture units and as much more as is required by the fact
that the colors employed do not transmit 100 per cent of the light of
the wavelengths used. That this brings up a rather serious problem
for large theaters goes without saying. If it can be solved, however,
the way appears to lie open for better color than has ever before been
shown in production, and at a relatively low cost.

It might be added that there has been an equally definite trend
toward the belief that the public will be won only by a process using
three fundamental colors, rather than the two now available by com-
mercial methods. The Kodacolor film has already been commercially
released by Eastman Kodak as 16 mm. and it is understood that it is
being constantly improved. Technicolor, Brewster Color, and others
are also examples of concerns working to this end. The last
two concerns are reported to be working on subtractive methods.
Any three-color subtractive methods as yet available have appeared
to be rather costly as regards sensitive materials and equipment
necessary. If these items could be cut down, it is quite possible that
a subtractive method having nearly, if not equally, as good color as
any additive one, would be the most satisfactory, since no changes of
importance in the projectors would be necessary.

In any case, those who have been fortunate enough, in the privacy

794 COLOR COMMITTEE

of their laboratories, to see how beautiful the most ordinary sets
can be made by the use of color photography will never give up the
belief that, in the not too distant future, a process will be developed
that will make the movies so attractive that, if people will not pay
more, at least more people will pay as much to be entertained by them.

NEW COLOR PROCESSES

Vocolor. — This process uses color wheel projection which draws
down one picture at a time, but which by optical means shows two
pictures superimposed on the projection screen. Black-and-white
films may be shown on the same projector without affecting the sound,
as the speed of the film through the projector is standard.

The negatives are made in the usual Kinemacolor manner, one
exposure at a time, and fringing is noticed.

Colorfilm. — This method uses double-coated film for the positives.
The film is first printed and developed in the usual way. Both sides
are toned red with uranium. All treatments so far are done by im-
mersion. The side that is to be blue-green is then passed over wicks
that feed a solution of iron and acid to one side of the film, converting
the red tone to blue (U..S. Patent No. 1,633,652).

W. V. D. KELLEY, Chairman
J. CALVIN BROWN F. E. IVES

JOHN G. CAPSTAFF H. W. MOYSE

W. T. CRESPINEL R. M. OTIS

RALPH M. EVANS WM. H. PECK

ARTHUR WADDINGHAM

SOCIETY OF MOTION PICTURE
ENGINEERS

OFFICERS
1931-1932

President
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y.

Past-President
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y.

Vice-Presidents

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J.
E. I. SPONABLE, Fox Film Corp., New York. N. Y.

Secretary
J. H. KURLANDER, Westmghouse Lamp Co., Bloomfield, N. J.

Treasurer
H. T. COWLING, Eastman Kodak Co., Rochester, N. Y.

Board of Governors

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada

H. T. COWLING, Eastman Kodak Co., 343 State St., Rochester, N. Y.

J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y.

P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y.

O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y.

A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New

York, N. Y.

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd.,

Los Angeles, Calif.
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio

E. I. SPONABLE, Fox Film Corp., 850 Tenth Ave., New York, N. Y.

795

THOMAS A. EDISON

Thomas Alva Edison was the first motion picture engineer. Before
him there were many who thought of motion pictures and who made
endeavors toward producing them. After him came many who
labored on improvements and elaborations of the motion picture.
None have had for the motion picture, or have brought to it, a broader
concept than did Mr. Edison.

THOMAS A. EDISON

He viewed the problem of the motion picture as the making of a
machine, a machine tool in the service of the art of expression. He
was personally interested in it chiefly as a maker of a mechanism,
which he delivered to the world to do with as it might.

The motion picture was a set of dawdling experiments and a haze of
day-dreams when Edison assigned himself the problem of bringing it
into a practical working existence, sometime in the year 1887. The
motion picture is in a very real sense the offspring of the Edison
796

THOMAS A. EDISON 797

phonograph. It was in 1887, in a bit of a lull in the laboratory work,
and in a day, too, when the commercial affairs of the phonograph
were annoying, that Edison took a bit of playtime to spend casually
on a machine "that should do for the eye what the phonograph did
for the ear."

Edison set a staff to work on his preliminary drawings, locked up in
the secrecy of room five at the West Orange (N. J.) works. His first
picture machine was a spiral record of microscopic pictures photo-
graphed on a cylinder like a phonograph, actuated with an inter-
mittent motion and viewed under a microscope. He had filled a room
with sound from a needle in a tiny groove and he was out to fill it with
pictures in a somewhat similar manner. In time, he decided upon a
machine that would feed pictures the size of postage stamps upon a
flexible tape moving past a lens, for viewing them either directly by
magnification, or by projection. By the mid-summer of 1889,
he had achieved such a machine, but had no satisfactory tape. He
demonstrated the machine with strips of collodion varnish that went
to bits and failed immediately. In the autumn he heard of the
coming of George Eastman's flexible medium for roller photography
in the Kodak. He sent to Rochester for a sample and put a trial
strip fifty feet long through his machine. It worked, and the motion
picture was an accomplished fact.

Interestingly enough, Edison's concept was a talking picture, and
in 1889-90 he built a talking picture machine, a twin phonograph
peep-show device.

It was not until late in 1892 that a promoter chanced upon the
motion picture machine in a corner of the West Orange plant and
prevailed upon Edison to let him put it on the market. The machine
in its peep-show form went out into the world, and all over the world,
beginning April, 1894. That was the Edison Kinetoscope. It pre-
sented film of the same dimensions, using the same sprocket holes and
other physical characteristics as the motion picture film of today.
The very size, which remains the same today, despite many experi-
ments then and since aimed at greater areas, was determined by the
covering power of the objective of a microscope that happened to be
about the plant in 1887 when the experimentation began.

For a complexity of commercial reasons which we, as engineers, are
not concerned about here, Edison wanted to keep the motion picture
in the peep-show for a while. But all over the world showmen were
demanding a machine that would show pictures to a whole room full of

7C8 THOMAS A. EDISON

paying patrons at one time, and so a score of inventors took the
Kinetoscope and set about the task of wedding it to the magic
lantern. Most of the technical tangles and patent wars of the indus-
try since have resulted from these parallel efforts. The history of the
motion picture industry in every nation in the world, and of every
motion picture corporation now in existence, can be traced to an
Edison Kinetoscope, be it in London, Paris, Berlin, Stockholm, or
Shanghai.

It is of incidental interest in this day of the talking picture to recall
that it also was Edison's exploration of the properties of the double-
filament incandescent lamp that led to the radio valve of today with
all its sound-picture functions and applications. It is coincidental
that William Kennedy Laurie Dickson, the same laboratory assistant
who worked on the motion picture job for Edison in room five, was
also the assistant who made the galvanometer tests of the "Edison
effect" in the twin filament lamps. They had sound and the radio
there, too, filed away in the notes of an unexplored region. One
lifetime was not enough in which to cover all that vast world of
technology that came within the range of Edison's vision.

Mr. Edison's records and correspondence of the day reflect a
recognition that the motion picture should present the sound, the
color, and the perspective of reality ; and that it was destined to serve
as a major successor of all the prior arts of expression, in entertain-
ment, in advertising, in education, and as an instrument of record.
He made it a tool, and left it largely to others to use and apply it.

Edison was concerned with what he deemed the great important
work of the world and the mechanisms with which to do it. He was
a maker of machines that worked. He brought processes and
methods across the dim borderland from the dreamers and the
experimental laboratories into the factories of modern, working fact.

TERRY RAMSAYE

SOCIETY ANNOUNCEMENTS

SPRING, 1932, CONVENTION

The program of the Washington Convention followed substantially
the Tentative Program mailed to the membership a month or so ago,
with some alterations in the order of presentation of the papers.
The attendance was unexpectedly large, particularly in view of exist-
ing conditions, and the interest shown in the proceedings was indeed
very gratifying. Great credit is to be given Mr. W. C. Kunzmann,
Chairman of the Convention Arrangements Committee; Mr. O. M.
Glunt, Chairman of the Papers Committee; Mr. N. D. Golden,
Chairman of the Local (Washington) Arrangements Committee;
Mr. H. Griffin, Chairman in Charge of Projection, assisted by Mr.
J. Frank, Jr.; Mr. J. I. Crabtree, who arranged the motion picture
exhibitions; Mr. W. Whitmore, Chairman of the Publicity Com-
mittee; and all those who assisted in arranging the details of the
Convention.

Interesting features of the Convention were the sessions devoted to
sixteen millimeter sound-on-film and to the problems of the release
print.

The proceedings of the Convention were divided into the following
sessions: General Session ; Committee Reports and Society Business ;
Photographic Session ; Symposium at the new building of the Depart-
ment of Commerce ; Projection Session ; Release Print Session ; and
Theater Operation Session. Of particular interest among these
were the sessions on projection, the release print, and theater opera-
tion, the attention being paid to these subjects indicating the rapid
broadening of the interest of the Society in the field of the theater.
The General Session, held on the first day of the Convention, included
a symposium on sixteen millimeter sound-on-film, a subject of out-
standing interest at the present moment.

On Wednesday, May llth, the Society was entertained by the
Department of Commerce in the auditorium of the new Department
Building, Mr. C. J. North presiding. Very interesting addresses
were given by Dr. Julius Klein, Assistant Director of the Department
of Commerce; Mr. F. M. Feiker, Director of the Bureau of Foreign

799

800 SOCIETY ANNOUNCEMENTS [j. s. M. p. E.

and Domestic Commerce; Mr. T. E. Robertson, Commissioner of the
Patent Office; and Mr. W. M. Steuart, Director of the Bureau of the
Census. Thanks are due the Department of Commerce for a highly
interesting session, and particularly Messrs. North and Golden for
arranging the session.

The semi-annual banquet of the Society was held on Thursday,
May 12th, at the Wardman Park Hotel, the Hon. W. P. Connery, Jr.,
Congressman from Massachusetts, acting as Master of Ceremonies.
Addresses were delivered by Mr. J. M. Gibbs, of the U. S. George
Washington Bicentennial Committee, and by the Hon. Robert
Ramspeck, Representative from the Fifth Georgia District.

Acknowledgment is to be made of the courtesies extended to the
Society during the Spring Convention by the following organizations:
Bausch & Lomb Optical Company, du Pont Film Mfg. Corp.,
Greater National Capital Convention Bureau, International Pro-
jector Corp., National Carbon Co., National Theater Supply Co.,
(Washington Branch), Raven Screen Co., RCA Photophone, Inc.,
Strong Electric Co., and Washington Projectionists Local No. 224.

The following Washington theater circuits are to be thanked for
honoring in their theaters the identification cards of members of the
Society: Fox Theater Corp., Loews Theaters, Inc., Radio-Keith-
Orpheum, and Warner Bros.

Thanks are also due to the following exchanges for supplying films
for the entertainment of the members: Paramount Publix Corp.,
Metro-Goldwyn-Mayer Pictures, Warner Bros. Pictures, RKO
Exchange, First National, United Artists, Universal Pictures Corp.,
Educational Film Exchange, Columbia Pictures, Pathe Exchange, and
the Motion Picture Division of the U. S. Department of Commerce .

BOARD OF GOVERNORS

At the meeting of the Board of Governors held at Washington on
May 8th, prior to the opening of the Spring Convention, final arrange-
ments for the Convention were completed, and action was taken upon
a number of amendments of the By-Laws of the Society, to be pre-
sented for voting at the open meeting on May 9th. Revised forms of
the Constitution and By-Laws, in accordance with these alterations,
will be available shortly to all members upon request. Among the
matters of immediate interest that were acted upon favorably by the
Board and subsequently by the membership at large, were the follow-
ing:

June, 1932] SOCIETY ANNOUNCEMENTS 801

Reduction of Admission Fees. — Acting upon the recommendations
of the Ways and Means Committee and of the Board of Governors,
a reduction of the admission and transfer fees was effected at the open
meeting of the Society on May 9th by unanimous approval of the
following amendment of the By-Laws of the Society:

By-Law VII. Dues and Indebtedness.

Section 1. The admission fee for applicants to the grade of Active membership
shall be ten dollars, and to the grade of Associate membership, five dollars.

Section 2. The transfer fee from the Associate grade to the Active grade shall
be the difference between the above-mentioned fees, or five dollars.

The complete schedule of membership fees is therefore as follows:

Admission fee to Active membership $10. 00

Admission fee to Associate membership 5. 00

Annual dues for Active membership 20. 00

Annual dues for Associate membership 10. 00

Transfer fee, Associate to Active 5. 00

Advertisements in the Journal. — In order to assist in the financial
operations of the Society, the Board of Governors, at a meeting held
on May 8th at Washington, authorized the Editor-Manager to solicit
advertisements for the JOURNAL, beginning immediately. Information
concerning the placing of advertisements in the JOURNAL may be
obtained from the General Office of the Society, together with a
schedule of rates and other data concerning the Society and its
activities. The assistance of the membership in securing advertise-
ments for their JOURNAL will be greatly appreciated.

COMMITTEE ON MOTION PICTURE EXHIBITION

In order to spread knowledge concerning the S. M. P. E. among
exhibitors, to promote the support and approval of the exhibitors of
the activities of the S. M. P. E., to obtain from exhibitors statements
of their problems, and to seek and propose solutions for these problems
and ultimately to bring about an organic relation between motion
picture engineers and exhibitors, the Board of Governors decided to
establish a committee to be known as the Committee on Exhibition,
whose function it would be to deal with the problems outlined above.
Announcement of the personnel of this committee will be made at a
later date.

802 SOCIETY ANNOUNCEMENTS

SUSTAINING MEMBERS

Agfa Ansco Corp.
Bausch & Lomb Optical Co.

Bell & Howell Co.
Bell Telephone Laboratories, Inc.

Case Research Laboratory
Du Pont Film Manufacturing Co.

Eastman Kodak Co.

Electrical Research Products, Inc.

Mole-Richardson, Inc.

National Carbon Co.

RCA Photophone, Inc.

Technicolor Motion Picture Corp.

HONOR ROLL

OF THE

SOCIETY OF MOTION PICTURE ENGINEERS

By action of the Board of Governors, October 4, 1931, this Honor Roll was estab-
lished for the purpose of perpetuating the names of distinguished pioneers who are
now deceased:

Louis AIME AUGUSTIN LE PRINCE

WILLIAM FRIESE -GREENE

THOMAS ALVA EDISON

GEORGE EASTMAN

AUTHOR INDEX, VOLUME XVIII
JANUARY TO JUNE, 1932

AALBERG, J. O.

BARTON, F. C.
BURKE, BURT S.

CARROLL, B. H.

(and D. HUBBARD)
CHAMBERS, G. A.
CHRISLER, V. L.

(and W. F. SNYDER)
CLARK, L. E.

COLE, D. M.

COOK, A. A.
CRABTREE, J.

CRABTREE, J. I.

(and H. D. RUSSELL)
CURTIS, A. M.

(and T. E. SHEA

and C. H. RUMPEL)
DAVIS, R.

(and G. K. NEELAND)
DAVIS, R.

(and G. K. NEELAND)
DEPUE, O. B.

DUBRAY, J. A.

(and A. S. HOWELL)
DUNOYER, Louis
EDGERTON, H. E.

FRAYNE, J. G.

(and H. C. SILENT)
FREDERICK, H. A.

Issue Page
Studio Projection and Reproduction

Practice May 652

Victrolac Motion Picture Records April 452

Thermionic Tube Control of Theater

Lighting Jan. 90

The Mechanism of Hypersensitization May 600
Process Photography June 782

Measurements with a Reverberation

Meter April 479

Sound Recording for Independent Pro-
ductions May 659
Sound in the Los Angeles Theater — Los

Angeles, Calif. Mar. 365

Optics of Projectors for 16 Mm. Film April 461
Directional Effects in Continuous Film

Processing Feb. 207

The Reducing Action of Fixing Baths on

the Silver Image Mar. 371

The Rapid Record Oscillograph in Sound

Picture Studies Jan. 39

Variation of Photographic Sensitivity

with Different Light Sources June 732

Variation of Photographic Sensitivity

with Development Time June 742

A Machine for Printing Picture and Sound

Simultaneously and Automatically May 643

Proposed Change in the Present Stand-
ards of 35 Mm. Film Perforations April 503

Lighting of Sound Films Jan. 3

Stroboscopic and Slow-Motion Moving
Pictures by Means of Intermittent
Light Mar. 356

Western Electric Noiseless Recording May 551

Vertical Sound Records: Recent
Fundamental Advances in Mechanical
Records on "Wax" Feb.

141

803

804

INDEX

[J. S. M. P. E.

GEOGHEGAN. G

GOEHNER, W. R.
(and T. E. SHEA
and W. HERRIOTT)

GOLDEN, N. D.
(and C. J. NORTH)

HALL, V. C.

HERRIOTT, W.
(and T. E. SHEA
and W. R. GOEHNER)

HOLDEN, E. C.

HOWELL, A. S.

(and J. A. DUBRAY)
HUBBARD, D.

(and B. H. CARROLL)
IVES, H. E.

JONES, L. A.
JONES, L. A.
KELLY, M. J.

LEAHY, W.

MAY, R. P.
NEELAND, G. K.

(and R. DAVIS)
NEELAND, G. K.

(and R. DAVIS)
NIXON, I. L.

NORTH, C. J.

(and N. D. GOLDEN)
OLSON, H. F.

RUMPEL, C. H.

(and A. M CURTIS
and T. E. SHEA)

RUOT, M.

RUSSELL, H. D.

(and J. I. CRABTREE)

SCHLANGER, BEN

Some Color Problems May 619

The Principles of the Light Valve June 697

The European Film Market — Then and

Now April 442

The Decibel in the Motion Picture In-
dustry Mar. 292

The Principles of the Light Valve June 697

Silica Gel Air Conditioning for Film

Processing April 471

Proposed Change in the Present Stand-
ards of 35 Mm. Film Perforations April 503

The Mechanism of Hypersensitization May 600

The Problem of Projecting Motion Pic-
tures in Relief April 417

Photographic Sensitometry, Part III Jan. 54

Photographic Sensitometry, Part IV Mar. 324

Vacuum Tube and Photoelectric Tube
Developments for Sound Picture Sys-
tems June 761

Time-and-Temperature vs. the Test
System for Development of Motion
Picture Negatives May 649

16 Mm. Sound Film Dimensions April 488

Variation of Photographic Sensitivity

with Different Light Sources June 732

Variation of Photographic Sensitivity

with Development Time June 742

Optical Instruments and Their Applica-
tion in the Motion Picture Industry Mar. 304

The European Film Market — Then and

Now April 442

Recent Developments in Theater Loud
Speakers of the Directional Baffle
Type May 571

The Rapid Record Oscillograph in Sound

Picture Studies Jan. 39

The Motion Picture Industry in Japan May 628
The Reducing Action of Fixing Baths on

the Silver Image Mar. 371

Utilization of Desirable Seating Areas in
Relation to Screen Shapes and Sizes
and Theater Floor Inclinations Feb. 189

June, 1932]

INDEX

805

SCHROTT, P.

SCHWARTZ, R. P.
(and H. B. TUTTLE)

SHEA, T. E.

(and A. M. CURTIS
and C. H. RUMPEL)

SHEA, T. E.

(and W. HERRIOTT
and W. R. GOEHNER)

SHEPPARD, S. E.

SILENT, H. C.

(and J. G. FRAYNE)
SNYDER, W. F.

(and V. L. CHRISLER)
SPENCE, J. L.

STRONG, H. H.
STOKOWSKI, LEOPOLD
TUTTLE, CLIFTON

TUTTLE, H. B.

(and R. P. SCHWARTZ)

TUTTLE, W. N.

VICTOR, A. F.

WALKER, V.
WESTERBERG, J. F.

WHITE, D. R
WHITE, D. R.
WOLF, S. K.
WOOD, R. V.

The Selenophon Sound Recording System May 622
Advantages of Using 16 Mm. Super-
sensitive Panchromatic Film in Mak-
ing Medical Motion Pictures May 609
The Rapid Record Oscillograph in Sound

Picture Studies Jan. 39

The Principles of the Light Valve

June 697

Resume of the Proceedings of the Dresden

International Photographic Congress Feb. 232
Western Electric Noiseless Recording May 551

Measurements with a Reverberation

Meter April 479

Mechanical Advantages of the Optical

Intermittent Projector May 593

A Reflector Arc Lamp for Portable Pro-
jectors June 752

Sound Recording — From the Musician's

Point of View Feb. 164

On the Assignment of Printing Exposure
by Measurement of Negative Char-
acteristics Feb. 172

Advantages of Using 16 Mm. Supersensi-
tive Panchromatic Film in Making
Medical Motion Pictures May 609

A Method of Measuring Directly the Dis-
tortion in Audio Frequency Amplifier
Systems Feb. 199

The Animatophone — A New Type 16

Mm. Synchronous Disk Reproducer April 512

Special Process Technic May 662

Size of Image as a Guide to Depth of

Focus in Cinematography May 655

Two Special Sensitometers Mar. 279

Gamma by Least Squares May 584

The Acoustics of Large Auditoriums April 517

A Device for Printing Sound Films June 788

CLASSIFIED INDEX, VOLUME XVIII
JANUARY TO JUNE, 1932

Acoustical Measurements.

Measurements with a Reverberation Meter, V. L. CHRISLER and W. F. SNYDER,

No. 4 (April), p. 479.

The Acoustics of Large Auditoriums, S. K. WOLF, No. 4 (April), p. 517.
Ail Conditioning.

Silica Gel Air Conditioning for Film Processing, E. C. HOLDEN, No. 4 (April),
p. 471.

Amplifiers.

A Method of Measuring Directly the Distortion in Audio Frequency Amplifier
Systems, W. N. TUTTLE, No. 2 (February), p. 199.

Vacuum Tube and Photoelectric Tube Developments for Sound Picture Sys-
tems, M. J. KELLY, No. 6 (June), p. 761.

Apparatus, General.

Thermionic Tube Control of Theater Lighting, BURT S. BURKE, No. 1 (Janu-
ary), p. 90.

A Method of Measuring Directly the Distortion in Audio Frequency Amplifier
Systems, W. N. TUTTLE, No. 2 (February), p. 199.

Two Special Sensitometers, D. R. WHITE, No. 3 (March), p. 279.

Optical Instruments and Their Application in the Motion Picture Industry,
I. L. NIXON, No. 3 (March), p. 304.

Western Electric Noiseless Recording, H. C. SILENT and J. G. FRAYNE, No. 5
(May), p. 551.

Architecture.

Utilization of Desirable Seating Areas in Relation to Screen Shapes and Sizes
and Theater Floor Inclinations, BEN SCHLANGER, No. 2 (February), p. 189.
Arcs, Projection.

A Reflector Arc Lamp for Portable Projectors, H. H. STRONG, No. 6 (June),
p. 752.

Color Photography.

Some Color Problems, G. GEOGHEGAN, No. 5 (May), p. 619.
Report of the Color Committee, No. 6 (June), p. 792.

Committee Reports.

Color, No. 6 (June), p. 792.

Journal and Progress Awards, No. 3 (March), p. 410.

Progress, No. 2 (February), p. 252.

Projection Practice, No. 1 (January), pp. 107 and 135.

Projection Practice, No. 4 (April), p. 539.

Projection Screens, No. 2 (February), p. 242.
806

INDEX 807

Projection Screens, No. 4 (April), p. 538.

Projection Theory, No. 1 (January), p. 113.

Sound, No. 3 (March), p. 410.

Sound, No. 4 (April), p. 526.

Standards and Nomenclature, No. 2 (February), p. 273.

Standards and Nomenclature, No. 3 (March), p. 409.

Studio Lighting, No. 5 (May), p. 666.

The Motion Picture Industry in Japan, M. RUOT, No. 5 (May), p. 628.

Ways and Means, No. 2 (February), p. 274.

Composite Photography.

Special Process Technic, V. WALKER, No. 5 (May), p. 662.
Process Photography, G. A. CHAMBERS, No. 6 (June), p. 782.

Defects in Film Resulting from Processing.

Directional Effects in Continuous Film Processing, J. CRABTREE, No. 2 (Feb-
ruary), p. 207.

Development of Motion Picture Film.

Directional Effects in Continuous Film Processing, J. CRABTREE, No. 2 (Febru-
ary), p. 207.

The Mechanism of Hypersensitization, B. H. CARROLL and D. HUBBARD, No.
5 (May), p. 600.

Time-and-Temperature vs. the Test System for Development of Motion Picture
Negatives, W. LEAHY, No. 5 (May), p. 649.

Variation of Photographic Sensitivity with Different Light Sources, R. DAVIS
and G. K. NEELAND, No. 6 (June), p. 732.

Variation of Photographic Sensitivity with Development Time, R. DAVIS and
G. K. NEELAND, No. 6 (June), p. 742.

Electrical Machinery and Equipment.

Thermionic Tube Control of Theater Lighting, BURT S. BURKE, No. 1 (Janu-
ary), p. 90.

Emulsions, Hypersensitization of.

The Mechanism of Hypersensitization, B. H. CARROLL and D. HUBBARD, No.
5 (May), p. 600.

Exhibition.

The European Film Market— Then and Now, C. J. NORTH and N. D. GOLDEN,
No. 4 (April), p. 442.

Film, Photographic Characteristics.

Photographic Sensitometry, Part III, LOYD A. JONES, No. 1 (January), p. 54.

On the Assignment of Printing Exposure by Measurement of Negative Char-
acteristics, CLIFTON TUTTLE, No. 2 (February), p. 172.

Directional Effects in Continuous Film Processing, J. CRABTREE, No. 2
(February), p. 207.

Photographic Sensitometry, Part IV, LOYD A. JONES, No. 3 (March), p. 324.

808 INDEX [j. s. M. P. E.

Gamma by Least Squares, D. R. WHITE, No. 5 (May), p. 584.

Variation of Photographic Sensitivity with Different Light Sources, R. DAVIS

and G. K. NEELAND, No. 6 (June), p. 732.
Variation of Photographic Sensitivity with Development Time, R. DAVIS and

G. K. NEELAND, No. 6 (June), p. 742.

Fixing of Motion Picture Film.

The Reducing Action of Fixing Baths on the Silver Image, H. D. RUSSELL and
J. I. CRABTREE, No. 3 (March), p. 371.

General.

Sound Recording — From the Musician's Point of View, LEOPOLD STOKOWSKI,
No. 2 (February), p. 164.

Resume of the Proceedings of the Dresden International Photographic Con-
gress, S. E. SHEPPARD, No. 2 (February), p. 232.

The Decibel in the Motion Picture Industry, V. C. HALL, No. 3 (March), p. 292.

The European Film Market — Then and Now, C. J. North and N. D. GOLDEN,
No. 4 (April), p. 442.

The Motion Picture Industry in Japan, M. RUOT, No. 5 (May), p. 628.

Sound Recording for Independent Productions, L. E. CLARK, No. 5 (May),
p. 659.

Home Motion Picture Equipment.

A Portable Non-Intermittent Cine Projector, No. 1 (January), p. 101.
Optics of Projectors for 16 Mm. Film, A. A. COOK, No. 4 (April), p. 461.
16 Mm. Sound Film Dimensions, R. P. MAY, No. 4 (April), p. 488.
The Animatophone — A New Type 16 Mm. Synchronous Disk Reproducer,
A. F. VICTOR, No. 4 (April), p. 512.

Hypersensitization.

The Mechanism of Hypersensitization, B. H. CARROLL and D. HUBBARD, No.
5 (May), p. 600.

Illumination in Photography.

Stroboscopic and Slow-Motion Moving Pictures by Means of Intermittent

Light, H. E. EDGERTON, No. 3 (March), p. 356.
Report of the Studio Lighting Committee, No. 5 (May), p. 666.

Illumination, Projection.

A Reflector Arc Lamp for Portable Projectors, H. H. STRONG, No. 6 (June),
p. 752.

Illumination, Sound Recorders.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.

Illumination, Theater.

Thermionic Tube Control of Theater Lighting, BURT S. BURKE, No. 1 (Janu-
ary), p. 90.

Incandescent Lamps for Recording.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.

June, 1932] INDEX 809

Index.

Author Index, Volume XVIII, No. 6 (June), p. 803.
Classified Index, Volume XVIII, No. 6 (June), p. 806.

Laboratory Apparatus.

The Rapid Record Oscillograph in Sound Picture Studies, A. M. CURTIS, T. E.
SHEA, and C. H. RUMPEL, No. 1 (January), p. 39.

A Method of Measuring Directly the Distortion in Audio Frequency Amplifier
Systems, W. N. TUTTLE, No. 2 (February), p. 199.

Directional Effects in Continuous Film Processing, J. CRABTREE, No. 2 (Febru-
ary), p. 207.

Optical Instruments and Their Application in the Motion Picture Industry,
I. L. NIXON, No. 3 (March), p. 304.

Silica Gel Air Conditioning for Film Processing, E. C. HOLDEN, No. 4 (April),
p. 471.

A Machine for Printing Picture and Sound Simultaneously and Automatically,

0. B. DEPUE, No. 5 (May), p. 643.

Lenses.

Size of Image as a Guide to Depth of Focus in Cinematography, J. F. WESTER-
BERG, No. 5 (May), p. 655.

Light Valves.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.
The Principles of the Light Valve, T. E. SHEA, W. HERRIOTT, and W. R.
GOEHNER, No. 6 (June), p. 697.

Loud Speakers.

Recent Developments in Theater Loud Speakers of the Directional Baffle
Type, H. F. OLSON, No. 5 (May), p. 571.

Medical Photography.

Advantages of Using 16 Mm. Supersensitive Panchromatic Film in Making
Medical Motion Pictures, H. B. TUTTLE and R. P. SCHWARTZ, No. 5 (May),
p. 609.

Obituary.

George Eastman, No. 4, (April) p. 539; No. 5 (May), p. 685.
Thomas A. Edison, No. 6 (June), p. 796.

Optical Intermittents.

Mechanical Advantages of the Optical Intermittent Projector, J. L. SPENCE,
No. 5 (May), p. 593.

Optics.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.
Optics of Projectors for 16 Mm. Film, A. A. COOK, No. 4 (April), p. 461.
Optical Instruments and Their Application in the Motion Picture Industry,

1. L. NIXON, No. 3 (March), p. 304.

Size of Image as a Guide to Depth of Focus in Cinematography, J. F. WESTER-
BERG, No. 5 (May), p. 655.

810 INDEX [J.S.M.P.E.

Oscillograph.

The Rapid Record Oscillograph in Sound Picture Studies, A. M. CURTIS, T. E.
SHEA, and C. H. RUMPEL, No. 1 (January), p. 39.

Perforation.

Proposed Change in the Present Standards of 35 Mm. Film Perforations, A. S.
HOWELL and J. A. DUBRAY, No. 4 (April), p. 503.

Photoelectric Cells.

Vacuum Tube and Photoelectric Tube Developments for Sound Picture Sys-
tems, M. J. KELLY, No. 6 (June), p. 761.

Printers.

A Machine for Printing Picture and Sound Simultaneously and Automatically,

O. B. DEPUE, No. 5 (May), p. 643.
A Device for Printing Sound Films, R. V. WOOD, No. 6 (June), p. 788.

Printing.

On the Assignment of Printing Exposure by Measurement of Negative Charac-
teristics, CLIFTON TUTTLE, No. 2 (February), p. 172.

Process Photography.

Special Process Technic, V. WALKER, No. 5 (May), p. 662.

Variation of Photographic Sensitivity with Different Light Sources, R. DAVIS

and G. K. NEELAND, No. 6 (June), p. 739.
Variation of Photographic Sensitivity with Development Time, R. DAVIS and

G. K. NEELAND, No. 6 (June), p. 742.
Process Photography, G. A. CHAMBERS, No. 6 (June), p. 782.

Processing.

Directional Effects in Continuous Film Processing, J. CRABTREE, No. 2 (Febru-
ary), p. 207.

The Reducing Action of Fixing Baths on the Silver Image, H. D. RUSSELL and
J. I. CRABTREE, No. 3 (March), p. 371.

Progress.

Organization of Progress Committee Work, No. 2 (February), p. 252.
The Motion Picture Industry in Japan, M. RUOT, No. 5 (May), p. 628.

Projection, General Information.

Report of the Projection Practice Committee, No. 1 (January), pp. 107 and

135.

Report of the Projection Theory Committee, No. 1 (January), p. 113.
Utilization of Desirable Seating Areas in Relation to Screen Shapes and Sizes

and Theater Floor Inclinations, BEN SCHLANGER, No. 2 (February), p. 189.
Report of the Projection Screens Committee, No. 2 (February), p. 242.
Projection Screens Committee, No. 4 (April), p. 538.
Projection Practice Committee, No. 4 (April), p. 539.
Studio Projection and Reproduction Practice, J. O. AALBERG, No. 5 (May), p.

652.

Projectors, Continuous.

A Portable Non-Intermittent Cine Projector, No. 1 (January), p. 101.

June, 1932] INDEX 811

Mechanical Advantages of the Optical Intermittent Projector, J. L. SPENCE,
No. 5 (May), p. 593.

Projectors, Portable.

A Portable Non-Intermittent Cine Projector, No. 1 (January), p. 101.
The Animatophone — A New Type 16 Mm. Synchronous Disk Reproducer,
A. F. VICTOR, No. 4 (April), p. 512.

Projectors, Special Type.

A Portable Non-Intermittent Cine Projector, No. 1 (January), p. 101.

The Problem of Projecting Motion Pictures in Relief, H. E. IVES, No. 4 (April),

p. 417.
Mechanical Advantages of the Optical Intermittent Projector, J. L. SPENCE,

No. 5 (May), p. 593.

Reduction.

The Reducing Action of Fixing Baths on the Silver Image, H. D. RUSSELL
and J. I. CRABTREE, No. 3 (March), p. 371.

Reverberation of Auditoriums.

Measurements with a Reverberation Meter, V. L. CHRISLER and W. F. SNYDER,

No. 4 (April), p. 479.
The Acoustics of Large Auditoriums, S. K. WOLF, No. 4 (April), p. 517.

Screens.

Report of the Projection Screens Committee, No. 2 (February), p. 242.
Projection Screens Committee, No. 4 (April), p. 538.

Selenophon.

The Selenophon Sound Recording System, P. SCHROTT, No. 5 (May), p. 622.

Sensitometry, Methods and Instruments.

Photographic Sensitometry, Part III, LOYD A. JONES, No. 1 (January), p. 54.
Two Special Sensitometers, D. R. WHITE, No. 3 (March), p. 279.
Photographic Sensitometry, Part IV, LOYD A. JONES, No. 3 (March), p. 324.
Gamma by Least Squares, D. R. WHITE, No. 5 (May), p. 584.

Sixteen Millimeter Equipment.

Optics of Projectors for 16 Mm. Film, A. A. COOK, No. 4 (April), p. 461.

16 Mm. Sound Film Dimensions, R. P. MAY, No. 4 (April), p. 488.

The Animatophone — A New Type 16 Mm. Synchronous Disk Reproducer, A. F.

VICTOR, No. 4 (April), p. 512.
Advantages of Using 16 Mm. Supersensitive Panchromatic Film in Making

Medical Motion Pictures, H. B. TUTTLE and R. P. SCHWARTZ, No. 5 (May),

p. 609.

Sound as an Art.

Sound Recording — From the Musician's Point of View, LEOPOLD STOKOWSKI,
No. 2 (February), p. 164.

Sound as a Science.

The Decibel in the Motion Picture Industry, V. C. HALL, No. 3 (March), p. 292.
Sound Committee, No. 3 (March), p. 410.

812 INDEX [j. S. M. P. E.

Sound Installations in Theaters.
Sound in the Los Angeles Theater — Los Angeles, Calif., D. M. COLE, No. 3

(March), p. 365.
Measurements with a Reverberation Meter, V. L. CHRISLER and W. F. SNYDER,

No. 4 (April), p. 479.

The Acoustics of Large Auditoriums, S. K. WOLF, No. 4 (April), p. 517.
Sound Recording, Disk.

Vertical Sound Records: Recent Fundamental Advances in Mechanical

Records on "Wax," H. A. FREDERICK, No. 2 (February), p. 141.
Victrolac Motion Picture Records, F. C. BARTON, No. 4 (April), p. 452.
Sound Recording, Variable Density Method.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.

Western Electric Noiseless Recording, H. C. SILENT and J. G. FRAYNE, No. 5

(May), p. 551.
The Principles of the Light Valve, T. E. SHEA, W. HERRIOTT, and W. R.

GOEHNER, No. 6 (June), p. 697.
Sound Recording, Variable Width Method.

Lighting of Sound Films, Louis DUNOYER, No. 1 (January), p. 3.
The Selenophon Sound Recording System, P. SCHROTT, No. 5 (May), p. 622.
Sound Reproduction, Disk.

Vertical Sound Records: Recent Fundamental Advances in Mechanical

Records on "Wax," H. A. FREDERICK, No. 2 (February), p. 141.
Sound Recording — From the Musician's Point of View, LEOPOLD STOKOWSKI,

No. 2 (February), p. 164.

Victrolac Motion Picture Records, F. C. BARTON, No. 4 (April), p. 452.
The Animatophone — A New Type 16 Mm. Synchronous Disk Reproducer,

A. F. VICTOR, No. 4 (April), p. 512.
Sound Reproduction, Film.

16 Mm. Sound Film Dimensions, R. P. MAY, No. 4 (April), p. 488.
Sound Reproduction, General Information concerning.

Sound Recording — From the Musician's Point of View, LEOPOLD STOKOWSKI,

No. 2 (February), p. 164.
The Decibel in the Motion Picture Industry, V. C. HALL, No. 3 (March),

p. 292.
Recent Developments in Theater Loud Speakers of the Directional Baffle

Type, H. F. OLSON, No. 5 (May), p. 571.
Sound Recording for Independent Productions, L. E. CLARK, No. 5 (May),

p. 659.

Sound Reproduction, Studio Installations.
Studio Projection and Reproduction Practice, J. O. AALBERG, No. 5 (May),

p. 652.

Standardization.

Proposed Change in the Present Standards of 35 Mm. Film Perforations, A. S.

HOWELL and J. A. DUBRAY, No. 4 (April), p. 503.

Resume of the Proceedings of the Dresden International Photographic Con-
gress, S. E. SHEPPARD, No. 2 (February), p. 232.

June, 1932] INDEX 813

Standards and Nomenclature.

Standards Committee, No. 2 (February), p. 273.
Standards Committee, No. 3 (March), p. 409.
Stereoscopy.

The Problem of Projecting Motion Pictures in Relief, H. E. IVES, No. 4 (April),

p. 417.
Stroboscope.
Stroboscopic and Slow-Motion Moving Pictures by Means of Intermittent

Light, H. E. EDGERTON, No. 3 (March), p. 356.
Studio Equipment.

Studio Projection and Reproduction Practice, J. O. AALBERG, No. 5 (May),

p. 652.

Report of the Studio Lighting Committee, No. 5 (May), p. 666.
Technical Motion Picture Photography.

Stroboscopic and Slow-Motion Moving Pictures by Means of Intermittent

Light, H. E. EDGERTON, No. 3 (March), p. 356.
Theater Design.

Utilization of Desirable Seating Areas in Relation to Screen Shapes and Sizes
and Theater Floor Inclinations, BEN SCHLANGER, No. 2 (February), p. 189.
Theater Design and Equipment.

Sound in the Los Angeles Theater — Los Angeles, Calif., D. M. COLE, No. 3

(March), p. 365.
Trick Photography.

Special Process Technic, V. WALKER, No. 5 (May), p. 662.
Process Photography, G. A. CHAMBERS, No. 6 (June), p. 782.

Vacuum Tubes.

Vacuum Tube and Photoelectric Tube Developments for Sound Picture Sys-
tems, M. J. KELLY, No. 6 (June), p. 761.

Society of Motion Picture Engineers

33 WEST 42ND ST.
NEW YORK. N. Y.

APPLICATION FOR MEMBERSHIP
APPLICANT'S RECORD

Name Age

Mailing Address ,

Permanent Address

Present Occupation

Employer

A complete account of the applicant's qualifications and accomplishments is
required before an application can be submitted to the Board of Governors.
The applicant should describe any inventions and improvements he has made
in the art, as these are considered of more importance than a mere record of
experience or the names of positions the applicant has filled.

Education .

Record of Accomplishments.

Motion Picture Experience.

Grade Applied For

REFERENCES

1 3.

2. 4.

The undersigned certifies that the above statements are correct, and agrees,
if elected to membership, that he will be governed by the Society's Constitution
and By-Laws so long as his connection with the Society continues.

Date 19. .. Signed

( Use a separate sheet of paper for complete record of accomplishments)

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Journal of the Society of Motion Picture Engineers"

See other formats