33 rf
Journal of the
Society of Motion Picture Engineers
VOLUME 51 JULY 1948 NUMBER 1
PAGE
Brightness and Illumination Requirements H. L. LOGAN 1
Light Modulation by P-Type Crystals . GEORGE D. GOTSCHALL 13
Portable 16-Mm Sound Projector H. H. WILSON 21
Optical Problems in Large-Screen Television. . . .1. G. MALOFF 30
Developments in Large-Screen Television . RALPH V. LITTLE, JR. 37
Discussion — Large-Screen Television 47
Motion Picture Theater Air Conditioning . D WIGHT D. KIMBALL 52
Air Purification by Glycol Vapor J. W. SPISELMAN 70
Ultraviolet Air Disinfection in the Theater. . .L. J. BUTTOLPH 79
Service and Maintenance of Air-Conditioning Systems
W. B. COTT. 92
Discussion — Ventilating and Air Conditioning 94
Display Frames in the Motion Picture Theater. .LESTER RING 101
Society Announcements 104
Book Review:
"Developing — Technique of the Negative," by C. I. Jacobson
Reviewed by Joseph S. Friedman 105
Current Literature . . 106
ARTHUR C. DOWNES HELEN M. STOTE GORDON A. CHAMBERS
Chairman Editor Chairman
Board of Editors Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions anc1
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOUBNAL must be obtained in writing from the General Office of the Society.
Copyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC . . . EXECUTIVE SECRETARY
OFFICERS
1947-1948
PRESIDENT EDITORIAL VICE-PRESIDENT
Loren L. Ryder Clyde R. Keith
5451 Marathon St. 233 Broadway
Hollywood 38, Calif. New York 7, N. Y.
PAST-PRESIDENT CONVENTION VICE-PRESIDENT
Donald E. Hyndman William C. Kunzmann
342 Madison Ave. Box 6087
New York 17, N. Y. Cleveland, Ohio
EXECUTIVE VICE-PRESIDENT SECRETARY
Earl I. Sponable G. T. Lorance
460 West 54th St. 63 Bedford Rd.
New York 19, N. Y. Pleasantville, N. Y. .
1948-1949
ENGINEERING VICE-PRESIDENT FINANCIAL VICE-PRESIDENT
John A. Maurer James Frank, Jr.
37-01— 31st St. 18 Cameron PI.
.Long Island City 1, N. Y. New Rochelle, N. Y.
TREASURER
Ralph B. Austrian
247 Park Ave.
New York 17, N. Y.
Governors
1947-1948
John W. Boyle Robert M. Corbin Charles R. Daily
1207 N. Mansfield Ave. 343 State St. 5451 Marathon St.
Hollywood 38, Calif. Rochester 4, N. Y. Hollywood 38, Calif.
David B. Joy Hollis W. Moyse
30 E. 42d St. 6656 Santa Monica Blvd.
New York 17, N. Y. Hollywood, Calif.
1948
William H. Rivers S. P. Solow R. T. Van Niman
342 Madison Ave. 959 Seward St. 4431 W. Lake St.
New York 17, N. Y. Hollywood, Calif. Chicago, 111.
1948-1949
Alan W. Cook Gordon E. Sawyer
4 Druid PI. Lloyd T. Goldsmith 857 N. Martel St.
Binghampton, N. Y. Burbank, Calif. Hollywood, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Brightness and Illumination
Requirements*
BY H. L. LOGAN
HOLOPHANE COMPANY, INC., NEW YORK 17, NEW YORK
Summary — This paper analyzes the problem presented by the con-
tinuous lighting of motion picture theater auditoriums; gives data on screen
brightness with various types of film running; relates auditorium brightness
to average screen brightness with film running; proposes a specific arrange-
ment of brightnesses from screen background to theater lobby; and suggests
a practical method by which this arrangement of brightnesses may be
attained with standard lighting equipment.
THE CONTINUOUS LIGHTING of motion picture theaters is a rather-
special problem. The tentative and controversial rules sug-
gested in Report No. 1 on l 'Brightness and Brightness Ratios" of the
Illuminating Engineering Society do not seem to apply, as they are
for spaces in which critical seeing tasks occur, and in which the object
of regard is a detail seen at close range.
The screen is the object of regard in a motion picture theater and is
seen at an average minimum distance1 of 22.5 feet. This distance is
greater than the minimum at which the ciliary muscles tense in order
to bring an object into focus.2 That is, the observers' lenses are in a
relaxed state at all screen-viewing distances in the average motion
picture theater, which is characteristic of distant, rather than close-
range vision.
The three-to-one rule3 would be impossible to apply in any event as
diffusion from the interior of the house would put an overlay of light
on the screen that would greatly reduce contrasts. Mimimum screen
brightnesses, with film running, as measured by the writer (see Table
I), ran as low as l/w of the average brightness. Still lower bright-
nesses were frequently encountered but they were below the range of
the instrument used and could not be accurately measured. The
author estimates that minimum brightnesses of 1/w of the average
brightness are common.
* Presented April 21, 1948, at the Atlantic Coast Section in New York.
JULY, 1948 JOURNAL OF THE SMPE VOLUME 51 1
2 LOGAN July
If the interior of the house was lighted to a brightness of one third
the average screen brightness (with film running) , and in such a way
that no direct light could reach the screen, the screen could still re-
ceive an overlay of light by diffusion from the house surfaces that
would approach one third of the average screen brightness. In this
case details that had a brightness less one third of the average screen
brightness would tend to be washed out, and brighter contrasts would
be diluted.
• TABLE I
MEAN BRIGHTNESS MEASUREMENTS OF MOVING PICTURE SCREENS
— Brightness In Foot-Lamberts —
Screen - — With Film Running —
Film Blank Minimum Average Maximum
Black-and-white news
15.0
0.35
1.08
2.50
Black-and-white feature
15.0
0.04
1.09
5.10
Black-and-white news
13.6
0.26
1.42
4.20
Cine-color short
15.0
0.53
1.40
2.70
Technicolor feature
10.1
0.18
1.27
5.63
Technicolor travelog
15.0
1.20
2.40
3.70
Self-colored animated cartoon
10.1
0.36
1.65
4.73
Self-colored animated cartoon
15.0
0.53
2.80
5.90
Such washing out and dilution of screen detail would not be accept-
able and so the subject must be approached from another angle. It
is for this reason that the author undertook to measure screen bright-
nesses with film running. Until such brightnesses were known, engi-
neers would be guessing at permissible brightnesses in the rest of the
observer's field of view, as such brightnesses obviously had to be con-
siderably less than screen brightnesses in order not to wash out
screen detail with an overlay of diffused light originating from the
walls, ceiling, and floor of the illuminated auditorium.
Measurements of screen brightnesses with film running were not
possible until the development of the instrument shown in Fig. 1.
This instrument has a specially shaped mirror that picks up the same
field of view as an observer. The mirror is viewed by a photosensi-
tive electronic cell and the results read on a milliammeter. When the
instrument is so located with respect to the screen that the screen sen-
sibly fills the fteld of the instrument, and no other light approaches
from any other part of the instrument field, the needle deviation is a
measure of the total light coming from the screen at any instant. The
1948
BRIGHTNESS AND ILLUMINATION REQUIREMENTS
instrument is equipped with filters to correct the response ol the elec-
tronic cell to that of the standard observer, the action of the
instrument being independent of human judgment, and automatic.
The instrument measures "steradian foot-lamberts," which can be
converted into other units when the distance of the instrument from
the source of light is known, or when there is some reference condition
to tie to, such as the brightness of the blank screen (which was sepa-
Fig. 1 — Logan fluxmeter.
rately measured by two observers with two different Luckiesh-Taylor
Brightness Meters in this investigation).
The writer has adopted the average screen brightness with film
running as the basic reference criterion in the design of motion picture
theater lighting. Previous investigators, having no way of arriving at
instantaneous average brightness for entire screen with film running,
have used stills and measured the "white" and the "blacks" which
they have then used as reference values.4- 5 The eye may spend as
LOGAN
July
BLACK AND WHITE "NEWS"
BLACK AND WHITE FEATURE "LURED"
TIME
NORMANDIE THEATRE N.YC
MIN BRIGHTNESS 0.04 FT-L
\AVER. 1.09
MAX. 5.10
,
H
ft
UH
r^~
Jl/ V
^ \/\i
TIME
BLACK AND WHITE "NEWS"
SELF-COLORED ANIMATED CARTOON
PARK AVE. THEATRE N.YC.
MIN BRIGHTNESS 0.26 FT-L.
AVER. "
MAX. 4.20
TIME
NORMANDIE THEATRE N.YC. /\
MIN. BRIGHTNESS 0.53 FT-L/ \
TIME
CINE COLOR "SHORT"
NORMANDIE THEATRE N.YC.
MIN BRIGHTNESS 0.53 FT-L
AVER. 1.40 *
MAX 2.70
TIME
SELF-COUORED ANIMATED CARTOON
PARK AVE. THEATRE N.YC.
MIN BRIGHTNESS 0.36 FT^L
AVER. n " 1.65
MAX. In 4.73
TIME
TECHNICOLOR "TRAVELOG"
NORMANDIE THEATRE N.YC.
MIN BRIGHTNESS 1.20 FT-L
AVER. Z40
MAX. 3.70
C\^
M \l^
\
^^
TIME
FIG. 2 — Sequence of brightness changes for various types of film.
1948
BRIGHTNESS AND ILLUMINATION REQUIREMENTS
S3
sil
^>
little as 5/ioo second on a fixation point, and its fixation pauses nor-
mally average 15/ioo second.6 Motion picture film is moved along at a
rate that is based on this fact and its complement, the persistence of
vision. As a result the eyes in screen viewing seldom have an oppor-
tunity to rest on a fixation point long enough to adapt to it. The
author believes that with the rapidly fluctuating brightness of the
screen with picture running, the eyes have no choice but to adapt to
the mean screen brightness. See Fig.
2 for sequence of brightness varia-
tions on screen as a whole, for films
measured.
If we adopt the lowest mean bright-
ness, as measured in this investigation,
that is likely to be met. namely that
of black-and-white newsreels, as our
reference value, we start with 1 foot-
lambert for the screen in action. We
can allow one tenth of that,7 so long as
it is very uniformly distributed, as the
steady brightness of the walls, ceiling,
and floor of the auditorium. That is,
these surfaces may have a brightness
of 0.1 foot-lambert. This brightness
should be carried right up to the edge
of the screen. Fabir Birren reports
to the writer that when the space be-
tween screen and proscenium arch was
lighted in experiments in the Walt
Disney studios the illusion of great
depth was created in the pictures.
This area should be lighted so as to
appear as a pale gray mist. The
lighting preferably should be abso-
lutely uniform, but if that cannot be
accomplished because of job condi-
tions, then the brightness should be
least near the screen, and rise to the
brightness of Vio foot-lambert of the
auditorium walls. This is not much
light but it is close to what the Park
4
c — -
6 LOGAN July
Avenue Theater, New York City, provides for full house lighting
with no picture running, for example. Measurements in this modern
theater with house lighting on full gave the following :
Foot-Lamberts
Walls 0.120
Balcony face 0.085
Ceiling 0.070
Offhand, it would seem that the house lights in this theater therefore
could be operated at all times with great improvement to the ease,
comfort, and safety of the patrons. Perhaps the principal reason why
this may not be true, is that the auditorium is largely lighted by coves.
It is impossible to control the action of coves. The wall area in the
immediate neighborhood of these coves has a brightness of 10 foot-
lamberts. m This means that the cove running around the proscenium
arch, for example, which is one of the main sources of light in this
auditorium, is nearly ten times the brightness of the screen when a
black-and-white newsreel is running; and 500 times the minimum
screen brightness that can be expected. Jones5 found that a bright-
ness of 3 foot-lamberts was the highest that could be tolerated toward
the front of the auditorium.
In addition to this, the ceiling roundels that helped provide the
house lighting had a brightness of 600 foot-lamberts. This is not so
• important as the cove brightness because these roundels are not in-
cluded in the field of view of most people on the auditorium floor and
would be disturbing principally to occupants of the balcony. The
significant point is that the brightness of these ceiling roundels is a
function of the size of lamp used, and the latter is larger than neces-
sary because the reflection factors of the various surfaces reached by
the light are too IOAV. If higher reflection factors for floor walls, and if
the backs of seats were used, the lamps could be reduced in proportion,
and the brightness of the roundels could be dropped to more reason-
able figures.
It is evident that in order to attain satisfactory house lighting
while the screen is in action, the distribution of light must be very
carefully controlled, as the brightness level of 0.1 foot-lambert must
be the actual maximum brightness at any point within 30 degrees of
the line of sight of a patron watching the screen (see Fig. 3). This
control not only will involve the careful selection of location of the
light sources, but also the careful choice of materials for walls, ceiling,
1948
BRIGHTNESS AND ILLUMINATION REQUIREMENTS
and floor, to reflect the proper quantity of light efficiently, and so per-
mit the use of small lamps in order to have equipment of very low
brightness.
This problem will ease somewhat as screen brightnesses become
higher. If, for example, colored films replace black-and-white en-
tirely, house lighting can about double, as indicated by the figures in
the last column of Table II.
That there is a need for house lighting is evident. Patrons of mov-
ing picture theaters are coming and going constantly. The lack of
pjihSPECTIVji OF ' A HXPOTHiTICALLX ,
VERAGE-, MOTION PICTURE HOUSE
Fig. 3 — All that is included in the field of view of a patron seated in the
standard observer's position.
light handicaps safe movement and causes inconvenience, not only to
the patrons who are moving in and out of seats, but to those with
whom they interfere because they cannot see sufficiently well to move
with the least disturbance.
Illumination levels in moving picture theaters with the house lights
out are somewhere between starlight and moonlight, and much closer
to the former than the latter. Under these conditions accidents will
average seven times the theoretical minimum rate.8 There is no
possibility that accidents can be reduced to the theoretical minimum
rate in the foreseeable future as a general brightness of the field of view
of 6 f oot-lamberts would be required, or sixty times the maximum that
8
LOGAN
July
present screen brightnesses will permit. However, the adoption of a
general brightness level of 0.1 foot-lambert would tend to reduce acci-
dents by about 43 per cent over present experience, which would be a
worth-while gain.
Second, the lack of sufficient general light invites undesirable con-
duct on the part of some members of the audience, particularly chil-
dren and young people.
TABLE II
COMPARISONS OF AVERAGE BRIGHTNESS MEASUREMENTS OF MOTION PICTURE
SCREENS
Brightness of Screen
with Film Running as
Average Brightness a Percentage of Blank
in Foot-Lamberts Screen Brightness
Screen Film Group
Film Blank Running Net Average
Black-and-white news
15.0
.08
7
Black-and-white feature
15.0
.09
7
8
Black-and-white news
13.6
.42
10
Cine-color short
15.0
.40
9
Technicolor feature
10.1
.27
13
Technicolor travelog
15.0 J
J.40
16
15
Self-colored animated cartoon
10.1
L.65
16
Self-colored animated cartoon
15.0 J
J.80
19
Third, it reduces the comfort of patrons. Vision is tolerable in
most moving picture theaters but all authorities agree that, with a few
notable exceptions, it is far from comfortable Television manufac-
turers have been faced with this comfort factor and have found it
necessary to raise television screen brightness so that the screen can
be viewed in domestic interiors with house lights on.
Last, house lighting inspires and aids better "housekeeping," which
patrons find inviting.
A brightness of 0.1 foot-lambert is too low for people to adapt to
quickly when coming from outdoors, unless a long foyer is available in
which the lighting can drop steadily as the people move along. How-
ever, some improvement in the situation can be brought about in any
case if the brightness of the foyer surfaces at the theater end is set at 2
foot-lamberts. This should be succeeded by a brightness of 1 foot-
lambert for the surfaces of the extreme rear of the auditorium behind
1948 BRIGHTNESS AND ILLUMINATION REQUIREMENTS
>>»»• ^GttTune'-L'l
. /• u\
SECTION
OBSERVERS POSITION FOR
ILLW1INEERING PERSPECTIVE
PLAN
Fig. 4-^-Section gives sight lines of closest and farthest observer, and of
observer in standard position. It also shows the 30-degree angle with sight
lines and that no luminous part of any lighting unit comes within 30 degrees
of sight line of any observer. Finally, it shows that no direct light from a
lighting unit can reach the screen. Figures 0.1^ etc., show recommended
brightness levels in foot-lamberts.
10 LOGAN July
the last row of seats (the crossover) . This 1 f oot-lambert brightness
should drop to 0.5 f oot-lambert on the aisle floors within 10 feet of the
rear end of the aisle, and to the prevailing 0.1 f oot-lambert within 20
feet. From then on, up to the front of the auditorium, the floor
brightness of the aisles should remain at 0.1 f oot-lambert. This ar-
rangement of brightnesses is illustrated in Fig. 4, with a suggested ar-
rangement of lighting outlets to accomplish it.
Higher house brightnesses would be possible if motion picture thea-
ters were designed to permit them. This might sometimes require
the screen to be louvered or hooded (after the fashion of traffic lights,
or the miniature screens used for sales promotion in camera stores).
Offhand this would appea r to reduce the number of seats by narrowing
the angle of view, but this would not be necessary as the principal
louvering would be against the ceiling to prevent direct light from the
ceiling lights striking the screen. The actual amount of hooding re-
quired could also be reduced by sinking the lights into the ceiling, so
that the depth of the coffer acted as a louver against the screen (see Figs.
3 and 4) . This would have the advantage of hiding the main ceiling
lights from the balcony patrons. An examination of Figs. 3 and 4
will show that such ceiling lights in the main ceiling cannot come into
the field of view of any patron on the main floor as long as none n re
placed in the forward 30 per cent of the ceiling. This prohibition also
prevents stray light from any of the ceiling sources reaching the
screen. Figs. 3 and 4 also show that the lights in the main ceiling,
and those in the balcony soffit, when recessed in properly design e 1
coffers, are hidden from most patrons. In the few cases where the
lens can become visible it is at the upper edge of a patron's field of vi:» w
where it is farthest from the line of sight, and the least effective1 in re-
ducing visual efficiency and comfort. The sides of the light coffers
should be painted a dark gray to prevent them being bright enough to
disturb patrons. Concealed downlights also could be used. In
many theaters the use of lens units in coffers, or downlights, would be
sufficient, in combination with the fact that screens are usually placed
6 to 10 feet behind the proscenium arch, to make special louvering of
the screen unnecessary.
In addition to this it would be desirable to give all surfaces that are
parallel to and face the screen, such as the balcony face and rear wall
of the theater, a low reflecting finish (about 20 per cent), to make their
contribution to screen brightness neglibible. Surfaces parallel to, but
that face away from the screen, such as the backs of the seats, should
1948 BRIGHTNESS AND ILLUMINATION REQUIREMENTS 11
be given a reflection factor about equal to floor, or 30 per cent.
Finally, the walls and ceiling should be sloped away from the screen as
far as possible and given a ribbed surface. One side of each rib should
face away from the screen and be given a reflection factor of 50 per
cent. The other side of the rib that would face in the general direc-
tion of the screen'could be dark gray with a 10 per cent reflection fac-
tor. This permits light to be accepted by these surfaces without its
getting back to the screen. Floor coverings should have a reflection
factor of 30 per cent and ceiling lights should begin no closer to the
screen than about one third the depth of the house. They would be
arranged over the remaining two thirds of the ceiling over the aisles
so that the aisles would get the benefit of the principal illumination and
no patron could be directly under a light to receive a high light on
back of head and shoulders that might be disturbing to others. This
would usually also light up the walls owing to the side aisles running
along the walls. Where there are no side aisles the lights should also
run in such relation to the walls as to light them uniformly.
Similar lights should be repeated under the balcony over the aisles,
and along the back crossover.
Illumination on the ceiling would come from diffusion from the
walls and floor. If ceiling was finished white it would acquire a
brightness about equal to moonlight.
It would be desirable to raise this brightness to Vio of a foot-lam-
bert but most attempts to do this raise more problems than they
solve. Where the scale of the interior permits, as in Radio City
Music Hall, it can be done by a similar series of well-designed stepped
coves. This is impractical in the average motion picture auditorium,
and it is better to let the ceiling remain dark than to run into the great
brightness variations that accompany most attempts at ceiling
illumination.
It is easier to meet the visual requirements of continuous motion
picture theater lighting with incandescent lamps than with fluores-
cent, as the extraordinary degree of control required is difficult with
fluorescent. Fluorescent can be used for the decorative and inter-
mission lighting. Further, the incandescent equipment can be
dimmed easily so that after the computed installation is made, the
exact point at which the house lighting no longer handicaps the
screen can be determined by experiment.
In conclusion, the brightness level of 0.1 foot-lambert suggested in
this paper for house lighting (about three times full moonlight),
12 LOGAN
would call for an illumination level of from 0.3 to 0.4 foot-candle, on
the basis of the reflection factors recommended This can be secured,
from 60-watt, incandescent lamps on about 15-foot centers average,
in controlled, coffered, direct-lighting equipment. Where the bright-
ness level is to rise; as at the rear stretch of the aisles, and the rear
crossover, the lights should be spaced proportionately closer. If
these incandescent lamps were used behind a Controlens the glass
surface could have an off-axis brightness of as low as 60 foot-lamberts,
instead of 600 which is present practice. Even this low equipment
brightness of 60 foot-lamberts would not come into the normal field of
view, owing to the coffer shielding, as previously explained.
Now that the nature of the problem is understood, designers can be
depended upon to come forth with a variety of layouts and equipment
that will meet the conditions.
REFERENCES
(1) "Report of screen-brightness committee," /. Soc. Mot. Pict. Eng., vol.
50, pp. 260-274; March, 1948.
(2) "I.E.S. Lighting Handbook," Illuminating Engineering Society, New
York, New York, 1947, p. 2-2.
(3) "The brightness ratio of the visual task" (ex. gr., the screen) "to its im-
mediate surroundings should be no greater than three." From "Brightness and
brightness ratios," Report No. 1 of the Committee on Standards of Quality and
Quantity for Interior Illumination of the Illuminating Engineering Society,
Ilium. Eng., vol. 39, pp. 713-723; December, 1944.
(4) F. M. Falge and W. D. Riddle, "The lighting of motion picture audi-
toriums," /. Soc. Mot. Pict. Eng., vol. 3, pp. 201-212; February, 1939.
(5) L. A. Jones, "The interior illumination of the motion picture theater,"
Trans. Soc. Mot. Pict. Eng., no. 10, pp. 83-96; 1920.
(6) M. L. Luckiesh, "Reading as a Visual Task," D. Van Nostrand Company,
Inc., New York, New York, 1942, p. 23.
(7) This relationship of 10 to 1 for screen brightness to surrounding brightness
(or auditorium brightness), is based on Lythgoe's research as reported by Parry
Moon in "The Scientific Basis of Illuminating Engineering," McGraw-Hill
Book Company, Inc., New York, New York, 1936, p. 441.
(8) H. L. Logan, "The role of lighting in accident prevention," Elec. Eng ,
vol. 62, pp. 143-147; April, 1943.
Light Modulation by P-Type
Crystals*
BY GEORGE D. GOTSCHALL
BRUSH DEVELOPMENT COMPANY, CLEVELAND, OHIO
Summary— A means of modulating a light beam with a flat response
which greatly exceeds the audio range is described utilizing the linear
electrooptic effect in P-type** crystals. The unit uses parallel polarized
light and although it requires relatively high voltages, it draws essentially
no current and has no moving parts. From the basic aspects it seems very
promising for an efficient yet sturdy variable-density recording system.
INTENSITY MODULATION of a light beam may be accomplished in one
of three or more ways: (1) mechanical shutter or diaphragm, (2)
modulation of intensity of the light source proper, or (3) phase
retardation of polarized light passing through an electrooptic medium.
The first two methods are quite well known, but the third, although
not new in principle, is relatively new in application. The quadratic
electrooptic effect, also known as the Kerr effect, has been known to
scientists for some tune, especially in such polar liquids as nitro-
benzene. However, the effect is small, voltages are extremely high,
the light beam is restricted in width, and the cell is liquid in form.
The effect to be described herein is the so-called linear electrooptic
effect which is a property characteristic of piezoelectric crystals and
which overcomes the Kerr-cell disadvantages wholly or in part.
The basic symmetry relationships of the linear electrooptic effect
were first discussed by Pockels,1 who made detailed measurements on
quartz, sodium chlorate, tourmaline, and Rochelle salt. This work
is still basic for our knowledge of the electrooptic effect in crystals,
although during the last fifteen years several German scientists have
published articles on this phenomenon with emphasis on the applica-
tion of zinc sulfide.
* Presented November 19, 1947, at the Atlantic Coast Section Meeting in
New York.
** The term "P-type" has been given to a group of isomorphic crystals which
include the primary phosphates and arsonates of ammonium, potassium, or
rubidium.
JULY, 1948 JOURNAL OF THE SMPE VOLUME 51 13
4 GOTSCHALL July
The requirements which a crystal must meet in order to be suitable
electrooptically are many, and as a result only a few crystalline sub-
stances are satisfactory. Extensive research at the Brush Develop-
ment Company led to the discovery of a substantial electrooptic
effect in ammonium dihydrogen phosphate crystals as well as other
crystals of the primary phosphate family, termed P-type. It was
recognized that the symmetry relationships in this crystal family
were particularly favorable for the occurrence of a sizable parallel
electrooptic effect, i. e., with the light beam parallel to the electric
field. Among the outstanding features of these crystals are the facts
they can be artificially grown to large sizes, have excellent trans-
mission characteristics, satisfactory physical properties, and possess
the highest electrooptic constants.
To explain briefly the nature of the parallel electrooptic effect,
first consider the mechanics of light transmission through a trans-
PARALLEL LIGHT SOUJBCC CRYSTAL ANALYZER (P£)
\f , I
1 I
Ij JL | t _-z— LJGHT BEAM
/ /
POL ABIZEB (P,) V* BETABDA TlON SHEET
(WHEN US£O)
Fig. 1 — Simple electrooptic, system.
parent substance. Transparent media are composed of atoms or
molecules carrying electric charges or dipoles. The refraction (or
velocity) of light is then determined by the disposition and electro-
magnetic interaction of these particles with the electric-field com-
ponent of the incident light wave. The application of an electric
field to the transparent medium will produce displacement or deforma-
tion of the particles and thereby influence the refractive properties
of the material. In the case of the P-type crystals, a certain cut and
alignment allows light to pass with equal velocities for all planes of
polarization. However, the application of an electric field reduces
the phase velocity in one specific plane of polarization and increases
it in a perpendicular plane of polarization, thus introducing a phase
difference in the two components. The parallel electrooptic effect
may then be described as a linear change of the refractive indexes
(or phase velocities) for a giyen applied electric field. This change is
1948
LIGHT MODULATION BY P-TYPE CRYSTALS
15
at best in the order of one part per million for a field of one kilovolt
per centimeter, but even this minute change is satisfactory for a
device based on phase shifts.
To explain the behavior a bit more fully, let us consider a simple
system (Fig. 1) where we polarize a parallel light beam so that the
vibration direction is parallel to one of the crystalline axes, X or
Y, (mutually perpendicular to the Z or optic axis in P-type crystals) .
This light beam can be resolved into two vector components along-
secondary axes X and ? (also called mechanical axes) (Fig. 2). In
A
X
45
A'
Y
Fig. 2— Z-0 Plate of "PN" crystal showing position of
polaroids and secondary axis.
this figure, the polarizer (P^) is shown as parallel to the Y axis and
the analyzer. (P2) parallel to the X axis. Initial conditions will
permit no light to pass, for we have the analyzer at 90 degrees to the
polarized beam. However, under the influence of an electric field,
the velocity of the vector components along X and Y are changed,
one being increased and the other decreased, the effect on each
depending on the polarity of the voltage. This results in incomplete
extinction of the light, which will now be visible through (P2).
The electrooptic effect will be maximum for the light polarized
parallel to one of the crystalline axes, X or Y (as shown in the above
16
GOTSCHALL
July
diagram) and minimum for light polarized parallel to one of the
secondary axes, X or ?.
For a field of one kilovolt per centimeter, the difference in the
two refractive indexes was found to be 2.9 X 10 ~6, or equivalent to
a phase difference of about Tr/20 radians per centimeter for green
light. The intensity of the light passed is found to be proportional
to the sine squared of half the angular phase difference. For a
"PN"* (ammonium dihydrogen phosphate) crystal as discussed we will
get maximum transmission for a retardation of one half wavelength
which requires approximately 9000 volts for green light (Fig. 3).
For the system as described, thickness of the crystal does not enter,
VOL TA GE APPL /ED
O.C. K^LOVOLTS
Fig. 3 — Light intensity versus direct, kilovolts for "PN"-Nesa unit using
green parallel light and crossed polaroids.
as doubling the thickness for a given voltage will result in half the
change of refractive index but twice the light path.
The value of the change in refractive index as given above for "PN"
is the total electrooptic effect which is composed of the "clamped"
electrooptic effect and the "indirect" electrooptic effect. This latter
effect is created by the piezoelectric deformation of the crystal lattice
which causes a refractive index change through the elastooptic effect.
In the P-type crystals these effects are of the same sign, and recent
experimental data have shown the division in "PN" to be 58 per cent
"clamped" and 42 per cent "indirect."
* Copyright, Brush Development Company, Cleveland, Ohio.
1948
LIGHT MODULATION BY P-TYPE CRYSTALS
17
Theoretically, the "clamped" electrooptic effect is frequency-
independent practically to the infrared region because displacements
in the order of only intermolecular distances are involved. The
"indirect" effect has resonant frequencies dependent upon the phys-
ical dimensions. It is to be expected then that this effect would
become negligible above the range of natural resonant frequencies of
the crystal. The frequency curve is shown in Fig. 4 and it is to be
noted that the slight increase in response at the high end is due to
approaching resonant frequency. Mechanical damping will elimi-
/o.ooo 20.000
Fig. 4— Frequency response of light modulator using a
150-volt, root-mean-square, sine wave. White, parallel,
direct-current light source was used with X/4 plate inserted
between the crossed polaroids.
nate this rising characteristic. The linearity of modulated output
versus modulation voltage is shown as Fig. 5.
When using the unit as a light-beam modulator, it is usually de-
sirable to use an optical bias of a quarter-wave retardation. As
shown in Fig. 3; this transfers the initial operating point from the
origin to point X with no voltage applied. Operating from this point
of the curve it can be seen that we get maximum modulation and
minimum distortion for a given applied audio voltage. In most of
the frequency experiments, a modulating voltage of 150 volts, root-
mean-square (425 volts peak-to-peak) was used with very satis-
factory results.
A convenient size of crystal for experimentation was found to be
18
GOTSCHALL
July
1
2 inches square and about Vie inch thick. After polishing to an
optical finish, transparent conducting electrodes, Nesa,* are at-
tached by means of a special transparent conducting cement. The
Nesa electrodes are l3/]6 inches square, thus providing a nice margin
of crystal to act as high-voltage insulation. This margin is treated
to prevent damage by moisture and consequent leakage or voltage
flashover; The unit easily accommodates a 1-inch diameter parallel
light beam. The static capacity of such a unit is under 100 micro-
microfarads, diminishing rap-
idly in the resonant frequency
region. The direct-current
shunt resistance is in the order
of 150 to 350 megohms.
The transmission of the
over-all system depends
chiefly upon the transmission
of the polarizing material and
the transparent electrodes as
the transmission of "PN" is
practically 100 per cent from
2000 angstrom units to 1.5
microns. The transmission of
a recently completed unit
composed of "PN" crystal
and two Nesa electrodes ce-
mented onto place was about
75 per cent whereas the Polar-
oid** had 28 to 32 per cent
transmission when it ideally
should be 50 per cent. This
results in an over-all maxi-
mum transmission value of
about 15 per cent for half-wave retardation which is more than enough
to activate a photoelectric cell or expose film.
The dark-to-light ratio depends upon many factors mentioned
above plus alignment and beam parallelism and color. However,
dark-to-light ratios as high as 300 to 1 have been obtained with the
application of 9 direct kilovolts. Alignment of the crystal with
* Made by the Pittsburgh Plate Glass Co., Pittsburgh, Pa.
** Made by the Polaroid Corporation, Cambridge, Mass.
so
/OO0 £O0O 3OOO
VOLTAGE APPLIED
A.C. VOLTS (P. TOP.)
Fig. 5 — Linearity of modulator using
"PN" crystal driven at 1000 cycles.
Direct-current parallel light with X/4
plate inserted.
1948
LIGHT MODULATION BY P-TYPE CRYSTALS
19
SIGNAL 'VOLTAGE—
-LAMP POWER
Fig. 6— Assembled view of experimental light modulator.
relation to the light beam is very critical and is done with the assist-
ance of a projection lens.
A complete system including the light source and lens system was
constructed in a housing 21/2 X 4 X 5Y2 inches. As a light source,
a 10-watt Western Union concentrated arc lamp2 was used and
r~ ANALYZED AND
APZ&TU8E
Fig. 7 — Exploded view of experimental light modulator.
20 GOTSCHALL
operated from two heavy-duty 45-volt B batteries in series. A
portable high-voltage supply was first used to strike the arc, and
later to apply voltage to the crystal unit. Since the lamp will remain
lit at reduced current, a stand-by switch reduces the current by 80
per cent except when actually in use. A photograph of this unit is
shown in Figs. 6 and 7. Although the point light source is conven-
ient in size, in the laboratory suitable parallelism was obtained from
an inexpensive Central Scientific light source having a 6-volt, 18-
ampere tungsten filament. Direct current from a bank of storage
batteries was supplied for frequency measurements while 60-cycle
alternating current was used for all other tests.
The unit as described has a flat response in excess of audio fequire-
ments, has an excellent dark-to-light ratio, is simple and sturdy,
has no moving parts, has a high impedance input, and has good
transmission characteristics for application to variable-density sound-
on-film recording. The adaptation is very simple as light emanating
from the aperture of the modulator as shown in Fig. 7 is parallel,
and a cylindrical lens could then focus this circular beam of modu-
lated light into a small slit on the sound track.
REFERENCES
(1) F. Pockels, "Electro-Optisches Verhalten Piezoelektrisher Kristalle,"
Goettingen Abhandlungen, vol. 39, 1894.
(2) W. D. Buckingham and C. R. Deiberf, "Characteristics and applications
of concentrated-arc lamps," /. Soc. Mot. Pict. Eng., vol. 47, pp. 376-400; Novem-
ber, 1946.
FORTY YEARS AGO
Talking Pictures in Rochester
This is the way the press agent describes the animated pictures: "To
hear the voice, to catch every sound and the intonation of every word
and see the people in life size moving before your eyes, and yet realize
there is not a single person there — it seems like some phantom of the
brain, an hallucination, and one is almost tempted to rush to the stage
and grapple with the ghostly actors as one is moved to cry out in the
vividness of a dream. Such is the wonderful spell that is cast over the
spectator on his first view of the marvelous talking, singing, dancing
moving pictures which Manager Parry of the National will introduce
for the first time in Rochester to-morrow afternoon."
— The Moving Picture World, May 30, 1908
Portable 16-Mm Sound Projector
BY H. H. WILSON
AMPRO CORPORATION, CHICAGO, ILLINOIS
Summary — Some of the problems encountered in the design and pro-
duction of a quality 16-mm sound projector designed to meet the require-
ments of the school, church, and industrial fields, are analyzed. Screen
illumination, sound reproduction, operational controls, film handling and
film protection, noise factors, styling, maintenance, and weight limitations
will be discussed. Manufacturing problems and their solution on the basis
of mass production will be considered.
THE PROJECTOR described was designed to meet the requirements
of the school, church, and industrial fields. First, the general
conditions under which the equipment will be used; second, the
consumers' requirements; and third, the actual design and manu-
facture of a projector to meet these requirements will be described.
The Application — The school, church, and certain phases of in
dustrial applications can be considered as a group so far as the design
of this type of equipment is concerned because of the technical simi-
larity of their application^ of the equipment. In most cases the
projector will be used to project sound films before small audiences
although occasionally it will be used before audiences of 500 to 600
people.
Classrooms usually have hard-surfaced walls and frequently cannot
be completely darkened. Industrial showings may be made in
many types of rooms ranging in size and acoustical properties from
the private office to the product display room.
Consumer Requirements — Since we have considered the general
conditions under which this equipment will be used, let us now turn
to a more detailed analysis of what will be required in the projector
in order to make it most useful as an audio-visual tool.
The distance from the last row of seats to the front of the typical
classroom is about 25 feet; therefore the screen width should be
approximately 4 feet (a 39- X 52-inch screen is commonly used).
This screen has an area of 14 square feet, consequently, at least
165 lumens will be required to provide adequate screen illumination.
* Presented April 25, 1947, at the SMPE Convention in Chicago.
JULY, 1948 JOURNAL OP THE SMPE VOLUME 51 21
22 WILSON July
a few watts of audio power would be required for the class-
room, but additional power will be required for larger rooms. In-
creased power can be obtained at a reasonable cost, so it was con-
sidered advisable to provide an amplifier having 10 to 15 watts'
output. Sound quality should be as good as space, weight, and
cost factors will permit.
The projector frequently will be operated by elementary or high-
school pupils; therefore, the controls should be as few and as acces-
sible as possible. It will be necessary to provide for both sound and
silent speeds, reverse operation, and the projection of a single frame.
Simplicity of threading is required as well as maximum film
protection.
Due to the excessive area of hard-surfaced Avails, the mechanism
must be very quiet in operation or the attention of the audience will
be distracted from the subject matter being presented.
Naturally, audio-visual equipment will be used only for those
applications where either it can do a better teaching job or a more
economical one than can be accomplished by other means; conse-
quently both the initial cost of the equipment and the maintenance
cost must be kept as low as is practicable. The design of the pro-
jector should be such as to provide maximum film protection because
first, the cost of procuring or replacing prints is a factor in the' cost
of a visual-education program, and second, the teaching film differs
from the entertainment film in that the contents of the teaching
film does not become obsolete for many years; consequently, film
wear is the principal cause for retirement of a print.
For ease of carrying and storage it is desirable that the unit be as
small and compact as possible and that neither the projector nor
speaker case, when packed, weigh more than 50 pounds.
General Design — In the design of the Premier-20 projector great
effort was made to meet the consumer's requirements. The com-
plete design cannot be said to be entirely new; those units, which
past experience indicated to be satisfactory, were retained in this
model, some units have been redesigned, and the methods of manu-
facturing other units have been revised in order to maintain closer
tolerances and thereby improve their performance.
Light Source and Optics — The projector was designed to use pro-
jection lamps of the medium prefocused base type. A 750-watt,
25-hour lamp is standard equipment although a 1000- watt lamp may
be used. The lamp socket is mounted in such a manner that the
1948
PORTABLE 16-MM SOUND PROJECTOR
23
lamp can be moved either vertically or horizontally. This type of
mounting provides two advantages, one, positive alignment of the
lamp filament with the optical axis can be obtained, and two, the
lamp position can be adjusted in order to correct for the shifting
of the filament which may result from use.
The reflector and condensing lenses are mounted on the right-
hand cover of the projector and the holder is properly positioned on
the cover at the time of assembly. The cover assembly and its optical
components may be removed for cleaning and replaced without dis-
turbing their alignment with the aperture.
Fig. 1 — Intermittent and shutter.
The projection-lens mount is solidly attached to the mechanism
head by means of a dowel and screws in order to maintain positive
alignment of the lens with the aperture plate. Coarse focusing of
the projection lens is obtained by sliding the lens in the holder^ fine
focusing, by rotation of the threaded lens barrel. The lens is locked
in position by means of a cam-and-spring assembly which clamps
the lens in position without producing any tendency of the lens to
shift in the holder. A coated //1. 6 lens of 2-inch focal length is
standard equipment. Lenses of other focal lengths also may be used.
Intermittent and Shutter — (Fig. 1) The intermittent unit consists
of three cams and a shuttle. Two of the cams make up the lateral
24 WILSON July
cam assembly (6) which causes the teeth of the claw to engage the
perforations and also holds the shuttle (2) and claw in the retracted
position while the shuttle makes an idle stroke. The purpose of the
idle stroke is to allow the vertical cam (8) to be revolved at 2880
revolutions per minute (at sound speed) thereby providing a com-
paratively short film-transport time without the use of a vertical
cam havdng such a small working angle that the life of the inter-
mittent might be seriously affected. The film-transport time is
4.68 milliseconds, therefore, approximately 1/B of the complete
projection cycle is required for film advance. Each lateral cam is
ground individually and then paired with the mating cam by selective
assembly. The vertical cam is rough-ground, then attached to the
camshaft; and finish-ground in order to maintain the throw within
close limits.
The action between the claw and the film perforations is quite
similar to that of a rack-and -pinion mechanism. The radius of the
pinion (distance from shuttle pivot (1) to film plane) is quite long
and the tooth form has been corrected in order to eliminate any
tendency of the claw teeth to drag across the edges of the perforations
during the entrance or retraction periods. This type of movement is
very quiet and produces very little film slap; it also has the advantage
that when used in conjunction with the gate mechanism, which will
be described, it will not damage the perforations if the lower loop
is lost, and the loop can be reset without stopping the projector.
The two-blade barrel-type shutter (4) revolves at 2160 revolutions
per minute at sound speed and provides three interceptions of the
light beam per frame. The shutter is attached to an adjusting disk
so that exact timing with the intermittent can be obtained. The
shutter is driven by means of a quill through which the safety-shutter
shaft passes. The safety shutter (5) consists of a single curved blade
having a radius shorter than the intercepting shutter and pivoted
on the axis of the intercepting shutter. The single hole pierced in
the safety shutter acts as an optical stop and maintains a low aperture
temperature when still pictures are being projected. The safety
shutter is actuated by means of a centrifugal clutch mechanism which
is built into the shutter-gear assembly.
Framer and Film Gate — Framing is accomplished by the vertical
movement of the framer plate which slides in a milled channel directly
behind the aperture plate (Fig. 2). In assembly the aperture plate
is aligned with the aperture in the framer plate and is attached to
1948 PORTABLE 16-MM SOUND PROJECTOR 25
the mechanism head by means of six screws. The right-hand edge of
the aperture plate is used for film-positioning and springs located on
the opposite edge of the plate hold the film against the right-hand
edge.
The pressure shoe (1) is attached to the pressure-shoe carrier
(2) which slides in ways milled in the carrier yoke (3). The carrier
yoke is pivoted near the front of the projection-lens mount and is
locked to the lens mount by means of a latch (4) located near the
rear of the lens mount. Moving the gate lever to its forward position
moves the pressure-shoe carrier and the pressure shoe away from the
Fig. 2 — Lens-mount assembly.
aperture plate and simultaneously opens the sprocket shoes. Press-
ing the gate latch toward the aperture releases the carrier yoke so
that the yoke and pressure shoe can be swung outward approximately
90 degrees. This makes both the pressure shoe and the aperture
plate readily accessible for cleaning.
Pressure is applied to the pressure shoe by means of two coil
springs of conical form. Control of the shoe pressure is obtained by
movement of the carrier stop (5). The contact surface of the shoe
is 3Vs inches in length; this results in a low pressure per unit of area;
consequently, if the lower loop is lost, both the film and the shoe are
26 WILSON July
pushed away from the aperture by the claw and the film is not
damaged. Lateral alignment of the pressure shoe is obtained by
moving the mounting plates (6) by means of which the shoe is at-
tached to the carrier.
Bearings, Lubrication, and Gearing — All shafts in the mechanism
are parallel; consequently, the bearing bosses can be reamed in a
single fixture thereby simplifying the manufacturing procedure and
maintaining a high degree of precision. "Oilite" bearings are used
for all shafts and are burnished to size after they are pressed into the
mechanism casting.
All moving parts of the mechanism receive oil from a central oil
well. Oil is distributed to various parts by means of oil tubes which
are sealed into the oil well and bearing bosses. The shuttle and
bearings that are not parts of the mechanism housing receive oil
through oil passages drilled axially in the shafts and holes drilled
radially in the shafts at the proper points.
The gearing is of the helical type. All transmission members
lie within four parallel planes which are closely spaced. This type
of design produces a very compact transmission assembly in which
end thrusts can be kept low, a high degree of precision can be main-
tained, and gear noise and vibration can be reduced to a minimum.
Motor and Drive Unit — The drive motor is of the universal type.
A Lee governor is used to control both the sound and silent speeds.
The speed-selector switch and reversing switch are attached to the
motor by short leads and are located on the amplifier-housing wn.ll
adjacent to the motor. The complete motor assembly can be re-
moved by disconnecting two leads from the terminal board on the
motor, removing the motor-retaining parts, the switch-retaining nuts
and the fan.
The main drive system consists of a flat, neoprene-impregnated,
fabric drive belt running on pulleys having synthetic rubber facings.
The driven pulley is mounted on an "Oilite" bearing and also serves
as the driving member of a single-plate automotive-type clutch. The
clutch mechanism is actuated by means of a knob on the right-hand
cover of the projector. The purpose of the clutch is to disengage
the projector mechanism from the motor in order to project single
frames.
Ventilation — A radial-blade centrifugal fan is mounted on the
motor-armature shaft. The fan .is 3Vs inches in diameter and oper-
ates at the motor speed of 5200 revolutions per minute when the
1948 PORTABLE 16-MM SOUND PROJECTOR 27
projector is running at sound speed. Air is drawn across the ampli-
fier and motor and into the fan intake thereby providing ventilation
for the amplifier and motor. A fan-scroll reversing vane is used to
maintain substantially the same air delivery regardless of the direction
of fan rotation (1).
Feed and Take-up Mechanisms — The feed-reel arm is permanently
attached to the mechanism housing and is pivoted in order that it
may be swung back over the top of the mechanism when the pro-
jector is placed in the case. When the projector is operated in
reverse a ball-type clutch contained within the spindle cup auto-
matically engages and the spindle is driven in a counterclockwise
direction by means of a spring belt. High-speed rewinding is ac-
complished by engaging a dog- type clutch located on the left-hand
side of the mechanism. A pulley formed as an integral part of the
clutch, by means of a spring belt, drives a pulley attached to the
left-hand end of the reel spindle. Two thousand feet of film can be
rewound in two minutes.
The take-up-reel arm also is pivoted in order that it may be swung
up in front of the mechanism housing when the projector is placed
in the case. The take-up spindle is connected to the spindle-drive
pulley through a ball-type clutch which automatically disconnects
the spindle from the pulley when the projector is operated in reverse.
For rewinding, the take-up belt is shifted to a loose pulley by means of
a manually operated belt shifter.
Soundhead—The soundhead is assembled as a separate unit and is
attached to both the mechanism housing and the amplifier housing.
Direct scanning is accomplished on a rotary sound drum mounted
on ball bearings and stabilized by means of a flywheel. Loop vibra-
tion is controlled by passing the film between the pressure and tension
rollers. The pressure-roller arm is pivoted and its travel controlled
in such a manner that it can be used to reset the lower loop without
stopping the projector.
Amplifier and Speaker — The three-stage amplifier consists of a
6J7 voltage amplifier, a 6J5 driver, and push-pull 6V6 output stage.
The output is 15 watts with less than 5 per cent distortion. The
standard output impedance is 15 ohms; a simple adapter (available
as an accessory) makes necessary connections to the 7.5-ohm tap
of the output transformer for operation of two speakers. A 6V6
is used as a radio-frequency oscillator to provide 6 volts at 1 ampere
for the exciter lamp.
28
WILSON
July
The volume control for sound on film controls the intensity of the
exciter lamp; a potentiometer in the grid circuit of the 6J7 controls
the volume of the microphone or phonograph. A slotted-shaft type
of control, located in the rear control panel, controls the polarizing
voltage applied to the phototube. The tone control is of the inverse-
feedback type.
Plug and receptacle connectors are located in the rear control panel
for connecting a converter or inverter if required. The standard
GATE RETRACTOR
LEVER (I)
PRESSURE
ROLLER (2)
Fig. 3
amplifier was designed to operate on 50 or 60 cycles, 105 to 125 volts
alternating current. A 25-cycle model is also available.
Removal of the main nameplate provides access to the tube sockets
for voltage checks. The amplifier may be removed by taking off the
bottom cover, disconnecting four leads from terminal boards, and
removing eight retaining screws.
A 12-inch permanent-magnet dynamic speaker is used. The
magnet weight is 4x/2 pounds, and the voice-coil impedance is 15
ohms. The speaker is mounted in a case 16 inches high, 16 inches
1948 PORTABLE 16-MM SOUND PROJECTOR 29
wide, and 93/4 inches deep. Fifty feet of speaker cable is standard
equipment.
Simplicity of Threading — Since the sprocket shoes open automati-
cally when the film gate is opened it is only necessary to move the
gate-retractor lever (1) to the forward position and move the pressure
roller (2) to the rear position in order to prepare the projector for
threading (Fig. 3).
Size and Weight — The above-described projector is packed in a
case 21 Y2 inches long, 16 inches high, and 93/4 inches wide. The
projector alone weighs 33V4 pounds. Packed in the case with
accessories its weight is 50 pounds. The speaker case which also
contains a 1600-foot reel and the power and speaker cables, weighs
26 pounds.
ACKNOWLEDGMENTS
The writer wishes to acknowledge the assistance of Messrs. A.
Shapiro, T. J. Morgan, A. S. Dearborn, and T. R. Neesley in the
preparation of this paper.
REFERENCE
(1) A detailed description of the design and operation of this unit is con-
tained in the paper "Design progress in an 8-mm projector," by Thomas J. Mor-
gan, J. Soc. Mot. Pict. Eng., vol. 49, pp. 453-463; November, 1947.
FORTY YEARS AGO
Moving Pictures in Schools
Moving pictures, as an aid to education, are now being utilized in the
National Preparatory School, in the City of Mexico, where a machine
of the latest pattern has been- installed. The pictures will illustrate sub-
jects in geography, history, physics, morals, and manual training.
Mexico is the second country to adopt the cinematograph as an educa-
tional factor, Germany having been the first.
— The Moving Picture World, May 16, 1908
Optical Problems in Large- Screen
Television*
BY I. G. MALOFF
RCA VICTOR DIVISION, RADIO CORPORATION OF AMERICA, CAMDEN,
NEW JERSEY
Summary— Optical problems in large-screen television are enumerated
and present-day solutions of these problems are discussed. Details of one
prewar and two postwar models of RCA large-screen projectors are described.
OPTICAL PROBLEMS involved in producing large-screen television
include: (1) choice of suitable source of picture; (2) choice of
suitable optical projection system; (3) choice of a suitable screen to
fit a particular auditorium; and (4) selection of proper ambient
lighting in the auditorium.
In the past a great number of solutions to the above problems have
been suggested, some of them tried, and some demonstrated. Among
these are various types of light valves, supersonic light cells, and
mirror and lens drums, also Mangin mirrors, refractive and reflective
optical systems, lens and mirror-type viewing screens, and many
others. Also the so-called "intermediate" or "zwischen" film method
has been proposed" and tried in the early thirties in Germany, aban-
doned, and now is again under development in this country at the
Radio Corporation of America and other laboratories. In the
intermediate film method a television picture appearing on the face
of a cathode-ray tube is photographed on motion picture film, quickly
processed, and reproduced through a regular film projector. Of
course there is a certain delay and a certain amount of instantaneity is
lost.
The llC A technical staff, while having investigated and tried most
of the proposed methods, has directed its large-screen television de-
velopment mostly along the lines of the combination of (1) high-
voltage cathode-ray tubes; (2) reflective or "Schmidt" . optics ;
and (3) directional viewing screens tailor-made to fit particular audi-
toriums; in other words, "instantaneous" systems.
* Presented October 23, 1947, at the SMPE Convention in New York.
30 JULY, 1948 JOURNAL OF THE SMPE VOLUME 51
LARGE-SCREEN TELEVISION
31
Before the war, RCA produced and publicly demonstrated in a
regular theater in New York City a .large-screen television picture on
a 15- X 20-foot screen. The equipment utilized a 7-inch projection
cathode-ray tube operating at 70 kilovolts. The optical system was
of the reflective or Schmidt type, using a 30-inch spherical mirror
and a 221/2-inch aspherical correcting lens. The general appearance
of this equipment is shown in Fig. 1.
Fig. 1 — Prewar television projector utilizing 31-inch mirror.
Since the war the RCA organization, basing its work on previous
experience, chose to continue developments along the lines of the
prewar prototype. Recent developments resulted in two types of
large-screen television systems. The first, the auditorium type,
utilizes a cathode-ray tube 7 inches in diameter operating at 50 kilo-
volts. The optical system consists of a 21-inch spherical mirror and
a 14V2-inch aspherical correcting lens. This system, having approxi-
mately a 6- X 8-foot screen, was publicly demonstrated at the con-
vention of the National Association of Broadcasters in Atlantic
City last September and is being demonstrated at this Convention.
32
MALOFF
July
Fig. 2 — Postwar television projector utilizing
21-inch mirror.
The second system, the theater type, makes use of a 15-inch cath-
ode-ray tube operating at 80 kilovolts. The optical system consists
Fig. 3 — Postwar television projector utilizing 42-inch mirror.
LARGE-SCREEN TELEVISION
33
Fig. 4 — Principle of reflective projection system.
of a 42-inch spherical mirror, and a 30-inch aspherical correcting
lens. At present it is the largest Schmidt-type system in the world,
since the 72-inch Schmidt telescope of Mount Wilson as yet is not in
operation. Two 42-inch RCA-Schmidt systems have been completed,
tested, and found to be up to expectations. These systems give
pictures of 18 X 24 feet in size and will be publicly demonstrated in
Fig. 5 — Machine for grinding 42-inch mirror.
34
MALOFF
July
the next few weeks. The general views of the two systems are shown
in Fig. 2 and Fig. 3.
In general, the optical problem of large-screen television is to pro-
duce on a given size screen a picture of sufficient high-light brightness,
resolution, and tone gradation, so that nothing contained in the
incoming signal is lost. The word "sufficient" has often been re-
Fig. 6 — Polishing 42-inch mirror.
placed by "maximum obtainable." It is a pleasure to state that with
the new projector the standard of the Society of Motion Picture En-
gineers of 7 to 14 foot-lamberts of high-light brightness has been met.
The general principle of reflective or Schmidt optics, as used in
projection, has been described in several publications.1' 2 In Fig. 4
the essential features of it are shown. Here a thin aspherical lens
placed at the center of curvature of a spherical mirror introduces an
1948
LARGE-SCREEN TELEVISION
35
amount of spherical aberration equal to that of the mirror but
opposite in sign.
The construction of the 42-inch mirrors (which was done at the
RTA Camden plant) involved the development of a special machine
shown in Fig. 5. This figure gives a general view of the grinder
having a 53-inch turntable. A 42-inch mirror blank is being lowered
into a cradle by an electric hoist operated by the author and an assist-
ant. The weight of the blank is 350 pounds. A view of the polisher
Fig. 7 — Aluminizing equipment and finished 42-inch mirror.
hi operation is shown in Fig. 6. With polishing completed the mirror
was aluminized in the tank shown in Fig. 7. A mirror already alumi-
nized can be seen at the left. Such large mirrors having relatively
short focal lengths can produce weird optical effects such as shown
in Fig. 8.
The construction of aspherical correcting lenses has been done
essentially by the methods described in cited publications. They
were made of glass, an inherently costly process. Eventually, how-
ever, these lenses may be molded from plastics just as in the case of
correcting lenses for home-projection-television receivers. These
lenses are being manufactured by the thousands at a cost of a few
36
MALOFF
dollars each. One of the advantages of plastic lenses is that they are
practically unbreakable.
Fig. 8 — Close-up of 42-inch mirror.
ACKNOWLEDGMENTS
The author acknowledges with thanks the able assistance of
Messrs. R. F. Leuschner'and M. Di Lorenzo in the construction of the
optical systems described.
REFERENCES
(1) I. G. Maloff and D. W. Epstein, "Reflective optics in projection tele-
vision," Electronics, vol. 17, pp. 98-105; December, 1944.
(2) D. W. Epstein and I. G. Maloff, "Projection television," /. Soc. Mot.
Pict. Eng., vol. 44, pp. 443-456; June, 1945.
Developments in Large-Screen
Television*
BY RALPH V. LITTLE, JR.
RCA VICTOR DIVISION, RADIO CORPORATION OF AMERICA, CAMDEN,
NEW JERSEY
Summary — An experimental large-screen program is being carried on to
determine the requirement for theater use. The governing factors: the
light source, the optical system, and the screen are discussed. Photographs
show equipment built for an experimental program.
THE HIGH DEGREE of excellence achieved in the production and
reproduction of sound motion pictures has placed this art above
all others in popularity and entertainment value. With high stand-
ards already established, large-screen television makes its debut
in the entertainment field, not as a competitor, but as an ally, an
ally \vith mutual interest and, we believe, vast possibilities.
Large-screen television is still in the experimental stage but con-
siderable progress has been made during the past two years. Ex-
perimental equipment has been built and demonstrated with ex-
cellent results. This equipment, which will now be described, will
form the basis for determining specific requirements and future design.
There are three major elements in a large-screen projection system
which are combined to produce the over-all result viewed on the
screen. The first is the source of light and picture, the projectioy
kinescope, which translates the video information into a pattern of
light on the tube face by the scanning process; second, the optical
system, the function of which is to collect the light rays from the
face of the kinescope and direct them to the screen, properly focused,
as an image of desired size; and third is the screen from which the
picture is viewed. These three elements must each be designed for
their best efficiencies in a co-ordinated system, in order to make
possible the best in picture quality and brightness. We shall examine
each element of such a system in order to understand their limita-
tions and discuss the problems common to each.
Presented October 23, 1947, at the SMPE Convention in New York.
JULY, 1948 JOURNAL OF THE SMPE VOLUME 51 37
38
LITTLE
July
The projection kinescope is similar to that used in the direct-
viewing table-model television receiver and requires the same video
amplifier, deflection, and high-voltage functions as required for the
receiver; the differences are those of magnitude in order to obtain
the very bright picture required. We st?e in Fig. 1, the diagram of a
typical projection kinescope tube; the electron gun here emits a
stream of electrons which are focused by an electron lens and ac-
celerated by the high-anode potential, which is 50 kilovolts for the
7-inch tube, to the screen where it causes the phosphor coating of the
face to emit light in accordance with the density of the electron beam
which is controlled by the video signal. The deflection yoke sur-
rounds the neck of the tube and is provided with suitable currents
SECOND ANODE CONTACT
CON "POL GMD
1 — Cross section of projection kinescope.
to make the scanning raster necessary to form a picture-image pattern.
Television has no satisfactory method of using a supplementary high-
intensity light source, such as a carbon arc, which might be controlled
At video frequencies; so high-light brightness is a function of phosphor
efficiencies. The method then of obtaining high-light output from a
projection kinescope, as compared with a home-receiver kinescope,
is to provide high accelerating voltage on the tube. This permits the
phosphor to be bombarded with electrons of high velocity which
produces more light while the current remains low.
The relative voltages used on typical kinescopes are : for the 10-inch
home receiver, 9000 volts, while 50,000 volts is used for the 7-inch
projection kinescope, and for the larger 12-inch and 15-inch projection
kinescopes, 80,000 volts accelerating potential is used. Although
the high voltages are used, the current requirements are small and
are generated in safe radio-frequency power supplies which have very
1948
LARGE-SCREEN TELEVISION
39
lo\v stored energy. Future developments will be centered on the
improvement of the phosphors and the electron optics of the kine-
scope. The typical projection kinescope high-light brightness is
about 3000 foot-lamberts.
The second requirement of the projection system is the lens and
since Mr. Maloff has discussed this subject in detail, only a brief sum-
mary will be made. You are familiar with typical motion picture
lens which may nominally be an //2.0 with a 5-inch focal length and
uses elements 21/z inches in diameter. A refractive lens of this type
for a kinescope of 7-inch diameter would require a lens equal to the
face diameter to gather sufficient light from the picture and in a
Fig. 2 — Block diagram of simplifier.
practical design could not exceed a speed of //1.5. The //1.5 lens
would have a gain of 1.75 over the conventional, thus leaving much
to be desired in efficiency. Television engineers soon realized that
the lens system was one place where more gain might be realized.
The reflective optical system of the Radio Corporation of America
was devised and gave effective speeds to //0.6 with the 42-inch mirror
system which has been completed for use with a 15-inch kinescope
to produce an 18- X 24-foot picture. The relative speed as com-
pared with the //2.0 system would then be eleven times the gain in
light, a truly remarkable increase. Reflective optical systems are
characterized by short focal lengths which are necessary in order to
produce the fastest lens speed or smallest / number. The projector
for the 6- X 8-foot screen uses a 21-inch mirror, a 14-inch correction
40 LITTLE July
lens, a 7-inch projection kinescope, and has a throw distance of 15
feet. The projector for the 18- X 24-foot picture uses a 42-inch mir-
ror, a 36-inch lens, a 15-inch kinescope, and requires a throw distance
of 40 feet.
The screen then forms the final link in our over-all system and
affords another opportunity to improve the gain in picture brightness.
Consideration has been given to the various types of screens available.
Experience obtained in our experimental work indicates that some
form of directional-viewing screen gives the best compromise in high-
Fig. 3— Rear quarter view of 6- X 8-foot screen
projector.
light brightness and viewing field. The beaded screen has given good
results in this respect and is used with our demonstrations, but further
development is in order. Developments in screens promise to permit
greater gains and it is expected that gains as high as 3 may be obtain-
able to the advantage of the over-all system.
The Society of Motion Picture Engineers recommends an optimum
screen brightness of 7 to 14 foot-lamberts. Television is approaching
this requirement of the theater; developments under way will pro-
vide a picture equivalent to the recommended standard of brightness.
In Fig. 2 is a diagram of the essential elements of the large-screen
projector, the projection kinescope, the RCA reflective optical system,
1948 LARGE-SCREEN TELEVISION 41
the video amplifier, which modulates the kinescope to produce the
light and dark areas corresponding to the television-camera image,
the deflection circuits to produce the synchronized scanning raster,
the radio-frequency oscillator and rectifier to supply the 50 kilovolts,
and the necessary power supplies.
The projector for the 6- X 8-foot screen is shown in Fig. 3 and is a
completely self-contained unit which operates from a video signal
Fig. 4— Control panel.
supplied from a coaxial cable or a television receiver. The unit
measures 53 X 32 X 60 inches and weighs 1200 pounds and requires
18 amperes at 115 volts 60-cycle alternating current or approximately
2 kilowatts. The cabinet houses the optical barrel containing the
RCA reflective optical system, the 7-inch projection kinescope, and a
50-kilovolt high-voltage rectifier unit. Aligned on each side of the
cabinet are the video amplifier, deflection units for vertical and hori-
zontal scanning voltages, the radio-frequency oscillator which drives
the high-voltage rectifier, and the necessary regulated power supplies.
The control panel (Fig. 4) is located at the rear and has all operational
42
LITTLE
July
controls accessible. It is noteworthy that most of these controls
would not be made available on a commercial design, but would be
on the associated chassis units since frequent adjustment is not
required. The required controls would be found consisting of the
contrast, brightness, optical focus, and electrical focus. A photo-
graph of the optical barrel is presented in Fig. 5 to show the mounting
of the mirror, the bottom tank of which houses the high- voltage supply
Fig. 5 — Optical barrel showing mounting of
mirror.
and effectively shields it from radiating. This unit must be very
rigid to hold the optical system in precise alignment.
An interior view of the left side (Fig. 6) shows the orderly arrange-
ment of the electrical equipment used. The left-hand panel is the
synchronizing circuit panel, the major unit on the right is the de-
flection chassis which drives the deflection yoke on the neck of the
tube. The other chassis are direct-current power supplies.
The other side (Fig. 7) shows the video amplifier on the right.
The radio-frequency oscillator below it drives the high-voltage supply.
1948 LARGE-SCREEN TELEVISION 43
i
Fig. 6 — Left side of 6- X 8-foot projector showing chassis.
m
Fig. 7 — Right side of 6- X 8-foot projector showing chassis.
44 LITTLE July
On the left we have a power supply for the units just described and a
fuse panel for the protection circuits.
A unique design feature of the projector is the high-voltage power
supply, Fig. 8. As previously mentioned, it is driven from a power
oscillator operating at 20 kilocycles. Energy at this frequency is fed
to a step-up transformer which develops 25 kilovolts peak-to-peak
alternating current, Avhich is then doubled in the special rectifier
1 1
Fig. 8 — 50-kilovolt rectifier.
circuit to furnish the 50 kilovolts required. The figure shows the
actual supply designed for this equipment, and its unique features
include the self-contained filament transformers built into the socket
of each -tube.
A similar high-voltage rectifier is shown in Fig. 9, but with a
quadrupler to supply the 80 kilovolts required for the projector using
the 18- X 24-foot screen. Tne radio-frequency voltage generated
in the coil is impressed on each rectifier tube. These tubes for direct
current are in series so that four times the voltage is realized across
1948
LARGE-SCREEN TELEVISION
45
the output resistor and kinescope. These unique power supplies
employ a circuit developed by O. H. Schade of the RCA Victor Divi1-
sion, Harrison, and a mechanical arrangement devised by Fred G.
Albin, recently of Camden Engineering and now at the Hollywood
office.
Fig. 9— 80-kilovolt rectifier.
Fig. 10 shows the projector built for the 42-inch optical system and
is the largest unit of its kind ever attempted. It will throw an 18-
X 24-foot picture on the screen from the face of a 15-inch kinescope.
The mechanical and electrical problems were of great magnitude as
46
LITTLE
was to be expected. Any resemblance of this unit to another nation-
ally advertised product is purely coincidental. This unit represents
the accumulation of many years of effort on the part of many engi-
neers in the RCA Victor Division located at Camden, Harrison, and
Lancaster, and the RCA Laboratories of Princeton. Credit is due
them for their contribution to the over-all project as well as to F. G.
Albin who co-ordinated the design of the equipment described hero.
Fig. 10 — Large theater projector for 18- X 24-foot screen
using 42-inch mirror and 15-inch kinescope.
ACKNOWLEDGMENTS
In closing, I wish particularly to thank Mr. Earl Sponable of
Twentieth Century-Fox for making available the equipment which
we are demonstrating this evening. This equipment, together witli
the larger projector, form a part of a co-operative venture in theater
television which the Radio Corporation is making with Warner
Brothers and Twentieth Century-Fox as recently announced to the
press.
DISCUSSION
Note: Chairman Larsen requested that discussion on the two
preceding papers be held until after the Large-Screen Television
Demonstration. Therefore, the following discussion concerns both
papers.
MR. J. I. CRABTREE : Does the aspherical lens need cleaning very often? If so,
being plastic, do you not impair the optical properties in cleaning it?
MR. I. G. MALOFF: Not especially in the large-screen projector. If it is cleaned
with the antistatic compound, we do not have to clean it very often. In the home-
projection receiver, we make a hood that protects it from collecting dust. The
normal cleaning with a soft cloth does not spoil it, because we use the hardest
available plastic.
DR. E. W. KELLOGG : I should imagine the audience might be interested if Mr.
Maloff would give us the figures of the optical speed or effective / number that is
attainable in the Schmidt system, and also the field size in degrees so that they
might compare it with what is possible' with camera lenses or projection lenses.
MR. MALOFF: The / number as such loses its meaning at, I would say, about
//1. 4. So the best figure is the efficiency of the lens. By defining efficiency as the
ratio of the number of lumens delivered to the screen, to the lumens produced by
the tube, we arrive at a figure between 30 and 40 per cent with the reflective optics,
with a very large magnification. The figures for the//2 lens, for the same magni-
fication, run close to 4 or 5 per cent. I cannot tell you the field angles, offhand.
CAPTAIN A. G. D. WEST: How many lumens do you project in this projector
and how many do you expect to project in the new 42-inch mirror projector?
MR. MALOFF: Suppose we turn the answers to your question around. The
prewar projector gave us high-light brightness somewhere between 1 and 2 foot-
lamberts. That brightness was found in a number of theaters around the country
by the Committee on Screen Brightness of the Society of Motion Picture Engi-
neers, which report was published around 1936 or 1937. The size of our screen in
the New Yorker theater in 1940 was 15 X 20 feet. The screen gain was 2: By
gain of a screen, we mean the ratio of brightness normal to the screen, to the inci-
dent illumination ; that is, how many foot-lamberts' brightness are obtained for 1
1'oot-candle illumination or 1 lumen per square foot.
What we are doing now is this: We went to. a 15-inch tube and increased the
area of the emitter four times, roughly. Then we increased the voltage some-
what, and we used an aluminum-backed screen. Before the war we also used an
aluminum-backed screen, but it was of an amorphous type, which did not have a
mirror reflecting the light that was going back toward the gun. It was an absorber
of that light. We put it on only to maintain the luminescent material at the
second-anode voltage. There is such a phenomenon known as ''sticking" of the
luminescent material. That means it does not quite reach the voltage put on the
second anode, never rising above the "sticking potential."
This gives us a gain of approximately eight times. There are a few other small
gains, for example, in higher light output from the phosphor.
This would give us, with a perfectly diffusing screen having a gain of 1, a net
gain of four times the prewar screen brightness.
We do not^pTopose to use, with theater television, screens that illuminate ceiling
47
48 DISCUSSION July
and floor. We want the light to fall where the people are. Therefore, we propose
to build screens that will throw the light only where the audience is. We have
done this to a certain extent with the home projection receivers. We hope to do
so with the theater projection receivers. The screen in the home projection re-
ceiver has a gain of 6. I doubt if we can get that kind of gain for a theater, but we
ought to be able to get a gain between 3 and 4, and we are working hard at it.
CAPTAIN WEST: As to the answer about how many lumens, I believe the pre-
war, that is, 1941, projector gave about 200 or 300 lumens.
MB. MALOFF: On the New Yorker installation we ran close to 500 microamperes,
average beam current. That gives us a peak, say, of 2 milliamperes. At 70 kilo-
volts, it is 140 watts. Now, 140 watts at 2 candle power per watt gives you 280
candle power. Assuming that we emit from the face of the cathode-ray tube ac-
cording to Lambert's law, we multiply that by 3. That gives us somewhere in the
neighborhood of 600 or 700 lumens. The same arithmetic applies again now, ex-
cept that we are getting between 4 and 5 candle power per watt from the lumines-
cent screen, this new aluminized screen. The new screen has mirror aluminum;
it is not amorphous aluminum. We coat the screen with organic material. We
fill all the little holes, the little depressions in the luminescent material, and the
coating leaves a shiny surface. Then we evaporate aluminum on that shiny sur-
face, and by baking and evacuating with pumps, we exhaust all the organic ma-
terial. So we have left a shiny aluminum surface over the luminescent material.
In this way we more than double the efficiency of the luminescent material.
CAPTAIN WEST: That is our practice, of course. However, I think we expect
to get 1000 lumens from our 40-inch projector. You remember I mentioned about
Dr. Zworkin's being in Paris. After he returned from that visit, I heard that he
was achieving 12,000 lumens and 40 foot-lamberts on that size screen. So that
rather depressed Professor Fisher, who was working on the other system.
I should have liked to bring a projector here to compare with the one used in the
large-screen television demonstration, but it was not possible. However, in Lon-
don we are projecting on a larger screen. It is very difficult to make a comparison,
but, first of all, I should say my impressions of the picture are exceedingly good.
My first impression is an impression of the color. It is a better color than we are
having at the moment for a larger screen. I like the blue white and the bluish
white in the home receiver.
Second, there seems to be good interlacing, which we do not have at home.
The contrast range was very good. Was the center part film transmission?
MR. LITTLE: I believe certain portions of that program were from film.
CAPTAIN WEST: The transmission of the British Broadcasting Corporation
suffers from a good deal of shading. Generally speaking, I am very favorably
impressed. I think it is a very good picture, indeed.
DR. K. PESTRECOV : I think we need a committee on standardization of screen
terminology. Recently we heard a report on screen-brightness measurements.
At that time the ratio of foot-lamberts to foot-candles on the screen was called
efficiency of the screen. As I remember, the efficiency would run from about
50 per cent to about 90 per cent. I believe Mr. Maloff prefers the term "gain."
If it is really the same quantity, then a gain of 2 would correspond, as was defined
a day or two ago, to an efficiency of 200 per cent. That is the first question.
Second, if television engineers can design a screen, or hope to design a screen,
1948 LARGE-SCREEN BRIGHTNESS 49
with a gain of 2, or an efficiency of 2'00 per cent, the screen also should be suitable
for general motion picture projection. Perhaps, it will be a real advance so far
as obtaining brighter pictures in general, because for theater television you are
not inclined to use one screen and another screen for motion pictures.
MR. MALOFF: The first question is on efficiency and gain. There has not
been any standardization in that field, so far, except among ourselves. Television
engineers have a clear distinction between the two terms.
The one term, efficiency of the screen, is simply determined by putting a
photometer on the other side and determining how much light at all angles gets
through that screen. I am mostly talking about transmission screens, but the same
applies to reflective screens. However, when we talk about gain, we measure
this by comparing the light with what would come from a perfectly diffusing
screen according to Lambert's law. We concentrate the returned light into a
narrow pyramid, more or less. Horizontally it is wide. What we are trying to
do is to get 60 degrees width from the screen, completely uniformly, with a sharp
cutoff beyond that. Vertically, we are trying to get a 20-degree spread.
Theoretically, you can get close to a gain of 12 if you collect light that went to
various places before. However, you can never get efficiency of the screen of
over 100 per cent, because you absorb some light.
Before very long, we shall all have to get together and straighten out this
matter, at least among us television engineers. Then we might have either con-
version factors to translate to motion picture practice, or perhaps we can adopt
the same terminology and the same definitions. Such is the case of resolution
right now. When we talk about resolution in television, we say "500 lines."
When an optical man looks at it, he will say it is only "250 lines," because we
count every line, white and black, whereas he is counting only the black lines.
As to the second question, whether such screens as we are using now in the
television industry are suitable for motion picture projection, we have various
reflective screens. One concern is putting in a reflective screen with a gain of
just about 6. It was demonstrated in New York and in other cities. Screens
with a gain of 12 were demonstrated. That particular screen, however, has too
narrow a vertical angle, and they have put in one with a lower value of gain.
So, all screens, both of the translucent type and the reflective type, could be
used in theater projection of motion pictures. However, in some of the theaters
the angles are so wide that you cannot use a directional screen; tha"t is, where
there is a second and third balcony. That is why we could not use a very high-
gain screen in the New Yorker theater before the war.
In an auditorium like this one we should use a curved type of screen. There
is an exhibit right outside the door of a curved screen, which definitely can give
you a different directional distribution, vertically and horizontally. However,
the problem is not so acute for the motion picture engineers as it is for television
engineers. You start with such high values of light that you can waste it. If
you can put a few extra seats here and there, you do so. The light goes down,
but you still hold within your standard; that is, if it drops from 10 to 7 foot-
lamberts, you do not mind that very much if you have a few extra seats.
We barely reach sufficient brightness. We cannot waste it, and we might have
to waste a, few seats in the theaters in order to show theater television.
DR. PESTRECOV: Thank you, Mr. Maloff. I purposely meant to provoke
50 DISCUSSION July
the discussion, because I have had discussions with Mr. Maloff before,
and I more or less knew what he was talking about when he mentioned
the term "gain." However, I believe that perhaps many people here do not
know that term. As a matter of fact, I did not know it about two years ago,
and many people in the optical industry and the motion picture industry still do
not know that term. The point is that you get gain when you narrow the angle of
reflection; is that correct?
MB. MALOFF: That is correct.
DR. PESTRECOV: So, perhaps, it really might be better to employ that term,
make it standard, and then we shall not talk so much about the efficiency of the
screen. As to the committee that reported on the brightness of the screen, what
the committee actually measured at that time was the brightness of the screen in a
certain direction. They did not measure the total light reflected, 1 believe. Per-
haps some time in the future we can introduce that term and really talk about gain
of the screen. In this particular case probably it does not have much meaning,
but when we start to talk about television screens in the motion picture industry,
then we have to use that term, and I think it should be more or less explained.
Maybe if you explain it when you write this paper for publication, I think it would
be very useful; at least, we shall have a definite and authoritative reference.
MR. LITTLE: Captain West might be able to answer a question on screens.
In his paper he mentioned the lenticular screen which they were using in England,
which gives a gain of 3. He also showed slides showing the distribution through-
out the house, and that screen gave excellent coverage. Captain West, would you
care to give us some explanation of the type of screen that you use?
CAPTAIN WEST: We are not using that screen at the moment. The one I re-
ferred to there under the heading of lenticular, which makes a large gain, was first
demonstrated by Dr. Muller in Berlin at an exhibition. It consisted of a series of
mirrors like a cat's eyes you see when you are driving on the road at night, looking
out at you. It was very carefully arranged. I tested it very clearly on this tele-
vision projector, which was similar to the one I illustrated on Tuesday night, of
the pipe-shaped tube of the lens. I must say that if you were sitting at the end
of the row and got out of your seat into^the gangway, the picture vanished; and
when you went back into your seat, it appeared again. The idea is that all the
light was reflected back into the seats, and not all over the theater. That amount
of light was conserved. It was always very expensive to make and had to be
tailored for every theater.
There is an intermediate type of lenticular screen which we have been using.
We have not used it so much as we wanted to, because of the shortage of metal and
other materials. It is similar to what was described by Mr. Maloff for the home,
except that it is a reflecting screen instead of a transparent one. I think that it
corresponds to the screen in my diagram, which I referred to as a "stippled- metal
screen." That gives a reflection factor right down the center of about 2l/2to 3 times.
One more question. I suppose you are getting a good show for one particular
reason, in that it is all coming down from that little tower up there, is that right?
MR. LITTLE: Yes, that is correct. The program came from the Empire State
Building, but I might add that the proximity really causes a great deal of diffi-
culty. Tonight, about twenty minutes after seven, I would have said tonight's
show was not going on, because we had a great deal of interference which appar-
ently was cross modulation in the receiver between frequency-modulation and
1948 LARGE-SCREEN BRIGHTNESS 51
television signals. We were very much discouraged about putting on the show
Some of the difficulty you did see in the picture during the show was caused in the
receiver and not in the projection equipment. It was unforeseen, I assure you,
and normally that type of interference is not present.
CAPTAIN WEST: That should please you very much, because we find that when
everything goes wrong, before a demonstration, it is usually all right.
There is one other thing I would like to mention which helps very much, in
presentation of television on the screen. That is the sound. We are doing ex-
periments in theaters now which in the last eight years have had a little disturb-
ance around them, not fit for the public to enter; in fact, all the seats had been
taken out, parts of the roof were down, and that sort of thing. The sound is very
bad in our television presentation. The sound was very good here tonight. I
am absolutely certain that if you get good sound you get a much better picture.
MR. BEN SCHLANGER: If we can call this theater television a baby, I wonder if
we are not making this baby run before it creeps. From what I can see, you are
limited to a screen characteristic which throws the light back in a very narrow
angle. Is it not better to take theater television and put it into shelters which are
made for theater television? You are overstepping your bonds in trying to show
television in existing theaters, where 50 per cent of the location will be inadequate
and will not show the job off as well as it could be. The way you light the interior
of the motion picture theater, there would be too much light in competition to the
amount of light that you can get off the screen with television.
MR. LITTLE : I hasten to point out the remark during the paper, that the pres-
ent equipment is the basis of an experimental program. We do not know what
form television theaters will take, or what form television programming will take.
We do not know what form television equipment, as such, will take. We are just
embarking on this field, and we hope to get the answers. We certainly do not
know them and as manufacturers we do not propose to give the answers to the in-
dustry. We are trying to help the industry find the answers. You gentlemen are
part of the industry, and we expect the answers to come from you. We cannot
give them. We can build you the equipment if you can tell us what you want.
MR. SCHLANGER: All'the demonstrations and all the tests have been in existing
theaters. It has never been given a really fair trial in a room that would really
show it off the way it should.
MR. LITTLE: Maybe those limitations are inherent, but we do not believe so.
We are certainly looking for an answer.
MR. R. B. AUSTRIAN: Mr. Maloff, in describing the screens and assigning the
values to them which you did, I understood you to make a statement that there
was no reason why they could not be used interchangeably for regular motion
picture projection. Do I assume that the screens you worked with were non-
porous; and if you had to perforate them for proper sound presentation as to be
acceptable today, would that not change some of your reflection characteristics?
MR. MALOFF: Yes, very definitely, if you use directional screens. By the
time you perforate it, you probably lose part of the effect that you gain. Maybe
your sound effect will not be as good as you would like to have it. I do not think
a perforated screen is an important item, but the industry probably thinks differ-
ently. If we perforate a directional screen, depending upon the percentage of
the holes to the rest of the screen, we shall lose that much more light.
Theater Engineering Conference
Ventilating and Air Conditioning
•
Motion Picture Theater
Air Conditioning*
BY DWIGHT D. KIMBALL
CONSULTING ENGINEER, NEW YORK, NEW YORK
Summary — Air conditioning as now defined involves four basic ele-
ments. These are a definite controlled temperature, the maintenance of the
desirable relative humidity, a predetermined rate of air movement, and air
filtration. In a properly installed air-conditioning system these elements
can be predetermined and independently controlled, but temperature,
humidity, and air movement must be controlled with a definite relation one to
the other.
BASICALLY modern air conditioning is but the ultimate develop-
ment of ordinary ventilation. For example, the ventilating
system installed in Carnegie Hall more than 50 years ago was made an
air-conditioning system by adding to the existing ventilating system
the necessary cooling equipment without changes in the fan equip-
ment or duct system.
The generally accepted standard of summer theater atmospheric
conditions is 80 degrees on a day of 95 degrees outside temperature,
50 per cent relative humidity, and approximately 12 to 15 feet per
minute air movement within the seating area. An excessive relative
humidity will more than anything else lessen the sensation of comfort
of the occupants of the theater.
The conditions as above stated should extend to every seat in the
theater and should not be merely the average over the entire theater
seating area. Herein lies the importance of correct air distribution.
Seventy degrees inside of the theater with 95 degrees outside (a
difference of 25 degrees) definitely may cause a serious shock to some
people, especially the aged and those not in the best of health, and is a
source of discomfiture to most people. Physiological tests have
shown that a 12-degree difference between inside and outside temper-
atures is the desirable maximum.
* Presented October 24, 1947, at the SMPE Convention in New York.
52 JULY, 1948 JOURNAL OF THE SMPE VOLUME 51
AIR CONDITIONING 53
Frequently it is asked when the air-conditioning industry will pro-
duce new type of equipment that will substantially reduce the cost of
theater air conditioning. There is little or nothing of this nature in
sight at this tune.
The next advance to be anticipated is a means of independently
providing for dehumidification. At present dehumidification is
effected by first lowering the temperature of the air supply, to a point
lower than that actually required for the cooling of the theater in
order to extract the necessary amount of moisture from the air sup-
plied to the theater, then raising the temperature of the air upon leav-
ing the cooling coils by the use of the auditorium return air by-pass or
reheat steam coils to raise the temperature of the air to a point at
which it may be admitted to the theater.
Independent dehumidifying equipment is now available but it is
expensive, space-consuming, and requires high-pressure steam or gas
for the regeneration (or drying) of the moisture-absorbing material.
The major features of a modern air-conditioning system are air
supply and its distribution ; cooling equipment, such as refrigeration
or well water; and treatment of secondary spaces, including the pro-
jection room, lounges, toilets, foyer, and lobby.
AIR SUPPLY
The volume of air supply, as well as the capacity of the cooling
plant, is determined by calculating the heat load. within the theater
including transmission of heat from without through walls, floor, and
roof; heat and moisture given off by the theater occupants, including
standees; and electric load.
This calculation is so made as to determine separately the sensible
heat and the latent heat. With total internal sensible heat load
determined, a temperature differential between the desired room tem-
perature and the temperature of the air admitted to the theater is
selected, this depending upon the rate of air movement desired, the
height of the air-supply diffusers above the floor, and the distribution
of the air diffusers. This temperature diffusion difference may vary
from 12 to 18 degrees.
Sometimes the resulting determination of the amount of the air
supply to the auditorium will be found to be equivalent to 18 to 20
cubic feet of air per minute per occupant. But this is not the final
answer because not 100 per cent of the air supplied can be applied
directly tp the benefit of the theater's occupants. Some of the air
54 KIMBALL July
supplied may short-circuit to the return air and exhaust outlets, some
is lost through doors and otherwise, and more air than thus deter-
mined is required to assure its distribution to all portions of the seating
area.
Over a period of years it has been proved definitely that the air
supply to the theater proper should be not less than 24 cubic feet of
air per minute per occupant. For a de-luxe installation 30 cubic feet of
air per minute per occupant may well be provided.
In determining the capacity of the main supply fan, the air filters,
and heating and cooling coils, there must be added to the air to be
supplied to the theater proper the amount of air which must be sup-
plied to the lounges, foyer, lobby, and other parts of the theater.
DISTRIBUTION OF AIR SUPPLY
Quite as important as is the volume of air supplied to the theater
is its distribution therein. This is determined by the number, size,
and location of the air-supply diffusers which should be so determined
as to, shall we say, spray the air over the entire occupied area of the
theater, including the standee space, thus serving every person in the
theater. In this matter the area under the balcony must not be neg-
lected. Invariably the air should be admitted to the theater from the
ceiling, and in the case of balcony houses also from the balcony soffit
for the seating area under the balcony and the standee area.
The ceiling cannot be designed for the exclusive benefit of the ceil-
ing diffusers but the utilitarian value of the ceiling diffusers and the
inherent limitations upon the location thereof must be taken into
consideration in the design of the ceiling. Ceiling diffusers may take
various forms. There are the old-style plaster plaques, lacking the
desirable diffusing and induction effects, and not especially sightly.
Now more generally used are Anemostats, Aerofuse, or similar dif-
fusing outlets. This type of air-supply outlet has the very impor-
tant merit of producing a secondary air movement.
In determining the arrangement of the air distribution in a balcony
theater the theater should be considered as divided into three zones.
1. That portion of the orchestra floor in front of the balcony rail.
- 2 The orchestra floor area under the balcony.
3. The balcony.
The air distribution should be so designed as to provide a direct
supply of air and a withdrawal of return air in direct proportion to the
number of occupants in each of these zones. (See Fig. l.\
1948
AIR CONDITIONING
55
However small may be the seating and standee area beneath the
balcony it is essential that a direct air supply and return air outlets be
provided therefor.
The temperature of the air supplied to the theater during the winter
at no time should exceed 80 to 90 degrees. With a higher temper-
ature of the air-supply, stratification of the theater temperature will
become very serious and promote wasteful operation. There was
once found an air supply of 100 degrees temperature, a ceiling temper-
ature of 95 degrees, and a floor temperature of 54 degrees. All steam
was shut off from the air-heating coils for 30*minutes while 40-degree
RETURN
FROM BALCONY
Am SUPPLY To-
LINDER BALCONY
AREA
GRILLE
Fig. 1 — Diagrammatic longitudinal section through theater.
air was blown into the theater to eliminate the high-temperature air at
the ceiling, then 75-degree air was blown in for 30 minutes and the
floor temperature became 70 degrees.
RETURN AIR
It is the universal practice to design theater air-eonditioning sys-
tems upon the assumption that 75 per cent of the amount of air sup-
plied to the theater auditorium is returned to the fan room for recon-
ditioning and return (with 25 per cent of outside air) to the theater.
The supply of 25 per cent of outside air to the theater is desirable
and adequate for the elimination of odors in the theater.
A general distribution of the air-supply diffusers is essential to a
56 KIMBALL July
proper air distribution and equally essential is the general distribution
of the return-air outlets within the theater.
A general distribution of the air-supply diffusers, with the return-
air outlets limited to a restricted area, or vice versa, invariably proves
unsatisfactory, causing excessive temperatures and a sensation often
described as "dead" in those portions of the seating areas which are
neglected by either supply-air or return-air outlets.
The most satisfactory and generally used means of withdrawing re-
turn air from the theater is through standard mushroom outlets
located under the seats-, usually communicating with return-air tun-
nels under the floor. In the balcony either mushrooms or riser
grilles may be used, these communicating with the balcony void from
which the return air is drawn back to the fan room.
FAN-ROOM EQUIPMENT
The main fan room is the heart of the air-conditioning system.
The fan-room equipment includes the main air-supply fan, its driving
motor, heating coils, cooling coils, air niters, sheet-metal casings, and
piping connections.
The capacity of all of this equipment is determined by the total air
supply required for the theater and accessory spaces.
Main Supply Fan
Assuming a theater designed for 1000 persons including standees,
and an air supply of 24 cubic feet of air per minute per occupant, we
have a base figure of 24,000 cubic feet of air per minute. To this
must be added an amount of air equal to that which must be supplied
to lounges, foyer, lobby, and other parts of the theater, from which
spaces it is not customary to withdraw return air, some air to replace
that which is lost through duct seams, and a small amount of air to
assure a mild excess air pressure within the theater to counteract in-
filtration through the doors.
The finally determined capacity of the blower will be found to be
about 30,000 to 32,000 cubic feet of air per minute. The total re-
sistance, or static pressure, will be found to be I1/* to \l/% inches. A
motor of 15 horsepower will be required to drive the blower.
To avoid noisy operation the blower should be so selected as to
operate with an outlet velocity of about 1200 feet per minute, or under
some favorable conditions up to 1400 feet per minute.
1948 Am CONDITIONING 57
The pulleys of the V-belt motor and fan drive should be of the vari-
pitch type to make possible any desired correction of the blower speed.
The V-belts should be 25 per cent greater in capacity than that of the
motor for greater durability.
HEATING AND COOLING COILS
The heating and cooling coils should be selected upon the basis of an
air velocity of approximately 500 feet per minute through the coils,
which is the generally accepted standard.
* Heating coils using low-pressure steam will usually be two rows of
tubes in depth, assuming a proper utilization of the return air. In
general practice the number of rows of tubes required in cooling coils
when using well water of 50 degrees or less in temperature is four.
With well water above 50 degrees and up to 54 degrees six rows of
tubes usually are used. If the well water is somewhat above 54 de-
grees eight rows of tubes are recommended. Additional rows of tubes
serve no useful purpose in theater work.
With a well-water supply of 54 degrees and above the relative
humidity within the theater on hot and humid days, or with capacity
audiences, cannot be maintained at a level of comfort, say at 50 per
cent or at the very maximum 55 per cent, and one of the main purposes
of the air-conditioning installation is then defeated. Under such
conditions the installation of supplementary refrigeration equipment
is desirable. In fact, a condition frequently is found where refriger-
ation equipment may best be used to provide for all of the cooling.
Direct expansion cooling coils through which the refrigerant, usu-
ally Freon, is circulated through the cooling coils directly from the
refrigerating plant will be discussed hereinafter.
AIR FILTRATION
Among the forms of air filters most generally used are dry-cell
filters of the so-called "throw-away" type and similar types of wire-
mesh air filters, the cells of which are dipped in an oil preparation and
when dirty can be washed, redipped, and reused.
The problem of maintaining the efficiency of these air filters is a
troublesome one in either case. After long experience the author
adopted the use of the ' 'throw-away" type of filter because in most
cases the engineer or janitor may throw away and replace the dirt-
loaded cells but would much more often neglect to remove, wash, im-
merse in oil, and replace the washable cells.
58 KIMBALL July
At best the operator generally regards the air filters as a nuisance.
One engineer even removed the air-filter cells completely so that they
would not become dirty.
The above types of air filters if allowed to remain in use until they
become excessively loaded with dust and dirt may reduce the supply
of air to the theater as much as 50 per cent.
Also available are automatically operating oil-immersed air filters
in which the air-filter cells are attached to traveling motor-operated
chains designed automatically to immerse the filter cells in a bath of
oil contained in a liquid tank at the base of the filter unit, thus elimi-
nating the frequent replacement or washing of the filter cells.
A much more efficient and desirable method of air filtration, but
also much more expensive and space-consuming, is found in the elec-
trostatic type of air-cleaning equipment, such as the Westinghouse
Precipitron and the Raytheon Precipitator.
Wherever space may be made available and the considerably
greater cost of this electronic air cleaning is acceptable to the theater
owner its use is highly desirable. All replacement or washing of filter
cells is then eliminated.
Assuming a theater air-conditioning system using 30,000 cubic feet
of air per minute, the comparative cost of the different types of air
filters mentioned above will be found to be about as follows:
Dry-cell throw-away type $ 500.00
Air-mat type 1400.00
Oil-dipped reusable type
2 inches thick 870.00
4 inches thick 1170.00
Automatic oil-immersed 2400.00
Electronic or electrostatic . 5100.00
Clean air is desirable not only from a health standpoint but also as
a protection of the theater furnishings and decorations.
FAN-ROOM LAYOUT
As has been said the heart of the air-conditioning system is found in
the main fan-room equipment. For the purpose of this discussion a
plan of a typical fan room is shown in Fig. 2. Quite generally this
space is found available at an elevated level at the front of the theater
at one side of the screen platform.
In the interest of minimum total cost of installation of the heating
1948
AIR CONDITIONING
59
and air-conditioning systems it is desirable that the boiler room and
refrigerating machinery (or well-water pump) be located in the same
general area but in the basement to shorten the interconnecting pip-
ing lines.
This arrangement of the fan-room equipment is designed to give
access to each piece of equipment, including both sides of the heating
and cooling coils. This last is important for the inspection and clean-
ing of the coil tubes and for repairs thereto when necessary.
INTAKE
WEATHER PROOF LouvREb
4 iNbECT SCREEN-
RETURN AIR FLOOR GHATINO
OVER RETURN AIR Due T FROM
AUDITORIUM. DAMPER BELOW
Two PART DAMPER-
MINIMUM AREA AUTOMATICALLY
OPENS i CLObt-3 WITH FAN MOIOK
OPERATION
MAXIMUM AREA FOR Ust DURING
IN BETWEEN 5tAbONb MANUALLY
OPERATED
ACCESS DOOR
RtTimN AIR BY PASS
WITH AUTOMA
DAMPER IN DUCT BELOW
SUPPLY BLOWER
MOUNTED ON VIBRATION
ELIMINATORS
AUDITORIUM WALL-
Fig. 2 — Typical plan of fan room.
The outside inlet should be made of such a size as to admit an
amount of air at 500 feet per minute velocity equal to the capacity of
the blower so that 100 per cent of outside air may be utilized during
the "in-between" . seasons when neither heating nor cooling is
required.
Back of the weatherproof louvers there is provided a -louver-type
damper made in two sections. The smaller section is sized to admit
the amount of outside air required during the cooling and heating
60 KIMBALL July
season. This so-called "minimum outside air damper" is automati-
cally operated so as to open when the fan motor is started and close
when this motor stops.
The larger section of the outside-air-intake damper is manually
operated and is to be opened only during the period of fan operation
in the so-called "in-between" seasons. The use of this so-called
"maximum outside-air damper" will be found helpful during spring
and fall seasons to provide ventilation and some cooling effect in
mild weather with a resulting saving in operation costs.
A portion of the return air is admitted to the apparatus chamber
between the outside-air intake and the air niters. The remainder of
the return air is admitted into the fan chamber beyond the cooling
coils to mix with the air which has passed through these coils to serve
as a reheat medium to raise the temperature of the air coming off the
cooling coils, which has been lowered to the point required for the
elimination of moisture resulting in an air temperature too low for
admission to the theater. A by-pass damper provided at this point
should be controlled automatically to maintain the correct tempera-
ture of the theater air supply.
An important feature of this fan-room layout is the placing of the
blower in such a way that the air may freely enter the fan inlet, espe-
cially important in the case of a double-inlet fan to prevent an unbal-
anced or a noisy fan. Free access to all fan-room equipment is essen-
tial for inspection, cleaning, and repairs.
Years ago it was a frequent practice to include in theater installa-
tions an exhaust fan to remove the excess of outside air used. This
practice has long since been discontinued as serving no useful purpose
and involving an unnecessary cost.
AIR STERILIZATION
It was suggested that this paper should cover the subject of air
sterilization but inasmuch as that problem will be described separately
in other papers, the remarks thereon will be brief here.
Two methods of sterilization have been brought to the author's
attention : ultraviolet radiation and glycol-vapor treatment.
That method which not only treats the air passing through the air-
conditioning chamber but also carries the germicidal agent on into
the theater and to the theater occupants would appear to be the more
effective.
Glycol equipment, which is now available, requires but a small space,
1948 AIR CONDITIONING 61
is relatively inexpensive to install and operate, is odorless, nontoxic,
and is carried directly into the theater in the supply air stream.
The ultraviolet ray equipment involves the placing of lamp units
in the theater walls and a considerable amount of wiring. The author
is not at this time persuaded of the merits of this system for theater work.
So far as is known, no theater up to date has been provided with
germicidal-air-treatment equipment but this subject does seem to be
worthy of very serious consideration.
PROJECTION-ROOM VENTILATION
The ventilation of the projection room is governed by the rules of
the National Board of Fire Underwriters and in New York City by
certain provisions of the City's Building Laws. Some of these require-
ments are confusing and indefinite and are not specific in an engineer-
ing sense. Moreover, the situation is further confused by the varying
interpretations given by different inspectors or engineers representing
these Bureaus.
Fig. 3 shows the author's standard plan of projection-room ventila-
tion which appears to be acceptable to all authorities having jurisdic-
tion thereon.
The essential features are as follows: (a) a motor-driven exhaust
fan with ducts to remove the heat from the projection machines, ex-
hausting 50 to 100 cubic feet per minute from each machine; (6) a
second exhaust fan to ventilate the projection room, the rewind room,
the motor-generator room, and the toilet room. Twelve to 20
changes of air per hour may desirably be exhausted from these rooms
because of the heat released therein. A single fan may not be used to
serve both of these purposes, nor may either of these fans serve to
ventilate any other spaces. The discharge ducts from these two fans
should be carried directly to out-of-doors. No code specifically states
this but the author has had objections filed to the carrying of these
discharge ducts through other theater spaces; (c) two outside air
ducts directly from the roof (or upper part of the side walls) to serve
the machine rooms and the rewind room (New York City Code) ; (d)
film cabinets of a capacity of 50 pounds or more of film are required
to be provided with a vent directly to the outside of the building.
The rules of the National Board of Fire Underwriters specifically
state that the " Ventilation of the projection-room area shall not be
connected in any way with the ventilating or air-conditioning system
serving other portions of the building."1' 2
62
KlMBALL
LOBBY
July
The real problem in treating the lobby is that of counteracting the
wind blowing in through the outer doors which is sometimes found
to be blowing on into the theater resulting in annoyance to those per-
sons occupying the rear theater seats.
The use of radiators in the lobby is generally unsatisfactory because
a large amount of lobby radiation is required for which it is generally
1948
AIR CONDITIONING
63
found difficult to find space, the radiators are objectionable in appear-
ance even when recessed and grilled, and there will be times when the
cold air will blow through the lobby into the theater, and the radiation
is not immediately responsive to sudden demands for heat, such as
when the outer doors are opened.
A more efficient means of heating the lobby, and with the same
equipment supplying conditioned air thereto, is found in the extension
to the lobby of a branch duct from the theater air-conditioning supply
duct and interposing in this branch duct a booster air-heating coil and
FAN MOUNTED ON
PLATFORM PLATFORM
SUSPENDED FROM
SLABABOVE.
BOOSTER FAN-
5ERWCE
ROOM
(FROM CONDITIONED AlR
\ SUPPLY Ducr
•-BOOSTER COIL
E- F- G i H ARE TYPE
R-ZOAoiTAiR OUTLETS,
Fig. 4 — Lobby heating and air conditioning.
a small booster fan which will supply air to the lobby at a pressure
sufficient to counteract the pressure of the air blowing in through the
outer doors. By this method the lobby is evenly heated during the
winter and air-conditioned during the summer.
Fig. 4 shows a typical lobby treatment. The solid lines show the
preferred method; dotted lines show an alternate method.
Lobby temperature control, during the heating season, may be
accomplished by a thermostat operating a modulating steam-control
valve at the booster coil or face and by-pass dampers at the booster
heating coil.
64 KIMBALL July
LOUNGES AND TOILET ROOMS
All toilet rooms, whether or not having windows, should be provided
with mechanical exhaust ventilation. Windows are generally kept
closed during the cold weather but when opened they may serve to
admit air which will blow the toilet-room odors into the theater.
Inasmuch as the toilet rooms are generally entered from the
lounges the general practice is to supply conditioned air to the lounges
and to draw air from the lounges through louvers in the toilet-room
doors into the toilet room from which it is exhausted as above stated.
AUTOMATIC CONTROLS
The operation of a theater-heating and air-conditioning system
lacking automatic-control equipment involves an excessive amount of
attention on the part of the operator and a degree of skill which is
rarely available.
Automatic temperature and humidity control greatly lessens the
amount of attention required of the operator, automatically compen-
sates for varying occupancy and weather conditions, and lessens oper-
ating costs. Automatic control devices are delicate and should be
the subject of an annual checkup by the manufacturer of the equip-
ment used.
A simple system of automatic control would include the following :
For Winter Operation
A master duct-type modulating thermostat having its bulb located
in the stream of the outside-air supply. A submaster modulating
room thermostat located in the theater to control automatic valves
inserted in the steam connections to the air-heating coils. A modu-
lating duct-type thermostat in the main fan discharge duct designed
to prevent the supply of air to the theater at too low a temperature.
For Summer Operation
A proper relative humidity in the theater is of prime importance
because it more directly affects the comfort of the theater patrons than
does temperature. To control the relative humidity an automatic
modulating humidostat is located in the return-air duct near the fan
room to so control the operation of a modulating three-way water
valve inserted hi the cold-water mains connecting to the cooling coils
as to pass through the tubes of the air-cooling coils that amount of
1948 AIK CONDITIONING 05
cold water required to lower ' the temperature of the air passing
through the cooling coils to the temperature required to extract by
condensation on the coils all excess of moisture beyond that required
to maintain the desired relative humidity in the theater.
This may reduce the temperature of the theater air supply below
the temperature at which it may be discharged into the theater.
Then a portion of the return air is carried around the cooling coils,
thus utilizing the heat picked up in the theater for the raising of the
temperature of the air supplied to the theater. An automatic damper
is provided in this return-air by-pass controlled by the submaster
modulating room thermostat described under " winter operation."
A manually operated summer-winter switch is provided to transfer
the effect of the submaster room thermostat from the control of the
steam valves at the air-heating coils during the heating season to the
control of the automatic return-air by-pass- damper during the cooling
season. This switch also holds the three-way water valve closed dur-
ing the heating season and the steam valves closed during the cooling
season.
A simple means of opening and closing the minimum outside-air-
intake damper when the main supply fan is started and stopped is
desirable. This consists of a damper motor applied to the minimum
intake damper and an electric-pneumatic switch wired into the motor
circuit with a damper motor. *
A lobby thermostat of the modulating type with an automatic
valve in the steam connection to the lobby booster heating coil is
essential to the control of the lobby temperature. The usual chrono-
therm is assumed to be provided for the control of the heating-boiler
oil burner.
This system may be elaborated upon at will. It will include a small
air compressor and other incidental equipment. An electrical auto-
matic control system may be installed to accomplish the same ends.
It will be noted that reference is made above to the use of modulat-
ing thermostats. These may be used for heating controls only where
a vacuum steam-heating system is installed. Such a system is highly
desirable because it provides for the only successful method of control-
ling the temperature of the air supplied to the theater by making
possible the modulation of the supply of steam to the air-heating coils
in exact proportion to the demand for heat.
A vacuum steam-heating system is more economical in operation
than is a gravity-heating system and a substantial portion of the cost
66 KIMBALL July
of the vacuum-pump installation may be saved through a reduction in
steam- and return-piping sizes and a saving in the cost of pipe covering.
PRIMARY COOLING MEDIUM
Well Water
Where well water is obtainable at temperatures of 50 to 54 degrees
it will serve every purpose required of air conditioning and the well
and pump installation will cost but 20 to 40 per cent of the cost of the
required refrigeration installation.
If the well water is found to be above 54 degrees in temperature
either supplementary refrigeration equipment must be provided or a
higher relative humidity will occur in the theater with a sacrifice in
comfort conditions.
Refrigeration
The capacity of the necessary refrigeration installation is deter-
mined by generally understood if rather involved calculations. An
ample capacity in the refrigeration plant is a very comforting and
reassuring thing.
It has been the author's experience, including a survey of 100
theaters for one chain, that a general checking figure on the tonnage
of refrigeration required will run about as'follows:
600-seat theater .12 occupants per ton
1000-seat theater 14 occupants per ton
1500-seat theater 15 occupants per ton
1800-seat theater 16 occupants per ton
Larger theater 17 occupants per ton
Generally, Freon is the refrigerant used ina ir-conditioning work.
The refrigeration plant will include one or two compressors as re-
quired; motor; means of condensing, i. e., the cooling of the com-
pressed refrigerant; water chiller, a chilled-water circulating pump,
and a chilled-water circulating piping system, if a water-circulating
system is required or .used; and starters, automatic controls, and
other parts.
A single compressor may be installed if the tonnage is 50 tons or less
or if a centrifugal compressor is used. Interruptions of service are
very rare. Reciprocating compressors are very generally used,
especially in theaters of 1500 seats or less; a centrifugal compressor
may well be used for larger theaters.
1948 AIR CONDITIONING 67
The motor horsepower per ton of refrigeration will vary with the
actual operating conditions and motors of ample capacity which will
not become heated under maximum load conditions should be se-
lected. Generally speaking the motor capacity should be not lesS
than one and one tenth to one and two tenths horsepower per ton, or
slightly less sometimes with very cold condensing water.
Condensing, or cooling of the compressed refrigerant coming from
the compressor, may be accomplished by means of a standard shell
and tube condenser supplied with water from the street mains, from a
cooling tower, or by an evaporative condenser.
If the street water may be used the least cost of installation results.
In New York City the supply of city water for this purpose is limited
to 2,680,000 gallons of water per year and this is not sufficient for even
a 600-seat theater. Then a cooling tower must be used to provide
condensing water for a shell and tube condenser, or an evaporative
.condenser may be used instead of the shell and tube condenser and
cooling tower except in some cases an evaporative condenser cannot be
installed within restrictions of the New York City refrigeration code.
Where the indirect, or water-circulating system must be used, as in
New York City, a shell and tube water chiller, similar in construction
to the condenser, must be used with a chilled-water circulating pump.
Automatic high- and low-temperature controls, a low- temperature
compressor cutout to prevent freezing of the water in the chiller, and
usually compressor capacity controls are applied to the water chiller.
The well-water or chilled-water piping carrying the cooling water
to and from the cooling coils must be insulated with molded cork,
preferably 2 inches thick. An alternate to this indirect, or water-
circulating, system is the direct expansion' system in which the
compressed refrigerant after being condensed or cooled in the con-
denser is conveyed directly to and from the cooling coils.
This direct expansion system is less expensive to install because it
eliminates the water chiller and water-circulating pump and reduces
insulation costs. However, it cannot be installed in places of public
assembly in New York and certain other cities.
Where the main fan room with the air-cooling coils is located at a
considerable elevation above the compressors the direct expansion
system operates at a disadvantage, and there is a further disadvantage
in the direct expansion system in that it does not lend itself to a satis-
factory method of the automatic control of theater temperature and
humidity.
68 . KIMBALL k July
NOISE
Sometimes a small matter can be the cause of very objectionable noise
and still be difficult to locate. The most frequently found causes of
noise might be listed in the following order of frequency of occurrence :
1. Lightly constructed duct work having poorly made seams and
joints, insufficient bracing permitting the vibration of the duct sheets,
bad turns, loose dampers, and edges of metal projecting into the ducts.
2. Excessive air velocities through the ducts or through the supply
grilles and diffusers. Maximum duct velocities should not exceed
1200 feet per minute at the fan, or possibly 1400 feet under favorable
conditions, the air velocity being gradually reduced as branches are
taken from the main duct to about 600 feet in the individual branches.
3. The blower is by no means the most frequent source of noise
although usually first suspected. An excessive fan speed or an exces-
sive air velocity through the fan outlet will cause noise. The correct
blower speed will depend upon the size, type, and characteristics of the
blower. The air-outlet velocity from the blower should not be over
1400 feet per minute and 1200 or 1300 feet is better. No sharp bend
in the duct wrork should be made near the blower outlet. The blower
must be reasonably free from vibration. It must be so placed that
the distance from the blower inlet ring to the enclosing wall should be
approximately equal to the diameter of the fan inlet. This is espe-
cially important in the case of a double-inlet, double- width blower. A
canvas sleeve must be provided in duct connections to blowers so
that blower vibration wilfnot cause a vibration of the ducts. A loose
belt can cause much noise in itself and cause the blower to become
noisy. A blower wheel out of balance or with loose shaft collars may
cause pounding.
4. A motor is sometimes noisy, perhaps because of faulty setting
or even due to its construction. The motor may have to be replaced
by a quiet motor.
If the noise is caused by air travel in or out of the fan the simplest
cure may be the installation of acoustical duct lining for a distance of
about twenty feet.
MAINTENANCE
The installation of an air-conditioning system involves a very sub-
stantial investment of money and includes much equipment. Most
assuredly it is worthy of the utmost care. Neglect of this equipment
temporarily may save money but ultimately it will involve an expense
largely exceeding the amount saved in failing to maintain the
1948 AIK CONDITIONING 69
equipment properly. Moreover failure to maintain the equipment
properly in good condition inevitably will increase the operating costs.
COSTS
The costs of an air-conditioning installation will vary so widely with
the type of installation involved that no general figures may be given.
Prices have been changing so rapidly that it is difficult to keep up to
date on them. In 1942 bids were received on a theater air-condition-
ing installation which was not made* because of war restrictions.
Last month new bids were received upon the same plans and specifica-
tions. The new bids exceeded those of 1942 by 78 per cent. At least
half of this increase appears to have occurred within the last two years.
GUARANTEES
Do not buy air-conditioning systems upon the basis of guaranteed
theater conditions. Proving the facts in court is almost impossible.
The standard guarantee of results is that a condition of 80 degrees
and 50 per cent relative humidity shall be maintained within the thea-
ter when an outside temperature of 95 degrees dry bulb and 75 degrees
wet bulb prevails, with a 100 per cent occupancy of the theater, and
while using a predetermined volume of outside air. In thirty years
the author has yet to see these conditions simultaneously prevailing
for such a test.
CONCLUSION
In the foregoing it has been possible to give but a bare outline of a
theater air-conditioning system. Actually such an installation in-
volves a multitude of details, all of which are important, many of
which are highly technical, and all of which must be correlated care-
fully. To prove successful it must be a compound of theory and prac-
tical experience, with the latter predominating.
It has been said that he who serves as his own lawyer has a fool for a
client. Looking back over many years of experience in theater air
conditioning it seems that much the same thing may be said of one
who assumes to act as his own engineer in the installation of an air-
conditioning system. Competent advice will save the purchaser a
great deal of worry .
REFERENCES
(1) A. C. Dowries, "Gases from carbon arcs," /. Soc. Mot. PicL Eng., vol. 35,
pp. 32-47; July, 1940.
(2) P. Drinker and J. R. Snell, "Ventilation of motion picture booths," /. Ind.
Hyg. and Tox., vol. 20, p. 321; April, 1938.
Theater Engineering Conference
Ventilating and Air Conditioning
•
Air Purification by Glycol Vapor
BY J. W. SPISELMAN
Am PURIFICATION SERVICE, INC., NEWARK, NEW JERSEY
Summary — The germicidal activity of glycol vapor on air-suspended bac-
teria and viruses has been clearly demonstrated. The most suitable com-
pound thus far found for use in such air disinfection is triethylene glycol.
When dispersed in air as a true vapor in exceedingly small amounts it is
highly germicidal for pathogens of the respiratory tract, including influenza
virus. It is nontoxic, nonirritating, odorless, tasteless, invisible, and inex-
pensive. Satisfactory devices for the vaporization and regulation of bacteri-
cidal concentrations of glycol are now made and are in use.
DURING THE LAST fifty years great strides have been made in
protecting our people from the spread of disease through the
food we eat and the liquids we drink. Our water and milk supplies
are guarded with unceasing vigilance ; our foods and our drugs must
meet tests of purity laid down in a rigid code.
But what about the air we breathe?
A man eats about two pounds of food a day. He drinks, say, a
quart of liquids a day. But he breathes about 80 pounds of air per
day. That air contains germs, dusts, smoke, organic matter, pollen,
and noxious gases.
Man has through evolution and environment built up his ability
to withstand the onslaughts of these air-borne enemies; but that line
of defense is vulnerable, and is often beaten down. In fact, it has
been established that better than 50 per cent of all industrial sickness
absences are due to respiratory diseases.1 And these respiratory
diseases are due primarily to air-borne infection!
And the questions are quite properly asked — "What can be done
about it? What can be done to sanitize, or disinfect, or sterilize the
air which is being continuously contaminated with bacteria and
viruses dispersed into it as people around us cough, sneeze, or talk?
* Presented October 24, 1947, at the SMPE Convention in New York.
70 JULY, 1948 JOURNAL OF THE SMPE VOLUME 51
AIR PURIFICATION BY GLYCOL VAPOR 71
What can be done to protect us when we are congregrated in places of
public assembly, theaters, schools, and industry?"
This paper presents a picture showing how far one branch of air
sterilization has gone toward answering those questions.
Early in 1941, Drs. Robertson, Bigg, Miller, and Baker of the
University of Chicago announced2 that they had succeeded in steri-
lizing air by using certain glycols. They used propylene glycol, one
of a family of glycols, and obtained almost instantaneous sterilization
of the air in a test chamber infected with high concentrations of staph-
ylococcus and streptococcus germs.
In the latter part of 1941 just before the United States entered the
war, while selective-service camps were expanding, the Surgeon
General of the Army formed the Commission on Prevention of Air-
borne Infection and Control of Influenza and appointed the same
Dr. Robertson as chairman. Because of the extensive experience
The Research Corporation had acquired during the preceding years in
glycol air conditioning, the Corporation was asked to assign the
group, which now comprises Air Purification Service, Inc., to assist
Dr. Robertson's Commission in various phases of further develop-
ment and engineering.
The development since that early period has been extensive. At
the very beginning, it became apparent that it was not the fine mist
of glycol that was the active agent but it was the true gaseous vapor
of the glycol.3' 4 For the vast majority of cases triethylene glycol
rather than propylene glycol was more economical and efficient.5
Equipment was developed for the true vaporization of the glycols,6
and extensive field tests were performed.7
It also became apparent that the amount of triethylene glycol
necessary for such sterilization was fantastically minute. One cubic
centimeter of glycol liquid would sterilize 250 to 400 million cubic
centimeters of air. Visualizing it in another manner, all the air in a
building covering a full city block and six stories high could be steri-
lized by one pint of triethylene glycol. An air-conditioning system
using 15,000 cubic feet per minute of fresh air would require only
five ounces of triethylene glycol per hour for sterilization. The
actual quantity of vapor in the air approaches that of our most rare
gases; it is less than Vioo the quantity of neon in the air we breathe.
What are these glycols? Chemically they are kin to the alcohols;
physically they look quite like glycerin. Triethelyne glycol (TEG)
is a mildly viscous, colorless, and odorless liquid which, when vaporized,
74
SPISELMAN
July
It should be remembered that
at the stage of our present knowl-
edge, glycol vapor definitely had
been shown to be a preventive
medium, but not a cure, for air-
borne infection. Yet, recent work
in the aerosol field has indicated
that the glycols may also have
a therapeutic value by forming
antibiotics with the blood
serum.12' 13 This work undoubt-
edly will be followed up for
further knowledge on this point.
Glycol is vaporized through
the medium of heat and a pre-
heated stream of air. Triethyl-
ene glycol for instance, cannot
Fig. 2— Typical installation of
vaporizer.
Fig. 1 — Vaporizer.
be simply boiled to vaporize it,
since it boils at 550 degrees
Fahrenheit, but starts to decom-
pose chemically at approximately
300 degrees Fahrenheit. The
equipment for such vaporizing
is not complex; present equip-
ment available is foolproof, eco-
nomical of operation, and small
for the job it can do. The
present full-size glycol vaporizer
has a base of 15 X 15 inches
and is but 18 inches high, and
will treat 20,000 cubic feet per
minute of fresh air.
Its application to an air-con-
ditioning or ventilating system
is generally mechanically simple
after the proper engineering con-
siderations have been made. A
small quantity of air, roughly
about 20 cubic feet per minute,
is continuously preheated within
the vaporizer to the proper
1948
AIR PURIFICATION BY GLYCOL VAPOR
75
vaporizing temperature under close thermostatic control, and is then
induced over evaporating surfaces containing heated TEG. This
carrier stream, warm and laden with glycol vapor, is then injected
into the main stream of air in the ventilating system. The glycol
disseminates within that main stream and is thus carried through
the distributing ducts throughout the ventilated area. .Complete
permeation of every nook and cranny of the treated space is thus
obtained.
Fig. 3 — Installation within a plenum
chamber.
Fig. 4 — Installation across a heating
coil.
Figs. 1 to 5 show the glycol-vaporizing unit and typical examples of
applications to air systems.
Fig. 1 shows the vaporizer itself. A small stream of air enters the
top left-hand inlet, is preheated under thermostatic control, flows over
evaporating surfaces and an indicating thermometer, thence out
through the outlet port on the top right hand.. The output is based
on the temperature of the leaving air and is a logarithmic relationship.
The bottom of the vaporizer is a tank section holding five gallons of
glycol. There are no moving parts in the unit.
76
SPISELMAN
July
Fig. 2 shows a typical installation. The vaporizer is set conven-
iently close to the suction side of the main blower of a ventilating or
air-conditioning system. The small stream of air is induced through
the glycol-vaporizing unit by the suction of the main blower, and the
glycol-vapor-laden output stream flows directly to the blower to be
distributed throughout the system.
Fig. 5 — Size of unit.
Fig. 3 shows another type of installation in which the glycol vapor-
izer is placed within a plenum chamber on the suction side of a blower.
Here the difference in pressure between the outside of the plenum
and inside is used to induce the small stream of air through the
vaporizer.
Fig. 4 shows the unit operating by obtaining the necessary small
pressure drop by means of the pressure drop of a coil in the venti-
lating system. The pressure drop across the heating coil in this case
is sufficient to force the air through the vaporizer to insure proper
operation.
1948 AIR PURIFICATION BY GLYCOL VAPOR 77
Fig. 5 shows an installation similar to Fig. 2 and indicates the
relative size of the unit in the far background as compared to other
component parts of a ventilating system.
In all cases, the glycol-vaporizing unit is interlocked with the blower
motor, so that when the blower is shut down, the heating elements
of the vaporizer are also shut off and the vaporizer becomes inactive.
The cost of TEG is relatively low; a 1000-seat theater will use
about eight cents worth of TEG per hour of operation. The electrical
power required for the vaporizer is less than 750 watts, and would be
lost in the total electrical bill of such a house. An illustration of such
an installation in a large theater is in the Rivoli Theater, New York
City. With the close co-operation of Mr. G. P. Skouras and Mr.
Montague Salmon, managing director, operation costs and public re-
action are being carefully noted.
One of the results noted in a glycolized atomsphere is the feeling of
"freshness" which seems to pervade the air. The mustiness often
associated with air-conditioning systems is eliminated. Our ex-
planation for that is that the molds and similar organisms which will
collect in duct work and give off odors in their life processes are killed
by the glycol vapors, and the odor disappears.14 We believe also
that there is some control over odors generated in an occupied area.
In closing, it should be borne in mind that when a vapor is used
for the control of air-borne bacteria and viruses, it pervades the entire
atmosphere. It is where you want it when you want it at the source
of such contaminating organisms, the mouths and nostrils of all of us.
REFERENCES
(1) William M. Gafafer, "Manual of Industrial Hygiene," United States
Public Health Service, 1943.
(2) O. H. Robertson, E. Bigg, B. F. Miller, and T. Baker, "Sterilization of
air by certain glycols employed as aerosols," Science, vol. 93, no. 2409, pp. 213-
214; February, 1941.
(3) O. H. Robertson, "Sterilization of air with glycol vapors," The Harvey
Lectures Series, vol. 38, pp. 227-254; 1942-1943.
(4) O. H. Robertson, B. F. Miller, and E. Bigg, "Method of Sterilizing Air,"
United States Patent No. 2,333,124, November 2, 1943.
(5) M. Hamburger, T. T. Puck, and O. H. Robertson, "The effect of tri-
ethylene glycol vapor on air-borne beta hemolytic streptococci in hospital wards 1,"
J. Infect. Dis., vol. 76, p. 208; May, 1945.
(6) S. C. Coey and J. W. Spiselman, "Space Sterilization," United States
Patent No. 2,344,536, March, 1944.
78 SPISELMAN
(7) O. H. Robertson, "New methods for the control of airborne infection
with especial reference to the use of triethylene glycol vapor," Wisconsin Med. J.,
vol. 46, p. 311; March, 1947.
(8) O. H. Robertson, "Disinfection of air bygermicidal vapors and mists,"
Amer. J. Pub. Health, vol. 36, pp. 390-391; March, 1946.
(9) O. H. Robertson and T. T. Puck, "The lethal effect of triethylene glycol
vapor on air-borne bacteria and influenza virus," Science, vol. 97, p. 142; Feb-
ruary, 1943.
(10) T. M. Harris and J. Stokes, Jr., "Airborne cross infection in the case of
the common cold — a further clinical study of the use of glycol vapor for air sterili-
zation," Amer. J. Med. Sri., vol. 206, pp. 631-636; April, 1943.
(11) T. M. Harris and J. Stokes, Jr., "Summary of a three-year study of the
clinical applications of the disinfection of air by glycol vapor," Amer. J. Med. Sci.,
vol. 209, p. 152; February, 1945.
(12) S. J. Prigal, T. H. McGavack, F. D. Speer, and O. R. Harris, "Aerosol
penicillin," /. Amer. Med. Ass., vol. 134, pp. 938; May, 1947.
(13) S. J. Prigal, T. H. McGavack, and M. Bell, "The effect of propylene
glycol on the antibiotic activity of human serum," Amer. J. Med., vol. 3, p. 185;
August, 1947.
(14) M. Mellody and E. Bigg, "The fungicidal action of triethylene glycol,"
/. Infec. Dis., vol. 79, pp. 45-56; July, 1946.
FORTY YEARS AGO
How Moving Pictures Originated
A paragraph is going the rounds of the press giving the following ver-
sion of the origin of moving pictures :
Sir John Herschel after dinner in 1826 asked his friend, Charles Bab-
bage, how he would show both sides of a shilling at once. Babbage re-
plied by taking a shilling from his pocket and holding it to a mirror.
This did not satisfy Sir John, who set the shilling spinning upon the
dinner table, at the same time pointing out that if the eye is placed on a
level with the rotating coin both sides can be seen at once. Babbage was
so struck by the experiment that the next day he described it to a friend,
Dr. Fitton, who immediately made a working model. On one side of a
disk was drawn a bird, on the other side an empty bird cage; when the
card was revolved on a silk thread the bird appeared to be in the cage.
This model showed the persistence of vision upon which all moving
pictures depend for their effect. The eye retains the image of the ob-
ject seen for a fraction of a second after the object has been removed.
This model was called the thaumatrope. Next came the zoetrope, or
wheel of life. A cylinder was perforated with a series of slots and
within the cylinder was placed a band of drawings of dancing men. On
the apparatus being slowly rotated, the figures seen through the slots
appeared to be in motion. The first systematic photographs taken at
regular intervals of men and animals were made by Muybridge in 1877.
— The Moving Picture World, April 4,1908
Theater Engineering Conference
Ventilating and Air Conditioning
•
Ultraviolet Air Disinfection
in the Theater*
BY L. J. BUTTOLPH
GENERAL ELECTRIC COMPANY, NELA PARK, CLEVELAND, OHIO
Summary — Theater attendance, and the decrease during times of epi-
demic respiratory disease, involves a public-health and a theater-operation
problem possible of partial solution by an increase in ventilation, a sanitary
ventilation, probably effective only when provided in amounts physically and
economically impractical because of the power and duct capacity required to
heat and distribute outdoor winter make-up air. Ultraviolet air disinfection
provides a way of making the air in the upper third or half of theater auditoria
and accessory rooms as good as outdoor air, or a sanitary ventilation of the
lower air, equivalent to 50 to 100 air changes, resulting fr6m the usual random
vertical air circulation throughout the horizontal cross section of occupied
rooms. Any sanitary ventilation value in the make-up air of a duct-heating
and air-conditioning system, may also be increased five- to tenfold by using
ultraviolet energy to disinfect all recirculated air to the bacterial equivalence
of outdoor air. There are tabulated lamp requirements for upper-air and
duct-air disinfection and schematic installation sketches.
ANNUAL NEWSPAPER notices urging people to stay away from
crowds during times of epidemic respiratory disease call atten-
tion to the public-health problem of the motion picture theater.
Those who stay away may be benefited but the resulting decrease in
attendance is usually only enough to create an economic problem for
the theater operator without solving his health problem.
The only solution of this problem is basically one of ventilation, of
providing about ten times as much air volume per patron, or ten times
more air changes per minute or hour, than has been provided in the
past. Optical considerations of screen-viewing distance and angle, to
say nothing of the economics of building construction, make any
radical increase from the current practice of about seven square feet of
the floor area per patron out of the question. The alternative is that
* Presented October 24, 1947, at the SMPE Convention in New York.
JULY, 1948 JOURNAL OF THE SMPE VOLUME 51 79
80 BUTTOLPH July
of providing a greatly increased ventilation, a sanitary ventilation, in
contrast with the minimum past practice found essential for the dilu-
tion of body odors and the distribution of heat. In the northern half
of the United States, the heating of much more oudoor air than is
essential for the removal of the body heat of the theater patrons is
economically impractical.
The recently available process of ultraviolet air disinfection has
neatly solved the whole problem of sanitary ventilation directly in the
theater auditorium itself by providing throughout the whole upper
half or two thirds of the theater auditorium and the accessory rooms,
reservoirs of air as relatively free of disease-producing bacteria as is the
outdoor air brought into the theater by mechanical means. Recently
available air-sampling techniques1- 2 have demonstrated that in any
occupied theater auditorium properly equipped with germicidal lamps,
the internal air circulation induced by the ventilating system and by
the body heat of the patrons is such as to provide at the breathing level
a sanitary ventilation from the disinfected zone above, equivalent tc
one to two air changes per minute or 50 to 100 air overturns per hour,
in contrast with the five to ten practical by the mechanical introduc-
tion of fresh air.
To whatever extent there may be recirculation of air by the theater-
heating or air-conditioning system there is a similar reason for in-
stalling germicidal lamps in the air ducts to make the recirculated air
equivalent to outdoor air for sanitary ventilation. In so far as any
sanitary ventilating value may be attributed to the make-up air of the
usual air-conditioning system, that value can thus be increased five- tc
tenfold.
A detailed discussion of the unique germicidal effectiveness of the
2400- to 2800-angstrom wavelength of ultraviolet energy3"7 is beyond
the scope of this paper. Apparently, however, because the peak of the
absorption curve of the nuclear protein of bacterial organisms occurs
at a wavelength of about 2600 angstrom units, the resonance radiation
of electrically activated mercury vapor, wavelength 2537 angstrom
units, is, as has recently been pointed out by McDonald,8 "the most
lethal wavelength yet discovered. It is hundreds of times more lethal
to cells than high-voltage X rays. ..." Also Luckiesh9 has shown the
same energy to be hundreds and even thousands of times more lethal
than the ultraviolet and visible radiation in direct sunlight. It is for
this reason that it is possible to disinfect air with intensities and total
amounts of germicidal ultraviolet entirely practical to produce and
1948
ULTRAVIOLET AIR DISINFECTION
81
82 BUTTOLPH July
distribute in occupied places without risk or inconvenience to the
occupants.
A variety of germicidal lamps is commercially available. All of
them are basically low-temperature, low-pressure, electric-discharge
lamps containing mercury vapor. They are electrically and physically
identical with or similar to corresponding tubular fluorescent lamps
except that they are made with special glass tubes transmitting the
2537-angstrom energy with about the same efficiency with which, in
fluorescent lamps, phosphor powders convert this same ultraviolet
energy to longer wavelengths of visible light readily transmitted by
ordinary glass tubing. The energy-conversion efficiency of these
lamps is such that of the total electrical-energy input to the tube and
ballasting device from 10 to 20 per cent is emitted as germicidal ultra-
violet. For example, a typical commercially available 30-watt
germicidal lamp and its ballast taking approximately 40 watts elec-
trical input will produce 7 watts of germicidal ultraviolet.
An even greater variety of germicidal fixtures than of lamps is
commercially available. All of them provide for the proper electrical
operation of the lamp and, except when intended for duct use, are
equipped with enclosing reflectors, and sometimes louvers. Such
fixtures should be carefully designed to keep the ultraviolet energy
away from the occupants of a room and to prevent its ineffective dis-
sipation through short distances to near-by walls and ceilings. In
meeting such specifications these fixtures become ultraviolet-energy-
projecting devices providing a fanlike distribution of ultraviolet
energy in planes inclined 10 to 20 degrees above the level of the
germicidal lamps, Fig. 1. Since such fixtures may vary greatly in
their effectiveness, depending upon the reflector contours and ma-
terial, they should be chosen with care to suit their operating locations.
Of the total ultraviolet output of the bare germicidal lamps, such
fixtures may emit 25 to 50 per cent giving them an over-all efficiency
in terms of the electrical input of 5 to 10 per cent, over-all efficiencies
still considerably higher than are secured with incandescent lamps in
spotlighting equipment of comparable optical characteristics.
Air disinfection in the auditoria, the accessory small rooms, and in
the air ducts of theaters, can be done with germicidal lamps in accord
with theoretical investigations10 and engineering interpretations11
backed by considerable practical experience in medium-sized rooms.
The theater auditoria provide, however, unique opportunities to take
advantage of the fact that, because the air absorption of germicidal
1948 ULTRAVIOLET AIR DISINFECTION 83
ultraviolet is negligible, the effectiveness of germicidal lamps goes up
linearly with the room dimensions.
The installation of germicidal lamps in the motion picture theater
presents two problems not encountered in hospitals, schoolrooms,
offices, or even the theater presenting stage shows. These problems
result from the fact that along with the ultraviolet from germicidal
lamps there go, inseparably, 3 or 4 lumens of visible blue light per
watt of tube input which, aided by the Purkinje effect in a darkened
theater, becomes visible to an extent out of all proportion to the lamp-
tube brightness or the illuminated walls and ceilings.
The general auditorium- or balcony-installation practice is to place
the germicidal fixtures as low as possible on the side walls in or
slightly above a plane passing from the head level of standing patrons
in the back of the auditorum or balcony to the top of the projection
screen. With properly designed and sometimes louvered fixtures, such
a placement will keep the germicidal lamps themselves out of sight of
anyone in either the auditorium or balcony, although this results in
rather high placement of fixtures over the front balcony in the older type
of theaters with high balconies over a shallow auditorium. The usual
low-ceiling-room practice can prevail in the portion of the auditorium
under the balcony.
The problem of blue light reflected down into the auditorium and
down onto the projection screen from the ceiling and side wall is not so
easily solved. In theaterswhere an unusually dark auditorium is main-
tained during projection, light from the germicidal lamps scattered
from the side walls and ceiling may be objectionable, especially during
the proj ec tion of Technicolor pictures. This problem results primarily
from the light reflectance of the ceiling and side walls but is also de-
pendent upon the ultraviolet and light-distribution characteristics of
the fixtures used. Although all ultraviolet and light from the fixture
eventually must reach the ceiling and side \valls somewhere, the light
is much less objectionable on the side walls than on the ceiling; for
this reason, in theaters with light-colored walls or lower-than-usual
ceilings, or with a high balcony necessitating placing the fixtures rela-
tively near the ceiling, only louvered low-ceiling-type fixtures of the
spatial distribution shown in Fig. 2 should be used. In theaters
without a balcony, and especially those with side walls and ceilings of
light reflectances less than 25 per cent, the basic open type of generally
available fixture usually can be used.
In this connection it should be noted that in those theaters providing
84 BUTTOLPH July
sufficient illumination of the projection-screen surroundings to re-
duce contrast glare, in accord with recent good illumination practice,
the problem of blue light scattered to the screen may not exist. Even
with rather highly reflective ceilings and side walls, the total lumens |
of light in the theater is only that usually provided by the stand-by
illumination in theaters where patrons may find their seatswithout the
assistance of aisle lights or flashlamps. When the installation of
germicidal lamps is anticipated in a theater design or decoration, the
apparent amount of blue light can be almost completely controlled by j
the choice of wall and ceiling treatment, generally for reflectances of
less than 25 per cent.
Since the blue light from germicidal lamps is accompanied by about
the same amount of energy in the near ultraviolet, of the wavelength
frequently used in theaters for the fluorescent activation of carpets
and decorative wall treatments, there are unexplored possibilities of
using fluorescent wall treatments to replace relatively high reflectance
of visible light with very low-level fluorescence by the near ultra-
violet to produce just visible decorative patterns without objection-
able reradiated or reflected energy to the projection screen.
Fortunately, there can be considerable freedom of choice as to the
location of germicidal fixtures on the theater side walls as nonuni-
f ormity of ultraviolet distribution in space is amply offset by the ran-
dom circulation of the air during the process of disinfection. Every
effort should be made to fit the germicidal lamp into the architectural
and decorative features of its surroundings even to the extent of en-
closing stock fixtures in custom-made enclosures. It is often very
difficult to adapt germicidal fixtures to the conventional architecture
of older theaters but, fortunately, many fixture designs are adaptable
to the modern treatments of theater interiors.
Since the ozone-producing ultraviolet from germicidal lamps is
completely absorbed by a few inches of air, the small amount of ozone
they produce does not increase with the room dimensions as does the
air-disinfecting action of the unabsorbed germicidal ultraviolet. For
this reason ozone is not likely even to be detectable in a ventilated
theater nor at all objectionable for such an installation as has been
suggested for accessory rooms.
Experience indicates that a normal germicidal-lamp installation is
of value for odor control in the theater. The effect is easily observed
but difficult to measure. The ultraviolet may promote the oxidation
of odorous substances, usually of an unstable chemical nature
1948
ULTRAVIOLET AIR DISINFECTION
85
anyhow, either directly or by way of the very active form of oxygen
present in the air from the formation and decomposition of ozone.
The ozone itself also doubtless has a desensitizing action on the nose
analogous to the effect of certain sound and light waves, or of certain
flavors on the corresponding senses. It is interesting to note that this
odor suppression seems to be effective under conditions where the
ozone itself is barely if at all detectable.
Manufacturers of fixtures suitable for theater use provide installa-
tion tables based upon the room dimensions and ceiling heights, al-
though few such tables are extended to the dimensions of theater
auditoria. Table I is an attempt to consolidate in a single relatively
TABLE I
NUMBER OF 30- WATT LAMPS FOB 99 PER CENT UPPER- AIR DISINFECTION (BASED
ON 6.5 ULTRAVIOLET WATTS OUTPUT AT 100 HOURS)
Average Room
Dimensions
12-23
24-35
36-47
48-61
62-73
74-85
86-97
Over 25
1
2-3
4-5
6-7
8-9
10-11
•
15
1
2-3
4-5
7-8
9-11
12-13
o
^ bC
111
33*
10
7
5
4
1
1
1
1-2
1-2
2-3
2-4
3-5
3-4
.4-5
5-7
7-10
5-6
7-8
9-11
13-16
8-9
10-12
14-18
20-25
11-13
15-18
21-25
29-34
15-16
21-24
29-33
39^4
£
3
2-3
4-6
9-14
19-23
28-34
39^5
51-57
compact form a recommendation based on the most commonly used
germicidal lamp, the so-called hot-cathode 30-watt type. By "Aver-
age Dimension" is meant a figure obtained by dividing by two the sum
of the length and breadth of the volume under consideration. The
theater with a balcony should be broken up into three areas, the front
auditorium, the auditorium under the balcony, and the area above the
balcony. In such a large theater, the increased effectiveness of fix-
tures in the larger-dimensioned front auditorium is fully offset by
greatly decreased effectiveness under and over the balcony. It is also
important to note that under and over the balcony only louvered fix-
tures should be used and but one half as many as are specified by Table
I to secure a theoretical upper-air disinfection of 90 instead of 99 per
cent. Only an exceptionally dark treatment of the ceilings over these
areas will permit the use of the full number of lamps specified by the
table.
The simplest case of the theater or auditorum is that without a
86 BUTTOLPH July
balcony, with dimensions of about 70 by 100 feet, and with a 30- to
40-foot ceiling height. A seating capacity of 1000 may be considered
representative of such a theater. The average dimension of 85 feet
and the possibility of a 20- to 30-foot fixture-to-ceiling distance indi-
cates, Table I, the need of a total of 8 to 10 units (depending upon the
type) to provide a 99 per cent theoretical upper-air disinfection. This
is provision for a lower-air disinfection at a rate equivalent to about
100 air changes per hour.
In the case of a theater with a balcony the calculation should be
broken up into three parts, the open space above the front orchestra
section, the orchestra area under the balcony, and the space over the
balcony. A typical larger theater with balcony and seating about
2000 people would have an orchestra area about 90 feet wide and 90
feet deep, one half of it being under a balcony, 90 feet wide and about
60 feet deep. The ceiling height over the front orchestra would be
about 50 feet, over the back orchestra 12 to 18 feet, and over the
sloping balcony 25 to 10 feet.
Table I calls for 6 to 8 lamps for an average room dimension of 65 to
70 feet and a fixture-to-ceiling distance of over 15 feet, but for the«ame
area under the balcony, with an average fixture-to-ceiling distance of
7 feet, one half the listing of Table I for accessory rooms or 5 to 7
units should be used. Similarly, for the area above the balcony, with
an average dimension of 75 feet and an average fixture-to-ceiling dis-
tance of 10 feet, 6 to 8 units would be needed. The total lamp re-
quirement for the seating area of the theater thus would be about 20,
or one 30-watt lamp for 100 patrons.
Small accessory lounging rooms and wash rooms with low ceilings
may be handled in accord with the lower left of Table I.
To make the recirculated air carried by the theater-ventilating and
-heating duct system equivalent to outdoor air for sanitary ventila-
tion, the same germicidal lamps used for upper-air installation may be
installed directly in the ducts, but the number required is not so
easily determined as for the upper air of the room because of the high
and variable air speed and the great variations in duct shapes.
For ducts whose greater dimension does not exceed the lesser by
more than 50 per cent, and with nonreflecting walls, the maximum
lamp requirements for a 99 per cent disinfection may be read from
Table II. If the cubic feet per minute of air flow is not known it may
be calculated as the product of the duct cross section, in square feet,
and the air speed in feet per minute.
1948
ULTRAVIOLET AIR DISINFECTION
87
Note .in the following table that the cubic-feet-per-minute figures
in the body of the table are directly proportional to both the number
of lamps and the lesser dimension of the duct so that the tables may
be expanded indefinitely by direct proportion and by lamp addition.
For example, the requirement in 30-watt lamps for 220,000 cubic feet
per minute of air carred by a 120- X 150-inch duct would be 10 times
TABLE II
GERMICIDAL-LAMP REQUIREMENTS FOR 99 PER CENT DISINFECTION OF DRY AIR
IN NONREFLECTIVE CIRCULAR OR NEARLY SQUARE DUCTS,
CUBIC FEET PER MINUTE
Number of 30- Watt Germicidal Lamps
Q
1
2
3
4
5
6
7
5
6
100
120
200
240
360
cf
o
7
140
280
420
560
II
g 0
8
9
160
180
320
360
480
540
640
720
800
900
1,080
3^
10
200
400
600
800
1,000
1,200
1,400
;_i
11
220
440
660
880
1,100
1,320
1,540
1
12
240
480
720
960
1,200
1,440
1,680
I-3
13
260
520
780
1,040
1,300
1,560
1,820
14
280
560
840
1,120
1,400
1,680
1,960
15
300
600
900
1,200
1,500
1,800
2,100
Number of 30- Watt Germicidal Lamps
7
8
9
10
20
30
40
20
2,800
3,200
3,600
4,000
. .
25
3,500
4,000
4,500
5,000
q
30
4,200
4,800
5,400
6,000
12,000
o
35
4,900
5,600
6,300
7,000
14,000
'% VI
40
5,600
6,400
7,200
8,000
16,000
II
45
6,300
7,200
8,100
9,000
18,000
27,000
5«
50
7,000
8,000
9,000
10,000
20,000
30,000
HI
55
7,700
8,800
9,900
11,000
22,000
33,000
1
60
8,400
9,600
10,800
12,000
24,000
36,000
48,000
J3
65
9,100
10,400
11,700
13,000
26,000
39,000
52,000
70
9,800
11,200
12,600
14,000
28,000
42,000
56,000
75
10,500
12,000
13,500
15,000
30,000
45,000
60,000
80
11,200
12,800
14,440
16,000
32,000
48,000
64,000
The above duct ratings are for a duct-air temperature of about 85 degrees Fahren-
heit. These ratings should be decreased 10 per cent for temperatures of either 75
or 100 degrees Fahrenheit, by 20 per cent for 65 or 115 degrees Fahrenheit, and by
30 per cent for 60 or 125 degrees Fahrenheit.
88 BUTTOLPH July
the 9 lamps required for 10,800 cubic feet per minute in a 60-inch duct,
or 90 lamps.
For a 95 per cent disinfection but seven tenths the above number of
lamps may be used, for 90 per cent one half, and for a 70 per cent dis-
infection only one quarter as many. Installations to deal with bac-
teria that have been exposed to a relative humidity greater than 60
per cent and to deal with fungi require many more lamps than are
called for in the preceding tables and should be treated as special cases
for which engineering data are available elsewhere.
In the frequent case of flat ducts having one dimension two or more
times as great as the other there should be reference to more detailed
methods of calculation available elsewhere but the maximum lamp
requirements may still be determined from the preceding tables by
subdividing the duct, and the air capa'city, so that the dimensions of
the subdivisions fall within the range of the tables. For example, a
2- X 9-foot duct carrying 9000 cubic feet per minute should be
treated as 3 ducts each 2X3 feet and carrying 3000 cubic feet per
minute. The tables call for a maximum of eighteen 30-watt lamps,
but if the duct is calculated as a whole by a more adequate method the
number is reduced to 15 lamps.
The mechanical details of a germicidal-lamp installation follow
closely those of the dust-filter installation and it is anticipated that
manufacturers will provide similar standard-unit assemblies. Al-
though there are many ways of installing germicidal lamps in air
ducts, the best compromise on the mechanical and radiation factors
calls for placing them lengthwise on the duct wall, on 4- to 5-inch
centers grouped in the center half of the duct walls and out of the
corners of rectangular ducts. The duct walls near the lamps and the
duct width in both directions from them should be of polished chro-
mium plate or aluminum if the conditions are such that the reflective
duct walls can be easily cleaned whenever the lamps are cleaned.
Standard wiring-channel strips, such as are used with the correspond-
ing fluorescent lamps, may be attached to the outside walls of the duct
with the lamp sockets projecting through holes in the duct walls and
the reflector lining. Two-lamp assemblies using high-power-factor
ballasts and moistureproof lampholders, especially designed for this
type of installation are commercially available.
Since the germicidal lamps must be kept reasonably free of dust,
there must be convenient access for cleaning. This usually can be
arranged by hinged panels on the sides or the bottom of the duct, and,
ULTRAVIOLET AIR DISINFECTION
89
if necessary, the lamp may also be mounted on these panels as well as
on the stationary duct walls, Figs. 3A and C. Where the mechanical
conditions demand it, the lamps may, of course, be installed end to end
along the duct. In any case, the reflector lining should be used on all
walls of the duct and should extend beyond the ends of the lamps a
(B)
-12-
(C) (D)
Fig. 3 — Schematic of germicidal lamps in ducts.
distance twice that to the opposite side of the duct. If chromium-
plated sheet steel is not available aluminum-foil-surfaced building
paper or board, or certain special aluminum paints may be used as
substitutes.
In large ducts and plenum chambers germicidal lamps may be
assembled like the rungs of a ladder in vertical frames supported out
90 BUTTOLPH Jllly
in the center of the chamber in whatever series or multiple arrange-
ment best fits the local conditions and provides access for cleaning
and replacement, Fig. 3D. In very large ducts, where the air speeds
are relatively low, the lamps should be so placed, when possible, as to
provide a maximum average distance from the lamps to the duct walls
in directions perpendicular to the lamp tubes, and regardless of the
direction of air movement.
There is a special installation problem in case of flat ducts which
may have one dimension 4 to 6 times the other. Such a duct cross
section limits the effectiveness of the lamps not only in proportion to
the lesser dimension but also because but little of the duct volume be-
yond the actual location of the lamps is useful for air irradiation. In
such cases the lamps should be distributed only over the longer duct
walls to within the lesser dimension from the edges, Fig. 3B.
In spite of the desirability, for efficiency, of the longest possible
travel of the ultraviolet before the first reflection, it is sometimes de-
sirable to combine the bactericidal treatment of air with the humidi-
fying, filtering, and heating treatment it gets in an air-conditioning
system. In such cases, it is desirable, when possible, to provide the
bactericidal treatment at a point of average air temperatures, away
from very hot air or very cold make-up air to maintain germicidal
output efficiency. When possible, the lamps should be placed after
the filtering which reduces the lamp cleaning, but before the humidi-
fication which tends to increase the tolerance of bacteria for germicidal
ultraviolet and may, in extreme conditions, cause electrical trouble*
in lampholders and starters mounted in the chamber with the lamps.
The fact that statistical evidence as to health value to the individual
patrons from air disinfection cannot be secured, because of the small
amount of their total time spent in the theater, is obviously no reason
for not providing sanitary ventilation along with other recognized
sanitary precautions to reduce the possibility of the spread of respira-
tory disease in the theater.
REFERENCES
(1) M. Luckiesh, A. H. Taylor, and L. L. Holladay, "Sampling devices for air-
borne bacteria," J. Bact., vol. 52, p. 55; July, 1946.
(2) M. Luckiesh, A. H. Taylor, and T. Knowles, "Killing air-borne respiratory
micro-organisms with germicidal energy," «/. Frank. Inst., vol. 244, p. 267; Octo-
ber, 1947.
(3) W. W. Coblentz and H. R. Fulton, "A radiometric investigation of the
germicidal action of ultra-violet radiation," Sci. Paper, no. 495, Bvr. of Stand.,
Jour. Res., vol. 19, p. 641; 1924.
1948 ULTRAVIOLET AIR DISINFECTION 91
'(4) Alexander Hollaender, "Abiotic and sublethal effects of ultraviolet radi-
ation on microorganisms," Amer. Assoc. Adv. Sci., Symposium on Aerobiology,
publication no. 17, p. 156; 1942.
(5) L. R. Roller, "Bactericidal effects of ultraviolet radiation produced by low
pressure mercury vapor lamps," J. Appl. Phys., vol. 10, p. 624; September, 1939.
(6) H. C. Rentschler, Rudolph Nagy, and Galina Mouromsefif, "Bactericidal
effect of ultraviolet radiation," /. Bad., vol. 41, p. 745; June, 1941.
(7) W. F. Wells, "Bactericidal irradiation of air," /. Frank. InsL, vol. 229, p.
347; March, 1940.
(8) Ellice McDonald, "Progress of the bio-chemical research foundation,"
J. Frank. InsL, vol. 242, p. 435; January, 1947.
(9) M. Luckiesh and A. H. Taylor, "Determining and reducing the concentra-
tion of air-borne micro-organisms," Amer. Soc. Heat, and Vent. Eng., Journal
Section, Heating, Piping and Air Conditioning, vol. 19, p. 113; January, 1947.
(10) M. Luckiesh and L. L. Holladay, "Tests and data on disinfection of air
with germicidal lamps," Gen. Elec. Rev., vol. 45, p. 223; April, 1924.
(11) L. J. Buttolph, "Principles of ultraviolet disinfection of enclosed spaces,"
Amer. Soc. Heat, and Vent. Eng., Journal Section, Heating, Piping and Air Condi-
tioning, vol. 17, p. 282; May, 1945.
FORTY YEARS AGO
Moving Picture Operators Dread the Summer
Moving picture machine operators dread the approaching hot weather.
Already they have experienced some of the discomforts that the Summer
will bring. When the temperature commences to remind one of the good
old Summer time and the mercury starts to climb, the stuffy little pic-
ture booths become so hot and the air so stifling that it is almost impos-
sible to remain in them any great length of time without going out to
get a whiff of the fresh air. Even in the Winter time it is necessary to
keep revolving fans constantly in motion to overcome the heat gener-
ated by the powerful rheostats. In Summer the conditions are well-nigh
unbearable. Up to this Summer the machine owners adopted their own
methods of constructing their booths and ventilating them. Recent
State restrictions have compelled them to enclose the machines in as-
bestos fireproof booths of certain dimensions, and these are like sweat-
boxes while the carbons are burning, the heat from them and the rheo-
stats being intense.
— The Moving Picture World, May 16, 1908
Theater Engineering Conference
Ventilating and Air Conditioning
•
Service and Maintenance of
Air- Conditioning Systems*
BY W. B. COTT
WESTINGHOUSE ELECTRIC CORPORATION, NEW YORK 17, NEW
YORK
Summary — Because of shortages of raw material and parts in the air-
conditioning and refrigeration field, it is necessary that theater owners
maintain and place in operation and service the apparatus already installed.
THE OLDER TYPE refrigeration cycle installed prior to the develop-
ment of the Freon refrigerants and the more modern refrigeration
cycle is designed by the manufacturer and engineered by the installer
to operate under exacting conditions and must be kept in clean, lubri-
cated, and effective operating condition for satisfactory operation.
The product of the manufacturer of air-conditioning apparatus, such
as refrigeration compressors, condensers, water-saving devices, de-
humidifiers, coils, heating elements, fans, motors, switches and start-
ers, thermostats, and diffusers, is a result of painstaking research
and diligent effort to produce a lower-cost product that can be mar-
keted in a highly competitive business.
These products are assembled by an installer or contractor together
with ducts, wiring, insulation, and piping, for a purchaser into an in-
stalled air-conditioning system. The reliable installer will design an
air-conditioning system for low maintenance costs taking into con-
sideration motor horsepower required, hours of operation, cost per
kilowatt-hour, lubricants required, paint, accessibility of service
valves and switches, worn parts replacement, and countless other
factors. The final picture presented to the buyer by the reliable in-
staller is the total cost in dollars out of pocket to the owner over a
given period of time. Low first cost is not always the cheapest hi the
over-all picture.
Check the layout of your equipment room to see that a mainte-
nance man will have sufficient room to check and lubricate apparatus.
* Presented October 24, 1947, at the SMPE Convention in New York.
92 JULY, 1948 JOURNAL OF THE SMPE VOLUME 51
MAINTENANCE OF AIR-CONDITIONING SYSTEMS 93
•
Lubrication points, valves, and gauge parts that are not accessible are
seldom checked. The best of mechanical equipment breaks down
occasionally or must be overhauled and ample room will result in a
faster, better repair job with resultant low cost and the system placed
in operation quicker.
An air-conditioning system in a theater represents a sizable invest-
ment to the purchaser and replacement of apparatuses high in equip-
ment cost and delay involved in procuring parts together with quali-
fied installation labor. The owner of an air-conditioning system
must arrange service and maintenance of his plant to assist in prevent-
ing breakdowns that result from lack of attention to the entire air-
conditioning system including a check of the system for Freon leaks,
particularly at the compressor seal, inspection and cleaning of drains,
the inspection and adjustment of all belts, safety controls and temper-
ature-regulation devices, and the cleaning and adjusting of all water
valves, sprays, pumps, starters, and gauges, the lubrication of motors
and bearings, the cleaning or 'replacement of air filters, and the adjust-
ment of dampers. Prompt replacement of* worn parts is imperative
in view of required operation of a plant and the annoyance attendant
to a shutdown with loss of business. Periodic service and maintenance
checks will enable you to keep a full charge of refrigerant in your plant
and will locate leaks which may result in expensive repairs, loose fan
belts or sheaves, and dirty filters that result in inefficient operation.
Regular checks may reveal other defects prior to serious trouble.
Various engineering societies and trade associations and all manu-
facturers of this apparatus have drawn up service and maintenance-
check charts with accompanying reports and varicolored or marked
tags of plates to be attached to various check points to assist in check-
ing and servicing apparatus. They have also prepared simple service
and maintenance contracts for use in the trade. It is strongly urged
that you contact the manufacturer, or his representative, of your
refrigeration machinery and request his advice and recommendations
regarding competent service and maintenance people and institute a
periodic service and maintenance program. You will have many
more hours of operation with less over-all expense in following the
recommendations of the manufacturer and his accredited representa-
tive who can supply factory parts and lubricants and who receive
manufacturers' bulletins on products.
Average costs of maintenance and service contracts, on a yearly
basis, have been 19 to 27 cents per seat, dependent, of course, on the
amount of equipment involved and the length of travel to the job.
Theater Engineering Conference
Ventilating and Air Conditioning
Note: For the Theater Engineering Session on Ventilating and Air
Conditioning, Chairman Seider requested that all discussion be held
until after the Delivery of the last paper in the group. The material
which follows, therefore, is in the nature of a panel discussion and
deals with all four papers in this particular section.
DISCUSSION
MR. HUBERT: Mr. Kimball spoke about return diffusion wells in cooling sys-
tems to get rid of. the water. At Lowell we found out that when we returned the
water to the ground, depending, of course, upon the volume of water and the size
of the return well, the well is effective for a period of about six months. Then it
will take no more water.
The pump man surges it with acid and it. commences to take water perfectly
again. It lasts for three months this time. The next acid treatment lasts for
about six weeks. About the fourth time, it is effective for about a day.
Is there any way that you have had success in returning water 'to the well?
The reason we aje interested in it is because the city has imposed a sewer charge
on us; in other words, if you put water into your theater, then they charge you
50 per cent of your water bill as a sewer charge. In the case where you are using
well water, they meter the well and if you use 600 gallons a minute, they figure up
what the cost would be. If you brought that water from the city, they charge
you that much for a sewer charge.
If we can return that water to the ground, we would save ourselves a great deal
of money, but so far we have not been able to do it. Is there any way that you can
assure a successful operation of these return wells over a period of time?
MR. DWIGHT D. KIMBALL: You must be in a district where you have a peculiar
subsoil condition. There is in New York State a limited area, largely around
Long Island, where the State Conservation Commisssion requires return of water.
I have had return diffusion wells that have been used for years without, such
trouble. Once in a great while you get a condition where you have to surge a well,
but it stands up for quite some time after that:
I do not know what your problem could be, but you certainly must be in a strict
area where you get these charges, because you can drill all the wells you want to,
if you get permission of the State Conservation Commission and abide by their
rules, but you do not pay any water rate.
MR. HUBERT: As I understand it, the City spent about $150,000 in one year on
municipal sewers. Under this new arrangement, they are going to collect about
$2,500,000 a year on the water charge.
However, our wells there are gravel wells. In most cases, the gravel is a mix-
ture of anywhere from half an inch up to about two inches in diameter. The
reason they started these return wells is that during the war the rubber companies,
and similar plants, used an immense amount of water. They were using about
75,000,000 gallons a day in the war industries. Naturally, the recovery is about
94
VENTILATING AND AIR CONDITIONING 95
50,000,000 gallons a day. So they made them put the water in the ground, be-
cause they could not keep the water table up; but they ran into trouble on the
11s and the wells would not take it.
We were formerly using 800 gallons a minute in our cooling system, when it was
street-water job. We have now installed refrigeration, and we have four thea-
there. Two hundred a minute is the smallest and about 450 is the largest.
MB. W. B. COTT: We found in our experience, particularly in our own plant,
that if you alternate the use of waste wells, you will prevent the silting up of the
gravel there. Louisville has a very low water table. There is a range of hills
back of Louisville that lowers the water table ; and as you get down to the flat of
the river, alternate the use of the wells. As an example, we are drilling one well
to put the water in now. Another well to put the water, in fifty to sixty days,
helps considerably there. The wells have not been silting up so fast as they have
been in the past.
It might be well for you to examine the possibility, with your well contractor, of
drilling additional wells for their disposal; that is, alternate the use of wells, one
in this period of time and the other in the next period of time.
MB. HUBEBT: We have never tried that. The Deal Pump and Supply Com-
pany near Louisville used to drill quite a few of those wells, and now they have
given up the idea of drilling return wells for the people, because everything has
been tried to keep them from liming up, but they always do, except for a short-
term operation where you want to use it for six months, then it is all right and they
will drill a return well. However, if you have the idea of using it over a period of
time, it is a waste of money. The liming process takes place over quite a large
area, and forms a large cone there. When you surge it, you just push this liming
process away from the well. Then it limes up to the well again. When you treat
it again, it pushes it farther. Eventually, it gets so limed up that you cannot push
it out, and that is why the well quits altogether.
If you drill two or three wells and alternate them as you suggest, would not that
be a case of prolonging the process until your wells lime up again?
MB. COTT: I do not think so. The various whiskey distilleries there return
their condenser water and the processed water to alternate wells.
MB. PHELAN: Mr. Kimball, I was impressed by your costs on filters, but, in
mentioning a throwaway-type filter costing $500, that is only the initial cost.
MB. KIMBALL: Yes.
MB. PHELAN: Do you not think that we should consider what they will cost
over a period of time in comparison to the electric filter, where the operating cost
only amounts to what an electric light bulb would use?
MB. KIMBALL: The difficulty is that most theater people want to make their
investment returnable in matters of replacing material on a basis of something like
four to six years, and you do not come out even on that basis. If you can do it
over a longer period of years, you have a saving in favor of the electrical. How-
ever, they look too much not alone on this, but on their investment costs.
MB. PHELAN: When you enter into the cleaning costs, the drapery costs, and so
forth, I think that that would pull down the over-all cost on them, too.
MB. KIMBALL: It is pretty hard to get a theater man to take that into account.
I venture to say here in the Times Square district, you can go to theater after
theater and most of them have not been redecorated since they were built.
96 DISCUSSION July
MR. ROBERT LEWIS: Dr. Buttolph, I have been using ultraviolet disinfecting
lamps for some time. I have observed two factors. First, the reflectors on these
lamps, apparently, by virtue of their being pointed upward, act as sort of a catch-
all for dead bugs, silting, and other things. Second, apparently, there is a degen-
eration of the glass envelope, either by bombardment at the end or by a general
glass degeneration.
I am well aware that you are required to have decent reflector performance, but
the thing which struck us as peculiar was that it appeared to us after a period of
not more than a day or so, that these small dusty positions on the reflectors ap-
peared fluorescent. We wondered if you have any figures on that type of prob-
lem and efficiency, and, second, what is the average life you should expect from
glass envelopes.
DR. L. J. BUTTOLPH: The germicidal lamps themselves depreciate very rapidly
the first few hours and the first day of operation. As a matter of fact, they are
officially rated after 100 hours of operation, to offset that to some extent. After
that, they depreciate about the way fluorescent lamps do.
The problem of collection of dirt on the reflectors is exactly the problem you
have with lighting fixtures. If the installation has been engineered with an ade-
quate factor of safety, that is not too serious a matter, however; but it is impor-
tant that you originally specify two or three times as much germicidal ultraviolet
as is really necessary, just to take care of those variations.
MR. LEWIS: 'Perhaps I did not make the question quite so precise as I should.
It was our observation that the effectiveness of the reflector was zero after a day.
DR. BUTTOLPH: No, it is not that bad. We have measured many of them.
The ordinary dust that settles on the reflector acts as a neutral filter between the
particles. You can get dust absorption up to 25 or 50 per cent, but your instal-
lation should take care of that. Again, it is the same problem that you have with
an installation for illumination.
MR. LEWIS: I believe that ultraviolet of that wavelength is not the same prob-
lem as illumination. Otherwise, I think it is an answer.
MR. ALBERT STETSON : Mr. Cott commented on the fact that the cost of service
had been accelerated upward. He said that it was now running from 23 to 27
cents. I believe he means 23 to 27 cents per seat per year.
MR. COTT: That is correct.
MR. M. D. KICZALES: Mr. Cott mentioned the use of ammonia refrigerant in
air-conditioning systems. I am curious to know what states permit the use of
ammonia in air-conditioning systems.
MR. COTT: There are quite a few ammonia systems installed; in fact, I can
take you within seven blocks of the hotel you are in, and show you four.
MR. KIMBALL: In places of public assembly?
MR. COTT: Yes, sir, in old equipment. They are carefully trapped and they
are carefully watched by the City of New York. There' are several theater in-
stallations using ammonia in Chicago and several in New Orleans. However,
we have been trying to sell those people replacement equipment in the past. It
may amaze you to know that there are six installations within the city limits of
Manhattan, using methyl chloride, which is highly poisonous and highly dangerous.
As a matter of fact, it is equipment that we as a manufacturer have a responsibility
for now, because we purchased the company that made the methyl chloride.
1948 VENTILATING AND AIR CONDITIONING 97
MB. KICZALES: We agree that present codes in practically all states of the
Union do not permit the use of ammonia.
MR. COTT: That is true on new installations, but there are existing installations
using those poisonous refrigerants. Under the new codes the use of ammonia
refrigerants is not permitted.
MR. KICZALES: Mr. Kimball, this afternoon we had quite a session on acoustics
and the prevention of noise in the systems. Particularly, some recommendations
were made to prevent the transmission of noise from air-conditioning equipment.
One of the speakers recommended certain limitations in the design of air-con-
ditioning systems primarily, saying that the air-supply ducts should be set at 500
feet per minute, and the recirculating grill should be set at 250 feet per minute, in
order to be safe and remain within the 35-decibel allowance for the theater.
MR. KIMBALL: Those are, economically or from an engineering standpoint,
rather absurd limits, because I do not know of any jobs installed with those very
low velocities. If you take a large 1000-seat theater, particularly under the
present rate, we could not get space in the building in many cases. However, I
have used 1200, in some cases 1400 feet, for years without trouble; that is, in the
main larger ducts. You get smaller ducts, of course, but we have no trouble if
the duct work is substantially designed and built.
MR. KICZALES : I made a recommendation like that this afternoon, and it seemed
that certain architects and engineers did not hold to that stand, when I mentioned
that we were using 1200- to 1400-foot velocity starting at the fan, and reducing as we
go along down to the outlet. It went as far as 600 feet per minute at the outlet
itself, and we wanted to design 400 to 450 at the recirculating grills.
Mr. Kimball mentioned extending the air-supply duct into the lobbies in order
to prevent a back draft from the opening of doors into the theater. I have very
successfully made use of a concealed unit heater with a recirculating grill at the
bottom, and filtering the air across the lobby entrance with proper controls at the
ceiling and at the floor to give a proper temperature. That would heat the air be-
fore it moved down the lobby into the rear of the theater.
MR. KIMBALL: That is perfectly possible, but it lacks one advantage. If you
install it as I suggest, you not only get heat in the winter, but you get air con-
ditioning into the lobby in the summer.
MR. KICZALES: With the air-conditioning system in the summer time, you can
do with a lower air temperature in the lobby as it usually, passsing through, is
more or less a cooling-off chamber to prepare you for the lower temperature in the
theater proper. We use more or less the exhaust system from the lobby and pass
it on through ; I mean, from the auditorium to the theater and then on out. That
is the air you have to throw away normally.
MR. KIMBALL: In the case of ultraviolet treatment, where in the theater would
you place your lamp to meet the approval of the architect? Second, what would
be the approximate cost of such a treatment, say, in a 1000-seat theater?
DR. BUTTOLPH: In the old theaters, it is almost impossible to find any place
that would satisfy even the architect who designed it, to say nothing of the modern
ones. Fortunately, the modern theater designs are rather adaptable. We have
one or two installations where they are perfectly adaptable.. They are horizontal
wall treatments into which fixtures can be recessed, to be practically unnoticeable.
The installation cost runs about $1.00 per seat. The lamp replacement cost is
98 DISCUSSION July
about 10 cents per seat. That cost does not include the maintenance, which can
be thrown in with the maintenance of the illumination of the place, because there
is just the matter of dusting up whenever they clean up the theater.
MR. KIMBALL: I have had two occasions within the last two years of giving th<i
theater a designed air-conditioning system, and the architect gave me the pro-
nouncement that there should be no opening outlets in the ceiling or walls. If he
will not permit air outlets, how will he allow those light outlets?
DR. BUTTOLPH: That light outlet is a horizontal slot only about 6 or 7 inches
high, at the most, and 3 feet long per unit. So it is not too conspicuous. It
should be broken up by horizontal black louvers, and thereby mask the reflectors.
So, it can be designed into a new place rather easily.
MR. KIMBALL: The great problem in a theater is this: In a filled auditorium,
you do not like sitting next to somebody who is coughing violently and sneezing.
Will that ultraviolet treatment take care of such a condition?
DR. BUTTOLPH: No, particularly not the psychology of that particular situ-
ation. Glycol will handle that particular job, at least the psychology of it. I
do not know whether it works fast enough to catch the drop in its foot of travel.
MR. J. W. SPISELMAN : Dr. Robertson and his associates have recently pub-
lished a paper in which they actually, by advanced methods of collecting air samples,
have tested the exact killing rate. What they found was that the kill was so rapid
within the first second that they could take their first air sample and see that at
4east 80 to 85 per cent of the kill had been completed. That is within the first
second of the injection of droplets simulating that of a sneeze or a cough. Within
the second second, another 50 per cent of the remaining 15 per cent will be killed.
At the end of the third second, they were down to virtually a zero count as far as
the bacteria injected into the chamber was concerned.
Other evidence, such as the direct spray into hospital wards, which I had men-
tioned before, and into cages with mice and into other guinea-pig tests, has indi-
cated that the glycol reaction is an extremely rapid one.
I might point out that glycol has been used for years as a dehumidifying agent
in massive absorbers; in other words, in much the same way that lithium chloride
is used, the same way that silica jel is used. Triethylene glycol has been used
as a chemical humectant, in a dehumidifying agent.
In some installations, air flowing at the rate of 500 feet per minute over a distance
of only some 2 feet, we can almost calculate how rapidly it is dehumidified. Air has
been dehumidified from, let us say, 60 to 70 per cent humidity down to 25, indicat-
ing a very rapid absorption of water.
Conversely, the argument is that at that same rate of speed a moist particle will
pick up glycol, indicating that the actual pickup of glycol must be an extremely
rapid affair; and once the concentration has been formed on the bacterial particle1,
death will take place.
MR. KIMBALL: How quickly?
MR. SPISELMAN : As quickly as medical men have been able to pick up an air
sample. In this one particular piece of work, they feel that they picked it up with-
in one half of one second, which is the first one that they want. At that time there
was between 75 and 80 per cent of kill; in other words, they had that much less
than they had sprayed in. I have heard that when ultraviolet kills germs, it will
kill them just as quickly.
1948 VENTILATING AND AIR CONDITIONING 99
DR. BUTTOLPH: We think that the sneeze has been entirely overrated as a
spreader of disease. The probability that an adjacent person actually will be
able to inhale any considerable number of organisms from a particular sneeze is
surprisingly remote. The rate of diffusion even in a foot or two, is rapid. In
general, the inhalation is not so timed with the sneeze as to gather much of the
contamination. It is largely a psychological problem.
MR. KIMBALL: I was going to say you have a psychological problem, and you
have confirmed it.
MB. KICZALES: The American Society of Heating and Ventilating Engineers has
always felt that the determination as to what germs are really effective and detri-
mental in air-conditioning systems, was up to the medical profession itself and not
up to mechanical engineers, I believe at the last meeting held about a year ago at
Cleveland, there were some talks presented about the use of germ-killing means in
air-conditioning systems, and no definite conclusion was reached as to whether
they were needed or not, even in large air-conditioning systems.
However, in my opinion, since we are talking about theaters in this particular
meeting, where you are being exposed to germs for about two hours, I doubt
whether there is any need for any germ-killing means in an air-conditioning system
in a theater. From my small knowledge of the medical profession and germs, we
find that there are germs in the air, but they are not all disease germs; that they
will not attack the body. You can put a glass of water or a little globule of water
under a microscope and you will find it crawling with germs, but it is still con-
sidered pure water. They do not kill. They do not cause disease. I wonder how
many germs there are in an air-conditioning system that do spread disease;
whether it is economically sound to put in some kind of germ-killing apparatus.
. DR. BUTTOLPH: The Society of Heating and Ventilating Engineers, through one
of its committees on air disinfection and the Research Laboratory in Cleveland,
is working on some research projects for the Society itself. It recognizes bacteria
as one of the real contaminants of air, along with body odor and dust. There is
IK (question about the recognition. Both the Society of Bacteriologists and the
American Medical Association recognize that there is a problem. The Council
of the American Medical Association has a setup by which it examines equipment
for air disinfection. That does not happen to be on air ducts, but it does read on
the need for air disinfection.
MR. SPISELMAN: It has been part of the work that I have done, although I am
an engineer. I have set out dishes and I have collected some of these plates that
show the-amount of bacteria and germs that are in the air. I was interested in it,
very much the same along the lines that your were, and I had a few medical men,
bacteriologists, examine the plates. I was really amazed at the number of patho-
gens that will fly around in the air.
I have asked the same question: Just why doesn't it affect all of us? Quite
often, the answer is that there is a certain threshold level to which you can with-
stand the bacteria and the pathogens. Beyond that threshold level, which is de-
termined by the concentration of those bacteria in the air, they start working on
the various people, on some more than on others.
Moreover, in a recent issue of Science Newsletter, I read that the cold virus is par-
ticularly bad in that one respect : By the time you know you have gotten the cold,
it has been in your body for a long time and has incubated. As a matter of fact,
100 DISCUSSION
they are trying to determine the rate at which a person does pick up a cold, if he is
susceptible to it, and they have it down to within minutes once they have been ex-
posed to it. That is about all I can say in reference to your question.
MR. KICZALES : Has the medical profession ever stated that the air-conditioning
systems in theaters do cause disease? Have they come out point-blank and stated
that they should be provided with some germ-killing apparatus? No one has yet
determined that some disease is caught in a theater.
We cannot be too sure whether anyone caught the cold after he left the theater,
whether the contact was made in a streetcar coming home, or on the street, or in
the theater. I believe the purpose of the research being done by the American
Society of Heating Engineers is to determine that. True, it is a project, but no
definite determinations have been made by the committee as to what was needed.
MR. SPISELMAN : I do not know whether any public health outfit has come out
and said flatly that the theaters are a hotbed of disease or anything of that nature,
but time and again I have picked up papers during epidemic periods, and one of
the first places that you are warned to stay away from are theaters and places
of public congregation. That of itself, coming out from individual public-health
servants, quite probably shows what they must have in the back of their minds
as to where the probable focal points of any disease or any epidemic may start.
The same thing is applied to swimming pools and to other places of public
congregation. By and large, they do not leave out the moving picture houses
or the theaters. They usually see to it that those are included in the statements.
MR. KICZALES: Someone should combat these statements by the public
officials; that a theater owner should put in some sort of system just to advertise
that he has some sterilizing equipment in his air-conditioning system.
MR. NEIL WHITE: Mr. Kimball, in the average air-conditioning installation in
a theater, what is the period over which a complete air recirculation takes place?
MR. KIMBALL: It will vary to a certain extent with the density of the seating
and the height of the theater; in other words, the cubic feet of space per person.
However, they run around seven changes an hour on an average.
MR. WHITE: I have had a little experience with one unit of this ultraviolet
lamp. I seemed to detect a change in odor in the room, and as though there had
been ozone generated or some ionization had taken place.
DR. BUTTOLPH: All germicidal lamps, at least if they are built so that they are
effective at all, produce minute amounts of ozone. It is a manufacturer's problem
to prevent their producing too much. There is probably some odor masking due
to the ozone. Other than that, germicidal ultraviolet is a remarkable photo-
catalyst; that is, ordinary oxidation by oxygen goes on much more rapidly in the
presence of germicidal ultraviolet. Probably both those things are effective.
Practically, I believe there is no effective installation of germicidal lamps where
there is not a noticeable change in odor.
Recently, a number of companies started promoting the lamps purely for that
purpose. They are entirely comparable with these recently advertised chemical
substances for that purpose. We have chosen not to feature that, because we
think that is a minor job the lamps can do in the long run. It is incidental to
their more important use for air disinfection.
Theater Engineering Conference
Promotional Display
Display Frames in the
Motion Picture Theater*
BY LESTER RING
STANLEY DISPLAYS, INC., NEW YORK 18, NEW YORK
Summary — There is no need to enter upon the importance of displaying
advertising in a theater. The reasons are too well known by all; but, the
number of frames, type and size, are worth considering in planning the in-
stallation of display frames.
WHILE THERE EXISTS no history of the events leading to the
evolution of the display frame, it may have begun with a fan-
fare of trumpets, followed by a courier announcing a message, meant
to reach as many as possible ; and when in later years we learned to
read, the rescript was fastened to the side of a prominent building,
thus starting the oldest form of what we know today as "billposting."
There is still with us, on highways, barns, roof tops, and sides of
buildings, in the form of 24 sheets and smaller sizes of lithographs,
hand lettering, electric signs, and other forms of displays.
The earliest print, I have seen, of a theater with posters on each
side of the entrance, was the Globe Theater in London, where plays
were written and produced by Shakespeare.
Our own " Opera House" of yesterday used the three-sheet litho-
graphed posters, 40 X 80 inches in size, pasting them to "House
Boards" in front and around the theater. Such boards consisted of a
wood backing with trim molding around the perimeter. Posters were
pasted on over another, as each attraction played the theater.
This method was used by the first theaters showing motion pic-
tures; and, as producers and film exchanges started renting one-
sheet and three-sheet lithographs, 11- X 14-inch photographs, and
similar material, with a rebate upon their return in good condition, a
need was apparent for their display without pasting. Thereupon, this
* Presented October 24, 1947, at the SMPE Convention in New York.
JULY, 1948 JOURNAL OF THE SMPE VOLUME 51 101
102 RING July
was done by thumbtacking, and in order to protect them from
embryo artists and weather, a glass door was hung in place, the fore-
runner of today's display frames.
The number of frames is dependent on space available, as well as
the policy of the theater, whether playing single or double features,
and the number of changes per week. Frames, each side of entrance
to theater, are always for "NOW SHOWING." Additional frames, on
the front or side of the theater, may be used for "NOW PLAYING" or
"NEXT ATTRACTIONS," those in the vestibule are usually for "NEXT
ATTRACTIONS," and those in lobby and foyer for "COMING." It is im-
portant that all frames be equipped to take the same layout of ad-
vertising material, so that the advertising may progress from COMING
to NEXT ATTRACTION, to NOW SHOWING, without additional purchases,
or having some of it left over in the manager's office. The total num-
ber of frames required for any theater cannot be worked out by
formula, but from the foregoing, six frames, or two for each category,
such as COMING, is the minimum.
Types of frames are usually of wood or metal; wood frames should be
of hard wood, such as walnut, oak, or birch. Metal frames should be'
of such material as will obviate polishing, and aluminum should be
anodized to prevent oxidation and pitting.
With the indirect illumination of lobby and foyers with cove light-
ing and pinpoint downlights, it has become necessary that display
frames be illuminated from within. Contemplated theaters should
make necessary provisions for this by providing recesses, and carrying
electrical outlets' to them. In an existing theater, if cutting recesses in
the walls, or furring the walls to create room for shadow boxes is in-
advisable, display frames can be built out with suitable depth, creating
shadow boxes within the display frame itself.
The front or outside frames should be illuminated, even though the
marquee may furnish sufficient light for readability of advertising
matter; this is done to create a point of interest at all times, and
especially when the marquee ceiling is not lit, as, between the time
the box office closes and the break of the show. If fluorescent tubes
are used for outside, they should be the low-temperature ones, to in-
sure proper starting in cold weather.
Fluorescent lighting and cold cathode are the two best media, em-
bodying maximum illumination with less current consumption and a
minimum amount of heat. Where sufficient recess depth of shadow
box is available, approximately 12 inches, incandescent lights of 150
1948 DISPLAY- FRAMES 103
watts set in reflectors top and bottom 9 inches on centers, are very
effective. Fluorescent and cold cathode should be installed on all four
sides, for an even distribution of light.
The size of frames is dependent upon policy and the number of
changes for each theater. It is important, however, to use frames as
large as possible, consistent with architecture and ceiling height.
Frames should be a minimum of 40 X 60 inches and a maximum of 40
X 80 inches inside for the individual frame with hinging sash. One
opening sliding-glass frame can be 10 to 16 feet long and 72 inches high,
glass size. The latter type should have but one sliding glass to each
track, to prevent chipping and breakage.
Since advertising material today is well standardized, equipment
inside the frame to receive such advertising is easily arranged. Where
double features are played, it is desirable to equip a frame to take ad-
vertising of both pictures. Prominence can be given to one picture
with stills or 11 X 14's of the cofeature. An ideal layout is a 30- X
40-inch, date strip, cofeature title card, and two stills.
Poster exchanges and frame manufacturers will be pleased to work
out sizes of frames required for various layouts. Auxiliary stand
frames can be used in a prominent location, to advertise a coming
attraction, a list of future coming titles, or institutional copy. In some
localities, building and public-assembly bureaus frown on stand
frames, classifying them as hazards.
Another* type of display is the banner, or reader board; this is a
frame set above the entrance doors, and is used for COMING, NEXT
ATTRACTION; and when placed above the first set of doors on the
street side, NOW PLAYING. Lobby banner frames may have a trough
of fluorescent or cold-cathode strips, top and bottom, or both, for
greater visibility. To realize the maximum from this type of frame,
it is advisable to have more than one banner board to progress the
material.
Society Announcements
Convention Papers
Preparations are being made for the Fall Meeting of the Society which will be
held at the Statler Hotel in Washington, D. C., October 25 to 29, 1948, inclusive.
Authors desiring to submit papers for presentation at this meeting are requested
to obtain Author's Forms from the Vice-Chairman of the Papers Committee
nearest them. The following are the names and addresses :
Joseph E. Aiken N. L. Simmons
225 Orange St., S. E. 6706 Santa Monica Blvd.
Washington 20, D. C. Hollywood 38, Calif.
E. S. Seeley R. T. Van Niman
250 West 57th St. 4431 West Lake St.
New York 19, N. Y. Chicago 24, Illinois
H. L. Walker
P. O. Drawer 279
Montreal 3, Que., Canada
Technical Societies Council Elects Officers
On May 20, 1948, the Technical Societies Council of New York held its annual
meeting and election of officers. Those elected were the following:
PRESIDENT SECRETARY
C. S.-Purnell W. F. O' Conner
American Institute of American Chemical Society
Electrical Engineers
VICE-PRESIDENT TREASURER
O. B. J. Fraser M. C. Giannini
American Institute of American Society of Heating
Mining and Metallurgy and Ventilating Engineers
In addition, five of the six directors on the governing board were elected.
The Council was incorporated one year ago with local groups of -fourteen lead-
ing engineering societies, representing some 25,000 engineers in the metropolitan
area, as charter members. Each society has two delegates to the Council, which
serves as a medium for mutual professional betterment, more effective public
service, the furtherance of high professional standards and the advancement of
engineering and scientific knowledge.
Journal Exchange
To complete a set, copies are urgently needed of SMPE Transactions numbers
1, 2 (1916); 5 (1917); 6, 7 (1918); 8, 9 (1919); 16 (1923); 18(1924); and
30, 31 (1927). Will anyone who wishes to sell any of these numbers please
write to R. Kingslake, Eastman Kodak Company, Rochester, New York.
104
Book Review
Developing — Technique of the Negative, by C. I. Jacobson
Published (1948) by the Focal Press, Inc., 381 Fourth Ave., New York 16,
N. Y. 309 pages + xiv pages + 10-page index. 52 illustrations. 51/* X 7l/z
inches. Price, $3.50.
This book describes in detail the process of converting an exposed photographic
film into a negative. It makes no attempt to explain the why or wherefore of the
processes involved. A knowledge of a certain minimum amount of physics and
chemistry would be required were this included. Instead, word descriptions and
"practical illustrations" are used throughout, to make the subject matter under-
standable to the reader. As a result, one obtains a rather oversimplified picture
of the developer technique, but the picture serves very well to an operator whose
knowledge of science is limited. The more mature reader with a knowledge of
chemistry, can also read the book with profit, for he would obtain a bird's-eye
view of the entire field, one that can serve as an introduction for a later and more
detailed study.
The book describes the composition of the developer solution, the methods of
formulating it, and the properties of the ingredients involved. It contains a
somewhat extended discussion of the differences between the many concoctions
that are now in common use as developers, grouping them into three general
classes. This is a useful generalization as it enables the technician to choose a
specific solution for a specific purpose.
While the discussion of the developing solutions forms the -most important
part of the book, it also contains sections on the aftertreatment of the negative.
Not the least interesting of the extraneous matter is a chapter dealing with dark-
rooms and darkroom equipment. Apartment-house dwellers will be especially
interested in the section which describes how a lavatory can be converted into a
darkroom.
Several errors were noted, but these appear to be not too important. The most
glaring of these appears on page 207. There it is noted that Kodak's D-76 and
Ansco's A-17 are compounded with sodium carbonate as the energizer. These
developers use borax. However, on page 153 the correct formula is given for
D-76.
The binding and the paper appear to be of good quality, a noteworthy event
these days of inferior quality. The book will stand considerable thumbing, and
it is the type of book that asks for such treatment.
JOSEPH S. FRIEDMAN
Ansco
Johnson City, N. Y
105
Current Literature
In the March, 1948, issue of Steelways, there appeared, on page 9, a popular
article entitled "Report from Hollywood," by Hannibal Coons. Nails, steel tub-
ing, structural steel, and other items made of this metal, are considered in rela-
tion to their-use in the construction of motion picture studios and within the com-
pleted buildings.
Copies of Steelways may be obtained on request from
The American Iron arid Steel Institute
350 Fifth Avenue
New York 1, New York .
RCA Index— 1947
Recently the RCA Review issued a 24-page Index of substantially all published
English-language technical papers on subjects in the radio, electronics, and re-
lated fields, the author or a coauthor of which was associated with the Radio
Corporation of America at the time of the paper's preparation or at the time the
work described in the paper was performed. The full title of this booklet is
"RCA Technical Papers (1947)— Index— Volume II (b)" and it may be obtained
on request from the RCA Review, RCA Laboratories Division, Princeton, N..I.
EMPLOYMENT SERVICE
POSITIONS WANTED
CAMERAMAN: Twelve years' experience in industrial pro-
duction, three years as chief cameraman with commercial studio.
Familiar with all types of work, 16 and 35, studio and location,
black-and-white and color, sound and silent. Knows editing,
sound and laboratory problems. Single, willing to relocate.
Write P. O. Box 1158, Grand Central Station, New York 17, N. Y.
CINEMATOGRAPHER: A-l references, wants employment with
industrial company anywhere in the United States. Will travel
any needed time. Experienced documentary, 35-mm and 16-mm,
color or black-and-white. Active Member SMPE. Charles N.
Arnold, P. O. Box 995, Peoria, 111.
ENGINEER: Recent graduate B.S. in Mechanical Engineering
from The University of Texas. Desires junior engineering posi-
tion with a manufacturing firm in the motion picture industry.
Background in mechanical and electronic equipment design.
Write to A. Kent Boyd, 3308 Liberty, Austin, Texas.
106
Journal of the
Society of Motion Picture Engineers
VOLUME 51 AUGUST 1948 NUMBER 2
PAGE
Television Transcription by Motion Picture Film
THOMAS T. GOLDSMITH, JR. AND HARRY MILHOLLAND 107
Television Recording Camera
J. L. BOON, W. FELDMAN, AND J. STOIBER 117
Development of Theater Television in England . . A. G. D. WEST 127
Auditorium Acoustics J. P. MAXFIELD 169
Quieting and Noise Isolation EDWARD J. CONTENT 184
Behavior of Acoustic Materials RICHARD K. COOK 192
Continuously Variable Band-Elimination Filter
KURT SINGER 203
Society Announcements 211
64th Semiannual Convention 212
Book Reviews:
"Magic Shadows," by Martin Quigley, Jr.
Reviewed by John E. Abbott 214
"Photographic Facts and Formulas," by E. J. Wall and
Franklin I. Jordan
Reviewed by Howard A. Miller 214
Section Meeting 216
Current Literature. ...» 217
New Products.. 218
ABTHUR C. DOWNES HELEN M. STOTE GORDON A. CHAMBERS
. Chairman Editor Chairman
Board of Editors Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930,
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOURNAL must be obtained in writing from the General Office of the Society.
Copyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC . . . EXECUTIVE SECRETARY
OFFICERS
1947-1948
PRESIDENT EDITORIAL VICE-PRESIDENT
Loren L. Ryder Clyde R. Keith
5451 Marathon St. 233 Broadway
Hollywood 38, Calif. New York 7, N. Y.
PAST-PRESIDENT CONVENTION VICE-PRESIDENT
Donald E. Hyndman William C. Kunzmann
342 Madison Ave. Box 6087
New York 17, N. Y. Cleveland, Ohio
EXECUTIVE VICE-PRESIDENT SECRETARY
Earl I. Sponable G. T. Lorance
460 West 54 St. 63 Bedford Rd.
New York 19, N. Y. Pleajsantville, N. Y.
1948-1949
ENGINEERING VICE-PRESIDENT FINANCIAL VICE-PRESIDENT
John A. Maurer James Frank, Jr.
37-01—31 St. 18 Cameron PI.
Long Island City 1, N. Y. New Rochelle, N. Y.
TREASURER
Ralph B. Austrian
247 Park Ave.
New York 17, N. Y.
Governors
1947-1948
John W. Boyle Robert M. Corbin . Charles R. Daily
1207 N. Mansfield Ave. 343 State St. 5451 Marathon St.
Hollywood 38, Calif. Rochester 4, N. Y. Hollywood 38, Calif.
David B. Joy Hollis W. Moyse
30 E. 42 St. 6656 Santa Monica Blvd.
New York 17, N. Y. Hollywood, Calif.
1948
William H. Rivers S. P. Solow R. T. Van Niman
342 Madison Ave. 959 Seward St. 4431 W. Lake St.
New York 17, N. Y. Hollywood, Calif. Chicago, 111.
1948-1949
Alan W. Cook Gordon E. Sawyer
4 Druid PI. Lloyd T. Goldsmith 857 N. Martel St.
Binghampton, N. Y. Burbank, Calif. Hollywood, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Section Officers and Office Staff listed on page 220.
Television Transcription by
Motion Picture Film*
BY THOMAS T. GOLDSMITH, JR. AND HARRY MILHOLLAND
ALLEN B. Du MONT LABORATORIES, INC., PASSAIC, NEW JERSEY
Summary — The paper describes the electronic and camera equipment
for recording television sight and sound on film, the picture made directly
from the face of the cathode-ray tube. The application of this technique
will be discussed with regard to documentary recording, network syndication
use, and theater television. Representative films recorded in this manner
are available.
FOR OVER TEN YEARS Du Mont Laboratories have photographed
television programs from the face of the cathode-ray tube using
both still cameras and motion picture cameras. The early motion
picture recording employed conventional cameras which were non-
synchronous with the television system, and stroboscopic patterns
of blanking, overexposure, and underexposure were present on the
films. Later there was developed a synchronously driven camera
operating at 15 frames per second, thus exposing one entire frame of
television and skipping the next during pulldown. This camera,
however, produced a nonstandard film making it difficult to utilize
the picture either for regular viewing or for television rebroadcast.
Since the camera equipment is specialized, we then approached
Eastman Kodak Company to develop a commercial camera of this
style. Boon, Feldman, and Stoiber describe this special camera
developed for use in television transcription.** We shall discuss some
of the electronic problems which arise in television transcription, and
consider the use of transcriptions by the broadcaster, advertiser, and
theater.
Television transcription is accomplished by recording a program
on motion picture film directly from the face of a cathode-ray tube.
The sound-channel recording is done by conventional means but the
picture recording is rather complex in order to achieve high quality.
A major consideration is the fact that the television picture rate of
transmission is 30 complete frames per second. On the other hand,
the standard of motion picture recording is 24 frames per second.
* Presented October 23, 1947, at the SMPE Convention in New York.
** JOUKNAL OF THE SMPE, this issue, pp. 117-126.
AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51 107
108
GOLDSMITH AND MILHOLLAND
August
We record the television pictures on film at the rate of 24 frames per
second so as to allow reprojection of the film either in a conventional
projector for direct viewing or in the standard projector for rebroad-
cast by television.
A transcription recording console consists of a special monitor
receiver and a film camera, with associated sound-recording facilities.
The photograph of Fig. 1 shows a special monitor of this type and the
recording camera as constructed by Eastman Kodak Company.
Particular precautions must be taken in the design of the monitor
to eliminate as far as possible many of the fluctuations which are
readily tolerated in home television receivers. For example, a high-
Fig. 1
voltage supply of excellent regulation is required so as to avoid any
change in picture size with the variation of picture brightness in the
scene being televised. The screen material of the cathode-ray tube
must be very fine so as to be below the spot-size limit of the electron
beam. Obviously, the linearity of scanning is adjusted as well as
possible. A form of gamma correction is inserted so that to some
degree the chemical gamma factor of the film can be matched to
produce most faithful contrast gradations in the pictures. It is
customary to use a positive picture on the monitor, but in some cases
where speed is essential, a negative picture is produced on the monitor
by means of video reversal of the signals which drive the cathode-ray
1948 TELEVISION TRANSCRIPTION BY FILM 109
tube. Where the negative picture is used, it is necessary to generate
a reverse blanking signal in the equipment so as to suppress completely
the normal synchronizing pulses, and obscure the return trace lines
from the picture. If a system were being developed exclusively for
theater television, the synchronizing signals could be in the whiter-
than-white direction, and it would be unnecessary to have this com-
plication. Where the negative polarity picture is reproduced on the
cathode-ray tube, it is even more important to provide gamma cor-
rection by electrical circuit design. Use of the negative picture allows
direct photography on positive stock resulting in both increased speed
and reduced cost.
We found it desirable to utilize a 12-inch cathode-ray tube operated
at 25,000 volts. Because of the large aperture of the lens, it is cus-
tomary to scan an area of only 6X8 inches on the face of this large
tube in order that the full rectangle of the picture be substantially
flat and be exposed to the camera without any cutting of the corners,
thus keeping good focus both electrically and optically.
Fig. 2 shows a timing diagram which illustrates the phase and
frequency relationships between the television signals and the re-
cording camera. At the top line is a timing indication expressed in
intervals of YISO second. This interval is a subdivision of both the
30-frame-per-second television-picture interval and the 24-frame-per-
second film-picture interval. The next line indicates the television
blanking interval and the useful television-picture interval. The
actual picture signals and horizontal synchronizing signals occur in
the interval entitled "scan" in the second line. Here the television
field interval of VGO second provides half of the interlace picture, and
the succeeding Veo-second field interval provides the other half of
the television interlace. Accordingly, two fields of television scan-
ning vertically from top to bottom constitute one complete frame
of television picture in an elapsed time of Vso second. On the next
line there is shown the camera-shutter characteristic which must
be very carefully adjusted for proper interval. On the bottom line
the pulldown cycle is illustrated. The most critical characteristic
in the recording camera is the timing of the shutter blanking and
exposure interval. The absolute intervals are the most important,
and if they are appropriately adjusted, then the exact phase
relationship is not very critical. As shown in Fig. 2, the phase
relationship has been so adjusted that one of the lap-dissolve points
for opening and closing of the shutter has been tucked under the
110
GOLDSMITH AND MILHOLLAND
August
television blanking interval. However, the other lap-dissolve point
is shown approximately in the middle of the television field interval.
If this shutter is not adjusted correctly, then a bar of distortion is
likely to appear in the recorded film picture. Such a lap-dissolve
bar is noticeable as a flicker caused either by underexposure or over-
exposure on a few elements or lines of the picture.
It is customary to drive a recording camera by synchronous motor,
and where the television signal and the recording camera power are
both controlled by the same power mains, then the camera runs in
exact synchronism with the television synchronizing generator.
However, it is desirable in many cases to record programs in one state
TIME
( SECONDS!
120 ifo i?0 '?0 l?0 '20
I I I I I I
I I I I I
ELEVISION ft-*3* -0 •nTK
"L::rG ooo Evc,
SCAM.NG II.NTTR^cll.NTER^Ctll
SHUTTER
PULSE
FILM
PULL DOWN — / U
n
n
Fig. 2 — Television-transcription-camera timing diagram.
which have originated in another state, thus not necessarily involving
synchronized power lines. Many films which we have taken have
been recorded in the nonsynchronous manner, and thus it is highly
desirable that the shutter angles be appropriately adjusted so that
the lap-dissolve bars are eliminated. It is best to have a slight
double exposure on the lap-dissolve lines rather than an underex-
posure in order that the distortion be minimized.
On nonsynchronous power supplies the two regions of lap dissolve
are in slow motion up or down the picture at the rate of the difference
frequency between the 60-cycle supply controlling the synchronizing
generator and the 60-cycle supply driving the synchronous motor of
the field camera. Two such bars are present as seen in Fig. 2.
The camera shown in Fig. 1 uses 16-mm film and can employ a reel
having 1200 feet of stock thus allowing about 33 minutes of recording.
1948
TELEVISION TRANSCRIPTION BY FILM
111
In practice, a dual system is employed so as to record programs for an
indefinite period. The teletranscriptions may be used for rebroad-
cast, promotional advertising, criticism of program techniques and
content, and legal records.
Fig. 3 is a picture which illustrates some of the work that can be
done by recording from television. This is an original enlarged from a
frame of 35-mm motion picture film. Fig. 4 is a recorded television
transmission of the same scene. Fig. 5 is another original, of which
Fig. 6 is the television recording. These were not taken on a 16-
mm camera and do not indicate any of the banding.
Another very promising use of transcriptions is for theater tele-
vision. The equipment and films discussed here have been primarily
for use on 16-mm film, but the same principles apply to the project of
theater television, using 35-mm film and a process of rapid develop-
Fig. 3 — Original.
Fig. 4 — Photograph received
cathode-ray tube.
on
ment for immediate projection in a minute or less after reception.
Much thought is being given to the theater-television possibilities and
systems are being studied having other than the standard broadcast-
line and frame rates. However, the immediate demand for recorded
television programs • is so great that the 16-mm equipment here
described has been developed to a practical degree and is in active use.
Television networks eventually will rely upon radio relay or coaxial
cable to connect the various stations. However, until such facilities
can be provided and can be operated at a reasonable cost, it is possible
for the network syndication to be accomplished by television tran-
scriptions, recording the major programs in the originating stations
on film which may then be shipped to the subscribing affiliates.
Where this method is employed the recording-camera equipment will
employ a separate sound system for best quality, since there is time
112 GOLDSMITH AND MILHOLLAND August
for limited editing and time for the special processing of the two films
in preparation of the prints.
On the other hand, for some applications such as a documentary
record of transmission, which eventually might be required by the
Federal Communications Commission it is entirely adequate to em-
ploy a single sound camera whereby the general subject matter and
sound are recorded on a single film. Here it is probable that no
prints would be required, and therefore greatest economy should be
practiced. However, for rebroadcast the highest possible quality
of sight and sound is desired and a separate sound system is preferred.
For some forms of recording the picture may be photographed on
to negative stock as a negative picture. Then, a television system
Fig. 5 — Original. Fig. 6 — Photograph received on
cathode-ray tube. ,
may be employed for viewing this picture since we have a regular
means of reversing the picture polarity electrically in the television
chain. We have used this method in regular broadcasts to save time
in film processing, taking regular newsreel records on the original
negative and playing the negative film through the television film
camera, reversing the video polarity so as to transmit a positive picture.
Now a word about future possibilities with television transcription.
Already the sensitivity of television cameras using the image-orthicon
tube exceeds the sensitivity of most photographic emulsions. Thus,
we can say that the television-camera chain serves as a light amplifier,
extending the range of photographic recording to scenes having lower
lighting levels than can be successfully recorded by a film camera
alone. However, further development must occur before the sensitive
1948 TELEVISION TRANSCRIPTION BY FILM 113
television system can fully match the contrast gradation fidelity and
resolution and sharpness available by direct film camera methods.
The intermediate film technique has already developed to such a
degree that the transcription compares favorably with an original
film used as the subject matter for such a transcription.
The television camera promises excellent utility in a motion picture
recording studio. At present such a camera is primarily an aid to
the regular picture recording camera. Used as such, it gives the
Operations Personnel a means of seeing immediately the setup and
programming for the scenes, thus providing a control far more useful
than the customary waiting until development of the film can provide
for a means of analysis of what is being filmed.
Usually such a viewing takes place the day after the scene has been
photographed ; this unnecessary delay proves very expensive. The
television camera used in conjunction with the motion picture camera
can provide at once on television monitors the scenes being recorded.
Ultimately it may well be that the television camera head alone
may be used in the studio to pick up the scene and the complex film
equipment will be located in central permanent laboratories where
both sight and sound are recorded. This procedure of placing micro-
phones in the studio and the complex sound-recording equipment in
the laboratory is customary at present in many studios. It is entirely
possible that both the sound and the picture recording equipment
may be located in the central laboratories. In this way the scenes
can be monitored while they are being recorded, thus aiding the
co-ordination between directors, cameramen, and performers.
DISCUSSION
MR. J. G. BRADLEY: I am interested in the recording aspect of this because I
see in this recording a possibility of creating and preserving records as library
material. What losses do you sustain when you retelevise these pictures taken
by camera from a televised image? In other words, how does the retelevised
picture compare with the original televised image in sound and picture quality?
DR. T. T. GOLDSMITH, JR. : The film recording compares favorably with the
original televised image. The sound was degraded pretty much tonight. This
sound was dubbed in from various sources. The camera that you see here does
not have sound facilities with it. The sound, for example, on the President's mess-
age was recorded in New Jersey in my house, and the picture was recorded here
in New York, and the President was speaking in Washington. So it was a rather
peculiar combination which resulted in the film you have seen. Other parts of the
sound here were dubbed from other recordings.
We are putting together complete sound-and-picture recording apparatus all
under one control, where we expect the sound quality to be quite comparable to
the reproduction obtained over the radio channel by direct reception.
114 GOLDSMITH AND MILHOLLAND August
As to the picture quality, which I believe is a primary consideration that 'you
have, we have rebroadcast quite a number of the programs which were recorded in
this manner. We know there are many flaws left in this system. We know there
are many cures for some of the flaws. Some of them will, inevitably, degrade
the picture somewhat. However, even at the present status, we have had things
working as we want them for film recording. Then we have played over the air a
test film consisting of a section of 16-mm film and then a section of teletranscribed
film. It was very difficult to tell at the receiving point just where one stopped and
the other started.
There is a noticeable difference particularly for those who are looking for the
flaws and the bars that are present, but teletranscription does give you a quite
faithful recording of what is going on the air. I think the Eastman people will
agree with me that many of the tests we have made have recorded just about all
there is on the cathode-ray tube. Some of the films that you have seen tonight
were taken with mobile field equipment, which does not have the full resolution
and the resulting characteristics of the iconoscope cameras. Some of the shots
were recorded from iconoscope-camera signals, and they are better. However,
even those suffer from lack of depth of focus in the studio. The iconoscope is not
too sensitive. So we have to use a fairly large aperture lens on the pickup icono-
scope camera. As a result, the background is badly defocused, but we have means
under development for eliminating that characteristic and having good resolution
and at the same time good depth of focus in the studio equipment. Thus we
shall have something that can realize more nearly a full 525-line television system.
Obviously, you cannot go to much better resolution than that and still keep
the broadcast standards that we have today; but I do believe for many docu-
mentary records the 16-mm film recorded by television can compare very favor-
ably with 16-mm direct recordings, and certainly be equal to the average record-
ngs used for film work generally.
MR. BRADLEY: I think my question is more technical than the way you have
answered it. In preserving motion picture records, we have to look forward to
copying the film. Each time we copy it we lose some light, whether it is 5 per cent
or 10 per cent, but there is a limit to the number of times we can copy a motion
picture, until there is complete degradation of light. Suppose you televised a mo-
tion picture and copied it with a camera. Then you televise the copied picture
and copy it again. How many copyings can you get before this degradation?
DR. GOLDSMITH: There is a small degradation each time. Do you mean the
degradation by retelevising, or by printing a. lot of positives from one negative?
MR. BRADLEY: Degradation by rephotographing the televised image. I
understand you can televise a picture many times without any degradation,
but when you photograph it again from a television screen, what degradation
sets in there?
DR. GOLDSMITH: There will be some further loss, of course, because each time
you televise it you change the contrast fidelity a little, and that would penalize it.
However, I do not know just what conditions you would encounter that would
make it necessary to transcribe again and again, for example, down to the tenth
transcription. If you transcribe once to get the film record down permanently
on a first negative, then you can make positive prints in any number, with the
normal limitation of printing, 'distribute them that way, and thereby, say, have
1948 TELEVISION TRANSCRIPTION BY FILM 115
50 positive prints going out to 50 affiliated stations, each one suffering only one
degradation due to retransmission.
MR. COOK: I noticed in the presentation of the last motion pictures
that there was less disturbance of the picture than in any of the previous ones,
either the projected demonstration tonight or the pictures that Captain West had
the other night; in other words, the picture was extremely stable. Were these
broadcast or were these taken directly from a coaxial cable to your tube?
DR. GOLDSMITH: We have done both. If I remember these particular clips
that were put together here, most of these were done at the station, but many of
them were by radio-relay link; not many of them, but many that we have done
have been that way; for example, the baseball game came over by radio-relay
link on some of its paths. Of course, the President came from Washington to
New York on the 2.7-megacycle band with coaxial cable. There were various
transmission means employed on many of the shots that you saw tonight. Some
of the others were taken off what we call a studio line right from the studio
without any broadcast interference. The quality that you saw here is quite repre-
sentative of what you get in many good receiving locations around New York,
which is perhaps 50 per cent of the locations that have television sets.
MR. BLOOM: You mentioned you had difficulty with the power-line
frequencies being different at one point of transmission than at the point of re-
ceiving. Is it possible to use a synchronizing system, an automatic-frequency
system, to drive your synchronous motors to operate that?
DR. GOLDSMITH: We have done that. On some tests we have driven our power
equipment from a synchronous supply operated that way, but we feel that if we
can get this shutter phasing and shutter angle appropriate, particularly the
shutter angle, then it will be unnecessary to take that precaution, and it is a far
simpler set of equipment. You can run completely synchronous with the trans-
mitted signals by using a synchronizing process of vertical field control to lock
in your mechanical mechanism as well.
MR. HUGH CHAIN: Are you offering this commercially yet? Is it out of the
experimental stage? If it is, have you any idea of the rates for this service?
DR. GOLDSMITH: It is a little premature. We are building the units. We built
quite a few of them experimentally and we are getting the things into production,
but one of the major problems has been the commercial camera: The commer-
cial electronic equipment is pretty well along, too. The general plan, so far as we
know, is to sell these units to interested subscribers, primarily thinking that
the first people that would want them would be broadcast stations to use for regu-
lar transcription at the broadcast stations. The units are available.
MR. CHAIN: But are you offering the scripts and transcription service to ad-
vertisers to your stations?
DR. GOLDSMITH: Yes, that has been done for some time. We have been re-
cording programs for sale to advertising agencies, and so forth.
MR. CHAIN: Mr. Feldman, the Jack Kilty part of the transcription did not
seem to have high contrast. You mentioned to me that it was done on sound
positive film. Is that lack of contrast inherent in the film, or what was the situ-
ation there?
MR. W. FELDMAN: We have used a couple of different types of sound positive
films. As to the particular one of the Jack Kilty Show, it was on 5302.
116 GOLDSMITH AND MILHOLLAND
MB. ROBERT FRASER: That film was a work print, a very hasty print. As
soon as we saw the banding, we did not go any farther. So the quality is very
poor. However, the negative would be capable of a much better print.
MR. FELDMAN: Actually, this print was made up specifically to show banding,
and not for any picture quality. That was not the prime purpose.
MR. LEWIN: There was a statement made by Dr. Maloff that confused
me quite a bit, and I did not have an opportunity to question him at the time. I
wonder whether one of the authorities in television would help to clarify it.
Dr. Maloff said that in counting the number of lines the television engineer
counts both the black and the white lines. I have always been under the im-
pression that when you say there are 525 lines, you mean the electron beam
actually scans 525 lines in the 30th of a second. That would be 525 white lines.
Am I wrong in that or not?
CHAIRMAN PAUL J. LARSEN : That is correct.
MR. LEWIN: In other words, if you count the black and white lines, there
are 1050 lines.
DR. GOLDSMITH: I shall try to clarify that in this way: In television we use
scanning lines which are potentially either white or black, or some gray inter-
mediate tone, depending upon the signal that is being produced. In the motion
picture terminology of lines, they refer to a series of black lines ruled on a paper
with equally spaced white lines between.
In television it is true that if you look at a television picture you can see the
scanning lines in the high-light part of the picture as white strips of light across,
with a very thin space between that may be black. We try in television to make
that space as small as possible; in other words, the spot size is just the size of the
space between the centers of two lines, in the interlace system. So that the lines
that are white just merge with one another on their edges.
Theoretically, in a perfectly square spot-scanning system in television you have
no black lines in between, and there comes a difference between the motion pic-
ture terminology of black lines separated by white spaces and the television sys-
tem where the scanning processes, as to lines, can either be black or white or gray,
at the control of the grid in the cathode-ray tube.
MR. DAVID B. JOY: I understood the speaker to say that by using a 35-mm
camera, a picture could be taken of the television image, developed, and ready for
projection in the regular projection equipment of the theater within a few min-
utes after the picture was received, thus giving you as much light on your projec-
tion screen as you would get with an ordinary motion picture. If that equipment
is not too complicated, it might be one solution to the theater television.
DR. GOLDSMITH: That is right. It is not too complicated; in fact, that
few minutes that you talk about has been experimentally pulled down to less than
a one-minute delay. A film frame is exposed in front of the transmitted picture
at the cathode-ray tube, is run through a developing machine, having both de-
veloping and fixing, at high temperatures and high drying speeds, and threads
right on through into the projector in the theater and can be projected in that
manner. The equipment is rather complex in the developing and fixing process.
I do not know about the permanency of such records, but it does allow editing,
saving the film if you want to for longer times, or allowing almost instantaneity of
broadcast, almost as soon as received.
Television Recording Camera*
BY J. L. BOON, W. FELDMAN, AND J. STOIBER
EASTMAN KODAK COMPANY, ROCHESTER, NEW YORK
Summary — A 16-mm motion picture camera for recording television pro-
grams at sound speed from a monitor receiving tube is described. Basic
camera-design features include a 1200-foot magazine, which permits con-
tinuous recording of a half-hour program, separate synchronous-motor drives
for the shutter and film-transport mechanisms, an 8-tooth sprocket pulldown
actuated by an accelerated geneva star, an //1. 6, 2-inch focal length coated
lens, and a 72-degree shutter. Other features include a " bloop" light to pro-
vide registration with the sound-film recorder, a film loop-loss indicator, and
appropriate footage indicators.
Some general operating characteristics of the camera are included along
with a description of the pulldown system and the general problem of film
steadiness. The last is particularly critical because of the high accelerations
involved in the pulldown, in addition to the reaction of film to temperature
and humidity changes.
IN THE RECORDING of television-tube imagery on film, one is con-
fronted with the primary problem of reducing an image frequency
of 30 cycles per second on the tube to one of 24 cycles per second on
film. It is possible to record television images satisfactorily at 15
cycles per second or even at 71/* cycles per second. In conforming
with the standards of the American Standards Association of re-
cording sound on film, however, one is limited to a frequency of 24
frames per second. Since the ratio of the two frequencies is 5 : 4, it is
evident that the transition is most easily accomplished by omitting
one of every five scanning traces. More exactly, the following se-
quence of events occurs: With the camera and television tube syn-
chronized, film is exposed for one complete tube cycle, lace, and inter-
lace patterns. During the first quarter of the next complete cycle
the film is advanced one frame. Exposure takes place for the remain-
ing three quarters of this cycle and the first quarter of the next cycle,
after which the film is again advanced. Fig. 1 shows an exterior view
of the camera and in Fig. 2 may be seen the schematic representation
of the events.
As a direct consequence of the limited pulldown time indicated
above, the camera is restricted to a closed-shutter angle of 72 degrees.
In addition, the pulldown angle for the film must be less than 72
degrees by the angle which the aperture subtends.
* Presented October 23, 1947, at the SMPE Convention in New York.
AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51 117
118
BOON, FELDMAN; AND STOIBER
August
In our camera the pulldown action is accomplished by means of an
8-tooth sprocket which is indexed by an 8-point geneva star. It may
be readily seen from the geometry of the star and its driver that the
latter must rotate 135 degrees during the indexing operation. But
since this angle is greater than the permissible pulldown angle, it is
necessary to interpose an accelerating mechanism between the con-
stant-speed drive shaft and the geneva star. A variable-arm, spline-
and-slot movement, shown in Fig. 3 and treated mathematically in
the Appendix, is used to reduce the pulldown time from one which is
equivalent to 135 degrees to one
which approximates 57 degrees.
The effect of this accelerator is
to produce extremely large peak
film accelerations. As an ex-
ample, the peak acceleration
shown in Fig. 4, for operation at
24 revolutions per second, pro-
duces a linear film acceleration of
the order of 6 X 104 inches per
second. A standard 8-point ge-
neva star driven by a constant-
speed shaft at the same speed
reaches about one tenth the
above acceleration at its peak.
As may be expected, a large
force is produced on the inter-
mittent movement. In addition,
such problems as optimum pres-
sure-pad tension, film steadiness
in the gate, and a general increase
in noise level become more evident. In view of the large forces pro-
duced, it was deemed advisable to use a sprocket pulldown instead of
a claw since the life of the latter would not be long enough for the
heavy-duty operation required.
FILM STEADINESS
During the preliminary camera tests it was observed that film
steadiness was very erratic and was particularly dependent upon hu-
midity conditions. As the relative humidity reached 85 per cent or
higher, it was impossible to obtain any semblance of picture steadiness.
Fig. 1 — Exterior view of camera.
1948
TELEVISION .RECORDING CAMERA
119
This general effect is not new and has been attributed to an adhe-
sive action between the film and the gate and pressure pad, some-
what analogous to the behavior of two polished glass surfaces in con-
tact and moved parallel to each other. Since the ratio of static to
kinetic friction is large, the forces required to start the motion and con-
tinue it vary considerably. It was noted that so far as the film was
concerned, consideration had to be given to both surfaces, the base as
well as the emulsion. As a consequence, both the gate and pressure-
pad constructions were modified. Instead of permitting the film to
ride on a continuous track in the gate, six studs, three on each side,
were embedded in the tracks and then lapped so that the stud sur-
faces were three or four thousandths of an inch above the track sur-
face. The studs not only served to decrease the contact area between
STANDARD
'(1/60 ~SEC)
CAMERA FRAME
STANDARDT FRAME
DOWN
Fig. 2 — Timing cycle of television monitor and camera.
the gate and film, but also reduced emulsion pile-up and its attendant
effect on steadiness. It should be mentioned, in passing, that a
straight gate and pressure pad are being used.
Modifications made on the pressure pad consisted in replacing the
usual type of continuous chrome-plated track with a 6-studded track
which matched the studs in the gate. In addition, another pressure
pad constructed solely of nylon and with continuous tracks was used.
Both pads provided steady film registration. It was observed, how-
ever, that the former permitted the film to buckle slightly and so was
discarded in favor of the latter. To date, no life tests have been made
on the nylon, although no physical changes have been noted after
some 25,000 feet of film have been run through the camera. It is
anticipated that the nylon will provide satisfactory operation and a
reasonably long life.
120 BOON, FELDMAN, AND STOIBER August
GENERAL PROPERTIES
Two synchronous motors are used to drive the camera. One is used
solely for the purpose of driving the shutter, and the second one, a
larger motor, takes care of the entire film transport. It was felt-
necessary to isolate the two systems in order to prevent any "hunt-
ing" from affecting the shutter. A coupling which ties the two motor
shafts together permits continuous observation of synchronization.
Once this coupling is set, it should not be necessary to readjust the
two motor positions, since their operation will remain synchronized.
The importance of a separate motor drive for the shutter is evident
when one considers that it is necessary, on alternate exposure frames,
to record from the middle of one scanning trace, as shown in Fig. 1,
LOCUS OF SLOT MOTION
GENEVA DRIVE PIN
\
\3ENEVA
Fig. 3 — Slot-and-spline driving geneva.
and to complete the full cycle on the first half of the third scanning
trace. Unless the shutter is able to reproduce its action, the scanning
lines show up as not meeting or overlapping. The net effect is a type
of banding which is reproduced on the film as an alternately varying
density region. It is not anticipated that one will be able to synchro-
nize shutter operation so that recording takes place at precisely the
same position on alternate frames. It should be possible, however, to
reduce the banding to, say, one scanning line. In tests made at both
the National Broadcasting Company and Du Mont Studios, it was ob-
served oftentimes that it was possible to record with no trace of
banding. On other occasions, a form of banding, which resulted from
shutter unsteadiness or from shifting scanning lines, took place.
A 1200-foot, double-chamber magazine is supplied with the camera
and is a self-contained unit in that it permits housing both the un-
exposed and the exposed filrn and may be readily removed from the
1948
TELEVISION RECORDING CAMERA
121
camera. The take-up drive consists of a sprocket-and-chain move-
ment, which is driven from a clutch-controlled shaft on the camera
proper. Separate arms are used to guide the film on both the supply
and take-up sides of the magazine chamber. Appropriate light locks
are provided to permit changing of loaded magazines in a lighted room.
It was mentioned above that some difficulty in film steadiness in the
gate was encountered. In tests made on the pulldown sprocket, it
was noted that there was a need not only for a "snubbing" roller, but
also a stripper roller which would aid in stripping the film from the
DRIVE SHAFT • ROTATION
Fig. 4 — Acceleration curve for 57-degree ac-
celerated pulldown.
pulldown sprocket. The former was located in a position approxi-
mately three frames from the edge of the aperture, although it was ob-
served that within plus or minus 10 degrees its position was not too
critical. The stripper roller was located below the sprocket and in
such a position as to aid the film in following its natural stripping path.
Several tests were made in an effort to determine the optimum pull-
down sprocket diameter. However, since there seemed to be little
difference in their operation, a final choice was made of a 0.762-inch
diameter sprocket.
One other property which has been mentioned is the operation
noise level of the camera. The high accelerations which take place
in the pulldown mechanism, as well as the gear-reduction drives which
122
BOON, FELDMAN, AND STOIBER
August
ate used, all contribute to camera noise. The level is not high enough,
however, to be objectionable.
ACKNOWLEDGMENTS
Considerable credit and thanks are due to the technical staffs of
both the Allen B. Du Mont Studio, Station WABD, and the National
Broadcasting Company Studio, Station WNBT, We are especially
grateful for their many suggestions which have been offered and in-
corporated in the preliminary camera model and also for their con-
tinual assistance in providing studio time and facilities for the pur-
poses of testing the camera.
Fig. 5 — Eight-point geneva geometry.
APPENDIX I
Motion of Eight-Point Geneva Star
In Fig. 5 is shown the geometrical arrangement of the intermittent
geneva star with center at P and the driving pin with center at 0'.
Since we require that the driver move the star 45 degrees during its
period of engagement, certain well-defined relations are determined.
From the figure,
R = Ri tan 22.5 degrees (1)
x = R (1 - cos 6) (2)
where x is the horizontal component of displacement, R the distance
from 0' to the center of the driving pin, and 6 the angle of rotation of
the driving pin. Also,
R! - y = R sin 6, (3)
whence R cot 22.5 degrees - y = R sin 6, (4)
1948 TELEVISION RECORDING CAMERA 123
and y = #(2.41 - sin 0). (5)
Thus, ^ = tan <p = 2.41"- sin 0 ^
where v is the angle of rotation of the star. We have taken the origin
of motion at the initial point of engagement of the driving pin with
the star, but it turns out to be more useful if the origin is selected co-
incident with the maximum position of engagement of the star and pin.
Consequently, the shift in origin changes (6) to
t^^-^r-^f/M) (7)
where A is an arbitrary phase angle. We shall later take it to be 67*/2
degrees, consistent with the requhements stated above. Solving for
<p we obtain
- cos (A + 0)
r 1 - c
|_2.41-
sin(A
(8)
We shall now obtain the relations between the angular velocity and
acceleration of the pin and star. The first derivative with respect to
time yields
. = 2.41 sin (A +0) H-cosCA + 0) - 1 ,
7.82 - 4.82 sin (A + 0) - 2 cos (A + 0)
which reduces to
(cos A + 2.41 sin A} cos 0 -f (2.41 cos A — sin A) sin 0 — 1
^ ~ 7.82- (2 cos A + 4.82 sin A} cos 0 + (4.82 cos A -2 sin A) sin 0
and which, for A = 67*/2 degrees, becomes
2.61 cos 0 - 1 .
* " 7.82 - 5.22 cos 0
(11)
The angular acceleration of the star is obtained by a second time
differentiation.
. (7.82 - 5.22 cos 0) (-2.61 sin 0) - (2.61 cos 0-1) (5.22 sin 0) ,a
(7.82 -5.22 cos 0)2
2.61COS0-1 ..
r (7.82 -5.22 cos 0)
which reduces to
15. 18 sin 0 2.61 cos 0 — 1
(7.82 - 5.22 cos 0)2 ' (7.82 - 5.22 cos 0)
e. (13)
If the driver were moving at constant velocity, the second term in
(13) would vanish. Since, however, we shall be considering a system
124 BOON, FELDMAN, AND STOIBER August
where the driving pin is not moving with constant velocity, the second
term will be of significant importance.
APPENDIX II
Spline-and-Slot Accelerating System
At a fixed distance Rz in Fig. 6, about A as center, we drive a slot at
constant speed. The slot engages a spline whose driving center is at
0' and is tightly coupled to the geneva drive pin. An angle of ro-
tation a. in the constant drive produces a rotation 6 in the output.
LOCUS OF SLOT MOTION
\ SPLINE
Fig. 6 — Slot-and-spline geometry.
From the law of sines we obtain
-^— = -^ in triangle AOB'. (14)
sin a sm 0
Also C2 = (Rt + 6)2 + R2* - 2R2 (R, + 6) cos a. (15)
But (#1 + &) = (R* ~ a),
whence C2 = (R2 - a)2 + R2* - 2R2 (R2 - a) cos a. ^ (16)
From (14) and (16) we have
. fi _ _ Rz sin a _ x17^
= {(#2 - a)2 + RJ - 2R*(R* - a) cos a}1/*'
R2 sin a
tan " *
fi, cos a - (ft - a)'
and cot.» = °T'-. (18)
Rz sm a
1948 TELEVISION RECORDING CAMERA 125
Equation (18) provides the necessary relation between 6 and a, i.e.,
the output and input angles of rotation. Once Rz and a are selected,
one has a well-defined relation for the two angles. The zero point of
operation of the system has been taken with the slot engaging the
spline at the closest distance of approach to 0'.
From (18) we have
= cot-1 foot a. - (l - -j~\ csc «"]. (19)
The first derivative with respect to time provides us with an angular
velocity relation ; namely,
1 _ (l « ) CoS «
RJ — a. (20)
-[(*-!)- 0-t)> *
-i)1 -'('-*)«"•]'
To obtain the angular acceleration relation we differentiate a sec-
ond time, noting that a is a constant, and obtain
(21)
From (11) in Appendix I we obtain the geneva velocity as a function
of its input drive. But the input drive of the geneva is governed by
the output of the pin-and-slot drive since the two are coupled. Thus,
we are able to treat the geneva motion in terms of the constant-ve-
locity drive. Given
2.61 cos 9 - 1 ,
* ~ 7.82 -5.22 cos/
and replacing 0 by (20) we obtain
2.61 cos 0 - v > ^ (23)
* 7.82 - 5.22 cos 0\ ' -^ ~ < "' (Z6)
Substituting for cos 6 in terms of a gives
2.61 [Rt cos a - (Rt - a)] - Z
7.82Z - 5.22[#2cos « - (R2 - a)}
a (24)
126
BOON, FELDMAN, AND STOIBER
where
= \/2(#22 - 2R2a) (1 - cos «) + a2.
In a similar manner we may obtain the angular acceleration equa-
tion of the geneva in terms of the constant-speed shaft rotation.
From (13) in Appendix I we have
15.18 sin e 2.61 cos 6 - 1
where
(7.82 - 5.22 cos 0)2
>-('-*)
(7.82 - 5.22 cos e)
(25)
COS a
-[('-!;) -('-£
cos e
and sin e
cos a — (Rz — a)
sin a.
Performing the indicated replacements in (25), we obtain
.. _ [" -15. 187^2^ sin a "I
*' = U7.82Z - 522(R2 cos a - (R* - a)] }2J
X
1 ~ (l ~ l£) COS "
(26)
_ f 2.61 [R2 cos a - (Hz - a)} - Z "I
U7.82Z - 5.22[«2 cos a - (R2 - a)] }J
The appropriate constants foi a 57-degree pulldown angle have
been substituted in the final equation and one half of the symmetrical
acceleration curve of the geneva plotted in Fig. 3. For ease of ob-
servation, the curve has been inverted and shifted from its true posi-
tion on the other side of the zero line.
Development of Theater Television
in England*
BY A. G. D. WEST
DIRECTOR, CINEMA-TELEVISION, LTD., LONDON, ENGLAND
Summary — This paper will give a historical review of the progress of theater
television projection in Great Britain, both before and after the war, and will
describe the design and performance of the equipment which has been de-
veloped for distribution and projection of television programs. It will also
indicate the proposals now being made for the setting up of a theater tele-
vision service in England, first in London, and then throughout the country.
I. INTRODUCTION
THE ENGLISH cinema exhibitor is somewhat bewildered regarding
the subject of television and how it will affect him in the future.
Previous to the war, certain cinemas had large-screen television equip-
ment installed where programs of a topical nature transmitted by the
British Broadcasting Corporation were reproduced on the screen to
paying audiences. Results achieved indicated that with the normal
course of technical progress, television projection could, in time, pro-
vide a picture equal in quality to that given by normal film projec-
tion. But the practical problem of the use of television for cinema
entertainment, particularly in respect to how programs could be
built using the television medium, was the subject of much conjec-
ture, and sufficient experience was not available then, nor is it even
now, to enable the exhibitor to obtain a clear view as to how such
entertainment would be organized and presented to the public. I
have spent some time trying to get the entertainment industry to
study the practical problems concerned with the successful utilization
of television in its applications to the cinema industry. I have spent
much more time, naturally, endeavoring to press forward with the
solution of the technical problems, holding the view that, in the
achievement of an acceptable technical result, the technician will have
carried out his part of the bargain; it may be that as this state of
technical perfection is more closely approached the clarification of the
program requirement will be accelerated, and we shall find the means
of revitalizing the cinema industry in a way which will be a source of
* Presented October 21, 1947, at the SMPE Convention in New York.
AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51 127
128 WEST August
satisfaction to both technician and exhibitor. Large-screen tele-
vision has provided a means of interesting and attracting the cinema
patron on certain special occasions. On the other hand, we are not
yet satisfied that, either technically or in respect to program value, we
can yet retain the permanent interest of the public. We are only
part of the way through our job. Let us, therefore, take stock of the
present situation, and obtain clarification on some of our problems,
and how best to attack them.
In the advent of a new art — this was exemplified when sound, aris-
ing out of broadcasting, was applied to cinema technique — we must
find how the new art can help existing practice and vice versa. Our
chief problem today in the cinema industry is to study how tele-
vision can help the cinema, and also how the cinema can help tele-
vision. This is the vital moment, when television is just beginning
to show its head, for the cinema industry to take into account in its
planning for the future, both technically and commercially, the in-
valuable aid which television can provide in the field of theater enter-
tainment. Television enthusiasts (we shall call them "tele-vision-
aries") who have made a close study of the commercial possibilities of
the use of television for the entertainment of cinema audiences, have
forecast that, provided a broad co-operative view is taken by all the
various entertainment interests, including those which promote sport-
ing and similar events, opportunities for expansion in the entertain-
ment industry can be considerable, and would fully justify the wildest
dreams of the most imaginative exploiters in the entertainment field.
The following observations aim at giving a review of the position
of theater television in Britain^ and a summary of the aims of the
technician in preparing for full commercial use large-screen television
equipment, and the means whereby programs can be provided for
such an equipment, together with a statement of the various aspects
which will need to be considered in detail by the exhibitor, between
now and such time when commercial equipment will be available on
the market. The paper is concerned only with black-and-white pro-
jection. We have nothing as yet to show on color.
II. REVIEW OF PROGRESS UP TO 1939
Early History of Large-Screen Projection
Up to the beginning of the war in 1939, home and theater television
progressed side by side during those prewar years ; therefore, a few
words should be said on the development of the home television service.
1948 . THEATER TELEVISION IN ENGLAND 129
The beginning of official transmission of television in England was
due to the dogged persistence of John Logie Baird, who, as a result
of his experiments and demonstrations, over the period from 1923 to
1928, was able to get the British Broadcasting Corporation to radiate
vision signals, first in 1929 by an experimental service, and later, from
August, 1932, in the form of a regular program service. These televi-
sion transmissions provided over the normal broadcast channels a low-
definition picture on a 30-line basis. Such a coarse texture of picture
rendered the transmission of small detail impossible, and the program,
although interesting, had little entertainment value. But it started
the ball rolling, and, as you well know, from 1933 onward work was
commenced in many laboratories in England, America, France, and
Germany, toward the development of a higher standard of definition.
The low-definition broadcast service ceased in September, 1935, and
its place was taken by regular transmissions from the new television
station at the Alexandra Palace, London, on a 405-line interlaced
standard developed by the Marconi and Electrical and Musical In-
dustries companies.
Home and Theater Television
During the three years from August, 1936, to September, 1939,
some 20,000 home television receivers were sold in the London area;
the majority of these incorporated direct- viewing cathode-ray tubes
with a picture size between 8X6 inches and 13 X 10 inches.
During this period there was very rapid development in the type
of program material. Not only was studio space at Alexandra Palace
considerably enlarged to allow a variety of studio programs to be
transmitted, including plays and variety (vaudeville) productions,
which involved the use of multistudio technique, but the range of
outside events was increased by the laying of a ring cable of the co-
axial type, connecting the more important points of entertainment
and interest in the London area, and also by the provision of mobile
equipment which linked such events as Rugby football matches, on the
outskirts of London, and the Derby at Epsom Downs 20 miles away,
with Alexandra Palace, for rebroadcasting from that station.
The hours of transmission for home screens averaged 18 hours a
week, usually one hour in the afternoon and two in the evening, and
the improvement in programs, particularly in respect to outside
broadcasts and actualities (such as cricket, tennis, and boxing matches)
was so considerable in 1938 and 1939 that the home televiewer had
130
WEST
August
exceedingly good entertainment. (Figs. 1 and 2.) So far the
London area was the only favored area, but plans were in hand
for the extension of the service to other centers of population. How-
ever, television was brought to an abrupt conclusion in Great Britain
on September 3, 1939, and from that date onward no transmissions of
any type took place until the London Television Service was re-
opened in June, 1946, following the report of the Government Tele-
vision Committee published in April, 1945, which recommended early
resumption and expansion of the
television service in London and
in the provinces. Contracts have
been placed for the erection of
the Birmingham and Manchester
Stations, which will, in the first
instance, act mainly as relays of
the London program. Eventu-
ally four more provincial stations
will be built, providing a home
television service by 1952 which
will be available to 75 per cent of
the population of Great Britain.
As you well know, commercial
sponsoring of programs, both
sound and television, is not tol-
erated, and, therefore, the pro-
vision of the service, which at
the moment costs the B.B.C.
half a million pounds per annum,
in return for which they receive
25,000 pounds per annum, being
one pound per set per annum
for the 25,000 sets already in
operation in the London area, does not appear, at the moment, to
be an economic proposition. But the B.B.C. remains undaunted by
this problem, looking forward to a reversal of the economic picture,
when the country is covered with the television service, and when
there are sufficient material and labor available to manufacture enough
television receivers, at a reasonable price, -to satisfy all requirements.
All programs, of course, do not satisfy all tastes, but the B.B.C. is
doing wonderful pioneer work, considering the limitations of space
Fig. 1 — B.B.C. television cameras at
the trooping of the color, Horse Guards
Parade, London.
1948 THEATER TELEVISION IN ENGLAND 131
and equipment. I am perfectly satisfied to receive two good pro-
grams a week, for example, a good play or a good variety show, or a
good sporting event, for my one pound a year license fee. My satis-
faction is, of course, subject to the somewhat selfish provision that
I am not required to look at programs which I do not want to see;
we are beginning to realize that home television can be a remarkable
time-waster, if rigid self-control is not exercised in switching "on" or
rather "off" the receiver.
The progress of home television in Great Britain has been referred
to in some detail, because in many respects, and particularly in rela-
tion to the provision of a program, the service is quite different from
Fig. 2 — B.B.C. television cameras at the Oxford and Cam-
bridge boat race.
what you have here and this possibly may indicate slight differences of
approach to the application of the theater television technique.
From 1930 onward attention was paid to the possibilities of produc-
ing television pictures for demonstration to larger audiences. There
were three main lines of development each of which had a practical
result: mechanical systems, intermediate film projection systems,
and cathode-ray-tube projection methods. To these can be added a
fourth, light-valve systems, which were being thought about without
yielding anything to indicate possible practical results.
Large-screen television was first demonstrated to the public in
Great Britain by John Logie Baird in 1930 at the London Coliseum
Variety Theater, when he used a screen of 2100 lamps, operated by a
132 WEST August
mechanical commutator switch to provide a picture 30 X 70 inches in
size. This novelty was retained in the theater program for three
weeks, and, therefore, we are justified in saying that it excited con-
siderable interest, although the definition was crude, but 'the bright-
ness was adequate. An extension of this system was demonstrated
in Berlin by Karolus, who employed a bank of 10,000 lamps arranged
in a square frame of 100 horizontal rows, each containing 100 lamps.
These lamps consisted of miniature cathode-ray tubes arranged in
individual compartments in the screen, and the illumination was pro-
duced by the excitation of the fluorescent screen on the end of the bulb.
The operation of the lamps was controlled by electronic switches in
which an electron beam was rotated over a ring of 100 contacts.
At the same time the old mechanical methods were pushed to the
limit, and in June, 1932, Baird gave a demonstration of the Derby in
a London theater using a three-channel transmission over a distance
of 25 miles, each channel providing 10 lines of a 30-line picture 9X6
feet in size. The projector consisted of a mirror drum with Kerr-cell
modulation of the light. These events are mentioned because we
must not forget the work of the old pioneers. By their spade work
they were able to lay the foundations and excite the interest of the
public, and thereby find the means whereby progress could be made
and better methods developed.
Early Color Demonstrations
Before leaving the reference to these mechanical systems, we must
mention the first large-screen color demonstration in Great Britain,
which was presented by Baird as part of a variety program in the
3000-seat Dominion Theater in 1938. Looking back at that demon-
stration, in which a two-color process was employed in providing a
120-line interlaced picture, we find that the results were remarkable,
considering the state of the art at that time.
Realizing then the limitations of the mechanical methods, we had
before us two alternatives for providing a large-screen picture. First
(Fig. 3) the intermediate film method which consists in photograph-
ing the television picture reproduced on a small cathode-ray tube on
to a film, which after rapid development, fixing, and drying can be
projected as a standard film through the usual 35-mm projector.
Second (Fig. 4) there was the cathode-ray-tube projection involving
the stepping up of the faormal television receiver of the home to a
higher power basis, so that intensely brilliant images of a size approxi-
mately 6 inches in diameter can be projected by an efficient lens or
1948
THEATER TELEVISION IN ENGLAND
133
ACTUALITY PICK-UP.
THEATRE PROJECTION.
Fig. 3 — Delayed large-screen projection by the intermediate film process.
Television picture recorded on film at A. Film processed and dried at B.
Film projected by normal projector at C. Delay, 5 minutes.
mirror system to the full cinema screen size. This problem of the use
of one or other of these methods, or of both of them, still exists today.
First let us deal with the prewar studies of the intermediate film
process, which was developed both in Britain and Germany, and dem-
onstrated to theater audiences in 1935. This has the advantage
that it is possible to provide the normal standard of brightness on the
theater screen, because the processed film passes through a standard
projector. The degree of definition achieved was reasonably good,
but the method proved to be somewhat expensive, because of the high
film costs incurred; and the attempt by our associated company,
Fernseh A. G. in Berlin, to use a continuous loop of film, which was
cleaned and resensitized in a continuous process in the intermediate
film projector, was not attended with success. The 60-second delay
in reproduction, due to the time of processing of the film, was not re-
garded as a serious defect. Such equipment in practice, however,
needed a very high degree of supervision, and the maintenance of the
ACTUALITY PICK-UP.
THEATRE PROJECTION.
Fig. 4 — Instantaneous electronic large-screen projection.
134
WEST
August
processing baths and of the mechanical parts of the projector was re-
garded as being somewhat beyond the practical limitations imposed by
the day-by-day continuous service of cinema projection. Neverthe-
less, there were many who had, as many do now have, faith in this
method of television presentation, because in addition to the possi-
bilities of increased brightness and definition it has the additional ad-
vantage that by putting the received television picture on film, a per-
DEFLECTOR UNIT
VISION AMPLIFIER
TIME-BASE GENERATOR
TALK-BACK AMPLIFIER
TUBE ANODE
CURRENT METER
FOCUS CONTROL
\
VISION CAIN
Fig. 5 — 1938 Baird cathode-ray-tube projector for the
10- X 7V2-foot screen in the Tatler Theater, London.
manent record is made in the theater and this can be used over and
over again in subsequent performances.
However, further development of this process was dropped and
efforts were concentrated on the cathode-ray-tube projection method,
which appeared to offer the most scope for future practical develop-
ment. It formed the basis of the equipment developed by the Baird
Company for installation in 1938 and 1939 in the theaters of the
Gaumont-British Picture Corporation.
Theaters Equipped and Programs Provided
Early in 1938, a small projector was installed in the Tatler
1948
THEATER TELEVISION IN ENGLAND
135
Newsreel Theater (Figs. 5 and 6). It housed a cathode-ray tube op-
erating on 30,000 volts, and reproduced an intensely bright picture
(3X4 inches in size) on the screen of the cathode-ray tube, which was
projected by an //2.5 lens on to a screen 10 X 7l/z feet. The illumi-
nation on the theater screen was of the order of x/4 foot-candle, and the
brightness, using a semireflecting screen material, of the order of l/%
f oot-lambert ; and demonstrations were given of various actuality pro-
grams transmitted on the 405-line basis by the B.B.C. These were
mainly in the form of private demonstrations, and for a small theater
of that type with a total seating accommodation for 650 people, the
results were regarded as eminently satisfactory. The equipment was
Fig. 6 — Projection in the Tatler Theater, London,
1938.
entirely of an experimental nature and could not be handled by any-
one but a specialist.
These results gave encouragement for further work in larger thea-
ters, and early in 1939 the Marble Arch Pavilion with a seating ac-
commodation for 1290 persons was equipped with a higher power, dual
cathode-ray-tube projector, using the pipe-shaped tube with metal-
backed fluorescent screen, operating on 60,000 volts, with a Taylor-
Hobson 12V2-inch//1.5 anastigmatic lens. (Figs. 7-10.) This pro-
vided an illumination of l/z foot-candle on a screen 15 X 12 feet with
a brightness of 1 foot-lambert in the high lights. This equipment was
used for special programs on a commercial basis for paying audiences,
and we well remember a red-letter day in large-screen projection in
February, 1939, when a much publicized boxing match (the Boon-
136
WEST
August
Fig. 7 — 1939 Baird twin cathode-ray-tube projector at the
Marble Arch Pavilion, London.
Danahar fight) was reproduced to an excited and enthusiastic audi-
ence who had paid up to two guineas (ten dollars at that time) for
their seats in this theater for this particular event. The audience
stood up and cheered on the conclusion of this fight, which fortunately
went the full distance. Not a single person asked for his money
back! The success of the results achieved led to the Gaumont-
British Picture Corporation (whose President then was Mr. Isidor
Fig. 8 — 1939 Baird twin cathode-ray-tube projector
at the Marble Arch Pavilion, London. (Front view.)
1948
THEATER TELEVISION IN ENGLAND
137
Ostrer, a man of considerable vision, to whom we owe much for his
encouragement of television in the early days), ordering twelve equip-
ments for installation in the larger London theaters. By September,
1939, the following theaters had been equipped with these projectors.
Marble Arch Pavilion 1290 seats
New Victoria Cinema 2564 seats
Gaumont, Haymarket 1382 seats
Gaumont, Lewisham 3047 seats
Tatler Theater 650 seats
Fig. 9 — 1939 Baird twin cathode-ray-tube pro-
jector at the Marble Arch Pavilion, London. (Rear
view of controls and cathode-ray tubes.)
The incidence of war prevented the equipping of other theaters
and thus the plan to have selected television programs presented at
twelve London theaters to a total audience capacity of approximately
22,000 was never realized.
138
WEST
August
At the same period Scophony, Limited, with its optical mechanical
system with the supersonic light valve, equipped the Odeon Theater,
Leicester Square, 2116 seats (Fig. 11), and certain news theaters, and
were attracting full audiences for special programs.
III. REQUIREMENTS FOR A THEATER SERVICE
Before continuing with the historical development since the war,
I would like to discuss briefly the requirements for a theater television
service.
The Complete Theater System
It is the ultimate aim of the television engineer to provide the enter-
tainment industry with a complete television system which can handle
mr
I
Fig. 10 — High-tension units at the Marble Arch Pavilion.
and distribute all types of program material which will be of interest.
The system and the equipment utilized therein can be conveniently
divided as follows:
(a) Pickup equipment consisting of cameras and associated equip-
ment for synchronizing control for interior (such as studio and
dramatic presentations) and for exterior (outdoor scenes) together
with the necessary sound pickup, lighting, and power supply.
(b) Film-scanning equipment.
(c) Control-room equipment, for the purpose of selection and rout-
ing of programs.
(d) Distribution network, utilizing special cables or high-frequency
radio channels.
(e) Theater television projectors and loudspeakers.
1948
THEATER TELEVISION IN ENGLAND
139
Fig. 12 (a charter or ideal for British theater television engineers)
indicates a possible system of pickup, control, distribution, and thea-
ter reproduction which is capable of dealing with events taking place
mainly in the London area, and of distribution not only to theaters in
London but also in the provinces. At the same time it comprises
provincial program sources also.
Progress after the termination of the war has been concentrated
under all the above head- 2
ings, and will continue un-
til there is evolved a satis-
factory system which ex- JBlb^
hibitors will welcome as a
valuable contribution to-
ward their theater enter-
tainment. The aim of the
technician, who is primarily
concerned with this aspect
of television, will be to
secure perfection independ-
ently in each of the divi-
sions of work enumerated
above.
Comparison with Film
Projection
The overriding problem
is, of course, the develop-
ment of theater television
projection to a form cpm-
parable to the present-day
film projection.
Such a program of work
can conveniently be visual-
ized in two stages:
(1) The attainment of the utmost possible performance in each link
of the 400- or 525-line system; alternatively the maximum possible
to the 3-megacycle bandwidth limit.
(2) The full equivalent to film projection (say 1000-line basis or
20-megacycle bandwidth, or whatever it may be found to be).
Fig. 11 — Scophony supersonic light-valve]
projector installed at the Odeon Theater,
Leicester Square, London, 1939.
140
WEST
August
Satisfying the Exhibitor
The exhibitor or promoter is our customer, and he presumably is
capable of visualizing a true representation of what the public will
require. It is our duty to satisfy him, if he wants it, by providing:
(1) Instantaneous projection in theaters, from a given distribution
center, of items of entertainment, of interesting events and actualities.
(2) Delayed presentation from the distribution center. For ex-
ample, daily films of local interest which are applicable to the theaters
in a local area.
LONDON Fi|m
PROGRAMME Sfudios
SOURCES [
West End Wemble
Theatre:; Stadiun
(Stage) (Footba
y Lord's Harrinqay Alb<
i Ground Arena Ha
1 (Crcket) /Boxing) "ffuV
rt Mobile
L. ^ts
Cable
Cable
Cable 1
Cablef
Cable
Cable
Wimbledon
( Tennis
V
DISTRIBUTION
CENTRE
(LONDON)
LON
THE/
r •" '
FILMS
• ' > 1 STUDIO 1 |
CO NT
g|
|
Cable
DON
0-RES
l\
Radio Pr
Provincia
ogrammeSources
< ^
1 Radio I Links I
PROVINCIAL THEATRES
I I I
Birmingham Manchester Glasgow etc.
(D<2)<D (D@(D © ©
Fig. 12 — Proposal for nation-wide theater television
network.
(3) Delayed presentation in individual theaters where the program
planning is impracticable to admit of (1), or requires re-presentation
additional to that given by (2).
All these needs must be provided with the qualities of normal film
projection.
There may be, there are sure to be, other requirements as well, but
for the moment we, as technicians, have many problems to solve (even
in black-and-white only), and they will take time. However, I must
emphasize that the theater owners and exhibitors must also spend this
1948 THEATER TELEVISION IN ENGLAND 141
period usefully in studying the possibilities and limitations of the ap-
plication of television and in trying to decide how they. want to use it
as a means of entertaining, attracting, and even educating the public.
There is no doubt that we are up against this problem of what to do
with television in the theaters, and we are entitled to ask, and to re-
ceive, an answer to this question, while we are working on our purely
technical problems, the solution of which is inevitable in the course
of time. I should not, however, like to be too optimistic at this mo-
ment by saying that we have ready a system which we can present
immediately to the exhibitor in the form suggested earlier in the paper.
It may be three years or it may be more before we can provide the
brightness and definition in the quality of the picture which will be
necessary, for the exhibitor to mingle his television with his film pro-
gram; but it may be that he will find it profitable to consider an inter-
mediate step whereby the television program can stand on its own
merit without achieving the full technical results of the film projec-
tion, and by segregating the television from the film program, or the
television theater from the film theater, can give us an opportunity of
gaining practical experience in the new technique.
It may well be, on the other hand, that the new art will not be con-
stricted to such applications, but will break out, with success, in an en-
tirely new medium of application, which we have so far not visualized.
I look forward to a more careful consideration of all the points by
those who are responsible for the provision of public entertainment,
and by such people who have the imagination and initiative to make
practical use of the new tool which is now being forged.
IV. POSTWAR DEVELOPMENT
Progress Toward Setting up a Theater Service
The keynote of the resumption of work on television in the
autumn of 1945 was set by the British Government Television
Committee, which issued a report in that year setting forth its deliber-
ations regarding the reinstatement and development of the tele-
vision service after the war, with special consideration given to (a)
the extension of the service to the larger centers of population, (b) en-
couragement of research and development, and (c) guidance to manu-
facturers of equipment.
Although 2*/2 years have elapsed, and very little has been done
under (a), (b), or (c), I should say that the report was extremely good
and showed that full consideration had been given by the Committee
142 WEST August
to the various aspects of television — technical, commercial, and
political.
So far as it concerns us, I should like to quote a few sentences from
the report. Thus, under the heading "Television in the Cinema,"
after stating that "The Committee had not been unmindful of the
potentiality of cinemas for displaying television programs," the report
went on to say:
"Before the war certain firms were interesting themselves in the
production of apparatus for this aspect of television, and a few cinemas
had acquired equipment capable of projecting a picture of large size
on the screen from a position in the stalls. Such apparatus was used
with some success on occasions when events of outstanding public
interest were televised."
Then it goes on to say later :
"We are encouraged to believe that the cinema industry and the
British Broadcasting Corporation Working in co-operation and not as
competitors in the exploitation of television, will achieve consider-
able results of a character beneficial to both."
And further :
"Although television in the home would compete with the cinema
for the public's interest, the extension side by side of the two forms of
entertainment should on the whole prove mutually helpful rather
than otherwise, and home and cinema television are likely to have a
stimulating effect on each other."
And, finally:
"We recommend that close attention should be given to the possi-
bilities of the use of television by cinemas."
This report was accepted by the British Government and issued in
the form of a white paper, but since that date no official pronounce-
ment has been given indicating that these recommendations have been
implemented in any way, or that any steps have been taken to give
effect to them or to encourage the cinema industry in its work on these
problems. I shall deal further with this point later on, but for the
moment I should prefer to submit to you details of the work which
has been done by commercial companies, and in particular by my own
Company with the encouragement of Mr. J. Arthur Rank, and of the
results achieved in the two years since we started thinking about tele-
vision again in the autumn of 1945. During this period, we have
seen the development and application of many new types of equip-
ment which have an important bearing on our work ; for example, new
1948 THEATER TELEVISION IN ENGLAND 143
types of television cameras, of scanners for film and still pictures, new
means of distribution by radio or by cable, and theater projectors
either of the cathode-ray-tube type with lens and mirror systems or of
the intermediate film type, or of the storage type.
Comparison with Cinema Standards
Before considering these in detail, let us consider five main head-
ings (which possibly can be regarded as separate factors, but which in
practice are all interlinked), to provide a basis of comparison with the
accepted standards of the cinema.
(1) Picture definition, or detail of the reproduced picture.
(2) Picture quality or faithful reproduction of the tone values, from
black through the half tones to white.
(3) Brightness of the reproduced picture, and its color.
(4) Freedom from interference, flicker, spurious patterns and effects,
shading, and background noises.
(5) Cost of manufacture of the equipment and of its installation and
maintenance.
Performance of Equipment
Let us now make a brief review of the various types of equipment
already developed on both sides of the Atlantic, demonstrated in Eng-
land, and also able to be manufactured.
(a) Cameras. You are quite familiar with the operation and char-
acteristics of the various types of television cameras, so I need not go
into them in detail, except to say that with the iconoscope we acknowl-
edge its superiority in definition, but also its limitation in the produc-
tion of undesired shading effects which cannot be controlled. The
image iconoscope has the advantage of a little more sensitivity than
the iconoscope and less shading troubles. The orthicon with its even
field of picture rendering is free from shading, but loses detail; and
the image orthicon with its enormously increased sensitivity, suffers,
however, from background noise and great difficulties in manufacture.
(b) Film Scanners. There are those, like the Mechau continuous-
motion mechanism, installed at the B.B.C., which use the iconoscope,
and therefore also suffer from shading distortions of the picture grada-
tion. There is the Farns worth dissector film scanner which gives an
even field, but is difficult to set up to avoid geometrical distortion of
the picture. And finally, there is the cathode-ray-tube flying-spot
scanner, which can give, under controlled conditions, as good a
144
WEST
August
picture as you would wish to see, with excellent definition and quality,
and free from shading.
(c) Caption or Still-Picture Scanners. The same remarks apply.
(d) Means of Distribution. By radio links, which can carry the full
requirement of frequency range, which are flexible in setting up and
operation, but which may be subject to interference.
By cable, with limited frequency band and high capital cost.
Fig. 13 — Lens and mirror electronic projectors.
MIRROR ARCLICHT PROJE<
f
OPENING 6 CLOSING
OF APERTURES IN SCREEN
CONTROLLED BY CATHODE
RAY SCANNING BEAM.
LARGE
SCREEN
FOCUSING SYSTEM
CATHODE CUN
Fig. 14 — Electronic storage projection system.
(e) Projection. Cathode-ray-tube projectors, either using a wide-
aperture anastigmatic lens or a Schmidt-mirror system, with its great
advantage. (Incidentally, I remember testing a Schmidt projection
system in 1937.) (Fig-. 13.)
Intermediate film projectors, in the operation of which much ex-
perience still is to be gained.
Storage projectors (described in principle in Fig. 14) of which only
one type so far has been shown to be reasonably practicable, namely
the AFIF system developed by Professor Fischer of Zurich. (Figs. 15
and 16.)
1948
THEATER TELEVISION IN ENGLAND
145
(/) Screens. Types of screens with higher reflection coefficients
than the normal matte white screen, such as the established types of
beaded or silver screens; new types of screens coated with material
which is a combination of matte white and silver; and lenticular types
of screens varying from the crudely stippled metal screens to the opti-
cally designed lenticular screens which project all the received light
back into the audience seating only. (Figs. 17 and 18.) A screen
Fig. 15 — AFIF storage large-screen projector.
having a reflecting cone with a vertical angle of .40 degrees and a total
horizontal angle of 104 degrees would be ideal for the average theater.
(Figs. 19 and 20.)
Arising out of the consideration of the qualities of the various types
of equipment referred to (lack of time prevents me from going into-a
detailed study), we have in my Company evolved an experimental
405-line system which has already been the subject of practical tests,
and which for the present consists of :
146
WEST
August
Telecameras
The image orthicon or image iconoscope.
Telecine
Cathode-ray-tube flying-spot film scanner. (Fig. 21.)
Telecaption
Cathode-ray-tube flying-spot still-picture scanner. (Figs. 22, 23, and 24.)
Teledistribution
Radio links operating with a few watts on a frequency of 480 megacycles at
distances up to 12 miles.
Teleprojectors
' A Schmidt mirror projector (Figs.
25 and 26) having a 27-inch di-
ameter mirror, and an 18-inch di-
ameter plastic correcting plate;
with an aluminum-backed straight-
through cathode-ray tube, operat-
ing on an anode voltage of 50,000; !
mounted in the stalls, or on the
front of the balcony; remotely con-
trolled from a console installed at
the back of the stalls or in the pro-
jection box. (Fig. 27.)
Theater Screens
Of a type where the reflected
light is concentrated into the area
occupied by the audiences.
I must now state the prac-
tical results achieved in terms
of the fundamental points of
performance which I have
specified above.
(1) The definition over the
whole system is such that 3-
megacycle vertical bars are re-
solved in the picture without
any noticeable phase defects.
(2) The measured high-light
brightness on a 14- X 11-foot
stippled metal screen in the
direction normal to the screen
down the center line of the
theater, is 5 foot-lamberts, compared with the accepted film stand-
ard of 7 to 14 foot-lamberts, and the measured black brightness j
is 0.1 foot-lambert; the average contrast range during a succession of
pictures is 30:1. At 30 degrees off the center line the high-light
brightness is 2.5 foot-lamberts. The output of light from the projec-
tor with no picture, and running at a brightness corresponding to the
maximum usable high-light brightness for a good quality picture i:
Fig. 16 — AFIF storage projector.
16, 17— Arc lamp.
1 — Eidophore oil film scanned in
vacuum by the cathode-ray beam 8.
4,5 — Heat-absorbing bars.
19, 20 — Lens and mirror projecting
the picture on to the screen 21.
1948
THEATER TELEVISION IN ENGLAND
147
300 lumens. A new projector, almost completed, will provide a light
output of 600 lumens, adjusted for conditions for good quality picture
projection. By the time we get into the London theaters we hope to
project 1000 lumens on to the screen. The color of the picture is off-
white in the direction of cream. Fig. 28 indicates the progress of
definition and brightness over the years, in comparison with the de-
sirable results which are equivalent to the average characteristics of
film projection in theaters. The important point about these curves
is that the upward tendency continues and there is no sign yet of a
slowing up of progress, which might be indicated by a flattening of
the curves.
VMWT WHITE
CBVSTAL BEADED
Fig. 17 — Polar diagrams showing reflectivity
of types of screens at various angles with the
normal to the screen, compared with that of the
matte white screen which is represented by the
circle of radius 1.
(3) The estimated over-all quality curve is approximately linear over
the range from black to two thirds the high-light brightness specified
in (2) above, and flattens out above that figure. For example, we
have measurements indicating a brightness curve for the projector as
follows :
A gamma of l/z in the shadows, caused mainly by scattered light.
A gamma of 2 over the greater part of the curve up to 2/3 high-light
brightness.
A gamma dropping to 1/2 at the upper values of brightness due to
electron-beam defocusing at high current, saturation in the fluorescent
powder, and other causes.
This distortion, if measured correctly, can be mostly made good to
provide an over-all constant-gamma condition.
(4) With regard to freedom from interference, I must admit that there
148
WEST
August
is much to be desired with existing standards, and with relatively un-
controlled local noises. Under the best conditions these can be a rela-
tively unimportant factor, but on occasions the interference may be
troublesome and cause annoyance to the spectator.
Proposals for an Experimental System for the London Area
The complete system described above, and which is in practical
operation in an experimental form, and can be engineered in a form
suitable for the production of a serviceable instrument, is, in my
i A ; 5-!O
•I2'5
H und«f O-S
Fig. 18 — Pictorial representation of screen-brightness
distribution in a theater for various types of screen, subject
to a given value of illumination.
opinion, a first practical solution which we can offer to the cinema in-
terests. It is a long way ahead of the 1939 equipment. It is up to
them to decide how, where, and when they can use it to advantage.
Our recommendation is to set up a sample system in daily operation
for invited and paying audiences. Fig. 29 illustrates a plan of a pro-
posed experimental system which we hope to work put during 1948 to
1948
THEATER TELEVISION IN ENGLAND
149
give us this experience. You will see that programs are to be pro-
vided from three centers, the B.B.C. Studios at Alexandra Palace in
the north of London, the Pinewood Film Production Studios of the
Rank Organization to the west of London, and the Crystal Palace site
PROJECTOR
CINEMA SCREEN^
ELEVATION OF CINEMA
Fig. 19— Elevation of average theater showing the angle
required for screen reflection in the vertical plane.
Fig. 20 — Plan of average theater showing the angle required for
screen reflection in the horizontal plane.
on the southern side overlooking London, where we shall set up a cen-
tral receiving station and retransmitting station, and some local
scanners for the transmission of films, interviews, and announcements.
The radio links will be on frequencies just above and below 480 mega-
cycles. Retransmissions will be beamed, from the Crystal Palace,
with an angle of 10 degrees in the direction of certain theaters which
150
WEST
August
are suitable for the install ation of the pro j ec tion equipment . We have
in mind four West End theaters and two suburban theaters. One
beam will suffice to cover the London West End cinemas and a se-
lected northwestern suburban cinema, and another beam will cover a
suburban cinema in the southeastern area of London.
Fig. 21— "Cintel" flying-spot (35-mm) film
scanner. (Scanning tube in left-hand box, optical
system and gate in center, multiplier phototube in
right-hand box.)
Figs. 30 and 31 are elevations of two of the selected cinemas show-
ing our proposals for equipping them.
Audience Reaction
So far, nothing has yet been done, so far as I know, on either side
of the Atlantic, which would give the exhibitor some practical figures
1948
THEATER TELEVISION IN ENGLAND
151
and experience to gauge future public requirements. We badly need
experience on public reactions to a regular service beyond the stage
when television was just a novelty and used only on special occasions.
We recently invited a cross section of our employees to see a pro-
jected B.B.C. program (lasting l-1/4 hours) in the cinema which we had
equipped with the projection installation described above. These
Fig. 22 — "Cintel" flying-spot caption scanner. (Scan-
ning tube below, lens and phototube box in center,
monitor tube above.)
employees had been working some distance away in other factories,
and had not seen any large-screen pictures before. The entertain-
ment value of the program projected happened to be poor. This was
beyond our control, but we were surprised at the tone of the response
to the questionnaire which was circulated to each employee after-
wards, asking for impressions. The total number concerned was 264
and there was a good mixture of technical, clerical, and administrative
152
WEST
August
Fig. 23 — Still picture transmitted on 405-line basis by "Cintel"
caption scanner.
(nontechnical) staff, bench workers, wiring and assembly girls, and
glass workers, and the following is the analysis of the voting papers :
Fig. 24 — Still picture transmitted on 405-line basis by "Chite!"
caption scanner.
1948 THEATER TELEVISION IN ENGLAND 153
AUDIENCE REACTION FIGURES
1. The picture was generally: better than expected 129
as expected 83
not so good as expected 52
total 264
2. The picture was : adequately bright 165
not bright enough 99
total 264
3. The detail was: just sufficient 79
not quite enough 145
not nearly enough 35
nonvoters 5
total 264
4. The picture caused: some eyestrain 158
no eyestrain 101
nonvoters 5
total 264
5. The picture: is good enough 163
is not good enough 96
nonvoters 5
total 264
,o justify reproduction on large screens of certain events of general interest, i.e.,
he Boat Race, Test Cricket, football, etc., to paying audiences.
So much for what might be termed an average audience.
Now to come to an audience of enthusiasts, where there is no ques-
ion at all regarding the practicability of utilizing large-screen tele-
dsion up to its present degree of performance.
We were invited at the beginning of the month to assist at the Con-
servative Conference which was held at Brighton on the south coast of
England, in the large 3000-seater Dome, built in the oriental style, by
ing George IV in 1805. As an attendance of 4000 persons at the
Conference was anticipated, room had to be found for an overflow
neeting in the Dolphin Theater, a 1000-seater, and about 500 feet
iway from the Dome. We set up image-orthicon cameras, manufac-
,ured for us by the Du Mont Company, facing the platform in the
Dome, one for the close-up of the speakers and the other for a general
/iew of Mr. Churchill's Shadow Cabinet on the platform. The picture
vas reproduced on a 14- X 11-foot screen on the stage of the distant
•heater, on a 405-line basis, with a Schmidt projector operating on
154
WEST
August
40,000 volts, and giving a high-light brightness down the center of
the theater of 4 foot-lamberts. To enable the delegates to study their
agenda papers there was about */* lighting left on in the theater. We
had accepted this invitation to gain experience, and we certainly did
get that experience from the enthusiastic Party representatives, espe-
cially on the occasion when their Leader was speaking. Throughout
the three days of the Conference, the theater was filled to overflow-
1
Fig. 25— "Cintel" mirror projector
for 16- X 12-foot screen. (Front
view.)
Fig. 26 — "Cintel" mirror projector
for 16- X 12-foot screen. (Rear
view.)
ing, and many of the visitors preferred the close-up of the speakers on
the large screen to the more distant view in the large Dome. On
the last day, I sat at various points in the theater and took Leica
snapshots, at exposures of l/w of a second, using an f/2 lens and Super
XX film, of the large-screen results, and I am happy to be able to pre-
sent some of the results here, ahd I am more than happy that they
represent one who is, I believe, regarded throughout the world, anc
even in Britain also, as one of the greatest leaders of our time. (Figs
32 and 33.)
1948
THEATER TELEVISION IN ENGLAND
155
Installation and Regulations
Finally, there is one factor not to be ignored ; the installation prob-
lem, especially in relation to national or local regulations which, when
originally framed, did not envisage the use of television in theaters.
^
a
\
t
bfi
U
^5
a
/„
H
S
H
> S3
. LU
2
^
CZ
r
/
UJ
>
O
\
> "1
f a:
_1
> —
> — i
tti
o
a?
<
• a.
111
Z2
J
o
^O
i ...
2
^
•xv
J
1
u.
a:
U.
\
UJ
\ —
H
u
—7 —
is
a
X
LU
Of
I
a
j
3
M*
\
7p
. T- <
! H-
> ^.
> >
—J
a
I
01D3TC
o
0
UJ
Q
UJ
a
PTICA
oFOCUS
UJ
S
a.
II
It
UJ
a
ft
Z
0
tn O
Z
-
a
|
—
-4
:
11
_ O SH
Ot
a<
^ - "^
>
UJ
C
|i
£? 0*
Z u.
®
i
w
2
X
K
HI
H l|
s §,
D J
z
o
^
a.
3
UJ
O 1
^-H S
o ^
5lc
I
x
^4
,'
t
H
-^6
<
o
S
<
8
UJ
"•o
•!
®i
js>
2
u
]{
5d
1
i,
J
z
^ T3
O— i
>
1
•
tt-< cs
9
2 1
Ul <
O
at
Of
c
i!
Mj
4 1
®
S
> ^
s
J3»A
UC
Ul
c
iE5 o
3
UJ,
'S
_i
CO
Of
Or
/>
>
u
Oh
1
.
J
"o
a
B
5
x
LL
11
"
3
a
—^
E
3
<
t
1 —
"~" — - -
__.
t
n
_ _ _ _ __ _ _
S *
o •*•*"
|.|
UJ
SOUND
0 0
ECEIVE
1 VIDEO
o
Q
Z
<
Q
o (J)
u z
z E
1
FRAME
Tiimc 0
°BASE0
0
O
_J
ii^ll
.S3
1;?
CD
IfL
f.
u
I
a
UJ
i
J
IF
1
U
1*
1
_j
SOUND
H
u
.ri
j
3
8
1
•
S
1
U)
^4-
5
^
' r
••
«
E
'
0
0
00
S
(N
/^
^v
1
156
WEST
August
In Great Britain the authorities are busy drafting more and more
new regulations. Everything has to be regulated. The old original
Cinematograph Act of 1909 (amended only once since that date in
1923 and before the advent of sound, and still legally in force) would
close half the cinemas in the country if the letter of the law were ob-
served. A new amendment of the old Act is now in preparation, and
has been drafted, and would, according to the exhibitors, close all
cinemas. Clauses have been drafted in anticipation of the installa-
tion of television equipment, and in such a form (I should say without
malice aforethought) that would make it quite impossible to install
FOOT
LAMBERTS
NUMBEP
OF LINES
AVEPAGE CINEMA PROJECTION
800
600
500
4OO
250
CINEMA PROJECTION EQUIVALENT
1938 1939 1946 1947 1948 1949 S5O
1938 1939 1946 1947 1948 1949 I9SO
Fig. 28 — Progress of brightness and equivalent definition in large-screen
(16- X 12-foot) television projection.
television in cinemas. For example, the draft stipulates that a tele
vision projector set up in the theater must be completely surrounde
by 14-inch brick walls without any doors. We have visions of th
projectionist being built in with the projector and remaining there a
the rest of his life. But I must admit that the authorities are, how
ever, open to suggestions for improvements in the regulations.
In actual practice, we have never had any difficulty in satisfyin
local authorities from the points of view of safety and fire. We hav
found them most co-operative and as anxious to gain experience ir
the new type of equipment and its installation as we are.
1948
THEATER TELEVISION IN ENGLAND
157
V. PRESENT PROBLEMS "
(REQUIREMENTS FOR THE IMMEDIATE FUTURE)
I have dealt with the present state of the art in Great Britain. I
may have painted, perhaps, too rosy a picture, but I prefer to be an
optimist, recognizing that we still have a long way to go. Our present
problems are as follows :
(1) Technical
We have to improve detail, quality, projection brightness, and free-
dom from miscellaneous minor but irritating defects. I prefer to
PALACE
ision Studios
STUDIO
(Telecameras)(Telecin<)(Contiol>
Fig. 29 — Proposed experimental cinema-television-distribution plan
in the London area. (For 1948.)
group all these points together and to refer to some of the funda-
mental problems associated with all of them.
(a) Number of Lines for Theater Standardization. We have seen
many references to the 1000-line desirability. On the other hand, we
have often heard that our 405-line system at its best is enough. That,
of course, refers to a controlled local picture. Therefore, ignoring for
the moment all the excellent work which so far has been done in try-
ing to establish the minimum basis for either home or large-screen pro-
jection, we decided to start afresh and^make a practical investigation
with many observers, of the brightness-resolution — contrast relation-
ship in projected pictures.
158
WEST
August
Some of the preliminary conclusions are given in Fig. 34 which show
the result of observations made on line patterns of various dimensions
exhibiting varying degrees of contrast and illuminated at various
values of brightness. The curves in the diagram connecting bright-
ness resolution and contrast should be taken as indicative of the order
of magnitude involved where the unit of relative brightness represents
the normal high-light brightness of a projected picture, say approxi-
mately 10 foot-lamberts. The number N of test lines per picture
height is equivalent to the number KN of lines of television scanning,
where the factor K lies between 2 and 3. Curve A indicates that the
HIGH TENSION UNI
Fig. 30 — New Victoria Theater, London. Proposed television
installation with the projector in the orchestra stalls.
eye can appreciate up to something between a 950- and 1400-line pic-
ture at a brightness of 10 foot-lamberts; but in practice, according to
curve D, the result of observations of projected films, it is satisfied with
something between a 650- and 950-line picture at that brightness.
Arising from this, it appears to be desirable that we should aim at a
standard of something round about 900 lines for theater television, and
up to 1200 lines, if we wish to record a picture on film which will pro-
vide prints equivalent to normal film practice.
(6) Systems of Scanning. We have got too much into the way of a
tacit acceptance of double interlacing, based on a theoretical calcu-
lation of its advantages. I am not at all sure that practice has proved
this.
At the recent Cannes Conference it was generally agreed that the
1948
THEATER TELEVISION IN ENGLAND
159
time was ripe for a renewed investigation of sequential processes. In
fact, all the authorities there admitted, as a result of their practical
experience of results using interlaced scanning, that they would prefer
a 500-line sequential picture at 50 frames per second to a 1000-line
interlaced picture at 50 frames, 25 pictures a second.
The following defects are observed in interlaced scanning: line
crawling, interline flicker, spurious pattern flicker, line breakup on
movement, pairing or loss of interlace, unequal field brightness, ir-
regularities and irritating effects on vision, and complexity of circuits
•and equipment. •
PROJECTOR
REMOTELY CONTROLLED
-.FROM CONSOLE
HIGH TENSION UNIT
Fig. 31 — Gaumont Theater, Hay market, London.
Proposed television installation with the projector on
the front of the balcony.
Some of these also appear with sequential scanning, with the added
disadvantage for a given channel bandwidth of greater "lininess" and
lower definition.
The list of interlacing defects is formidable, and indicates the reason
for disquiet as to the future of interlacing in improved television sys-
tems. The advantages, however, such as terms of improved defi-
nition, are not to be lightly disregarded. The final choice, to interlace
or not, cannot be decided without further observational data.
A number of various comparisons can be made, but they all resolve
themselves into a choice between either a loss of definition or the pres-
ence of flicker and stroboscopic defects. Other factors which will
160
WEST
August
Fig. 32 — Mr. Churchill speaking by large-screen
projection (14 X 11 feet).
require attention in this investigation are the compromise between
vertical and horizontal resolution, and the value of artificial means for
line broadening to reduce "lininess."
In drawing your attention again to sequential scanning, I should like
to mention that recently we made an equipment to demonstrate the
Fig. 33 — Mr. Churchill speaking by large-screen
projection (14 X U feet).
1948
THEATER TELEVISION IN ENGLAND
161
principles of scanning a picture and reconstituting it, for the Science
Museum in London, in connection with the Electron Jubilee Exhibi-
tion. We employed a scanning of 100 lines sequential, and the repro-
duced picture had a remarkable element of stability; in fact the rigid-
ity of a lantern slide, and we were not unduly bothered by the limita-
tions of definition due to the low number of scanning lines. In my
opinion, in introducing interlaced scanning we have deliberately tried
500
NUMBER
OF
TEST
UNES
PtR
PICTURE
HEIGHT
300
A. LIMIT OF VISUAL ACUITY * (MEAN 7 OBSERVERS)
B. 35 M.M STANDARD RELEASE PRINT (tow POWER MAGNIFICATION)
C. AS B, SINGLE FRAME STILL PROJECTION VIEWED ON SCREEN
0 AS C, BUT STANDARD VICWIN6 DISTANCE *
# VIEWING DISTANCE « 4-X PICTURE HEISHT
—RELATIVE BRIGHTNESS
Fig. 34 — Investigation of brightness-resolution — contrast characteristics.
Data for center of field, with low-contrast test object. (Density difference
0.3, equal line width — line spacing.) Constant "average" brightness of picture.
to deceive the eye, and the eye will not stand to be deceived, and it is
in this connection that we shall find advantages when we come to
achieve any system of storage projection.
We have made some interesting tests, originally out of curiosity
more than anything else, to compare the results of projecting, one after
the other, an intermediate film picture and an interlaced electronic
picture on the large screen, of the same subject scanned with the same
number of lines, and we were remarkably surprised at the amount of
irritation, as you might describe it, produced by the interlaced pro-
jected television picture on the eye, in comparison with the steady
162 WEST August
restfulness of the projected intermediate film picture. I have an idea
that here we have a vital point regarding vision which needs much
more study; and, furthermore, that an electronic sequential picture
will occupy an intermediate place between the other two regarding
the general stability and freedom from irritation (and from consequent
headaches) desirable for large-screen projection.
(c) Channel Bandwidth. We have had to change our minds during
the last two years regarding the amount of intelligence which can be
carried on a 3-megacycle channel. Now we find that we are able to
squeeze much more apparent detail and quality into a channel with a
definite cutoff at 3 megacycles, and we have been remarkably sur-
prised at the general increase of performance which has been achieved
by correcting for response, phase, gamma, and other requirements
throughout the whole system within this limitation of frequency. Up
to now for the 1000-line transmission, the bandwidth of up to 20 mega-
cycles has been mentioned. We believe that we shall achieve all we
want to do by concentrating on obtaining the maximum value that
can be obtained on a channel up to 12 megacycles only.
(d) Quality of Picture. We have been in the past, I feel, content tc
have seen occasionally, when all conditions were right, a picture oi
good quality, and then to feel that we had achieved a result which
would be universally acceptable. It is only recently that a full study
has been made of the component and over-all linearity of the system
and that steps have been taken to correct errors in gamma. This
process of gamma control, which ensures that the relative brightness
of parts of the reproduced picture bear a linear relationship to the cor-
responding parts of the picture being scanned, is of vital importance
in ensuring a picture of first-class quality. It is only when a system
has been set up which complies reasonably well with this condition and
registers an over-all gamma of about 1 that one realizes the enormous
improvement in general quality of the picture. As regards projection,
I am convinced that so far no projector of any type complies with this
condition. As previously mentioned, there is a distortion of the
gamma curve, particularly in the high-light region, and this must be
corrected, first, by studying each element of the system in turn, and3
second, by applying -an over-all correction when each element has
been improved as far as it will go.
(e) Picture Brightness. In the cathode-ray-tube projector the curve!
connecting brightness with anode voltage on the cathode-ray tube;
and the curve connecting brightness with beam current, both
1948 THEATER TELEVISION IN ENGLAND 163
saturation, which begins at a certain high-light brightness on the
viewing screen. The problem of extending the brightness curves is
one of the most important that we have at the moment. This in-
volves the following studies:
(i) The development of optical systems of the mirror type to even
greater efficiency than the Schmidt.
(ii) The development of tube electronic characteristics so that de-
focusing is controlled with an increase of voltage and current.
(iii) The development of a fluorescent material and its application
to the face of the tube, studying in particular the problems of high-
current saturation, defocusing, and halation in the layer; and also its
color and life characteristics.
(iv) The development of the viewing screen providing more econom-
ical use of the light projected on it, so that it is reflected back where
it is required and not dissipated throughout the theater.
(2) Distribution Systems
Considerable study has been made of the relative advantages and
disadvantages of cable and radio means of distribution. On behalf
of the radio link, we find lower capital and running costs, more flexi-
bility in operation, and against it the scarcity of channels, and inter-
ference ; on behalf of cable, a clear and undisturbed channel (at
least we hope so), and secrecy; against cable, the high cost of in-
stallation, resulting in high rental charges, and the length of time be-
fore the installation can be carried out, due to higher priority for in-
stallation labor. In Great Britain, both radio and cable links are
controlled by the Postmaster General, and in the setup of a radio sys-
tem of a permanent nature, such a system would most likely not be
licensed for commercial operation, but would be taken over by the
Post Office to operate in whatever manner it thinks fit. However, the
exceedingly high charges for the rental of coaxial cables (something in
the nature of £600 per mile per annum for a 3-megacycle cable has
been quoted) with no definite date of availability promised within the
next five or ten years, makes it imperative to provide experimentally a
radio-link system, and the first steps in this direction already have
been described. In the meantime, the first link in the provincial dis-
tribution system of B.B.C. television programs has been started.
Work has commenced on a radio link between London and Birming-
ham to operate on 900 megacycles.
Although you are, for your own commercial cinema schemes,
164 WEST August
pressing for allocations of frequencies above 1000 megacycles for radio
links, we are pressing for the 500- to 1000-megacycle band, because this
range offers, in our opinion, advantages for wide-band television which
may not be possible in the regions above 1000 megacycles.
(8) Program
Here we have many problems, the majority of which are outside
our technical province. I have already referred to some of them.
Others causing us much thought in England are as follows :
(a) License to Operate Commercially. Over two years ago we asked
the government to consider giving us facilities to operate on a com-
mercial basis between our studios and theaters. The permission
is concerned with the means of transmission and distribution. In
other words, we ask for a license to use the ether or the facilities pro-
vided by Post Office cables. In this respect we are dependent on the
Television Advisory Committee (which has taken the place of the
original Television Committee), and this Advisory Committee has
been taking plenty of evidence during the last two years but has been
very slow in making the appropriate recommendations to the Post-
master General who would present them, if he agreed, to Parliament
for latification.
(b) Three-Cornered Interests. It may be that although the report
of the Television Committee advised that steps should be taken to-
ward the encouragement and establishment of a television service for
cinemas, the delay in the granting of a license to operate commercially
has been mainly due to the difficulty of getting together in agreement
the three interests who are mostly concerned :
(i) The B.B.C. and its home viewing audience.
(ii) The promoters of sporting events, some of which can be classi-
fied as being of a national nature.
(iii) The cinema interests.
Therefore, if these three could be got to work together in harmony,
with full co-operation in the provision and exchange of program ma-
terial, the authorization of a license which would give the cinema in-
terests a start in commercializing television might be forthcoming.
However, pressure in this direction is bound to come when technical
results are obtained, which justify in themselves that a perfected in-
vention of this nature should be utilized for the nation's benefit. In
any case, as the price of home television receiving sets is for the time
being higher than the purchase level of the majority of the population,
iy4o
THEATER TELEVISION IN ENGLAND 165
is not television in the cinema the average man's way of participating
in this form of entertainment?
(c) Place of Television in the Theater Program. What do theater
interests intend to do with television? This is a question which, as
mentioned before, needs very careful study of all factors by the enter-
tainment industry. I have not yet heard a balanced and well-
thought-,out reply to this problem.
Are we wrong in assuming that large-screen television and cinemat-
ographic projection can be made complementary to each other?
Can we show them both in the same program?
On the long or very long view, the answer is yes.
But in the meantime, those who have financed its development must
be thinking of some return. In which case, can we commercialize on
an intermediate stage either by (i) provision of specialized television
theaters; or (ii) provision of television in cinema theaters, but tele-
vision and film each taking a separate and independent program
period for itself.
(d) Instantaneous Versus Delayed Action. I am not at all clear as
to the relative uses of instantaneous electronic projection of television
in theaters, and the delayed-action presentation by using the inter-
mediate film process. There are so many factors controlling the tim-
ing of programs in theaters that it would be extremely difficult to
guarantee that all theaters taking a particular program would be
standing by at exactly the correct moment. On the other hand, I
cannot visualize the practical operation and maintenance of inter-
mediate film equipment in individual theaters.
There is one thing of which I am quite certain. I have many times
experienced the tenseness of an audience watching, as it is taking
place, on the cinema screen, a national event, the outcome of which is
unknown, and I am convinced of the enormous entertainment value
of such an item. The satisfaction which I personally have experi-
enced ori such an occasion has been acknowledged also by all those
present. The important point is that the event is being watched as it
is happening, and half the entertainment value would be lost with
delayed presentation.
I feel, however, that the best way out of this problem is not by writ-
ing and talking, but by setting out to obtain practical experience in
both methods over a period of time; such work to be done in close co-
operation between the technicians and the leaders of the entertainment
industry, and it is only by facing this problem fairly and squarely that
166 WEST August
we can really get a solution that will satisfy future requirements in
the provision and extension of the cinema television service.
(4) General Economic Problem
Although I am not qualified to discuss this subject, I feel that this
is a matter which must not be left unmentioned in a general survey of
this nature.
In looking ahead, as the technician must look ahead, toward the
future of the entertainment industry and the impact of technical
progress on it, we must attempt to visualize the various possibilities
which may arise, so that we can provide information for those whose
duty it is to study the economic trend in relation to the ever-changing
needs and tastes of the public served by the industry. Here we have
in large-screen television a new tool rapidly approaching the practical
stage where it can be of value for entertainment and education. It
is our duty to give guidance, as far as we can, so that it can be used to
the best public advantage. I hope that this paper will, in describing
past experiences, and in discussing present problems and future possi-
bilities, make some contribution in this direction.
VI. CONCLUSION
Finally, I should like to put to you a few questions, based on my
remarks, and eagerly await your considered replies :
(1) Do you agree that the presentation of large-screen television to
the public should be made in two distinct stages?
Stage (i) on the 400- or 525-line basis, or the 3-megacycle band-
width limit.
Stage (ii) the equivalent to film projection. Or should we wait until
Stage (ii) is an accomplished fact?
(2) What do you regard as technical requirements for Stage (ii)?
(a) Number of lines.
(b) Sequential or interlaced.
(c) Bandwidth.
(3) Should we really make a comparison with film projec-
tion? Should not public television develop as a different medium,
and to a different standard?
(4) What will be the comparative practical uses for :
(a) Film intermediary.
(b) Instantaneous electronic projection.
(5) How will theater television be used by the entertainment
interests?
1948
THEATER TELEVISION IN ENGLAND
167
ACKNOWLEDGMENTS
I must give credit and appreciation to the members of my team, who
work with unequalled enthusiasm and unity of purpose in endeavor-
ing to solve our problems: Messrs. T. M. C. Lance, J. D. Percy, T. C.
Nuttall, L. C. Jesty, L. R. Johnson, K. A. R. Samson, E. McConnell,
M. Morgan, J. E. B. Jacob, and many others; also to Dr. C. Szegho,
now with the Rauland Corporation; and to Dr. Starkie of Imperial
Chemical Industries (Plastics Division) who has carried out the opti-
cal work for projection; and Mr. Warmisham of Taylor Hobson for
the optical work on the scanning side.
NOTE : Following the delivery of his paper, Mr. West showed a
short film, divided into two parts: (1) the recording of B.B.C. pro-
grams with particular reference to faults encountered, and (2) the
recording of pictures transmitted by "Cintel" film and caption
scanners.
DISCUSSION
MR. FRASER: Was the film that was just shown a 16-mm film?
MR. A. G. D. WEST: Yes, reversal 16-mm film.
MR. ERASER: Has anyone attempted to count the number of scanning lines to
see if there are 405 or 202 Va?
MR. WEST: I think it is likely that there is quite a lot of pairing. As I have
mentioned, we are concerned about the difficulties of correct interlacing. We have
an idea that not more than 10 per cent of receivers interlace properly. Does that
hold on this side of the Atlantic?
MR. SIEGFRIED: What is the greatest projection distance that Mr. West has
employed with practical results, and what is the largest picture?
MR. WEST: The projection distance maximum is 40 feet from the screen. We
have not been farther back than that.
MR. BEN SCHLANGER: There was one of the theater diagrams which you showed
which had a television projector on the face of the balcony. It did seem to me as
though that was one of the best locations. The projector did not seem to obstruct
the view from any part of the audience. It seems to be the most practical job. Is
that true?
' MR. WEST: Yes. We fully agree with that. We should be very pleased to see
theaters which had balconies coming out to, say, 50 feet from the screen. That
would be the ideal position for the projector, but few existing theaters satisfy that
condition. Most of them are 70 or perhaps 90 feet back.
MR. SCHLANGER: In contemplation of building a new theater, might that be
the course to follow?
MR. WEST: Yes, sir. Obviously it would be preferable to have the projector
back in the box which is the right place for it, but in determining the best position
for the projector in the auditorium we are subservient to the economic cost. I be-
lieve that we could produce a large projector with a 40-inch mirror which could
168 WEST
be put in the projection box, and would provide sufficient brightness, but it would
be very expensive. The glasswork alone might cost about $16,000 and further-
more the production output would be very slow. I believe it is the same over here.
It might take two years to produce one only. It, therefore, appears to be neither
an economic nor a practical proposition.
We, therefore, have to compromise with a smaller mirror, smaller dimensions of
projector, and a smaller throw distance to secure the brightness for a given size
screen. If theaters are to be designed for the purpose of large-screen television,
then the balcony should be designed so that the projector can be mounted on the
front of it at a distance not more than 50 feet from the screen.
MR. PAUL J. LARSEN: I agree that the front of the balcony is a very nice
position for the projector, but there is, in my opinion, a very much better place
where the projector can be placed in theaters without disturbing the seating ar-
rangements or anything else, and that is by hanging it from the ceiling. It can be
supported there very rigidly and solidly, and projecting downward to the screen.
In that way you could have your control box located in the projection room or in
the balcony, and that would not be taking up any space in the orchestra stalls.
MR. WEST: That is an interesting point of view. I think that we are rather
afraid that our roofs are not strong enough to support the equipment. There is
the question of servicing the projector also.
MR. LARSEN: It could hang from the ceiling most of the time just like a chan-
delier, and it could be lowered to the floor by pulley rope when servicing is required.
MR. WEST: Would not roof vibration cause trouble?
MR. LARSEN: I do not believe that it would be serious. I do recall some tests
made some time ago in projecting still pictures that way. Naturally you would
not depend on a single rope but you may use a triangular rope arrangement which
would hold it quite steady.
MR. SCHLANGER: Would not that be in the line of the film projection in the
projection room?
MR. LARSEN : You would place it at an angle so that it would not be.
MR. SCHLANGER: That might require quite a steep angle from the television
projector to the screen in order to get above the regular beam of the motion picture
projector.
MR. LARSEN: It would not be any worse than trying to have it down in the
orchestra and trying to project it up on to the screen.
MR. SCHLANGER: That position in front of the balcony that I saw in Mr.
West's diagram was practically a straight throw.
MR. LARSEN: That holds true where a balcony is available, but where a bal-
cony is not available, then the only place you have is in the orchestra stalls, and
you have to project upward at quite an angle. With my suggestion you could
project downward.
MR. SCHLANGER: Every theater will present a separate problem.
Theater Engineering Conference
Acoustics
•
Auditorium Acoustics*
BY J. P. MAXFIELD
VAN NUYS, CALIFORNIA
Summary — This paper presents a review of the factors affecting the
acoustic properties of auditoria. Emphasis is placed upon not only having
these factors meet the previously accepted requirements for technical excel-
lence, but those factors which contribute to the esthetic or dramatic effect,
particularly as regards shape of the auditorium and the diffusion of reflected
sound.
THE DESIGN OF a theater or other place of amusement partakes
of both the arts and the sciences. The final result must be
"pleasing" and capable of permitting the performance to arouse, in the
audience, a maximum of esthetic or dramatic effect. Herein lies the art.
The mechanisms by which these results are obtained, such as the
acoustic properties, the clarity of vision, and others, represent the
factors which are directly amenable to the methods of engineering.
Fortunately, the correlation between some of the objective factors
and the esthetic value or "pleasingness" of the artistic result can be
determined and then can be used to guide the engineer and the
architect in the best use of the known objective factors.
The purpose of this paper is to review the acoustics of auditoria,
with special emphasis on the motion picture theater, and to outline
the best conditions for the dramatic and esthetic presentation of the
program.
Professor Wallace Sabine laid the groundwork for the many later
developments in architectural acoustics. He studied methods of
controlling the reverberant characteristics of auditoria and the
dependence of these acoustic properties on the nature and amount
of sound absorption present in the theater.
The introduction of radio broadcasting, electric recording and re-
production of sound, and later talking pictures gave acoustical engi-
neers considerable opportunity to study the requirements for "pleasing
* Presented October 24, 1947, at the SMPE Convention in New York.
AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51 169
170
MAXFIELD
August
acoustics." The "single-channel" transmission, as contrasted with
normal binaural listening, tended to accentuate all acoustic effects,
and therefore rendered them more amenable to detailed study.
The results of such studies indicated that "pleasing acoustics"
resulted when the following broad requirements were met :
(1) The magnitude1"3 of the reverberation time and its frequency
characteristics4' 5 must lie within reasonable limits.
(2) The first discrete reflections from surfaces close to the source
must be carefully controlled and dispersed. They should reach the
audience area from a number of relatively small splays rather than
from a few large flat walls.
160
200
800 1000
NUMBER OF SEATS
1200
1400
Fig. 1
(3) The decay of sound essentially must be logarithmic, but
modulated by a large number of intensity variations brought about
by the shifting interference patterns during decay.
(4) The reverberation should consist mainly of reflections which
reach the listening position indirectly and with relatively long time
lapses.6
The recent literature regarding requirement (1) is in reasonable
agreement.
Requirements (2), (3), and (4) imply the use of means to direct and
to disperse or diffuse the sound. Volkmann7 and others have carried
the idea of diffusion to an extreme by the use of a large number of
poly cylindrical surfaces.
In view of the work reported by Hanson8 in 1931 and the experience
supporting the desirability of essentially logarithmic decay, the
1948
AUDITORIUM ACOUSTICS
171
question arises: Is there an optimum amount of diffusion from the
point of view of "pleasing acoustics?" Hanson showed that a log-
arithmic decay devoid of fluctuations due to the shifting interference
pattern did not yield a natural sound. If, however, he modulated
such a decay with the intensity fluctuations of a shifting interference
pattern, made up of a large number of small intensity variations, he
obtained a sound which showed auditorium "character" and was
"pleasing" to the listener.
It would seem, therefore, that the time has arrived to examine the
control of the discrete first-order reflections, the amount and kind of
- viS "u
\
MAC NAIR VALUES
tU -jOJ
"
SB.
in"" £ 100
S
X
512 ~ V
'ALUE
FROtv
MU
1
N
= LET
SON
CHER
DATA
& -
0-1 < 20
1 11
0.2 0.6 1.0 2 46 10
FREQUENCY IN XIO3
CYCLES PER SECOND
20 40 60 100 200
VOLUME IN CUBIC FEET
1.4
1.0
0.6
0.2
600
12
t- z
i§
II
iu^
£5
£~
> m
UJ H
E<
160
Z 140
cr 120
- 100
80
2 o iG ^ o>
" O O O O
RATIO OF REVERBERATIO
TIME (AT 256rv) TO 512A/
IN PER CENT
MAC
NAIR
12€
>~ VALUE
^
"
C
ILx^^
•^
MAC
NAIR
25
6^
u v;
VLUE
^^^
51;
>~ VALUE
— =f=p
512~ VALUE
•*«*
S
1
•
0 20 40 60 100 200 500 200 100 60 40 20 1C
XlO3 VOLUME IN CUBIC FEET XlO3
Fig. 2
diffusion, and the possible desirability of permitting a suitable amount
of nonuniform distribution of energy among the natural modes of the
theater to give "character" and "pleasingness" to its reverberation.
However, there are other important factors in theater design which
must be considered in producing an esthetically satisfactory audi-
torium.
Four of the important factors in theater design are
(1) The basic shape, which deals with the general relationships
between the length, the width, and the height.
(2) The volume or size, particularly in relation to the seating
capacity.
(3) The general reverberation characteristic.
172
MAXFIELD
August
(4) The shape, size, and position of the individual internal sur-
faces to control the proper distribution and dispersion of the sound.
NOTE: The four broad requirements apply to the last two factors.
THE BASIC SHAPE
Frequently this is influenced by the size and shape of the land
available as the site of the theater. Fortunately, the acoustic re-
PLAN
Fig. 3— New Kleinhaus Music Hall, Buffalo, N. Y.
quirements allow the architect a considerable leeway in the shape of
the floor plan. Experience has shown that a ratio of length to width
which lies between the limits9 2 : 1 and 7 : 5, when combined with
proper internal surface design, will yield excellent hearing conditions.
AUDITORIUM ACOUSTICS 173
Where the length becomes greater than twice the width the design
ds to approach the so-called "shooting gallery" shape with the
ulting difficulties of avoiding multiple reflections between the side
walls and the high attenuation over the audience heads. On the
other hand, a ratio of length to width of less than 7 : 5 approaches too
close to a square, in which a number of the natural modes of the room
tend to have nearly the same frequencies.
This statement of limiting ratios does not imply that the floor plan
should be a rectangle but merely that its general average dimensions
should lie within these limits. In general, it is desirable to avoid a
rectangle.
The ceiling height is largely controlled by the choice of the number
of cubic feet of volume to be allowed for each seat. However, for the
7 : 5 ratio of length to width it should not exceed one half the width,
while for the 2 : 1 ratio it should be less than two thirds the width.
Having chosen the ratio of length to width and the volume per seat,
as discussed later, the method of computing the height suggested by
Rettinger10 can be applied.
VOLUME PER SEAT
For the motion picture theater it is desirable to keep the volume
per seat low in order to minimize the amount of acoustic treatment
and sound dispersion necessary.
Fig. 1, from C. C. Potwin,1 shows desirable limits of volume in
cubic feet as a function of seating capacity. These values are based
on controlling the sound reflections and hence the reverberation by
proper shaping of the internal surfaces. Regarding Fig. 1, Potwin1
says: "These limits have been developed as a result of empirical
practice, and assume (1) the use of upholstered seats with a spring-
or rubber-cushioned bottom and padded back, (2) fully carpeted
aisles, and (3) furred construction of walls and ceiling for low-fre-
quency absorption. The broken curve is considered generally from
past experience to be a maximum practical limit for the auditorium
structure. In most cases a small amount of acoustical material
properly distributed will be required for these larger volumes."
Volumes greater than shown by the dotted line tend to result in
large flat surfaces which must be broken to disperse the sound properly
and lead away from economy in the planning of the theater.
GENERAL REVERBERATION CHARACTERISTIC
The literature covering this phase of design is quite complete and is,
174
MAXFIELD
August
generally, in good agreement. Fig. 2 shows the reverberation times
for various frequencies as a function of theater size. This figure sum-
marizes the data published by Maxfield and Potwin.5- 1
INTERNAL SHAPING
In rectangular rooms or theaters the sound energy, during decay,
tends to concentrate in certain well-defined modes of vibration.
PLAN
SECTION
Fig. 4
While this tendency is less in nonrectangular shapes, some additional
dispersion is necessary to cause a relatively smooth logarithmic decay.
This additional dispersion tends to lower the mean free path, that is,
to decrease the time interval between successive reflections.
The nature of the successive reflections determines the so-called
"character" of the theater and the time interval between them is
interpreted as "size."
It is generally recognized that when a theater has a "pleasing char-
acter," the esthetic value of the show performed in it is enhanced.
1948
AUDITORIUM ACOUSTICS
175
This is particularly true of musical renditions although it also applies
to speech.
It follows, therefore, that an amount of diffusion which destroys
this character may be undesirable since this excess diffusion or dis-
persion also causes a decrease in the time interval between successive
sound reflections. Also, it acoustically decreases the apparent size
of the auditorium. It -has been the experience of the author that too
much diffusion produces auditoria or theaters which are "character-
less and cramped" and that such theaters are legs pleasing as places of
entertainment.
Fig. 5
Fortunately, this amount of diffusion is not necessary to obtain
good definition for speech even when the reverberation time is suffi-
ciently long for good music production or reproduction. By careful
internal shaping to reflect the higher frequencies from numerous small-
angled surfaces on the side walls and ceiling, the definition and "pres-
ence" can be maintained without using too low a time of reverberation.
Fig. 3 shows diagrammatically* a plan and elevation of Kleinhaus
Music Hall in Buffalo, N. Y. The author and the late C. C. Potwin
* For this figure see Arch. Acoust., Design reprinted from Arch. Forum, September,
1939.
176 MAXFIELD August
believed that the amount of dispersion used in this design represents
the maximum "sound break-up" consistent with maintaining good
"character."
Fig. 4 shows similar diagrammatic sketches* of a motion picture
theater to which this type of sound diffusion and control of first
reflections has been applied (Normandie Theater, New York, N. Y.).
Where the internal shaping can be carried out as completely as shown
above, for instance, in remodelling an old theater, effective diffusion
can be obtained by distributing the necessary acoustic treatment in a
random manner over the walls and ceiling. The use of small absorb-
ing areas well distributed is superior to the use of a few large ones.
Completely covering any one wall with absorbing material is bad
practice.
Fig. 5 shows a studio t in which most of the sound diffusion was
odtained by the intelligent random distribution of absorbing material.
* Also in same Arch. Acoust., Design reprinted from Arch. Forum, September,
1939.
t See Fig. 4 of "The control of sound in theaters and preview rooms," by C. C.
Potwin, J. Soc. Mot. Pict. Eng., vol. 35, pp. 111-126; August, 1940.
REFERENCES
(1) C. C. Potwin, "Control of sound in theaters and preview rooms," J. Soc.
Mot. Pict. Eng., vol. 35, pp. 111-126; August, 1940.
- (2) P. E. Sabine, "Acoustics of sound recording rooms," Trans. Soc. Mot.
Pict. Eng., no. 12, p. 812; 1928.
(3) V. O. Knudsen; "Architectural Acoustics," John Wiley and Son, New
York, N. Y., 1932.
(4) W. A. MacNair, "Optimum reverberation time for auditoriums," J. Acous.
Soc. Amer., vol. 1, p. 242; 1930.
(5) J. P. Maxfield and C. C. Potwin, "Planning functionally for good acous-
tics," J. Acous. Soc. Amer., vol. 2, April, 1940.
(6) C. C. Potwin and J. P. Maxfield, "A modern concept of acoustical design,"
J. Acous. Soc. Amer., vol. 2, July, 1939.
(7) J. E. Volkmann, "Polycylindrical diffusers in room acoustic design,"
J. Acous. Soc. Amer., vol. 13, p. 234; 1942.
(8) R. L. Hanson, "Liveness in rooms," J. Acous. Soc. Amer., vol. 3, p. 318;
1932.
(9) C. C. Potwin, "Building Types Section," Arch. Rec., July, 1938.
(10) M. Rettinger, "Applied Architectural Acoustics," Chemical Publishing
Company, Brooklyn, N. Y., 1947, p. 75.
DISCUSSION
MEMBER: Does the slide that showed the volume per seat still hold?
MR. JOHN VOLKMANN: Yes, we still adhere to that. It is very desirable to
1948 AUDITORIUM ACOUSTICS 177
keep the volume down, but not down too low. You want to obtain enough re-
flection from the walls and the ceiling to give an enveloping effect of the sound.
If you make the room too small, below 100 cubic feet per seat, you get the
ceiling down very low. As was mentioned in the paper, you can get a room that is
too long relative to the width of the room. You can also have a room that is too
long relative to its height. Then a number of problems enter into the picture,
not only the cutting down of the liveness of the room by making the volume
per seat too small, but you get into the difficulty of projecting enough sound to the
last rows of seats.
For all practical purposes, I believe that the analysis that was made from
that original data still holds. It is a very acceptable guide.
MR. JOHN K. MILLIARD : It has been our experience where these so-called smaller
theaters are being operated that there is this feeling of better over-all co-ordination
of the sound and picture, and we feel that both from the production and reproduc-
tion standpoint, it is highly desirable to hold the volume down well within these
shaded areas.
MR. NEILL WADE: There was a mention of smooth logarithmic decay of the
sound in connection with the shape of the auditorium. It is still not clear to me
whether the rectangular shape tends to cause this smooth logarithmic decay or
whether it is the departure from flat walls which causes this type of decay; and
second, is this smooth logarithmic decay accepted practice today, or do we find
that we need something else — a departure from that — to get 'a pleasing result for
the listener?
MR. MILLIARD: I think a departure from the rectangular room is necessary
for the sake of diffusion; as Mr. Volkmann brought out, and as Mr. Maxfield
indicated in his paper, it is necessary to provide sufficient random distribution of
the sound so that reflection from any one surface is small compared to that coming
from the loudspeaker at the screen. That gives us the so-called presence that we
talk about and desire, in other words, restrict each individual flat surface down
to an area where the total energy from this surface reaching the listener in the audi-
torium is small as compared to that directly from the loudspeaker.
COMMENT BY MAIL FROM AUTHOR: Mr. R. L. Hanson* has shown, some years
ago, that a completely smooth logarithmic decay of sound is unpleasant. Ex-
perience in theaters has demonstrated, however, that the logarithmic decay modu-
lated by a sound interference pattern consisting of a large number of low-intensity
modulations produces the most pleasing effect.
MR. CHARLES LEE: A theater auditorium in which you have a patchwork
series of absorbing and reflecting surfaces, such as that horrible looking example
flashed upon the screen, would not be acceptable to the theater audience but it
might be to the studio recording. I find it very difficult to follow the formulas
in one theater after another that would yield the ideal results for the auditorium.
We have had proposed, from time to time, a series of variegated surfaces, and if
we did use them once or twice, you then are confronted with the architectural
problem of not having complete repetition for every auditorium.
CHAIRMAN HARVEY FLETCHER: Are you asking how you can make the audi-
torium satisfactory when there is a small. audience and also when there is a large
audience?
* R. L. Hanson, "Liveness of rooms," /. Acous. Soc. Amer., vol, 3, 1932.
178 MAXFIELD August
MR. VOLKMANN: If possible, that means that we should try to keep down the
variations in absorption in the room as the audience increases. The obvious
thing is to get seats which are as absorptive as possible, so that when the occupant
comes in and occupies the seat, he covers up about the same amount of absorbing
material, you might say, as his clothing contributes to the room. That I think
is the ideal procedure. That means getting seats that are satisfactory, i.e., that
have an absorption value of about 3l/z Sabines. Then the 4.2 units of the person
do not make a very great difference in the reverberation time.
CHAIRMAN FLETCHER: Some of you may have been in our little auditorium at
the Bell Laboratories. When that is full, the characteristics are scarcely any
different than when you have one person in it, and it is arranged so that the
seats have the same absorption as the persons when they are sitting in them.
MR. P. B. ONCLEY: In defense of the last slide which was flashed on the screen
— you didn't read the title at the bottom. That was the photograph of the studio
before the installation of decorative covering material. It wouldn't have to look
quite that bad.
MEMBER: There has been a tendency in the early history of acoustical design
for theaters to be dictators of what the places are going to look like, and 1 think
that is what has bothered Mr. Lee and many of us. I think as time went on we
finally found out pretty much what that slide teaches us, that there are a lot of
panels in a room. When seen together they certainly are ugly looking, but I
think the successful method has been to cover them with an over-all masking
whose appearance is up to the architect. This masking covers both the absorptive
material and the nonabsorptive surface, so the layman does not know where the
absorption is at all.
MR. W. E. MACKEE: We are building a number of theaters of less than 600
seats, 570 or 580, and our auditorium will be standard. The auditoria will
be 50 feet wide. The actual seating capacity will be about 90 feet, height 27
feet. It is costing us, completely equipped, about 115 celnts a cubic foot or $30.00
a square foot. We have to be careful with our seat spacing, and we are going to
use 33 inches, and about 15 feet in the back for standing room.
I am checking up on my architect. I am just asking whether 50 feet wide and
about 90 feet deep and 27 feet high is an ideal size for a theater of something over
525 seats, depending on the space between the seats.
MR. HILLIARD: According to rough calculations, he has 202 cubic feet per seat.
An over-all average of 18,000 theaters in the United States shows 125 cubic feet
per seat, so it look,-, as if you have a few more cubic feet than you need for the
optimum performance.
MR. MACKEE: These are designed for exhibitors who will not do more than
$1200 or $1500 a week. When we add one foot to the theater in length, it costs
us about $6000. We have to watch carefully or the exhibitor will not be able to
pay us, and we are going to have a theater on our hands. We want to cut it
down to the limit, otherwise we shall be in the theater business, and we are in the
banking business.
MR. E. J. CONTENT: According to Mr. Rettinger's figures, an ideal house for
600 seats would be a house about 48 feet wide, 88 feet long, with a ceiling 18 feet
high average.
1948 AUDITORIUM ACOUSTICS 179
MR. MACKEE: You are talking about just the seating area? Is the lobby
included?
MR. CONTENT: Not including the lobby.
MR. MACKEE: We are pretty near right then.
MR. CONTENT: You have too much ceiling height.
MR. MACKEE : These are one-floor theaters, and the projection room is up on a
sort of balcony.
MR. CONTENT: You could drop it enough by proper study, to reduce the
ceiling height sufficiently to obtain a reasonable cubic content per seat.
MR. MACKEE : I could knock off $10,000 or $15,000.
COMMENT BY LETTER FROM AUTHOR: The value of 200 cubic feet per seat is
definitely too high for best acoustics in a motion picture theater. The author
agrees with the comments from Mr. Milliard and Mr. Content and believes that
by lowering the ceiling height materially at the screen end of the theater, and
properly designing its shape, the situation could be much improved. If the projec-
tion booth can be lowered, somewhat, without damaging projection, this would help.
MR. JAMES FRANK, JR: Mr. Volkmann made a remark about the absorption
characteristics of the seats, that ought to be amplified. There is no question
that the ideal seat should have the amount of absorption he stated, which I pre-
sume means a soft covering on both the seat and the back.
On the other hand, we have to be practical from the point of view of main-
tenance, and I imagine a very large majority of seats in theaters have an imitation-
leather covering, at least on the seat portion. We know of course that in the
smaller and less expensive theaters, they use plywood backs, but a good many
people insist on imitation-leather covering on the seat. Some people have the
habit of cutting the covering, and it is easier and less expensive for the theater to
maintain its seats by recovering with imitation leather than with mohair or some
other soft fabric.
However, that is a very definite factor in connection with the acoustical condi-
tion of the theater. Is it possible to give some relative proportion of a theater
seat with soft covering on both the seat and the back as compared with an imita-
tion-leather covering on the seat and a soft back?
MR. VOLKMANN: Any range from 1.8 to 3.5 does give fairly satisfactory condi-
tions to meet that problem of variation of reverberation time with an audience.
The figure of 3.5 I quoted was to obtain as near ideal as possible, with the type
seats that are available. In other words, a very de luxe house does put in seats
of that type, but the average is more nearly around 1.7 or 1.8, and in between
there, 2.5 is another type of seat that is very common in a de luxe type of house.
CHAIRMAN FLETCHER: What is the absorption value of this imitation-leather
seat, without the back being covered?
MR. VOLKMANN : Without the back of the seat being covered, but with a cloth
upholstering on the back, the value is about 1.8 units.
With reference to the architect's prerogative of covering the absorbing mate-
rials, I believe that Mr. Maxfield probably deliberately left off the covering on
that slide in order not to intrude on the architect's field. Furcher in regard to
that, I believe the architect has all kinds of opportunities to express the artistic
phases of his work, in doing more things with regard to these diffusing surfaces.
There has been a tendency in the studio field, to arbitrarily accept a great deal
180 MAXFIELD August
of curve paneling of the cylindrical type, which from the fundamental basis is
not the only shape that will give adequate diffusion for the reflection surfaces in
the room, and I should like to make the appeal that the architects do a lot along
the line of creating designs of conversely shaped surfaces for the diffusing parts of
the auditorium.
The other point I wanted to bring out is that many of our calculations on the
acoustics of the auditorium are based on the reverberation time, but we must not
forget the echo problems in the room, and the one which gives us the most trouble
is that the rear wall usually is the most offending surface from the viewpoint of
echo, and echoes can be very localized to the degree where a few seats in the
auditorium can be highly disturbed by echo, and the rest of the auditorium will be
all right. I can think of a number of examples in the theater where we have had
that kind of problem.
MR. CHARLES LEE: This touches on a very important item in connection with
auditorium design, and that is that wherever a soft absorptive material is used
below the height of five or five and one half feet, children attending the show
frequently take delight in attacking these surfaces. Can the experts offer any-
thing other than perforated material, such as we are getting tired of seeing, that
will give us the reflection and the absorption in the spots that they believe are
technically desirable, and give us a chance to get them out of the way of the
young folks?
CHAIRMAN FLETCHER: Are you asking whether they get a new kind of material
or a remedy for the children, preventing them from cutting it?
MR. LEE: I do not think we want either one. We should like to know how we
can place these materials so we will avoid the reflective spots he cautions against.
MR. CONTENT: There is very little use in putting absorptive material any
lower than five or six feet. The space above represents the largest absorptive
surface in the entire auditorium. Absorptive material should be used on the
walls above five or six feet and hard surfaces down below.
MR. P. H. THOMASON: I noticed that Mr. Maxfield's paper omitted the treat-
ment of the rear wall, that is, backstage areas. Is not that very important to
obtain good acoustics in an auditorium, where you have the reflection on t
backstage, also the projection-room ceiling?
COMMENT BY MAIL FROM AUTHOR: It has been well understood for some yeard
that the backstage requires some treatment. Frequently the nature of this
treatment depends specifically on the type of sound system used and its positioi
on the stage. Therefore it was deemed wiser to leave the specification of this
treatment to the engineers responsible for the installation of the sound system.
MR. HILLIARD: There has been no question throughout a number of year
about backstage treatment. To a person in the audience the sound coming fron
the loudspeaker should be high and the general reflected sound should be low bj
comparison. For that reason, it is almost 100 per cent necessary that the back
stage wall be treated to a very large extent, depending upon the distance or rela-
tion of the loudspeaker to the rear wall. This avoids what we call slaps. If th<
loudspeaker were a part of the rear wall, then this damping would not be necessary
However, the farther out from it the speaker is placed the more drastic must b(
the treatment.
MR. WETHERELL: The speaker said that as the stage wall recedes from th(
1948 AUDITORIUM ACOUSTICS 181
loudspeaker, more acoustic treatment becomes necessary; as a general rule,
smaller houses will just allow a little over four feet between the screen and the
back wall of the stage to take care of the speaker unit. I was wondering whether
that four feet is enough of a depth.
MR. HILLIARD: I should consider that some treatment would still be necessary.
Others feel that this distance might be within the minimum not requiring any
treatment.
MR. WETHERELL: Where you have suspended walls, is it not necessary to use
some sound-deadening material in the back of those walls to keep the walls from
vibrating?
MR. CONTENT: Sound-isolation construction?
MR. WETHERELL: No, furred walls and metal lath and plaster; where you have
no intermediate supports, more or less a suspended wall, you might say, is it not
necessary to treat in back of those walls with rock wool or sound-deadening ma-
terial to keep the walls from vibrating?
MR. CONTENT: Ordinarily no, they do not vibrate to such an extent, because
there is enough mass in the lath and plaster to prevent violent vibration. There
may be some absorption, but it will not be too selective.
MR. ONCLEY: Mr. Maxfield pointed out that such construction added to the
low-frequency absorption.
MR. JOSEPH J. ZARO: Could any of the experts comment on a practice which
seems to be rather new, of using materials such as we have draped here in this
auditorium, on furring strips or auditorium insulation, but without any other
means of sound reinforcement behind the material itself?
CHAIRMAN FLETCHER: You are asking what effect that will have, or is it de-
sirable to do it?
MR. ZARO: Do they have any figures or data on the acoustical value of using
that material in that fashion?
CHAIRMAN FLETCHER: Such as this room, without anything back of it?
MR. BARONIK: Most installations of that type are not too good. The cloths
that are hung are usually rather sheer, and the absorption that takes place is at
the high frequencies only. This means that in many of these rooms, especially
with parallel side walls, you get a booming effect. I do not know how heavy this
particular material is. In many installations I have seen, you get a booming,
oarrel effect from that sort of construction, and if you want to preserve more
iniformity, you should have a material which absorbs more uniformly with
Tequency.
MR. BEN SCHLANGER: In the paper given, and in past practice, there has been
i desire shown by acoustical experts to introduce surfaces and forms near the
screen which will reinforce the original sound source. This usually leads to a con-
stricting opening at the picture, and does not allow for future expansion of what-
iver might happen at the screen end of the auditorium, that is, to get angular or
xmcave surfaces that will reflect the sound out into the auditorium. You cannot
lelp but have that go inward toward the optical center of the auditorium. Is this
practice essential, and could something else be done so that the auditorium can
>e left as wide as possible at the screen end, with some other approach to the
>roblem?
MR. CONTENT: I believe that much work remains to be done in co-operation
182 MAXFIELD August
between the acoustics engineer and the architect. The engineer himself cannot
design a complete theater. He can only say what acoustic materials should go
into the theater, and if it is not done in absolute co-operation with the architect,
you do not end up with something esthetically as well as acoustically correct.
We shall find, by working together with the architect, that we can open up the
front end of the theater to provide the effect there.
UNKNOWN: I wonder why the architect would like to have a wide front in the
auditorium. It seems contrary to my conception; it makes the screen look much
smaller than it would look if the auditorium tapered down toward the screen. It
does not provide for extra seats that are of any value.
MR. SCHL ANGER: I agreed with you years ago, but the reaction from important
clientele indicates, and I am beginning to believe it myself, that anything that
tapers toward the picture makes you particularly conscious of the closure and gives
a constricting feeling, and a more abstract effect is obtainable if you do not ob-
viously point a funnel toward the picture.
The specific question was whether the surfaces that we have been introducing
are absolutely essential to the amount of reinforcement we get out into the audi-
torium, whether that reinforcement is replaceable by other means which can be
accomplished.
MR. VOLKMANN: In the case of sound motion pictures, I refer to Mr. Ryder's
comments this morning; he advocated treating the whole front proscenium area
with absorptive materials and making the live end of the room in the rear. In
the case of sound motion pictures then, where you have amplification, you can
eliminate the necessity for the side walls, if I have understood you.
MR. SCHLANGER: That is a new approach, because it has been in the other
direction. For years we have been told to reinforce the sound at the screen end.
CHAIRMAN FLETCHER: May that not be an economic problem, if you have to
use four or five times as much power to get the sound out there; if you put ab-
sorptive materials back there and do not have this reflection — you see that prob-
lem comes into it too. That is what you would have to do.
MR. HILLIARD: About ten years ago, there appeared in literature material in
which considerable emphasis was placed on getting the architect to make the
surface convex instead of the customary concave, and we still believe that this is a!
very potent factor. By making the surface adjacent to the screen diffusing, we
can still build up the sound level in the auditorium, but we do not create the
focal points that were present in some of the earlier types of construction where
the curvature was concave. Since then many theaters have been built with the
surface curved so as to diffuse the sound. We have had excellent results, and I
believe that as an architect you will find that there are many, many ways in which
you can accomplish this" purpose and still maintain the objective that we are
seeking. You can do it with small curves, large curves, and decorate it in a man-
ner so that it does not look like a funnel. You can blend that in with the front
part of the auditorium, so the audience does not appreciate that it is being done.
MR. ONCLEY: Most of the talk has centered around the motion picture thea-
ters. This is a very special problem. However I think we might point out that
in many of the music halls, especially the opera halls of Europe, you have good
acoustics with stages as wide as the whole front of the theater, stages as deep as
1948 AUDITORIUM ACOUSTICS 183
the whole body. I should not think it would be necessary, in fact, to have a nar-
row stage in the moving picture theater.
COMMENT BY MAIL FROM AUTHOR: The author notes with interest the remarks
by Mr. Ben Schlanger. It is true that in theaters and auditoria in which no
sound-reproducing or amplifying equipment is used, it is important to shape the
stage end of the theater to aid sound reinforcement in the audience area.
There are two reasons why the author has recommended the narrow front, in
the past, for motion picture theaters. First, it is of material assistance in obtain-
ing a low value of cubic feet per seat. Second, he believed, as Mr. Schlanger says
he once did, that the architects desired that form for esthetic reasons.
There is no definite acoustic requirement for narrowing the theater at the
screen end where sound-reproducing or amplifying systems are employed, but it is
highly undesirable to obtain this breadth by employment of front-wall surfaces
which are concave toward the auditorium. Also, by the proper shaping of these
end-wall surfaces it becomes unnecessary to use large amounts of sound-absorbing
treatment on the walls immediately adjacent to the screen.
Theater Engineering Conference
Acoustics
•
Quieting and Noise Isolation*
BY EDWARD J. CONTENT
STAMFORD, CONNECTICUT
Summary — The purpose of this paper is to describe some of the ob-
jectionable noises, their causes, some of their remedies, and to point out
that it is much easier to avoid these troubles in building a new theater than to
rectify them in an old one.
NOISE is ANY undesirable sound, but all sounds do not have the
same effect on a person, and low tones, of course, are not nearly
so objectionable as higher tones. Howard Hardy recently said in
The Frontier: "A sound source of many component fragments will
sound much louder than one of the same intensity which has a pure
tone." Considerable confusion exists among inexperienced ob-
servers about the particular psychological factors with reference to
noise. It has been shown that noise of a frequency below 500 cycles is
not nearly so objectionable as noise consisting of high-frequency tones
and harmonics.
The object in noise reduction in design is to shift the objectionable
sound from high to lower frequency, as well as to lower its intensity.
Loudness alone is not an indication of the annoying effect. People
do not object to noisy machinery as much as to erratic and unex-
pected sources of sound. Such things as high-frequency screeches
are definitely more disturbing than the low frequency of thuds or all
the lower tones.
There is no doubt that high noise levels in theaters require the
operation of the sound system at a higher level, and even though the
audience does not realize that the sound level is higher than other-;
wise would be the case, it does put them under a nervous tension, and j
if the noise is extremely high, the sound level has to be so much higher j
to overcome the noise that it really becomes annoying.
There are several misconceptions that should be explained. Many i
* Presented October 24, 1947, at the SMPE Convention in New York.
184 AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51
QUIETING AND NOISE ISOLATION 185
people do not differentiate between sound isolation and acoustical
conditioning. Sound isolation fundamentally consists of two things —
soundproofing of solid-borne noises such as shocks and machinery,
and the sound insulation of air-borne noises such as is provided by
thick walls and special construction, which prevent the sound from
being transmitted from one point to another.
On the other hand, acoustical conditioning consists of three factors,
the control of the reverberation time, controlled by the amount of
sound-absorbing material used; the control of reverberation charac-
teristics determined by the type of materials, and how they are used;
and by the elimination of sound focal points and standing waves,
which is done by the elimination of opposing parallel and concave
surfaces.
As stated above, all of these faults are much more easily avoided
in new construction than cured in old construction. Wherever a
theater is to be built, a noise survey of the site is absolutely necessary
to determine the noises in the surrounding area. Outdoor and traffic
noises in some areas may reach 85 to 90 decibels above zero reference
sound level of 10~16 watt per cubic centimeter.
The noise in the theater itself should be kept to a point where it is
lower than the audience noise (a good value of audience noise is about
30 to 35 decibels) , which means that the outside walls of the theater
may be required to have an insulation value of 55 to 60 decibels, which
if dependent upon mass alone, requires a brick wall two feet thick.
Fortunately there are other ways of doing this by special construction
which is much less costly than a two-foot brick wall.
There are some noises that are not under the control of the archi-
tect and engineer, such as street noises caused by automobiles, bus
traffic, airplanes, railroad trains, streetcars, subway and elevated
trains, and garbage and ash collectors.
Of course, there are audience noises about which nothing can be
done, except hope for a quiet audience. There may also be noises
from adjacent property — music, loudspeaker systems, juke boxes,
hand trucks, factory operations, and about the noisiest source is a
bowling alley.
Windows have no place in the theater, as they are always weak
points in any wall which allow sound transmission. If there must
'be windows they should be fastened so they cannot be opened, to
reduce sound transmission. In some cases, where there is noise on
adjacent property such as floors above or below the theater, it may be
186 CONTENT August
necessary to construct isolated ceilings and floors, and even isolated
suspended walls, as in broadcastTstudio construction.
Of course, there must be fire-escape doors, which should be fitted
tightly to provide as good sound isolation as possible. This also
helps in the operation of the air-conditioning system.
A good example of noise isolation occurred at radio station WOR,
with studios on the first floor, where there was a corridor at the rear
which led from the street to the freight elevators. Many times during
the day there were deliveries by hand trucks which were very noisy
and interfering noises were heard in the studios. The solution was
surprisingly simple. About an inch and a half of street-paving as-
phalt was laid on the floor of the corridor, which eliminated that
noise in the studios.
There are a number of controllable noises in a theater. First there
are the noises from the projection, the rewind, and generator rooms.
The best insurance there of course is to specify and get quiet operating
machines. As Mr. Hardy has reported in another paper, the acoustic
problem is being thoroughly considered in the manufacture and design
of projectors. The ceilings of these rooms should have noise-reducing
treatment, such as fireproof acoustical tile, or some kind of fireproof
treatment having as high a value of sound absorption as possible.
If the ceiling is high, acoustical absorbing materials should be used
on the walls also; to four or six feet from the floor. The projection
ports should be fitted with optical glass. However, as that represents
another maintenance problem, another way that the noise can be
retarded from getting into the auditorium is by lining the top and
sides of the ports with acoustical tiles. The viewing ports should be
fitted with plate glass which reduces the amount of transmission
through these openings.
Another source of noise is noisy electrical equipment. The only
precaution required is to buy good equipment, making sure that it is
operated in the most efficient manner by engaging the services of an
organization thoroughly competent in maintenance of sound equip-
ment for the theater, to maintain the equipment. There are reliable
organizations which provide that service.
Other noises originate in the lobby, the promenade, the ladies'
and men's lounges, and the rest rooms. The ceilings of all these spaces
should be treated with noise-reduction materials with carpets on all
spaces except the rest rooms.
The ventilating systems are sources of noise such as motor or fan
1948 QUIETING AND NOISE ISOLATION 187
noises, which can be both air-borne and solid-borne. One airborne
noise is the noise of the air itself in rushing through the ducts and
grills, and the solid-borne noises can be caused by the vibrations be-
ing transmitted through the ducts.
The best insurance against these noises is to operate an air-condi-
tioning system so that the air in the ducts flows at low velocity. At
the registers, both supply and return, the air velocity should not exceed
250 or 300 feet per minute. Metal registers, where the metal is
placed edgewise, are also a source of noise. At the higher air velocity
the fins may begin to vibrate. It is best to use a flat punched register,
with at least 50 or 60 per cent opening so the air is not constricted, and
will not appreciably increase in velocity in the openings.
In one theater which had air noise in the ducts, the trouble was
corrected by putting in sound baffle boxes, sound traps, in the branch
! ducts just before they went to the inlet registers. In this manner it
i was still possible to get a sufficient amount of air with the existing
|j motor and fan.
In a new system, install low-speed, quiet operating fans because
:j high-speed fans are prone to greater and higher frequency noises.
Another cause of noise is the water-supply system, such as knocking
• in the pipes when a faucet is turned on, and there may be vibration
in the pipes, especially with copper and brass pipe. The proper way
I to prevent the knocking is to have the proper air cushions installed in
; the pipes and to make sure the valves seat properly.
Water lines can be fastened in shock-absorbing mounts so that
I* vibration will not be transmitted to the theater structure. The
troubles encountered in a hot-air system are very much the same as in
I ventilating. A hot-water heating system is like any other water
|: system, but with steam heat, hissing valves should be replaced with
quiet operating vent valves on the radiators. A big source of noise
is steam systems is in the pressure-reducing valves. In one installa-
|i tion the noise was as high as 90 decibels within a foot or two of the
valve. The valve and steam-pipe line causing the noise and vibration
were isolated from the rest of the building to prevent transmission of
structural-borne noises and covered with alternate layers of various
materials to reduce the noise in the rooms where the pipes and valves
were located. The pipes on both sides of the valve should be hung in
shock-absorbing mounts for sufficient distance, so the vibration will
not be transmitted to the rest of the structure.
The rest rooms should be separated from the auditorium by at
188 CONTENT August
least two walls, not necessarily two walls built together, but at least
two walls separating the rest rooms from the audience.
Other sources of trouble are noisy reactors and transformers for
fluorescent and cold-cathode lighting. Wherever these are noted,
they should be corrected. Noisy electric switches cause clicks at
times, which can be eliminated by replacement with mercury switches.
Oftentimes vibrations are set up by different kinds of machinery,
motors, pumps, forced-draft fans, oil burners; they should all be lo-
cated on antivibration mounts to prevent the vibration from being
transmitted to the building structure. Quite often this vibration will
set up very serious noises in some other part of the building.
If there are any elevators in the same building, all the machinery,
the hoist drums, the controllers, the contactors, the motors, should
all be hung on vibration-isolation mounts, and the guide rails for the
elevators should be mounted on antivibration mounts.
A certain amount of trouble is caused by concavity of rear walls in
theaters. Oftentimes domes and other concave surfaces catch sounds
and retransmit them to other points, sometimes louder than they were
in the original location. The answer is to eliminate the use of concave
surfaces if possible. If it is necessary to use concave surfaces, make
sure that the distance from the focal point to the concave surface is
either at least twice or Jess than one half the distance to the populated
area of the theater.
In one theater the author collaborated with Mr. Schlanger in re-
vising the acoustical treatment. In this theater having this trouble
the dome was eliminated and replaced with a flat ceiling.
In conclusion two points should be emphasized for either new
construction or for alteration ; one, engage the services of a registered
architect who has had experience in theater construction as there are
points that the average architect will never encounter; two, always
see that he engages the services of a competent acoustical engineer.
DISCUSSION
MB. W. E. MACKEE: Seat men think the most important thing is a seat. The
carpet man thinks carpet, is most important. The man who sells lighting tells
you how to light your theaters. The acoustical people say the most important
thing is acoustics. I think you have forgotten the real purpose of a motion
picture theater.
First you must understand that there is a different type of audience in the small
motion picture theater today. Sixty per cent of these have less than 500 seats,
and 74 per cent have less than 750 seats. What is a motion picture theater?
Before the war the average audience was supposed to be 19 years of age and pic-
tures were made for them. Then the war started, and exhibitors thought they
1948 QUIETING AND NOISE ISOLATION 189
would go out of business for our younger people were all going to war. As a matter
of fact, we did more business. We got an entirely new audience; older people
are going to the movies, discovering the movies. The average is 32 or 33 years
of age. They go more often and they pay more, that is our audience today.
Why do they go to see motion pictures?
They go primarily for two reasons, rest and recreation. They can stay home
and listen to the radio; incidentally radio audiences have dropped, and motion
picture audiences have increased. They are not staying home and listening to
the radio. We draw these facts from the radio industry. The motion picture
audience of today is composed of older people. They demand a different type of
picture than we had before the war.
What comes out of the radio? Charlie McCarthy. Charlie McCarthy is not
successful in motion pictures. One picture that brought all the people in is one
you probably do not remember — "Mrs. Miniver." It was not designed for the
19-year-old audience. It was designed for older people. Just about that time,
the audience aged. More people are coming in and they are staying in.
In the motion picture industry we have to bring them out of their homes.
You have a nice living room, and you have your radio and your family, but you
go a mile or five miles to a motion picture theater. Why do you do that, and pay
real money for it? You do that because the modern theater today is just as good
as your living room. The seats are very comfortable, and the most important
thing in the theater is not sound. The most important thing is the picture image,
going back to the fundamental purpose of the motion picture. Sound is secondary.
In the pictures that come to us, too often sound predominates, but the audiences
go for rest and recreation and a series of pictures flashed on the screen. Sound is
just explanatory. Anything you can do to remove your noises is desirable, but
do not forget that the motion picture basically is a series of photographic images
flashed on the screen.
DR. RICHARD COOK: Recently in Washington the picture was flashed on the
screen, but for some reason the sound did not come on. There were catcalls from
the audience, "Where is the sound?" They wanted sound right away.
DR. E. W. KELLOG: All we try to do in the way of better sound is to improve
the value of the theater for seeing the pictures.
MR. M ACKEE : We do not want newsreels in the small theaters. The football
pictures are full of sound and noise. We can do without a newsreel. We keep
sound down as low as we can. Air conditioning, yes; and quiet, yes; and a nice
comfortable theater. We do not say you should eliminate sound, but we definitely
are finding out what these older people want, and that is the most important thing
as far as dollars and cents in the box office are concerned.
MR. JOHN K. HILLIARD : Why is it, especially in a dramatic sort of picture, when
the sound is low, the audience is sitting there listlessly, and if the sound is brought
to the proper intensity, not necessarily loud, the audience immediately reacts
and applauds a scene, where if the sound were below what it should have been,
there is absolutely no reaction from the audience.
MR. MACKEE: If it is so low they cannot hear, they start to clap. If they
can hear, we do not hear any comment. Some theaters have earphones. We
could not understand why people would put on earphones when they are not deaf.
We asked one or two, and they said it was because they can control the sound.
190 CONTENT August
MB. LEONARD SATZ: Sound is so much a part of motion picture presentation
that I do not see how you could improve one without improving the other. Maybe
some prefer to call it intangible, but when a patron comes into the theater and
sits down, it is primarily to enjoy himself. If the sound is not good, he will leave
the theater. He does not want to strain to hear the words. We consider sound
in the smaller theaters just as important as in the larger theaters.
MR. GONZALEZ: During the war, we operated 1200 theaters.' We had up-
wards of 12 to 14 theaters in some installations. We put in proper acoustic
material, with the proper absorption, and someone painted the material with
oil paint. Where there was reverberation, the sound was distorted. We found
that at these places where the sound was improper, the soldiers would walk a
half mile or a mile to go to a theater that had good sound, and attendance at the
improperly wired theater fell off, which proves that we needed good sound.
Mr. Content said that the recirculating velocity should be limited to 250 feet per
minute, and that the duct be limited to 500 feet. That is rather expensive in
design. We found that we could safely go as high as 1200-foot velocity without
interfering with the sound system in the theater. So long as the duct was prop-
erly designed so that there were no abrupt changes of air to cause sound or air
noises in the supply grills, we went as high as 600 feet velocity without any per-
ceptible increase in the noises.
MR. WETHERELL: When I design a motion picture house one of my main in-
terests is in the appearance of the finished product and its effect on the viewer.
I wonder if there is not one point that has not been touched on concerning sound.
Someone mentioned that sound is secondary. It seems to me it might be, to this
extent. The aim of the acoustics and sound engineer should be to produce sound
that is so natural it becomes the background of the picture. You come to the
theater, you see the picture, the sound is woven around the action. It should be
so natural and so keyed that you do not realize you are listening to artificial sound.
It is not the aim to have sound so natural that perhaps it is secondary, but
it is quite an art to produce sound that is natural and accurately follows in tone
and volume the action.
Mr. BEN SCHLANGER: I want to defend what the paper said about air noises.
There are certain dramatic sequences where the words are spoken softly and
during those periods air noises are very disturbing. Maybe in the louder se-
quences— I do not mean noise but higher levels and less dramatic — air noise
may not be disturbing, but you have to design for the more particularized
mood in the picture once in a while.
MR. E. J. CONTENT: There are parts in the picture where sound levels will
reach as high as 80 or 90 decibels, and certain parts which will be of low intensity
as low as 30 or 35 decibels. If you have noises masking the sounds, they are not
producing the desired effect upon the viewer. The only way you can keep back-
ground noises to 30 or 35 decibels is to operate the air-conditioning system at low
velocity as mentioned.
MR. D. G. BELL: We have the main trunk lines approximately 700 feet per
minute in the lines to the outlets. The principal noise from the duct system usually
originates in the fan, in the blower, and it has been necessary in many other
cases to add an acoustic absorbent in the duct for ten or twelve feet after the
1948 QUIETING AND NOISE ISOLATION 191
blower in the duct, immediately after the blower; and using those velocities which
are recommended by the American Society of Heating and Ventilating Engineers,
we have kept the noise in the theater down to 30 and 35 decibels.
MB. SHEPABD: I believe the point that Mr. Gonzalez brought out may not
have been followed perfectly. He wanted to.show that by proper designing, you
can have fairly high velocity without the introduction of noise which can be gen-
erated by any vibrating element in the system. It can be reduced by the proper
construction, the proper handling of beams, and possible use of acoustical materi-
als. In many of the installations in the theaters that I have visited where that
was done, I do not know what the actual velocities were but they were fairly high,
and the noises were not excessive; just where you would have to obtain the maxi- '
mum velocity, I do not known.
MR. SGHLANGER: It is ridiculous to try to save a few dollars by using a higher
velocity than you should when you have already invested so many thousands of
dollars in a theater.
MR. SATZ : Do you have any special preference for slab cork for vibration elimi-
nators, cement pit with cork, for heavy fans; would you say that one is more
efficient than the other?
MR. CONTENT: That all depends on the weight of the machine, the frequency
of the vibration, and other factors, such as the weight of the noise-making parts.
The isolating material must be loaded to a point where the transmission of the
vibration through the material is very low. It is possible to use vibration mounts
where there will be more vibration transmitted than if there were no isolation at
all. Each individual problem must be analyzed carefully.
MR. SATZ : Ho you find that glass fabrics are any less suitable than cotton?
MR. CONTENT: I see no reason to condemn one as against the other. Sound
will transmit through the glass fabrics, through the pores as well as through the
cotton or other materials. Sound as we hear it is a movement of air, and it will
get through the pores of the glass cloth just as well as the cotton.
MR. A. D. PARK: What is the recommended treatment for the rear wall of a
motion picture theater?
MR. CONTENT: The best treatment for a rear wall is not to make it concave.
Break it up. If you do that, you may not need sound-absorbing material. All
you want to do is to prevent echoes which are reflected from the rear wall from
reaching the audience. If that wall disperses sound sufficiently so it will not
produce echoes, if it is a concave surface, you must use an absorptive material
with a high coefficient, so that very little sound will be reflected back to cause
trouble in the audience.
MR. JOHN VOLKMANN: It depends considerably on how far the rear wall is
from the seating area, that is, the front of the room. In certain seating regions,
it is possible where, in addition to shaping the walls, you have to put a lot of
absorptive material on it too. I know of cases where we have had a lot of ab-
sorbing material on it, and we had to angle the rear wall down, as well as treat it,
purely because the surfaces which, as I say, were treated with rock wool, per-
forated and paneled and cloth-covered on top of that — they were so disposed that
they were on a curved surface, and because they were so disposed, they tended to
concentrate the sounol into localized regions and not until we angled them for-
ward did we get rid of the concentration effect.
Theater Engineering Conference
Acoustics
•
Behavior of Acoustic Materials
BY RICHARD K. COOK
NATIONAL BUREAU OF STANDARDS, WASHINGTON, D. C.
Summary — Theater architects and engineers need accurate data on
the performance of acoustic materials, which are used to control the acous-
tics of theaters. Descriptions of prefabricated materials and acoustic plas-
ters are given. The mechanism of the sound-absorption process in por-
ous materials is briefly described. There are two commonly used absorption
coefficients, the "random-incidence" coefficient, and the "normal-incidence"
coefficient. The experimental methods used for measuring the two coeffi-
cients are described. The significance and limitations of these coefficients
hi theater design are pointed out, and it is concluded that, at the present
time, the random-incidence coefficient is more useful in auditorium design.
Recommendations for painting acoustic materials are made, and illustra-
tions of the results of painting are included.
THE CONTROL OF THE acoustics of auditoriums and theaters, and
the quieting of noisy rooms, both require the installation of
acoustic materials. By an acoustic material is usually meant a
sound-absorbent substance which is fastened in flat patches to the
walls and ceiling. Recently, however, some absorbers have been
fashioned in the form of cylinders, cones, and spheres, and have been
suspended at a distance from the walls and ceiling of the room.
The principal function of acoustic materials is to absorb sound en-
ergy which originates within the room. Only incidentally do they pre-
vent the transmission of sound energy from one room to another.
Such transmission is better prevented by other techniques.
Architects and engineers are faced with the problems of deciding
what material should be used in an auditorium to secure the proper
amount of sound absorption, and deciding what is the most econom-
ical absorbent to quiet a noisy location. It is clear that they must
have accurate data on how acoustic materials absorb sound. Such
data have been available for many years.1 However, the special
uses to which a material is put will in general require more detailed
* Presented October 24, 1947, at the SMPE Convention in New York.
192 AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51
1948
BEHAVIOR OF Acoustic MATERIALS
193
information than can be obtained from the results of routine labora-
tory tests. Nevertheless, the results of such tests are of great im-
portance to the theater engineer, and he should understand their
significance and limitations.
II. DESCRIPTION OF MATERIALS
There are two main kinds of acoustic materials. One kind is pre-
fabricated, a familiar example being the one-square-foot tiles which are
commonly used in many public places. The other is the kind which
is manufactured, so to speak, at the moment of application, and in-
cludes acoustic plasters and sprayed-on fibrous materials.
There are various types of the prefabricated kind. A common type
SOUND
SOUND
PERFORATED
FACING
PORES
Fig. 1. — Some types of acoustic materials.
is the homogeneous porous absorbent consisting of wood fibers, or
glass fibers, or granulated material, held together with a suitable
binder. Another common type is the porous material having a hard,
nonporous surface which is perforated (Fig. 1) so that the sound waves
can pass into the porous region and be absorbed. The perforations
might be regularly spaced slots, or circular holes, or irregular fissures.
Another type is the porous material installed in blanket form, such
as glass wool or rock wool, and protected by a perforated surfacing of
wood, metal, or asbestos cement board. The principal advantage of
the prefabricated acoustic material lies in the uniformity of the prod-
uct. The manufacture can be carefully controlled, and in general
there are relatively small variations in the absorption coefficients for
a particular type.
194 -COOK August
Several different types of acoustic plasters are available. Some
consist of granulated inorganic substances which are mixed with a
foaming agent and a suitable binder, and are applied with a trowel.
Sometimes the plaster is stippled (in order to improve the absorption
of sound) before it has set hard. Other types consist of fibrous mate-
rial, usually rock wool or asbestos, which is mixed with a binder and
sprayed directly on to the wall by means of special equipment.
Acoustic plasters are generally difficult to handle, and careful control
must be exercised when they are applied. Occasionally, however,
there are economic advantages in favor of plaster and sprayed-on
materials.
A third kind of acoustic material has been used in Europe, and
consists of sponge rubber having a low modulus of elasticity, and
covered with a thin impervious skin. Such absorbents do not seem
to be commercially available in this country.
Acoustic materials can also be classified by the way in which they
absorb sound (Fig. 1). A knowledge of the mechanism of absorption
is important, especially if one is going to be faced with the problem of
painting the material, or keeping it decorated. In almost all cases,
the absorber is porous, and the absorption of sound is due largely to
the viscous damping of the motion of molecules of air in the pores.
Sometimes the absorption process is aided by vibration of the porous
material itself. In addition, the propagation within a porous material
is influenced by the tortuousity of the channels .(see the sketch to the
right in Fig. 1) and by thermal effects. In fact, it is difficult to say
which absorption mechanism predominates in a given porous ab-
sorbent without conducting a complicated investigation as to how
sound is propagated through a material. The sponge rubber de-
scribed earlier absorbs sound by frictional damping in the rubber.
Oddly enough, some very soft and porous fibrous materials, after
being covered with an impervious layer of paint, behave like sponge
rubber. As a general rule, however, it is very important to preserve
the porosity of an absorbent. This makes painting difficult, a point
which will be discussed.later.
III. MEASUREMENT OF ABSORPTION COEFFICIENT
A number of elements are important in the choice of an acoustic
material. The user is interested not only in sound absorption, but is
also concerned with light reflection, fire resistance, appearance,
strength, and paintability, all of which are important. However, we
:
1948
BEHAVIOR OF ACOUSTIC MATERIALS
195
shall discuss only the measurement of sound-absorption coefficients,
and the influence of painting on acoustic properties.
Suppose a beam of sound waves is incident upon an acoustic mate-
rial. The beam carries power. The absorption coefficient is the frac-
tion of the power which is absorbed. For example, if a material ab-
sorbs 65 per cent of the sound power incident on it at 512 cycles per
second, the absorption coefficient at this frequency is 0.65. An im-
portant point is that the absorption coefficient depends not only on
the physical properties of the material, but also on how it is mounted.
As a general rule, an air space between the absorbent and the rigid
wall on which it is mounted enhances the absorption.
Physicists and engineers have struggled long with the problems
of how to define better and how to measure the absorption coefficient
ACOUSTIC _
MATERIAL"/ SHORT TuB£
VIBRATING
DIAPHRAGM
Fig. 2 — Impedance tube for measurement of normal-
incidence absorption coefficient.
of a material. In Fig. 2 is shown a technique for measuring absorp-
tion when sound is incident normally, i.e., perpendicularly, to the
surface. The source is a vibrating diaphragm located to the right
in the diagram. The acoustic material is to the left. By measuring
the sound pressure at the center of the surface, we can deduce, from
a knowledge of the amplitude of vibration of the diaphragm, what
the absorption coefficient is for sound at normal incidence. Varia-
tions of this technique have been used. Sometimes the standing-
wave pattern in the tube is explored with a probe tube and micro-
phone. The important thing is that the general idea is always the
same, namely, we measure the absorption coefficient when sound is
incident perpendicularly on the surface. The coefficient obtained in
this way is called the "normal-incidence absorption coefficient."
196 COOK August
The basic idea of the other technique which is extensively used for
measuring absorption is to have the sound incident on the material
from all possible directions. To achieve this, the material is placed
on the wall or floor of a large, highly reverberant room. Sound of
the desired frequency is introduced into the room, and the distribu-
tion of sound energv is "randomized" by any one of several ingenious
methods. This means that sound rays strike the surface equally
from all directions. When the sound field has become thoroughly
Fig. 3 — National Bureau of Standards reverberation
room. Used for measurement of random-incidence ab-
sorption coefficient.
randomized, the source is turned off, and the rate of decay of the
sound energy is measured with microphones and a recorder. The ab-
sorption can be deduced from the measured rate of decay. The co-
efficient obtained with this technique is called the "random-incidence
absorption coefficient."
In Fig. 3 is shown the 15,000-cubic-foot reverberation room at the
National Bureau of Standards. The sample, which is usually 72
square feet in area, is on the floor. The loudspeakers which supply
the sound are on the vanes. The vanes rotate while measurements
1948
BEHAVIOR OF ACOUSTIC MATERIALS
197
are being made, and help to randomize the sound field. The micro-
phones which pick up the sound are not shown.
The important question is, how significant are the normal-incidence
and random-incidence absorption coefficients in practice? As was
pointed out earlier, the principal use of an acoustic material in an
auditorium is for control of the reverberation time. Many years of
experience seem to show that the reverberation time can be com-
puted correctly if the random-incidence coefficient for the acoustic
treatment in the auditorium is used in the calculations. On the other
hand, it is not really necessary to design a motion picture theater for
optimum reverberation time. The loudspeaker system can supply
ample acoustic power, and hence large amounts of absorption can be,
and usually are, installed in a motion picture theater. Even though
Fig. 4 — Simplified geometrical acoustics of an
auditorium having a sound-absorbent ceiling.
the normal-incidence absorption is more easily determined in the
laboratory, it is still difficult, and in some cases it is impossible, to
deduce the random-incidence behavior from laboratory measurements
with normally incident sound. On the whole, the conclusion at the
present time is that the random-incidence absorption coefficient is
more useful in auditorium design.
Some of the difficulties involved in deciding how to measure sound
absorption can be appreciated by referring to Fig. 4. The sketch
shows that the listener receives sound which is non-normally reflected
from the acoustically treated ceiling. If one wishes to compute the
intensity of the reflected sound, neither the normal incidence nor the
random incidence coefficients can be used! To make matters worse,
the angle "6" of the reflected sound is different in different parts of the
auditorium. The object in pointing out these things is to indicate
198 COOK August
the limitations on the absorption coefficients when an acoustic mate-
rial is being chosen for a theater.
IV. PAINTABILITY OF ACOUSTIC MATERIALS
Since the majority of commercially available acoustic materials
depends on porosity for sound absorption, it is clear that painting
will present a problem. There is always the possibility that excessive
painting will clog the pores and prevent absorption of sound.
In the case of porous materials having a mechanically perforated or
Fig. 5 — Fissured surface of a porous Fig. 6 — Same material as in Fig. 5
acoustic material. after application of four coats of brush-
applied paint.
fissured facing, there is no serious difficulty. Paint can be applied
so long as the perforations and fissures remain open.
The painting of a porous material without large holes or fissures
is more difficult. The paint must be applied as thinly as possible,
preferably with a spray gun. If it is brush-applied, care must be taken
to thin the paint and to get it on the surface without closing the pores.
The nonporous rubberlike materials can be painted, provided the
paint does not substantially increase the weight of the facing. Too
much paint will reduce the absorption of sound at high frequencies.
The effect of painting on some typical porous materials can be seen
BEHAVIOR OF ACOUSTIC MATERIALS 199
from the following data obtained from a paper by Chrisler.2 Fig. 5
shows a fissured material which had a noise coefficient of 0.55 before
painting. The noise coefficient is here defined as the average of the
random-incidence absorption coefficients measured at frequencies of
256, 512, 1024, and 2048 cycles per second. Fig. 6 shows the same
material after it was brush-painted four coats. The noise coefficient
fell to 0.45 after painting, which is not a serious reduction. The
reason for the success of the brush painting is that the material was
fissured. The fissures which lead down into the porous material
Fig. 7 — Granular surface of a porous Fig. S^Same material as in Fig. 7
acoustic material. after application of five coats of brush-
applied paint.
were not covered over, and a considerable amount of sound ab-
sorption remained after painting.
Fig. 7 shows a porous material consisting of organic and inorganic
granules held together with a binder, and having a granulated sur-
face. Before painting, the noise coefficient was 0.60. Fig. 8 shows
the same absorbent after five coats of brush-applied paint, when the
noise coefficient fell to 0.25. This is a too-familiar horrible example
of bad treatment of an acoustic material. The method of painting
in this case should have been either with a spray gun, or with a
thinned paint carefully brushed on so as not to close the pores.
200 COOK August
REFERENCES
(1) The results of tests made at the National Bureau of Standards have been
published in its Letter Circular LC-870, "Sound Absorption Coefficients of the
More Common Acoustic Materials."
(2) V. L. Chrisler, "Effect of paint on the sound absorption of acoustic mate-
rials," J. Res. Nat. Bur. Stand., vol. 24, p. 547; RP 1298, 1940. Reprints may
be secured for 10 cents from the Superintendent of Documents, U. S. Government
Fruiting Office, Washington 25, D. C. (stamps not accepted).
DISCUSSION
MR. WETHERELL: What difference in effect do you obtain with varying per-
centages of opening, or holes in the surface? The area of the holes forms a very
small percentage of the total area. What is the difference in efficiency for dif-
ferent percentages of opening? How does that work out?
DR. RICHARD COOK: So long as the area of the slots or holes is ten per cent or
more of the total available area, there will be no significant effect except at high
frequencies where the absorption might be reduced a little if the area of the holes
is reduced. It depends on the spacing of the holes. The openings should not be
spaced more than, say, half an inch or so apart.
MR. WETHERELL: I cannot visualize whether that would be true.
DR. COOK: Most people cannot.
CHAIRMAN HARVEY FLETCHER: It works out that way mathematically.
DR. COOK: I have an explanation. The sound wave comes up to a hole; the
particles of air move faster as the sound goes through, and it then spreads out on
the other side without appreciably reducing the amount getting through the hole.
If you listen through a perforated screen, you can hear almost as well as when the
screen is not present, which shows that the sound gets through without any
difficulty.
DR. LEO L. BERANEK: Some materials are designed to make use of those holes
in reinforcing the absorption in a certain given frequency region. If the layer
that is placed on the front of the material has a given thickness — if it is not a really
thin layer, which is what Dr. Cook was thinking of — but if it has appreciable thick-
ness, then the shape of that hole, the thickness and the diameter, may combine
with properties of the material and the air space behind it to give an enhanced
absorption at some frequency. That effect makes material like Celotex a good
absorber in the region of 500 cycles per second.
I wish to mention three kinds of materials touched on lightly or not mentioned
at all. First let us consider a plywood surface. If one takes two sheets of ply-
wood, either eighth-inch or quarter-inch, and bonds them loosely together by,
say, spots of glue and places them in a room either in the form of curved sur-
faces, or a flat layer spaced away from the wall, then you will find that you get
quite large low-frequency absorption out of such a combination. We built a «
small studio recently at the Massachusetts Institute of Technology, and the rever-
beration without the 'introduction of any absorptive material was around five
tenths of a second, and fairly constant with frequency.
CHAIRMAN FLETCHER: Without any absorption?
DR. BERANEK: Except the plywood panels, and it was fairly constant up to
1948 BEHAVIOR OF ACOUSTIC MATERIALS 201
2000 cycles per second. Such panels generally become quite reflective above 1000
cycles per second. You have to depend on the people in the room to provide the
high-frequency absorption. So if you have cases where you want both diffusion
of sound and you want absorption at low frequency, then these plywood sur-
faces can be very effective.
The rubberlike material which was mentioned, with an impervious facing on
it, can absorb sound very effectively at low frequencies. At high frequencies it
becomes reflective, and you have to depend on other means to provide absorption
of sound.
There are sound-absorbent cones and spheres but the cones are the only ones
I have seen manufactured. They consist of a pair of cones back to back, hollow
inside, made of half- or three-eighths-inch fiberboard of some kind. They are
hung in the room. We have tried them and have had quite good results. My own
opinion is that they are most useful in a room with a high ceiling, where if you
put on the usual material, say inch-thick tile on the ceiling, you do not seem
to get so much effect out of it. Sound bounces back and forth between the par-
allel vertical walls, and by hanging these absorbers in the room, it is possible to
get a great improvement in the acoustic results over those obtained by covering
the ceiling only. Of course, they look peculiar. We put some up in some of the
M.I.T. lecture rooms recently. We put them up in the Boston Art Museum and
there was quite an article about "What Is New in Art Museums."
MR. ZARO: Dr. Cook, in showing your slides, you showed the slides where
you had brush-painted surfaces. What painting material did you use? Did you
use a casein paint or a light mixture of linseed oil and oil paint?
DR. COOK: The paint was an oil paint which was applied in such a way as to
hide a black stripe painted on the surface. The application was not, shall I say,
scientific. We wanted to get an idea of what might be expected from routine
painting of the material.
MR. ZARO: Would you care to project any recommendations as to the use of
casein paint as against oil paint on such surfaces?
DR. COOK: According to experiments made in our Sound Laboratory at the
National Bureau of Standards, it seems that there is less tendency for a casein-type
paint to fill up the holes of a porous material. It is quite safe sprayed on, but one
must also be right after the painter to make sure he does not try to cover the fine
holes due to porosity.
MR. C. W. LUHRMANN: On the sprayed sample you had, was that sprayed
asbestos or sprayed wool? I am familiar with sprayed asbestos and it loses very
little sound absorption on painting.
DR. COOK: Of the two slides I showed, neither referred to a sprayed-on or
plastic-applied material. What you say is true. One still gets appreciable ab-
sorption from some such materials after painting.
MR. LUHRMANN: We recently put on a material of sprayed asbestos, and the
architect was skeptical about paintability. He said, "I still think that it will not
stand paint appreciably and still hold its sound absorption." So he requested that
we paint that material as often as he desired, and we kept painting it, always
using a spray coat until the paint began to drip off the ceiling. We had no equip-
ment for absorption measurement there, but you could not detect any loss of
sound absorption after the painting.
202 COOK
DR. COOK: I shquld not wish to comment on this unless I had seen the mate-
rial. So much depends on the manner in which the material is applied, upon the
density of application, and upon the painting.
MEMBER: There is another commercial material on the market that does not
seem to fit into the pictures that you had. It is called Kimsul, and consists of
sheets of paper, held together very loosely, and it is faced with a sheet of loosely
woven muslin. Is there some theory to explain its action?
DR. COOK: I know the material you have in mind. That would probably come
under the category of the rubberlike materials mentioned earlier.
MR. E. J. CONTENT: Dr. Cook, do you think that you can apply any kind of
paint on the surface of materials other than those that have perforations, and not
change the absorption characteristics in any way as a function of frequency?
DR. COOK: No, even for the rubberlike materials, repeated application of paint
progressively reduces the absorption.
MR. DUNBAR: Was the paint applied all at one time, or was it allowed to dry
between coats?
DR. COOK: The paint was allowed to dry a day or two between coats.
MR. BEN SCHLANGER: Dr. Cook spoke about convex-faced material and
said he was going to tell us more about it.
DR. COOK: I was referring to spheres, cylinders, and cones. Yes, I did promise
to say something about them. The sound can, so to speak, hit the material from
all directions, whereas, when it is on a wall, the sound comes only from a hemi-
sphere of directions. It appears that the greatest absorption obtains in the case of
spheres, for a given material.
MR. SCHLANGER: I do not favor materials that have to be painted, but during
the war we could only get white acoustic materials, that is, factory-fabricated,
and we did not want white. So what I did, instead of painting with a brush, was
to use a dry-brush application, so there was no flow of material, and I continued
to apply the dry-brush stipple effect, until sufficient coat-covering was achieved.
MR. COLE: Dr. Cook, at what frequencies were the coefficients taken that you
quoted with regard to the fissured material and the one following that?
DR. COOK: Those were average coefficients, usually referred to as the noise
coefficients.
•
Continuously Variable Band
Elimination Filter*
BY KURT SINGER
RCA VICTOR DIVISION, HOLLYWOOD, CALIFORNIA
Summary— A band-elimination filter continuously variable within a range
from 30 to 9000 cycles has been developed. This device has proved ex-
tremely useful for the elimination of interference frequencies in the produc-
tion of sound for motion pictures.
T THE REQUEST of one of the Hollywood motion picture studios, a
continuously variable band-elimination filter has been developed
which is capable of suppressing a very narrow frequency band any-
where in the range from 30 to 9000 cycles. The immediate need for
such a device was occasioned by the necessity to eliminate arc whis-
tles from film recordings of one of the latest Technicolor productions.
The arc whistles had been caused by commutator ripple modulation
of the carbon arcs which were used for set illumination and had been
picked up by the microphone. Because of the magnitude of the sets,
such a great number of arc lights had to be employed that the series
reactors, usually employed for arc-whistle suppression and which are
connected between the motor-generator sets and the arc lights, became
so badly overloaded that their attenuation was considerably reduced.
The members of the Sound Department were aware of the problem,
but since it was impossible to correct the condition at the time, they
went ahead and recorded anyway, leaving the solution of the problem
to be worked out during the re-recording, when various expedients
could be tried.
When the re-recording of the picture was started, tests with var-
ious types of filters were made in an attempt to eliminate the arc
whistles ; however, it was found that conventional filters did not cut
sharply enough to permit elimination of the arc whistles without del-
eterious effects on recording quality. Since the only solution was to
eliminate the disturbing whistles electrically, we were asked whether
it would be feasible to develop, on short order, a device suitable for
the elimination of arc whistles in the neighborhood of 700 cycles. We
* Presented October 24, 1947, at the SMPE Convention in New York.
AUGUST, 1948 JOURNAL OF THE SMPE VOLUME 51 203
204 SINGER August
were also told "if we would be in a position to do not only this, but
at the same time could elaborate on such a device and make it suitable
for the elimination of deleterious noises anywhere in the audio spec-
trum, then really we should be doing something for them and the
film industry in general."
The device which was finally developed to fulfill the studio's re-
quirements is shown and explained in the accompanying illustrations.
A zero-gain amplifier was used which consisted of four amplifier
stages. A three-terminal adjustable Wien bridge was used as the
coupling circuit between the second and third amplifier stages.
Twenty-six decibels of feedback from output to input of the
four stages narrows the bandwidth at cutoff frequencies. Fig. 1
shows the input transformer, the secondary of which is terminated
in a voltage divider which permits adjustment of the amplifier to zero
gain. This transformer is connected into the grid of a pentode-
connected 1620 tube which is resistance-coupled to another 1620 tube
used as a triode. The plate of this second stage is fed back to the
cathode of the first stage. This was primarily done to lower the im-
pedance of the second 1620 tube sufficiently to present a very low
generator impedance to the three-terminal Wien bridge which fol-
lows. This expedient was necessary in order to prevent gain changes
caused by impedance variations of the Wien bridge. It can be seen
easily that as the generator impedance approaches zero, the genera-
tor in this particular case is the plate impedance of the second stage,
no gain change caused by impedance change of the Wien bridge will
be experienced. Conversely, if the load impedance into which this
Wien bridge works is very high as compared to the impedance of the
Wien bridge itself, again no gain change will take place.
Following this second stage is the Wien bridge which is a more or
less standard three-terminal configuration. The audio spectrum
from 30 to 9000 cycles is divided into five bands. This is accom-
plished by changing the capacitance arms of the Wien bridge but
keeping the variable range of the resistance arms the same for all five
bands. By means of this arrangement, it is possible to cover ap-
proximately a 3-to-l frequency range in the four bands from 60 to
9000 cycles and slightly better than a 2-to-l range from 30 to 68
cycles. The output of this Wien bridge is fed into the grid of another
1620 tube, pentode-connected, which in turn is resistance-coupled to
a triode-connected 1620. The plate of this fourth stage is connected
to a suitable output transformer and a portion of the output voltage
VARIABLE BAND-ELIMINATION FILTER
205
206
SINGER
August
ATTENUATION CHARACTERISTIC
OF WIEN BRIDGE
NO FEEDBACK
500-500,000 n
000
FREQUENCY IN CYCLES PER SECOND
10,000 20. Ob
Fig. 2
is fed back, 180 degrees out of phase, to the cathode of the first stage.
This over-all feedback results, first, in a considerable decrease of
over-all amplifier distortion. Second, since it corrects for frequency-
response variations within its limitations, it narrows the bandwidth
of the band which is eliminated at the cutoff frequencies.
i
i
LJ
»-
0
10
20
30
40
50
t
NUATION CHARACTERISE
OF WIEN BRIDGE
26DB FEEDBACK
500-500,000 fl
" ATTE
1*—
1 , ...
1
0
100
1000
10,00
0 20,
FREQUENCY IN CYCLES PER SECOND
Fig. 3
1948 VARIABLE BAND-ELIMINATION FILTER 207
Figs. 2 and 3 give a better picture of what is meant by narrowing
the bandwidth. Normally, a Wien bridge when used between two
amplifier stages attenuates rather gradually and the slope becomes
progressively steeper. With the use of over-all feedback around the
Wien bridge, the biggest part of this gradual region is eliminated and
the sides of the eliminated band are made considerably steeper.
Two variable, vernier resistors in the Wien bridge permit obtaining
an exact null. Fig. 4 shows a front view of the filter panel. The
large center dial permits selection of any frequency within the specific
band which is selected by setting of the rotary switch directly below.
The two smaller dials to the left and right of this large dial are the
two verniers, which assist in obtaining an accurate null. There is
Fig. 4 — Front view.
an "on-and-off" switch to the right and an "in-and-out" switch to
the left. This "in-and-out" switch permits removal or insertion of
this filter in the recording line at the operator's discretion. No click
or other disturbance is created when this filter is switched in and out
of the circuit.
Fig. 5 shows the interior of the device and Fig. 6 shows the fre-
quency characteristics as obtained with various settings of the dip-
I control dials. The bandwidth of the rejected frequency band is
; approximately 15 per cent at 6 decibels attenuation and 3 per cent at
' 30 decibels attenuation as referred to the peak-attenuation frequency.
The practical operation of the filter takes place in the following
| manner. Usually, one determines the exact frequency of the inter-
I ference by beating it with an oscillator. Then the band-elimination
208
SINGER
August
filter is connected into the recording circuit and the same frequency
as the interference is injected into the recording channel directly from
the oscillator. The frequency derived from the oscillator is elimi-
nated by means of the proper dial adjustments on the filter. This is
first done by ear and the ultimate and lowest setting is obtained by
means of high-gain volume indicators. The minimum attenuation of
the interference frequency which can thus be obtained is 50 decibels.
After the proper setting of the filter has been determined, the oscil-
lator is disconnected and the recording channel repatched for normal
operation with the band-elimination filter in the circuit. If the in-
Fig. 5 — Chassis view.
terferences consisted of a single frequency only, this frequency has
been eliminated. However, if there were higher-order harmonics
present they, of course, have been retained. How objectionable j
these harmonics are, depends upon their magnitude. In the case of;
the previously mentioned arc whistles, it was possible to eliminate them
completely by the use of this filter just by eliminating the fundamental.
It has also been found possible to find further use for this filter byj
using it for the elimination of camera noise and motor-generator
noise. Some added explanation might be in order. Camera noise,
is, as the name implies, generated by the motion picture camera.]
Most studios use blimps, that is, soundproof housings over the cameraj
1948
VARIABLE BAND-ELIMINATION FILTER
209
which are supposed to remove camera noise completely. Some
blimps, however, are not so good as others and under certain condi-
tions it is impossible to use blimps. Then one has to rely on low-
noise cameras which are not always low noise, the noise being a func-
tion of maintenance and wear. While camera noise does not consist
of only a single frequency, it has been found possible to reduce its
objectionable effects almost completely by the use of this filter
by eliminating the predominating noise frequency.
Motor-generator sets used for lighting current on location are
usually quite noisy since they employ gasoline motors. While an
attempt is usually made to keep them as far from the microphone as
TYPICAL ATTENUATION
CHARACTERISTICS OF
Ml 10135 VARIABLE BAND
ELIMINATION FILTER
FREQUENCY IN CYCLES PER SECOND
Fig. 6
possible, there are practical limitations to this distance, with the re-
sult that some studios have motor-generator-noise interference on a
number of their location shots. There again it has been found that
the elimination of the predominating noise frequency is sufficient to
attenuate the generator noise satisfactorily.
In conclusion, the summary of the electrical characteristics of this
band-elimination filter are as follows: The gain of the filter is zero
decibels, the input impedance is 600 or 250 ohms, and the output
impedance is 600 or 250 ohms. Zero gain is maintained whether the
250- or 600-ohm input or output connections are used. The filter
operates from a heater supply of 6 to 12 volts direct current and B
supply of 250 volts direct current. Four RCA 1620 Radiotrons are
210 SINGER
employed. The maximum input level that can be applied is —2
dbm.* The distortion at an input or, for that matter, output
level of —2 dbm without the frequency-selective circuit, is about
0.15 per cent from 50 to 8000 cycles. With the frequency-selective
bridge in the circuit, it is somewhat difficult to express distortion in
terms of percentage of fundamental frequency, if the fundamental
frequency lies in the attenuation band, since the ratio of fundamental
to harmonic is determined primarily by the attenuation of the funda-
mental frequency. However, for fundamentals which are outside
the attenuation band, the distortion is not any more than 0.15 per
cent. The peak rejection frequency is continuously adjustable be-
tween 30 and 9000 cycles. At least 50 decibels rejection is obtained
at any peak rejection frequency. The frequency spectrum from 20
to 9000 cycles is divided into five overlapping bands. A sixth posi-
tion on the selector switch permits removal of the frequency-selec-
tive Wien bridge so that the device may be operated as a flat zero-gain
amplifier which sometimes is useful if one wants an isolation amplifier.
The mechanical construction of this filter, which is known as the
MI-10135, permits mounting in a standard relay rack. The front
panel dimension is 19 X 10V2 inches. All tubes are nonmicrophoni-
cally mounted and the frequency-selective Wien bridge, as well as the
entire wiring, are completely shielded against electrostatic and elec-
tromagnetic fields. A dust cover, which is removable from the rear,
is also provided.
* Decibels with respect to 0.001 watt.
DISCUSSION
MK. GEORGE LEWIN: Is there any noticeable effect whatever on the quality of
voice or music with this filter in operation?
MR. E. E. MILLER: I do not believe the effect on voice or music through the
insertion of this filter is any more than that you will notice in a studio on the
scoring stage where you move the microphone an inch and a half wavelength at any
particular frequency to pick up the peak or null of a standing wave that exists
in that studio. I do not believe you will find this is any more than that.
DR. HOWARD C. HARDY: . I do not have so much enthusiasm as the speaker had
for the fact that you could take out the noise of a motor generator or camera by
this filter; certainly the spectra of some of those instruments are very wide
bands, and no peak usually exists that is over 60 or 70 decibels above the whole
main spectrum. At the most, you could eliminate 60.
MR. MILLER: I think the author makes that very clear. He pointed out that
if a high order of harmonics exists, we can remove the fundamental but not the
harmonic if it is of a high order.
Society Announcements
Czechoslovak Film Standards
As of July 1, 1948, the standard projection speed for 35-mm sound film in
Czechoslovakia will be 25 frames per second rather than the American Standard
of 24 frames per second. Mr. Frantisek Pilat, president of the Filmovy Technicky
Sbor (Czechoslovak Motion Picture Engineering Committee), reports that this
change was made because of the increased use of synchronous motors in theater
projectors in that country and also because of the 50-cycle power-line frequency
that is in common use in most European countries. With synchronous drive,
speed-regulation problems cease to exist as long as the line frequency is constant,
and, according to Mr. Pilat, practical tests proved that the resulting higher pitch
of the reproduced sound created no practical problems since it was not observed
by spectators.
International Scientific Film Congress
The second congress of The International Scientific Film Association will be
held in London from October 4 to 11, 1948.
The Association was constituted last year in Paris by delegates from 22 coun-
tries who had accepted the joint invitation to the inaugural congress from The
Scientific Film Associations of Great Britain and France. The primary aim of
the Association is:
"To raise the standard and to promote the use of the scientific film and related
material throughout the world in order to achieve the widest possible under-
standing and appreciation of scientific method and outlook, especially in
relation to social progress."
This year's congress is being convened by The Scientific Film Association of
Great Britain, with the help of The British Film Institute, and invitations have
already been issued to countries throughout the world. The congress will open
with a formal reception to the delegates on October 4 and the following three
days will be devoted to business meetings of The International Scientific Film
Association. On October 8, 9, and 10 there will be a Festival of Scientific Film
when it is hoped to show many contributions from all the participating countries
to members of the general public. The congress will close with a general assembly
of the delegates on October 11.
The .widespread public interest in England in the scientific film as evidenced
by over 10,000 members of local scientific film societies, the introduction of
scientific films and other visual aids into the educational program in that country
and, in particular, the many pioneer activities of The Scientific Film Association
with its country-wide membership make it particularly appropriate that this
congress should be held in Great Britain. Visitors from overseas will have an
opportunity of studying the many contributions which England has made by
the use of films to the "widest possible understanding and appreciation of scien-
tific method and outlook."
Further details may be obtained from The Scientific Film Association of
34 Soho Square, London, W.I.
211
64th Semiannual Convention
Hotel Statler, Washington, D. C., October a5-29, 1948
-PAPERS PROGRAM
Preparations are being made for the Fall Meeting of the Society which will be
held at the Statler Hotel in Washington, D. C., October 25 to 29, 1948, inclusive.
Authors desiring to submit papers for presentation at this meeting are requested
to obtain Author's Forms from the Vice-Chairman of the Papers Committee
nearest them. The following are the names and addresses:
Joseph E. Aiken E. S. Seeley N. L. Simmons
225 Orange St., S. E. 250 West 57 Street 6706 Santa Monica Blvd.
Washington 20, D. C. New York 19, N. Y. Hollywood 38, Calif.
R. T. Van Niman H. L. Walker
4431 West Lake St. P. O. Drawer 279
Chicago 24, Illinois Montreal 3, Que., Canada
Author's Forms and summaries of papers must be in the hands of Mr. Aiken
by September 1.
-PRELIMINARY CONVENTION PROGRAM
Present plans for the 64th Convention Program include a number of special
features that will be of interest to all Society members. There will be a sym-
posium on High-Speed Photography now being organized by Mir. J. H. Waddell,
Chairman of the SMPE Committee on High-Speed Photography, and it is ex-
pected that a large group of interesting and related papers on the subject will be
presented.
—BUSINESS SESSION
The annual Society Business Meeting is scheduled for 3:00 P.M., Tuesday,
October 26. All members planning to be in Washington should attend this ses-
sion since there will be important items of business to be discussed and voted upon .
-AWARDS
Annual presentation of Society awards is planned and recipients are now being
determined by the SMPE Committees on Journal Awards, Fellow Awards,
Samuel L. Warner Memorial Award, and Progress Medal Award. Also, the
newly elected Society officers will be introduced to the members.
212
PRELIMINARY PROGRAM
Monday, October 25, 1948
9 : 30 A.M. REGISTRATION.
Capitol Terrace Room
12:30 P.M. GET-TOGETHER LUNCHEON
Congressional Room
3:00 P.M. TECHNICAL SESSION
Presidential Ballroom
8:00 P.M. TECHNICAL SESSION
Presidential Ballroom
Tuesday, October 26, 1948
9 : 30 A.M. REGISTRATION
Capitol Terrace Room
10 : 00 A.M. TECHNICAL SESSION
Presidential Ballroom
2:00 P.M. TECHNICAL SESSION
Presidential Ballroom
3:00 P.M. BUSINESS SESSION OF THE
SOCIETY
Presidential Ballroom
3: 30 P.M. RESUMPTION OF TECH-
NICAL SESSION
Presidential Ballroom
OPEN EVENING
Wednesday, October 27, 1948
9 : 30 A.M. REGISTRATION
Capitol Terrace Room
10: 00 A.M. TECHNICAL SESSION
Presidential Ballroom
OPEN AFTERNOON
8: 30 P.M. 64TH SEMIANNUAL BAN
QUET
Presidential Ballroom
Thursday, October 28, 1948
OPEN MORNING
2:00 P.M. TECHNICAL SESSION
Presidential Ballroom
8:00 P.M. TECHNICAL SESSION
Location to be an-
nounced later
Friday, October 29, 1948
10: 00 A.M. TECHNICAL SESSION
Presidential Ballroom
2 : 00 P. M. TECHNI CAL SESSI ON
Presidential Ballroom
5:00 P.M. ADJOURNMENT
-LADIES' ACTIVITIES
The Ladies' Reception Hostess and Mr. W. C. Kunzmann are arranging a most
interesting program for ladies who plan to attend the Convention or to visit
Washington during Convention week. The Potomac Room at the hotel will be
ladies' headquarters; further information about the special program will appear in
the September issue of the JOURNAL.
-RESERVATIONS
Excellent accommodations at the Hotel Statler have been arranged for by the
Convention Committee. Members of the Society will receive reservation cards,
which they will be requested to fill out and mail directly to the hotel in Washing-
ton. Each member must arrange these accommodations directly with the hotel;
be sure to mention the SMPE Convention if you are writing on your own letter-
head. Reservations should be made by September 15.
213
Book Reviews
Magic Shadows, by Martin Quigley, Jr.
Published (1948) by the Georgetown University Press, Washington, D. C. 161
pages + 14-page appendix + 8-page bibliography + 7-page index. 24 illustra-
tions. GVsXOVa inches. Price, $3.50.
Film historians, recording the origins of the motion picture, seem impelled to
begin their studies with the Altamira cave paintings and then, working up slowly
through Leonardo, Rog6t, and Plateau, they finally come to Muybridge, Marey,
and the Edison prescreen experiments. Actually, the relationship of their
historic discoveries and devices to the history of the film itself is more than a
little remote; Mr. Quigley has quite properly removed this chapter from the
film histories and expanded it into a book that has its own validity. "Magic
Shadows" carefully traces the slow accretion of scientific knowledge, the sudden
acceleration in the mid-nineteenth century as early principles found practical
application, and finally the simultaneous rush to the screen in France, England,
Germany, and the United States in 1895-1896. Through it all Mr. Quigley
stresses the internationality of the sources, the innumerable individuals who
contributed to the scientific study of optics, and the universal appeal, not merely
of films today, but of the more basic urge to project the shadow of reality. An
elaborate chronology at once traces the growth of prescreen knowledge and em-
phasizes this multiplicity of its sources.
That same multiplicity is further revealed in the extensive bibliography that
Mr. Quigley has appended to his book. Working intermittently on it since 1936,
he has had opportunity to examine original sources both here and abroad, has
covered printed material in Latin, French, German, and English, and translations
from Greek and Arabian. But "Magic Shadows" is no mere compilation. The
main lines of the study were laid down by the veteran film historian, Terry
Ramsaye. In following them, Mr. Quigley has produced a study that is as
readable as it is useful, as thoughtful as it is informative.
JOHN E. ABBOTT
Webb and Knapp, Inc.
New York 17, N. Y.
Photographic Facts and Formulas, by E. J. Wall and Franklin I.
Jordan
Published (1947) by the American Photographic Publishing Company, 353
Newbury St., Boston 15, "Mass. 353 pages + 10-page index + vii pages. 18
illustrations. &/4 X 9Y4 inches. Price, $5.00.
This book is literally crammed with a multitude of both facts and formulas.
The new revision represents a minor modernization of the 1940 edition to include
references to recent developments such as coated lenses and the new color processes.
The material for the most part is presented in a clear and readable fashion with a
continuity of subject matter that was not evident in the 1924 and earlier editions.
The publisher's claim, however, that it is a practical handbook of directions for all
214
Book Reviews
photographic operations in common use is not strictly valid. The increasingly
important field of color photography, for example, is glossed over in twenty pages,
less than half the space allotted this subject in the 1940 edition. On the other
hand, the preparation of lantern slides, which is currently something of a lost
art, is allotted sixteen pages, and a process as obsolete as the making and toning
of printing-out papers is treated in exquisite detail.
Black-and-white photography is quite fully and capable handled, and the
experimental photographic hobbyist will be delighted at the practical working
approach to such subjects as image toning; the sensitizing of leather, fabrics,
and wood; oil, bromoil, and other transfer processes; gum-bichromate printing;
and carbon processes. There is a tendency, particularly in the chapter on "Photo-
mechanical Processes," to pile up formulas and working directions without any
real description of the process involved. In general, the material appears to have
been drawn from a variety of sources without too careful an effort to unify it.
Such important fields as reversal processing and tropical processing are only
sketchily treated, and there is a regrettable tendency to retain obsolete terminol-
ogy in some of the older formulas — such as boracic acid and carbonate of soda.
Despite these objections, the book is a sufficiently useful compendium of photo-
graphic information to be a worthy adjunct to the photographer's library. How-
ever, full-scale revision rather than mere deletion and addition is overdue. In
view of the enormous amount of pertinent photographic material available to the
compilers, there is not space in a photographic handbook of modest size for an
entry on "How to Make Marine Glue" or for five pages on "How to Resilver
Mirrors."
HOWARD A. MILLER
Eastman Kodak Company
Kodak Park, Rochester, N. Y.
FORTY YEARS AGO
Political Subjects Desired
A correspondent of the "St. Louis Post-Dispatch" says: "I should
like to ask through your columns why the moving picture show com-
panies do not make arrangements for a reproduction of the proceedings
of the Republican and Democratic national conventions that are to be
held soon? It would be very interesting and instructive, and millions
who are unable to go to the convention halls would like very much to
see it. And other notable gatherings should be reproduced."
— The Moving Picture World, June 18, 1908
215
Section Meeting
Midwest
George W. Colburn, secretary-treasurer of the Midwest Section, presided at the
May 13, 1948, meeting, which was held on the sound stage of the Atlas Film Cor-
poration. Ninety-two members and guests were present.
The first paper, "The DM-2 and DM-4 Developing Machine," was presented
informally by R. Paul Ireland, president, Engineering Development Laboratories.
Mr. Ireland described the physical setup of the machines and elucidated on the
features of roller design and the principles of physics involved.
"The RCA Six-Position Re-Recording Console" by Everett Miller of the
Radio Corporation of America was read by Frank Richter, sound engineer.
Atlas Film Corporation. The installation described by the paper was inspected
after the final presentation.
Erik I. Nielsen, senior organic chemist, Armour Research Foundation, gave a
talk entitled "Recent Developments in Plastics." This subject dealt with plastics
as applied to optics anol covered the problems involved hi mass production of high
quality plastic lenses.
The meeting adjourned at 10:00 P.M. and was followed by an inspection of the
developing machines described in the first presentation, the re-recording console
described in the second paper, and a general tour of the Atlas Film Corporation's
facilities.
Correspondence
It is highly desirable that members avail themselves of the opportunity
to express their opinions in the form of Letters to the Editor. When of
general interest, these will be published in the JOURNAL of the Society of
Motion Picture Engineers. These letters may be on technical or non-
technical subjects, and are understood to be the opinions of the writers and
do not necessarily reflect the point of view of the Society. Such letters
should be typewritten, double-spaced. If illustrations accompany these
contributions, they should be drawings on white paper or blue linen and
the lettering neatly done in black ink. Photographs should be sharp
and clear glossy prints.
Please address your communications to
Miss HELEN M. STOTE, Editor
Society of Motion Picture Engineers
Suite 912
342 Madison Avenue
New York 17, N. Y.
216
Current Literature
THE EDITORS present for convenient reference a list of articles dealing with
subjects cognate to motion picture engineering published in a number of se-
lected journals. Photostatic or microfilm copies of articles in magazines that are
available may be obtained from The Library of Congress, Washington, D. C., or
from the New York Public Library, New York, N. Y., at prevailing rates.
American Cinematographer
29, 4, April, 1948
Television Field Opens for Cinema-
tographers (p. 120) E. Tow
Progress on 8-mm Synchronized
Sound (p. 135)
29, 5, May, 1948
Extremely Wide Angle Lens for
Aerial Mapping (p. 154)
Appreciating the Motion Picture
(p. 163) C. LORING
International Projectionist
23, 4, April, 1948
Theater Television: A General
Analysis (p. 21) A. N. GOLDSMITH
More on "Quality" vs. "Pleasing"
Sound Reproduction (p. 30)
23, 5, May, 1948
Optical Efficiency in Projection (p. 5)
R. A. MITCHELL
Screen Data: Types, Sizes, Illumi-
nation for 35- and 16-mm Film
Projection (p. 8)
Handling, Storing Cine Film (p. 12)
Theater Television: A General
Analysis (p. 15) A. N. GOLDSMITH
Tele-Tech
7, 6, June, 1948
Sound Measurements in BC Studios
(p. 38) W. JACK
Audio Engineering
32, 5, May, 1948
Loudness Control for Reproducing
Systems (p. 11) D. C. BOMBERGER
Factors Affecting Frequency Re-
sponse and Distortion in Magnetic
Recording (p. 18) J. S. BOYERS
Horn-Type Loudspeakers (p. 25}
S. J. WHITE
British Kinematography
12, 4, April, 1948
Metals hi Kinema and Related
Equipment (p. 109) A. B. EVEREST
and F. HUDSON
Back Projection and Perspective.
1. Interlocking and Film Steadi-
ness (p. 127) G. HILL
La technique cinematographique
19, April 1, 1948
Trente annees de technicolor.
(Thirty Years of Technicolor)
(p. 133) W. R. GREENE
Precedes d'enregistrement sonore sur
film. (Procedure of Recording
Sound on Film) (p. 135)
P. JACQUIN
Radio News
39, 6, June, 1948
The Recording and Reproduction of
Sound. Pt. 16 (p. 65) O. READ
Modern Television Receivers. Pt. 3
(p. 71) M. S. KIVER
EMPLOYMENT SERVICE
POSITION WANTED
CAMERAMAN: Experienced in 35-mm and 16-mm cinematog-
raphy, color, black-and-white. Active member SMPE. Three
years' overseas experience as Official Army Photographer World
War II. Will consider offer anywhere in U.S.A. References
available. Free to travel part time at least. Write Charles Arnold,
P. O. Box 995, Peoria, 111.
217
New Products
Further information concerning the material described below can
be obtained by writing direct to the manufacturers. As in the case
of technical papers, publication of these news items does not consti-
tute endorsement of the manufacturer's statements nor of his products.
Synchro-Link, Pulsing Drive,
and Dyna-Link
Yardeny Laboratories, 105-107
Chambers Street, New York 7, New
York, recently put on the market their
Synchro-Link, Pulsing Drive, and
Dyna-Link.
The Synchro-Link is an inexpensive
remote-positioning servo control, which
will position one or several distant
motors, according to the setting of the
master-control dial. The accuracy is
independent of the load.
This equipment works on the prin-
ciple of a self-balancing electronic
bridge, and will control the speed ad-
justment on variable-speed transmis-
sions, the setting of motorized valves,
volume dampers, engine throttles,
pumps, machine tools, and special
machinery.
The master-control dial can be lo-
cated any distance from the Synchro-
Link controller up to several thousand
feet. Only 3 wires of light gauge pass-
"ing small control currents connect the
master control to the Synchro-Link
controller.
218
The Pulsing Drive is a new device for
controlling electrical motors when ac-
curate positioning is important. It
responds to the operation of a single
knob, and when this knob is rotated in
one direction, the Pulsing Drive closes
selectively one of two circuits for very
short periods of time repeated at a rate
dependent upon the speed of the knob
rotation. It is suited for controlling all
standard types of electric motors or
magnetic valves.
The Dyna-Link is an electronic con-
trol device, designed for industrial
applications of variable-speed power
transmission. It consists of a master
control calibrated in revolutions per
minute, the Dyna-Link controller, and
a speed-measuring generator. When
the operator sets the master control to
the desired speed setting, the Dyna-
Link controller energizes the pilot
motor in the proper direction for ad-
justing the speed changer until the
actual output speed corresponds to the
master-control setting. If the drive
slows down because of an increase hi
the load, the Dyna-Link controller
automatically detects the difference in
speed and corrects the adjustment.
Film Counter, Audio Compensa-
tor, and Phase Converter
A Film Counter, Audio Compensa-
tor, and Phase Converter are three
new products which are now being
produced by Arlington Electric Prod-
ucts, 500 W. 25 St., New York, Ne\v
York.
New Products
Further information concerning the material described below can
be obtained by writing direct to the manufacturers. As in the case
of technical papers, publication of these news items does not consti-
tute endorsement of the manufacturer's statements nor of his products.
The Film Counter is designed for
use in motion picture viewing, dubbing,
recording, and narrating, wherever
footage and cuing information are
desired.
The unit can be located remotely
from a projector, recorder, or dubbing
head and will read elapsed time in
minutes and tenths of a minute and in
feet of film that have passed through
the film machine. The counter can be
wired to start automatically with the
projector or dubbing head and can be
stopped and started any number of
tunes during a thousand-foot reel.
The Audio Compensator is used
j where audio equalization is required,
! and is applicable in film recording, disk
I recording, and general broadcast-studio
work. Equalization characteristics
available consist of three steps each
lowering or raising low frequencies
and lowering or raising high frequen-
cies. Each channel contains a two-
stage resistance-capacitance amplifier
employing Type 1620 tubes; Power
and audio connections are made
through multiple plugs.
The Phase Converter is designed for
use where it is necessary to operate
cameras or recording machines with
three-phase driving motors from a
single-phase source of power.
The converter is portable and does
not use electronic tubes or rotating
machinery. The converter input is
115-volt, 60-cycle, single-phase alter-
nating current, and the output is 220-
volt, 60-cycle, three-phase alternating
current of sufficient power to run one
motor properly. A motor running
from this converter will have the elec-
trical characteristics identical to that
of commercial three-phase power and
will have a speed synchronous to the
single-phase line frequency.
FORTY YEARS AGO
Washington, D. C., Wants Picture Machines Inclosed
Fire Chief Belt has recommended to the Commissioners that moving
picture machines used in the five-cent theaters and the regular theaters
of the District be inclosed in fire-proof boxes.
— The Moving Picture World, May 9, 1908
219
SECTION OFFICERS
Atlantic Coast
Chairman Secretary-Treasurer
William H. Rivers Edward Schmidt
Eastman Kodak Co. E. I. du Pont de Nemours & Co.
342 Madison Ave. 350 Fifth Ave.
New York 17, N. Y. New York 1, N. Y.
Midwest
Chairman Secretary-Treasurer
R. T. Van Niman George W. Colburn
Motiograph George W. Colburn Laboratory
4431 W. Lake St. 164 N. Wacker Dr.
Chicago 24, 111. Chicago 6, 111.
Pacific Coast
Chairman Secretary-Treasurer
S. P. Solow G. R. Crane
Consolidated Film Industries 212—24 St.
959 Seward St. Santa Monica, Calif.
Hollywood, Calif.
Student Chapter
University of Southern Calfornia
Chairman Secretary -Treasurer
Thomas Gavey John Barnwell
1046 N. Ridgewood PI. University of Southern California
Hollywood 38, Calif. Los Angeles, Calif.
Office Staff— New York
EXECUTIVE SECRETARY OFFICE MANAGER
Boyce Nemec Sigmund M. Muskat
STAFF ENGINEER JOURNAL EDITOR
Thomas F. Lo Giudice Helen M. Stote
Cecelia Blaha Dorothy Johnson
Helen Goodwyn Ethel Lewis
Beatrice Melican
220
Journal of the
Society of Motion Picture Engineers
VOLUME 51
SEPTEMBER 1948
NUMBER 3
PAGE
Report of the President LOREN L. RYDER 221
Historical Sketch of Television's Progress L. R. LANKES 223
Report of SMPE Standards Committee 230
Errors in Calibration of the / Number. . . .FRANCIS E. WASHER 242
Projection Equipment for Screening Rooms H. J. BENHAM 261
The Gaumont-Kalee Model 21 Projector
L. AUDIGER AND R. ROBERTSON 269
Zoomar Lens for 35-Mm Film F. G. BACK 294
Parabolic Sound Concentrators R. C. COILE 298
Committees of the Society 312
64th Semiannual Convention : 323
Section Meeting 327
Book Review:
"The Preparation and Use of Visual Aids/' by Kenneth B.
Haas and Harry G. Packer
Reviewed by W. A. Wittich 330
ARTHUR C. DOWNBS
Chairman
Board of Editors
HELEN M. STOTE
Editor
GORDON A. CHAMBERS
Chairman
Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
<J42 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930,
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOURNAL must be obtained in writing from the General Office of the Society,
t/opyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC . . . EXECUTIVE SECRETARY
OFFICERS
1947-1948
PRESIDENT EDITORIAL VICE-PRESIDENT
Loren L. Ryder Clyde R. Keith
5451 Marathon St. • 233 Broadway
Hollywood 38, Calif. New York 7, N. Y.
PAST-PRESIDENT CONVENTION VICE-PRESIDENT
Donald E. Hyndman William C. Kunzmann
342 Madison Ave. Box 6087
New York 17, N. Y. Cleveland, Ohio
EXECUTIVE VICE-PRESIDENT SECRETARY
Earl I. Sponable G. T. Lorance
460 West 54 St. 63 Bedford Rd.
New York 19, N. Y. Pleasantville, N. Y.
1948-1949
ENGINEERING VICE-PRESIDENT FINANCIAL VICE-PRESIDENT
John A. Maurer James Frank, Jr.
37-01—31 St. 18 Cameron PI.
Long Island City 1, N. Y. New Rochelle, N. Y.
TREASURER
Ralph B. Austrian
247 Park Ave.
New York 17, N. Y.
Governors
1947-1948
John W. Boyle Robert M. Corbin Charles R. Daily
1207 N. Mansfield Ave. 343 State St. 5451 Marathon St.
Hollywood 38, Calif. Rochester 4, N. Y. Hollywood 38, Calif.
David B. Joy Hollis W. Moyse
30 E. 42 St. 6656 Santa Monica Blvd.
New York 17, N. Y. Hollywood, Calif.
1948
William H. Rivers S. P. Solow R. T. Van Niman
342 Madison Ave. 959 Seward St. 4431 W. Lake St.
New York 17, N. Y. " Hollywood, Calif. Chicago, 111.
1948-1949
Alan W. Cook Gordon E. Sawyer
4 Druid PI. Lloyd T. Goldsmith 857 N. Martel St.
Binghampton, N. Y. Burbank, Calif. Hollywood, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Report of the President
THIS REPORT of the President is the story of the Society of Motion
Picture Engineers, its activities during the last six months, and
\vhat takes place at the New York Headquarters' Office 3000 miles
from this, the 63rd Semiannual Convention.
I am proud of that "63rd" figure and the continuity of activity
which it represents. Few people realize that for 31 1/2 years this
Society has served this industry. Few people realize that much of
the early world- wide standardization for silent pictures was worked
out by SMPE members. As of this date this industry has estab-
lished more standards with the American Standards Association than
any other United States industry. This is important, for our market
is world-wide and dependent upon the existence and the retention of
standards under which our product can be played. The Society is
still active and a look into the future would indicate that television
will bring more and greater problems in standardization. It is im-
portant to note that the economic value of this standardization in-
creases rapidly with complexity of equipment, and even to one famil-
iar with television it is complex.
The 62nd Convention of the Society, which was held in New York
last October, included a Theater Engineering Conference and Equip-
ment Exhibit. It brought into our circle many theater people and
their technical contributions along with an appreciation on our part
of their problems. The good which has resulted from that conven-
tion will be of lasting value.
The period following the New York Convention has been marked
by great technical changes. Television has grown from the ten-inch
image of a home receiver to a reality on the theater screen. This is a
milestone in motion pictures. It may bring about even greater
changes than occurred with the advent of sound.
Our progress has not been limited to any one field. New color
processes are now in commercial use and the thinking which has taken
place is the forerunner to the great program of color papers at the
63rd Convention. The completeness of this color coverage is not an
accident. It is the result of a realization on the part of our engineers
that color offers an outstanding technical advantage which the motion
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 221
222 RYDER
picture industry has and can use in meeting the competition of tele-
vision broadcasting to the home.
Magnetic recording has arrived at a state of development where it
is now finding its place as a tool for the broadcast and motion picture
industries. The papers presented at this convention and the work
of our Standards Committee will aid in the best use of this note-
worthy scientific development. There are other fields in which
there has been marked advancement and I say with pride that these
advancements are well recorded in the JOURNAL of the Society.
Over the years the Society has progressively and solidly grown into
a large and businesslike organization. Our Headquarters' Offices are
located in the Canadian Pacific Building at 342 Madison Avenue,
New York 17, New York. We have a paid staff of eight persons
under Mr. Boyce Nemec, the Executive Secretary. The work which
they do is the background of all Society activity — world-wide. Our
sections are located in New York, Chicago, and Hollywood, with a
Student Chapter at the University of Southern California. All of our
membership records, business administration, and the publication of
our JOURNAL are handled by the New York Office. The New York
Office and the personnel in that office are there to serve you, the
members of this Society. Please visit the office when you are in
New York or write whenever we can be of assistance to you.
The records as of March 31, 1948, show that our membership has
now reached 2787 members. For the year 1948 we anticipate a
revenue and an expenditure of approximately $95,000.
In making a study of our membership we find that it has not grown
as anticipated. This seems to be largely due to the complexity of
the system of admitting new members. Steps have been taken at
the Board Meeting of May 16 to rectify this condition. All persons
interested in the art of motion picture making are eligible for Asso-
ciate membership. We want their affiliation and we want to advance
their status when, as, and if their qualifications and activity justify
such advancement.
I want you to know that we, the Officers of the Society, are ap-
preciative of the excellent support which we have received from the
membership. I want the sustaining members to know that their
support is sincerely appreciated and valued.
Respectfully submitted,
LOREN L. RYDER, President
Historical Sketch of
Television's Progress
BY L. R. LANKES
EASTMAN KODAK COMPANY, ROCHESTER 4, NEW YORK
Summary— This is a brief review of published material and, in its original
form, was an introductory part of a symposium on the various aspects of
television which will affect the photographic industry. It is not an attempt
to answer directly the question, "Who invented television?" for, as Waldemar
Kaempffert, Science Editor of the New York Times, has already pointed
out, Professor William F. Ogburn in his "Social Change" has listed 148
major discoveries and inventions which were made simultaneously and
independently by at least two workers in the particular field concerned in
each case; and if the list were to include developments of secondary im-
portance, it would undoubtedly have grown into a volume at least as large
as an unabridged dictionary. Rather, then, it should be construed as an
attempt to convey a general understanding of the subject by considering how
it was pieced together.
OF ALL THE PURSUITS to which one can turn his attention, perhaps
none has aroused a higher degree of curiosity, enthusiasm, and
hope than the development of television. It has been said that tele-
vision holds the promise of being the medium that can bring the
peoples of far places emotionally face to face with one another's man-
ners, customs, and problems, and thereby make them understand that
they are all essentially human. It co,uld be said that the motion
picture also holds this promise since television is essentially motion
pictures with radio as the means of conveyance. However, there may
be advantages in television's claim to immediacy: namely, that what
is being viewed at the receiver is occurring now at the transmitter.
Contrary to general opinion, the concept of television is not a twen-
tieth-century product. Even in Biblical times abstract thinkers pre-
dicted that it would be possible to develop the ability to see events
occurring beyond the horizon. However, the crystallization of specific
inventions which led to television as we know it today, began with the
transition of the eighteenth to the nineteenth century. The first
SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 223
224 LANKES September
items are Alexander Volta's electric battery, the voltaic pile; Profes-
sor Berzelius' isolation of the element selenium; Oersted's discovery
of the principle of electromagnetic induction ; and the efforts of Am-
pere, Ohm, and Faraday.
The middle of the nineteenth century might be said to have borne
the infant, television, for in 1842 Alexander Bain,1 an English phys-
icist, first proposed a device to send pictures from one place to an-
other by electric wires. Bain's plan was so correct basically that it
embraced the fundamentals of all picture transmission, having recog-j
nized the particular problems posed by the need for synchronization
between transmitter and receiver. In 1847, Bakewell2 devised a
"copying telegraph" employing an elementary scanning device.
Specifically, this was an instrument for transmitting writing or draw-
ings in the form of nonconducting shellac ink on tin foil. The foil was
then wrapped around a cylinder which rose as it rotated, thereby
tracing out a spiral with a fixed metal needle pressing against the foil.
At the receiver, a similar cylinder was covered with chemically treated
paper. In 1862, Abbe Caselli2 transmitted the first electric picture
from Amiens to Paris.
The latter part of the nineteenth century saw the groundwork for
the construction of the present video industry. The light-sensitive
properties of selenium were discovered in 1873 by a telegraph operator
named May.3 In a terminal station for the Atlantic cable on the
coast of Ireland, May observed the effect of sunlight falling on selen-
ium resistors in some of his circuits. This indicated that light values
can be converted into equivalent electrical values. In 1875, G. R.
Carey, in Boston, and Ayrton and Perry, in England, proposed to
build a large mechanical eye using a plate of tiny selenium cells as the
retina.3 Each cell would be connected by wire to a corresponding spot
on the receiver. Electromagnets connected to each of the small sec-
tions of the receiver plate were to regulate the amount of light on each
section. Many other suggestions, all very similar in principle, were
advanced through this period. These were followed by Sir William
Crookes' discovery of cathode rays in his famous vacuum tube. In
1880, Leblanc2 developed the complete principle of scanning wherein
a picture is divided into lines and each line into tiny segments. Hertz/
in 1886, confirmed Maxwell's theories of electricity and discovered the
photoelectric effect in 1887, when he noticed that a spark could be
made to jump over a gap more readily if one of the electrodes were
illuminated than if the event occurred in darkness. The German
1 1)4 s TELEVISION PROGRESS 225
Hallwachs4 later studied the photoelectric effect systematically and
concluded that light set free electrical particles from the electrode
surface. Sir J. J. Thompson identified them as electrons and Einstein
announced the theory of the photoelectric effect. The practical side
was advanced by Elster4 and Geitel4 who, as early as 1890, -built
practical photoelectric cells. Thus the method was defined by which
a television camera would turn a picture into electricity. *
As a noteworthy aside, Thomas Edison5 filmed his first motion
picture in 1889; and Marconi,6 in 1895, sent and received his first
wireless signals across his father's estate.
Coincidental with these latter developments came the invention, in
1884, by the German Nipkow4 of the rotating scanning disk. This
disk made use of the very significant technique, previously suggested,
of dissecting the scene to be transmitted into points of light which
would then be measured on a time scale in orderly fashion. Nipkow's
work ranks high in the history of the medium because he realized so
early a system which was not improved upon, basically, for nearly
fifty years.
In 1890, the Englishman Button4 proposed a system for a television
receiver which ranks in importance with Nipkow's system for the
transmitter. Button's apparatus used a scanning disk and a light
source controlled by a Kerr cell. This method of reassembling the
image was likewise remarkable in that it was used widely in practical
television systems for nearly forty years.
At the turn of the century, Sir J. J. Thompson,7 in his work to de-
termine the charge-to-mass ratio of the electron, showed that the
cathode ray was in reality a beam of high-speed electrons. His
methods involved the application of both electric and magnetic de-
flecting forces. At about the same time, Professor Braun8 built a
3old-cathode-ray tube. With it he could show the effect of magnetism
3n electron beams in tracing their paths on a fluorescent screen. From
the viewpoint of television, this was to be the means of scanning con-
trol for Crookes' cathode rays- Amplitude control, on the other hand,
was to come later.
By the end of the first decade of the twentieth century, Professor
Boris Rosing2 had patented a television system, using a receiver re-
sembling the modern set, based on the Braun cathode-ray tube. In
1911, A. A. Campbell Swinton,3 a man of great imagination and fore-
sight, saw the possibility of television communication with variations
Rosing's cathode-ray tubes at both transmitter and receiver.
226 LANKES September
Recent years have shown that Swinton actually predicted television
apparatus as used today, having developed the theory of a cathode-ray-
tube camera. Meanwhile, Knudson2 had sent the first drawing by
radio.
Only a few of the early discoveries and inventions are directly em-
ployed in modern television. However, the original work and inven-
tions gave impetus to experiments in demonstrating that light could be
converted into electrical impulses which, in turn, could be transmitted
and later reconverted. Fortunately for television, the development of
the radio and electrical arts coincided with the advanced phases of
research in the fields of optics and vision.
World War I delayed progress universally, for the next important!
date is 1923 when Zworykin filed patent application on the first elec-j
tronic television camera tube, the iconoscope, wherein the means fo
scanning control, as well as for picture signal-amplitude control, were
all self-contained on a completely electronic basis. While the idea
had been proposed early in the art, this was the first practical mean
of achieving it.
At this time J. L. Baird4 in England, and C. F. Jenkins4 in the
United States, working independently, produced and demonstratec
systems of television based on mechanical scanning through the use o
the Nipkow disk or something similar to it. The disk carried hole
along a spiral in such a way that a scene, when viewed through a por
tion of it, would be broken into parallel lines or arcs, thereby providing
the means of measuring light values along the short tune-base whicl
represented the frame interval. The pictures were mere shadow
graphs at first, but Baird soon demonstrated television transmission o
half-tone pictures as well as infrared television.
This method of scanning, having serious limitations in definition, is
not in use today, nor is the receiving system that reconstructed the
picture by reversing the process. While the low-definition (less thai]
60-line) images of those days may seem to have little bearing on tech
niques which produce present-day, continuous-tone pictures in a 525
line system, much of -the theory which makes present equipment pos
sible was proved during this mechanical era.
In 1927 the Bell System demonstrated the transmission of televi
sion over substantial distances; between Washington and New York
over wire line, and between Whippany, New Jersey; and New Yort
over radio link. With this was published an analysis, thorough foi
the time, of the transmission problems facing television, particularly
1948 TELEVISION PROGRESS . 227
the frequency bandwidth requirements which have become so char-
acteristic of the art.9
The decade 1925 to 1935 produced many developments in steady
succession. These began with the National Broadcasting Company's
first radio network and Warner Brothers' "Vitaphone" sound-on-disk
system synchronized with motion pictures. Concurrently, Congress
established the Federal Radio Commission; progress continued with
Bairds' first trans- Atlantic television picture and his first crude sys-
tems of color and stereoscopic television; Farnsworth's system and
Zworykin's system of all-electronic television were introduced em-
ploying special cathode-ray receiver tubes called kinescopes; Bell
Laboratories demonstrated television in color, delivering a picture
of postage-stamp size; theater television was shown on screens as
wide as 10 feet; two-way-wire television-telephone demonstrations
were made by Bell; improved photoelectric cells and electronic tubes
were introduced; an extensive program of field tests by the Radio
Corporation of America was initiated starting with 240-line all-elec-
tronic television employing radio relay, to continue right through
the period of commercial operation ; and, finally, the 1935 announce-
ment of the principle of frequency modulation by Edwin Armstrong.
Through the efforts of men like Zworykin, Engstrom, and Gold-
smith of RCA; Farnsworth; Ives and others at the American Tele-
phone and Telegraph Company; Alexanderson of General Electric;
Dumont; and Goldmark of the Columbia Broadcasting System, well-
planned and well-executed programs made public participation in the
United States possible in 1934.
The Philips Company of Holland built the first iconoscope in
Europe in 1935. Television transmitters appeared in places such as
the Eiffel Tower and Stockholm. As the advance continued, A. T.
and T. successfully demonstrated the capabilities of coaxial cables in
I 1936. Such cables were laid from New York to Philadelphia, from
Paris to Bordeaux, and from Berlin to Nuremberg. The first patent
on coaxial cable was granted in England at this time and cables were
laid from the British Broadcasting Corporation transmitter to Buck-
ingham Palace and Victoria Station for the first direct televising of
coronation-procession street scenes.10"12
In 1938 television signals from London, on ultra-short waves, were
picked up on Long Island, although badly distorted.
The point was reached wherein one saw the telecasting of plays
from theater stages, the New York World's Fair, major-league
228 . LANKES September
baseball, and professional football. Meanwhile RCA introduced an
improved television camera tube, the orthicon. It is beyond the scope
of this paper to enumerate the many developments from that point
to date.
The lack of uniformity in choice of number of lines for the picture
structure was never satisfactory to the nontechnical observer who was
quick to compare television with motion pictures. Because of this,
and in keeping with the steady advances, "definition" was standard-
ized at 343 lines in 1935. Later this was raised to 441. In 1940 it was
increased to 525 where it remains as today's standard.
Although World War II brought an apparent period of inactivity,
an abundance of knowledge and technical personnel grew out of
government-sponsored radar and guided-missile programs. Acceler-
ated research and development produced items such as the high-sensi-
tivity image-orthicon and phosphors to withstand the bombardment
of highly accelerated electron beams, for brighter pictures.
The highly controversial issue of color versus black-and-white
television brought the industry to a virtual standstill. After this was
settled early in 1947 in favor of black-and-white, the prospective
broadcaster, the equipment manufacturer, and the receiving-set pur-
chaser appeared ready to invest in the fast-growing business. By
December 31, 1947, the score totaled 12 cities with television service;
18 stations operating and 55 licensees; 287 sponsors; 142,400 re-
ceivers in private homes; 27,600 receivers in public places; 195,000
total receiver production; and an estimated audience of 1,200,000
with assurance of nationwide networks in the reasonably near future.
BIBLIOGRAPHY
(1) RCA Institutes, Inc., "Radio Facsimile," vol. 1, 1938.
(2) American Television Society, Inc., "American Television Directory,"
1st Ann. Ed., 1946.
(3) Lee de Forest, 'Television To-day and Tomorrow," Dial Press, Inc.,
New York, N. Y., 1942.
(4) J. Porterfield and K. Reynolds, "We Present Television," W. W. Nor-
ton and Co., New York, N. Y., 1940.
(5) Deems Taylor, "A Pictorial History of the Movies," Simon and Schuster,
New York, N. Y., 1943.
(6) New York World Telegram, "Chronology of radio and television," Source:
The National Broadcasting Company, The World Almanac 1945, p. 650.
(7) J. J. Thompson, "Cathode rays," Phil. Mag., vol. 44, p. 293; 1897.
(8) F. Braun, "Ueber ein Verfahren zur Demonstration und zum Studium
des zeitlichen Verlaufs variabler Strome," Ann. Phys. und Chemie (Wied. Ann),
New Series, vol. 60, p. 552; 1897.
1948 TELEVISION PROGRESS 229
(9) H. E. Ives, F. Gray, J. W. Horton, II. C. Mathes, H. M. Stoller, E. R.
Morton, D. K. Gannett, E. I. Green, and E. L. Nelson, "Television Symposium,"
Trnnx. Amer. Inst. Elec. Eng., vol. 46, pp. 913-962; June, 1927.
(10) British Patent No. 284,005.
(11) K. Lake, "The coaxial cable," Telev. and Short Wave World, (known
as Television (London), prior to 1939), vol. 10, p. 202; April, 1937.
(12) "Special television cable," Elec. Rev. (London), vol. 120, p. 889; June
11, 1937.
(13) 0. E. Dunlap, "The Future of Television," Harper Brothers, New York,
N. Y., 1942.
(14) William C. Eddy, "Television— The Eyes of Tomorrow," Prentice-Hall,
New York, N. Y., 1945.
(15) D. G. Fink, "Principles of Television Engineering," McGraw-Hill
Publishing Company, New York, N. Y., 1940.
(16) P. C. Goldmark, J. N. Dyer, E. R. Piore, and J. M. Hollywood, "Color
television," J. Soc. Mot. Pict. Eng., vol. 38, pp. 311-353; April, 1942.
(17) R. W. Hubbell, "4000 Years of Television," G. P. Putnam Sons, New
York, N. Y., 1942.
(18) M. S. Kiver, "Television Simplified," D. Van Nostrand and Company,
NVw York, N. Y., 1946.
(19) E. J. G. Lewis, "Television" (Dictionary), Pitman Publishing Company,
New York, N. Y., 1936.
(20) National Television System Committee, "Television Standards and
Practices," McGraw-Hill Publishing Company, New York, N. Y., 1943.
(21) Radio Corporation of America, "Collected Addresses and Papers on the
Future of the New Art and Its Recent Technical Developments," vol. 1, 1936;
vol. 2, 1937; vol. 3, 1946; vol. 4, 1947.
(22) "Televiser," /. Telev., vol. 4, November-December, 1937.
Report of SMPE
Standards Committee
THE BYLAWS OF OUR SOCIETY wisely provide that the chairmen of
committees "shall not be eligible to serve in such capacity for
more than two consecutive terms." The first of this present year con-
stituted that limit for the writer's service as chairman of the Com-
mittee on Standards, and so made appropriate this final reporting of
the events of that period. At the same time it is hoped that this re-
view, including as it does a description of the terminal status of the
various standardization projects which were being conducted under
the writer's general direction, may be of service to the new chairman
and members of the Committee on Standards. Then too, the acceler-
ating influence of the wartime period on standardization activities has
stimulated a good deal of thinking with regard to the development of
sound peacetime practices in this field, so that I have ventured to in-
clude a certain amount of philosophizing in that connection.
The prewar pace of the Committee on Standards was quite a
leisurely one, determined in part by limited secretarial assistance from
the Society office, but conditioned also by the general feeling that such
a pace was altogether appropriate. In the 10-year period prior to
Pearl Harbor, the parent Committee held an average of about three
meetings a year. During the war the pace slackened to only one or
two meetings a year, and it has continued at this reduced rate to the
present time. The most important reason for this slackening of ac-
tivity during the war was the establishment through the War Produc-
tion Board of a number of War Committees of the American Stand-
ards Association, which operated at an unusually high rate of speed
and effectiveness in the development of War Standards in specific
fields defined by joint committees of the Armed Forces. The sub-
committees as well as the parent War Committee on Photography and
Cinematography, Z52, were staffed in large measure with members of
our Society, and with members of our Committee on Standards in
particular. This war committee considered a total of 72 proposals for
standardization, of which 61 were completely processed as War Stand-
ards in a two-year period ending with the termination of the project
* Presented May 17, 1948, at the SMPE Convention in Santa Monica, by F. T.
Bowditch, retiring chairman.
230 SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
STANDARDS COMMITTEE 231
in February, 1946. This high rate of activity may be compared with
the long-time prior achievement of 44 SMPE Recommended Prac-
tices, of which 33 had been advanced to American Standard (Z22) just
prior to the organization of the Z52 War Committee in December, 1943. 1
It was altogether proper that this high-priority war project should
absorb all the standardization talent and energies of our SMPE mem-
bership during these busy years, and it left us with a heavy portfolio
of postwar projects for consideration as American Standards. Thus it
was that in a meeting of ASA Sectional Committee on Motion Pic-
tures, Z22, held in October, 1945, 22 projects were referred to the
Committee on Standards, calling for the most part for revision of pre-
war American Standards in view of the many changes found necessary
in the preparation of the corresponding War Standards. This com-
paratively large task, judged by the prewar speed rate, could not be
handled effectively by the parent committee sitting as a whole, and so
was assigned to six subcommittees. These subcommittees, as always,
were appointed for the length of time necessary to complete their
assignments, and since they are still active but now under the direction
of a new Committee on Standards, the following detailed report seems
in order.
SUBCOMMITTEE ON CUTTING AND PERFORATING RAW STOCK
A Subcommittee on Cutting and Perforating Raw Stock was es-
tablished on November 8, 1945, under the chairmanship of Dr. E. K.
Carver with Messrs. F. L. Brethaner, A. W. Cook, and D. R. White
as members. This subcommittee was asked to review the following
five American Standards from the standpoint of (a) method of presen-
tation and (b) possible tightening of the limits for 16-mm film.
Z22.5-1941, Cutting and Perforating Negative and Positive Raw Stock (16-
Mm Silent)
Z22.12-1941, Same— (16-Mm Sound)
Z22.17-1941, Same— (8-Mm Film)
Z22.34-1941, Cutting and Perforating Negative Raw Stock (35-Mm Film)
Z22.36-1941, Cutting and Perforating Positive Raw Stock (35-Mm Film)
Four of these projects have since been completed, and were finally
approved by the American Standards Association on July 16, 1947.
The remaining one, Z22.34, was given preliminary approval by the
subcommittee but was then held open at the request of the Research
Council of the Academy of Motion Picture Arts and Sciences, and has
since been reassigned to the subcommittee.
232 STANDARDS COMMITTEE September
The Subcommittee on Cutting and Perforating Raw Stock now has
two projects of critical importance on its agenda. One of these has to
do with a consideration of dimensional standards for 32-mm film,
first, with regard to the reconciliation of conflicting practices with re-
gard to the location of the perforations, and second, with regard to the
effects produced by inaccuracies in slitting, such that the resulting
16-mm film is edge-guided erratically in projection. On account of the
critical effects on sound quality so produced, this project is being
jointly considered with the Committee on Sound.
The other project is a more extensive review of the Z22.34-1941
standard on cutting and perforating 35-mm negative raw stock, in
view of the difficulties pointed out by the Research Council in securing
accurate registration between Standard Positive and Negative film in
the printing of color motion pictures. Because of the new fields of in-
terest thus disclosed, Dr. Carver's subcommittee has been enlarged
by the addition of Messrs. E. A. Bertram, A. F. Edouart, E. Fehnders,
A. M. Gundelfinger, and W. E. Pohl. This group has been doing an
excellent job, and I want to express my sincere appreciation to them,
and, in particular, to Dr. Carver, who is always a tower of strength
wherever standardization accomplishment is needed.
SUBCOMMITTEE ON PHOTOGRAPHIC DENSITY AND SENSITOMETRY
This subcommittee was appointed, under the chairmanship of Dr.
D. R. White, to review the following ASA standards:
Z22.26-1941, Sensitometry
Z22.27-1941, Photographic Density
Messrs. R. Kingslake, G. A. Mitchell, and M. Sweet constitute the
other members of the subcommittee. The Density project has been
successfully processed, and was approved September 26,. 1947, by the
American Standards Association. The Sensitometry project ran into
greater difficulties, but Dr. White is presently optimistic with respect
to an early agreement. In view of this one remaining project, this
group is now titled the Subcommittee on Sensitometry.
SUBCOMMITTEE ON FILM SPLICES
This committee, under the chairmanship of Mr. W. H. Offenhauser,
Jr., was asked to revise the following ASA standards:
Z22.24-1941, Film Splices— Negative and Positive (16-Mm Silent)
Z22.25-1941, Film Splices— Negative and Positive (16-Mm Sound)
19 IS STANDARDS COMMITTEE 233
Messrs. E. A. Bertram, M. R. Boyer, T. R. Craig, A. W. Cook, C.
E. Ives, L. E. Jones, M. W. Palmer, Lloyd Thompson, M. G. Towns-
ley, and E. H. Unkles are the members of this subcommittee. A good
deal of preliminary work has been done, leading to the publication of a
very complete report2 in the Society's JOURNAL in July, 1946. Com-
ment accumulating as the result of this publication is to form the basis
of the subcommittee's final action.
SUBCOMMITTEE ON 16-MM AND 8-MM PROJECTOR SPROCKETS
Under the chairmanship of Dr. Otto Sandvik, a subcommittee con-
sisting of Messrs. H. Barnett, J. A. Maurer, L. T. Sachtleben, and M.
G. Townsley has been assigned the task of revising the following
American Standards :
. Z22.6-1941, Projector Sprocket (16-Mm Film)
Z22. 18-1941, Eight-Tooth Projector Sprockets (8-Mm Film)
It was agreed that the 8-mm and 16-mm fields were so different
from the 35-mm field, covering as they do a wide range of performance
quality in both amateur and professional equipment, that this project
could not be handled most effectively by the existing subcommittee on
35-mm sprockets. It was agreed that the 1941 standards specified
dimensional values which might better be left to the originality of in-
dividual designers, and that the new standards might therefore specify
a design practice insuring good performance with film. With this ob-
jective in mind, Messrs. J. S. Chandler, D. F. Lyman, and L. R. Mar-
tin did a very fine piece of work which resulted in the preparation of a
paper3 published for comment in the JOURNAL. However, basic dif-
ferences of opinion in the subcommittee have so far prevented the rec-
ommendation of a standard, the situation being substantially iden-
tical to that described in a report of the Committee on Standards4
submitted by Dr. Carver over ten years ago. It. is there stated that
"Many members feel that the Committee should not standardize
sprockets of any sort but that their design should be left to the pro-
jector and camera designers to achieve the best results with standard
film." That argument still rages, and has recently precipitated a good
deal of basic consideration as to the proper field of industrial stand-
ardization, a subject which this present report will discuss later on.
In the meantime, the Subcommittee on 16-Mm and 8-Mm Sprockets
is endeavoring to determine whether that part of the original project
essential to the interchangeability of film can be recommended for
234 STANDARDS COMMITTEE September
standardization. The suggestion has also been advanced that a
method of test for the film-handling ability of a sprocket might be
standardized, leaving full freedom for individual design, but permit-
ting the consumer to evaluate it on a sound basis.
SUBCOMMITTEE ON PROJECTION REELS
This subcommittee, under the chairmanship of Dr. D. F. Lyman
and including Messrs. H. Barnett, L. W. Davee, John Forrest, Lee
Jones, M. G. Townsley, and R. T. Van Niman, has been assigned the
task of reviewing the following ASA standards.
Z22.4-1941, Projection Reels (35-Mm Film)
Z22.11-1941, Projection Reels (16-Mm Film)
Z22.23-1941, Projection Reels (8-Mm Film)
It was suggested that these standards should be rewritten in the
style employed in the dimensional aspects of the American War Stand-
ard Specification for 16-Mm Motion Picture Projection Reels and
Containers, Z52.33-1945. This war standard recognized certain de-
sirable design considerations not included in the 1941 standards, such
as the ratio of core-to-flange diameters, the specification of a flange
diameter great enough to hold the rated film length with safety, and
the specification of a flange separation which will neither hold the film
too tightly nor permit lateral wandering. Difficulty is being ex-
perienced in reconciling these requirements with the characteristics of
equipment presently on the market, since no manufacturer is anxious
to vote for the obsolescence of his present goods and production equip-
ment. ' This, then, is another instance of the basic difficulties en-
countered in design standardization, particularly in a well-developed
field where many individual design practices are well established.
Perhaps here too the solution will be found in the specification of a
test method. Certainly this is to be preferred over a design standard
which does no more than specify the range of present trade practices.
SUBCOMMITTEE ON 8-MM AND 16-MM CAMERA AND PROJECTOR
APERTURES
This subcommittee consists of Mr. J. A. Maurer, chairman, with
Messrs. M. G. Townsley, L. T. Goldsmith, H. J. Hood, W. C. Miller,
and L. T. Sachtleben, and has been assigned the task of reviewing the
following ASA Standards :
1948 STANDARDS COMMITTEE 235
Z22.7-1941, Camera Aperture (16-Mm Silent)
Z22. 13-1941, Camera Aperture (16-Mm Sound)
Z22.19-1941, Camera Aperture (8-Mm Silent)
Z22.8-1941, Projector Aperture (16-Mm Silent)
Z22. 14-1941, Projector Aperture (16-Mm Sound)
Z22.20-1941, Projector Aperture (8-Mm Silent)
It was suggested that consideration should be given to war stand-
ards procedures in this field, since these went much farther in de-
nning good engineering practice and in explaining the reasons for
certain dimensional choices. The subcommittee has held several
meetings and at present is hopeful of an early successful conclusion of
this assignment.
In addition to the six subcommittees working on ASA assignments
growing out of wartime standardization, the Committee on Standards
has two other active subcommittees, and a number of projects now
under preliminary consideration. These are as follows :
SUBCOMMITTEE ON THE PHOTOMETRIC CALIBRATION OF CAMERA
LENSES
This very able subcommittee is under the chairmanship of Dr. R.
Kingslake, with a membership consisting of Messrs. F. G. Back, E.
Berlant, J. W. Boyle, L. E. Clark, C. R. Daily, I. G. Gardner, G.
Laube, E. B. Levinson, J. A. Maurer, A. E. Murray, J. Thompson,
M. G. Townsley, and G. C. Whitaker. This project was suggested by
the need of cinematographers for a lens transmission calibration of
some sort which could be used in combination with set-lighting infor-
mation to determine proper film exposures. The present //number
markings are not sufficiently indicative of the light transmission of a
lens, particularly since the advent of lens coatings. In joint meetings
with the Research Council, it was agreed that SMPE would study this
problem from a calibration method and apparatus standpoint, while
the Research Council and the American Society of Cinematographers
would continue their already active program of practical evaluation of
proposed calibration methods.
Dr. Kingslake has aggressively prosecuted this work to the point
where a proposed calibration method and system of lens markings has
been agreed to by the Eastern members of his committee, and these
are now under review by the West Coast studio group. An early
agreement is hoped for, which will constitute a very real technical
service to the motion picture industry.
236 STANDARDS COMMITTEE September
SUBCOMMITTEE ON 35-MM PROJECTION SPROCKET DESIGN
This subcommittee has Dr. E. K. Carver as chairman, with a mem-
bership consisting of Messrs. H. Barnett, M. H. Bennett, M. R. Boyer,
L. W. Davee, J. L. Forrest, C. F. Horstman, L. B. Isaac, and H.
Rubin. They have completed a very creditable job5 with respect to
the standardization of a larger sprocket diameter (0.943-inch instead
of 0.935-inch) which results in greatly reduced film wear, an item of
substantial economic benefit to the trade. American Standard Z22.35-
1947, was finally approved on July 16, 1947, incorporating this find-
ing. At the same time, the ASA-Z22 Committee has recognized that
other aspects of this standard are also in need of revision, so the sub-
committee is being retained to extend its study of this sprocket.
At the first of the year, therefore, the Committee on Standards had
eight actively functioning subcommittees working on projects of in-
terest to the industry. The following additional projects were in the
status indicated :
PROPOSED STANDARD FOR 35-MM FLUTTER TEST FILMS
After their publication for comment,6 these two proposals were re-
cently presented to the Committee on Standards by the Committee
on Sound, through the chairman, Dr. F. G. Frayne. Following their
consideration (in December, 1947) in a meeting of the Committee on
Standards, these proposals were returned to the Committee on Sound
with suggestions for minor changes. A formal ballot on the recom-
mendation of these proposals for final standardization is anticipated
soon.
16-MM AND 35-MM DAYLIGHT LOADING SPOOLS
It has been suggested that the field of "16-Mm and 35-Mm Day-
light Loading Spools" is in need of standardization. The Engineering
Vice-President, Mr. J. A. Maurer, has agreed that this is a proper sub-
ject for study, and has undertaken to secure the co-operation of a firm
to carry the chairmanship burden of a subcommittee on this project.
35-MM FILM CANS
Present 35-mm film cans are of varying diameter and embossing, so
that they do not stack conveniently nor otherwise handle to advantage
in processing laboratories. The need for standardization is indicated
and the Staff Engineer is conducting a preliminary survey to define
the task better before assigning it.
1<)4S STANDARDS COMMITTEE 237
.•
FILM-CUING MARKS
Film-printing laboratories, particularly 16-mm, have indicated the
need for a uniform system of film-cuing marks, and it has been sug-
gest ed that a representative group be organized to collect and analyze
suggestions from the various laboratories. This project was pending
in this form at the end of the year, and possibly should be assigned to
the Laboratory Practices Committee.
16-MM PRINT LENGTHS AND REEL SIZES
The Division of Motion Pictures and Sound Recordings of the
Nat ional Archives has suggested a need for the standardization of a
limited number of print lengths and reel sizes to facilitate storage.
This project is presently under review by the Staff Engineer.
35-MM SOUND-TRACK DIMENSIONS
A request has been received for the consideration of the five dif-
ferent types of sound tracks from the standpoint of possible standard-
ization. This has been referred by Mr. Maurer to the Committee on
Sound for first consideration.
FILM PRESERVATION
A proposal has been made that a standard procedure for the pres-
ervation of film for historical purposes should be worked out. An
early proposal from the Motion Picture Project of the Library of
Congress is anticipated, and this will be studied by the Engineering
Committee on the Preservation of Film for recommendation to the
'Committee on Standards.
GLOSSARY OF MOTION PICTURE TERMS
In 1943, Dr. John Andreas of the Technicolor Motion Picture Cor-
poration, submitted to the Engineering Vice-President a very exten-
sive Glossary of Motion Picture Terms. A number of copies was
prepared, and these were circulated to members of the Committee on
Standards for consideration as standards. This informal procedure
gave rise to a great many comments as 'to the detail of the wordings,
and in the submission of new terms for consideration at a rate faster
than we were able to secure agreement on the definitions for the old
ones. After trying several procedures, each dependent upon the dona-
tion of a great deal of individual and member-company effort, largely
238 STANDARDS COMMITTEE September
*
of an uninspiring routine nature, it was recognized that the project
could only be advanced at a satisfactory rate with the aid of consider-
able organizational supervision by the Society office. The project
was thus transferred to the Engineering Secretary in 1946, and re-
moved from the agenda of the Committee on Standards at that time.
Since then, it has been decided that the estimated cost in time and
money required to complete the Glossary is greater than can reason-
ably be assumed at this time in view of other commitments. This proj-
ect is thus altogether inactive at present. It is mentioned here so that
the following suggestion may be presented for general consideration.
Assuming that some sort of standard format could be agreed upon,
the Glossary project could be established on a permanent basis
through the formation of a Glossary Subcommittee in each Engineer-
ing Committee of the Society. These might be correlated through a
subcommittee of the Committee on Standards, the membership of
which would consist of the chairmen of the other subcommittees.
Since a Glossary is not a fixed thing, but changes from year to year as
new arts and new usages develop, Glossary subcommittees would al-
ways be in existence, although the personnel would be subject to
change with each new term of the appointing officer.
CONCLUSION
This leads naturally into the concluding section of this report,
which has to do with suggestions for a more effective Standards Com-
mittee procedure, based upon both the achievements and the disap-
pointments of the past. These comments have been arranged numer-
ically, in an order determined by the development of the ideas pre-
sented for consideration.
1. The membership of the retiring Committee on Standards was
chosen in an effort to secure a wide representation of technical ability
in all fields of engineering interest so that competent subcommittees
could be formed within the Committee membership. This led to the
formation of a large, unwieldy parent committee, consisting for the
most part of key members of the other Engineering Committees. Sub-
committees of the Committee on Standards were thus barely dis-
tinguishable from the associated Engineering Committees, so that it
came to be realized that a needless duplication of organizations was
occurring.
As a matter of principle, therefore, it was agreed that it is in general
unwise to form a subcommittee of the Committee on Standards in a
1948 STANDARDS COMMITTEE 239
field served by another Engineering Committee. If the Committee on
Standards agrees that a technical study is needed in a particular case,
the project should be assigned to the appropriate Engineering Com-
mittee, unless there is some good reason to believe that its completion
will be unduly delayed in this way. Since all major fields of interest
are served by Engineering Committees composed of the top experts in
their respective fields, the existence of an active group of such com-
mittees would require the formation of very few subcommittees of the
Committee on Standards.
2. The Committee on Standards should therefore be chosen pri-
marily as a policy-making group, to determine the type of study which
each proposal for standardization is to receive. This will require the
careful preparation of a preliminary analysis of each project, setting
forth the nature of the present art, and discussing frankly the eco-
nomic factors which may determine conflicting viewpoints. For in-
stance, in a great many cases, present practices cannot all be con-
ducted within a single standard, so that compliance is certain to cause
hardship to whoever is outside any final agreement, while a perform-
ance standard is naturally opposed by those who wish to operate in a
highly competitive amateur market rather than in the higher quality
professional field. In full recognition of such factors, it should then be
the primary duty of the Committee on Standards to decide either (a)
that the advantage to the industry is sufficient to require the prepara-
tion of a standard, (b) that a further study is desirable in order to
define the advantages of the proposal and the area of disagreement
better, or (c) that the proposal is not a proper one for present considera-
tion. As a matter of general policy, it is suggested that matters hav-
ing to do strictly with interchangeability, nomenclature, and methods
of test properly belong in the first category. On the other hand, a
performance specification should be approached with considerably
greater caution.
3. The task itself should then be assigned to a presently function-
ing Engineering Committee, if one exists in the field of interest. Only
where this is not the case should a special subcommittee of the Com-
mittee on Standards be established. As has already been indicated,
this is because the personnel best suited to the task likely are already
members of the appropriate Engineering Committee, and needless
confusion exists if these persons are asked to serve in two capacities.
At the same time, it is fruitless to attempt to interest the diverse rep-
resentation required of the entire Committee on Standards in the
240 STANDARDS COMMITTEE September
details of a proposition confined to a single technical field, nor can a
sound conclusion be reached in such a way.
4. The task group thus established should be charged in accord-
ance with the policy decision reached by the Committee on Standards.
If the determination of a standards recommendation is decided upon,
then the task group should be given the authority necessary to resolve
controversial discussions. A two-thirds vote rather than a substan-
tial unanimity might be established as the determining factor in such
a case. If only a study is to be made the primary duty of the task
group should be the preparation of a detailed report, completely de-
fining all controversial aspects, with a majority recommendation with
respect to further action. If, in the course of this study, substantial
unanimity is reached with respect to a definite standards proposal,
this can then be considered by the Committee on Standards. If no
such proposal is made, then the Committee on Standards should de-
cide whether (a) the new facts justify a policy decision that a standard
should be established, or (b) that the task group report be given suit-
able publicity as defining the present state of the art for the guidance
of the industry.
5. The Committee on Standards should seldom if ever attempt to
review the details of a task-group recommendation ; only the broader
implications as to the benefit of such a standard to the industry as a
whole should receive consideration. Nothing is so discouraging to a
capable task group as to have its hard-won compromises discarded,
and its recommendation rejected because the same old arguments
flare up in the parent committee. The chairman should be in a sound
position to refuse to reopen these arguments through assurance that
all viewpoints have already had their day in court in the task group.
This, of course, requires the formation of a truly representative task
group, which is essential to a worth-while result in any case.
6. Finally, if such an ideal routine should become operative,
the Committee on Standards would be in the very desirable position
of devoting most of its energies to the development of a sound philoso-
phy in the field of standardization. Viewpoints such as many of us
have recently been expressing in correspondence would be exchanged
around a conference table, and a basic policy developed which could
be specially adapted to each proposal for standardization. Such an
opportunity should result in the development of a sound attitude
which would gain for the Society the respect and compliance of the
industry in so far as standardization authority is concerned.
1948 STANDARDS COMMITTEE 241
It is realized that the foregoing is not all new, nor is it all my own.
It merely represents present ideas of a good program for the future,
based, as has been said, both upon the achievements and the disap-
pointments of the past. As ex-chairman of the Committee on Stand-
ards, my complete lack of authority in this field is fully recognized;
and I hope that these suggestions will be considered simply for what-
ever good may be derived from them in setting up a more definite and
aggressive program for the future, in keeping with the splendid
growth, not only in the size, but also in the technical responsibilities
of our Society.
REFERENCES
(1) "Recommended Practices of the Society of Motion Picture Engineers,"
«/. Soc. Mot. Pict. Eng., vol. 38, pp. 403-456; May, 1942.
(2) "Report of the Subcommittee on 16-Mm Film Splices," /. Soc. Mot. Pict.
Eng., vol. 47, pp. 1-11; July, 1946.
(3) "Proposals for 16-Mm and 8-Mm Sprocket Standards," J. S. Chandler,
D. F. Lyman, and L. R. Martin, /. Soc. Mot. Pict. Eng., vol. 48, pp. 483-520;
June, 1947.
(4) "Report of the Standards Committee," /. Soc. Mot. Pict. Eng., vol. 28,
pp. 21-23; January, 1937.
(5) Report of the Subcommittee on Projector Sprocket Design," J. Soc. Mot.
Pict. Eng., vol. 45, pp. 73-75; August, 1945.
(6) "Proposed Standard Specifications for Flutter or Wow as Related to
Sound Records," /. Soc. Mot. Pict. Eng., vol. 49, pp. 147-162; August, 1947.
Standards Committee 1947
F. T. BOWDITCH, Chairman
J. M. ANDREAS IRL GOSHAW W. H. OFFENHAUSER, JR.
M. F. BENNETT HERBERT GRIFFIN J. D. PHYFE
E. A. BERTRAM A. C. HARDY W. L. PRAGER
M. R. BOYER R. C. HOLSLAG G. F. RACKETT
F. L. BRETHAUER J. K. HILLIARD A. C. ROBERTSON
F. E. CARLSON D. B. JOY L. T. SACHTLEBEN
M. K. CARVER W. F. KELLEY OTTO SANDVIK
J. S. CHANDLER* R. KINGSLAKE C. R. SAWYER
A. W. COOK P. J. LARSEN J. A. SCHEIK
E. D. COOK C. L. LOOTENS R. R. SCOVILLE
L. W. DAVEE G. T. LORANCE J. H. SPRAY
A. A. DURYEA D. F. LYMAN MONROE SWEET
A. F. EDOUART PIERRE MERTZ LLOYD THOMPSON
P. C. GOLDMARK W. C. MILLER M. G. TOWNSLEY
A. N. GOLDSMITH H. W. MOYSE D. R. WHITE
L. T. GOLDSMITH H. E. WHITE
* Advisory Member.
Errors in Calibration of the / Number
BY FRANCIS E. WASHER
NATIONAL BUREAU OF STANDARDS, WASHINGTON, D. C.
Summary — The present system of marking the diaphragm stops in terms
of the geometric /number is subject to serious deficiencies so far as uniform
performance for lenses set at the same marked stop opening is concerned.
Decisions regarding the proper exposure time to use at a selected stop open-
ing may be in error by ±10 per cent for a lens whose surfaces do not have
antireflection coatings, and by even greater amounts for a lens whose
surfaces do have antireflection coatings. These errors arise from differences
in the reflection and absorption losses in the lens elements themselves, de-
partures of the measured from the nominal focal length, and departures of
the measured diaphragm openings from the nominal diaphragm openings.
A method is described whereby a lens can be calibrated by a light meter
in terms of an ideal lens so that the variation in axial illumination in the focal
plane need not exceed ±2 per cent in using different lenses set to the same
calibrated stop opening.
PREFACE
IN PROBLEMS OF photography where the accuracy of lens markings
is critical in determining the proper exposure, the various errors
to which these markings are subject is of considerable interest. The
present report gives the magnitude of such errors that were found to
exist in a representative group of 20 lenses having focal lengths that
range from 1/2 to 47.5 inches. In addition, the results of calibration
of these lenses by a photometric method that permits compensation
of light losses resulting from absorption, reflection, and scattering are
given. Values of lens transmittance for these lenses are shown. A
method of plotting results of nominal, true, and calibrated / numbers
is given that permits quick evaluation of the magnitude of the over-all
error in terms of fractions of a stop.
I. INTRODUCTION
With the advance of photographic technology, a need has developed
for more precise information on the light-transmitting characteristics
of photographic objectives. In particular, a specific need exists for a
more accurate means of marking or calibrating the lenses which em-
ploy a variable stop for adjusting the lens speed. The usual method,
at present, of calibrating a lens is to inscribe a scale of /numbers on the
242 SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
CALIBRATION OF / NUMBER 243
diaphragm control. These / numbers are based upon certain geo-
metric properties of the lens, and neglecting errors of marking, pro-
vide a satisfactory means of varying the speed of the particular lens
by definite integral steps. Unfortunately this system of marking
takes no cognizance of differences in light-transmitting properties that
occur among different types of lenses and in addition those differences
that result between lenses of the same type when the surfaces of one
have been treated to reduce reflection losses.
This problem has been under vigorous attack for the past ten years
and numerous methods 1-11 have been devised for the rating of lens
speed with respect to some standard. These methods differ in such
matters as type of light source, comparison lens or standard aperture,
and type of light-registering device. The theoretical aspects of the
problem have been discussed by McRae8 and by Gardner1' 2 who pro-
posed several possible methods for calibration of a lens. In the present
article, one of the methods described by Gardner is verified experi-
mentally. The experimental technique is described and the variations
in performance for 20 lenses, having focal lengths that range from 0.5
to 47.5 inches, are shown. Attention is given to sources of error in the
existing marked / number. Lastly, a process is described for deter-
mining the transmittance of a lens from data obtained in the source
of calibration.
II. APPARATUS AND METHOD OF MEASUREMENT
The apparatus consists essentially of a broad uniform source of
white light, a sensitive light-measuring device, and a holder which can
be used interchangeably either for mounting the lens under test or one
of a series of standard diaphragms, each of which has a centrally
located circular opening of known diameter. The arrangements of
these elements is the same as that suggested by Gardner.1- 2 The rel-
ative lens speed is determined by a comparison of the quantity of
light flux transmitted by a lens with that transmitted by a circular
opening. By making an appropriate series of measurements and by
proper interpretation of their significance, the lens can be calibrated in
terms of an "ideal" lens having 100 per cent transmittance.
1 . Procedure for a Lens
A lens is mounted in the holder and its axis is aligned with the center
of the broad uniform source and the center of the small circular open-
ing in the baffle covering the sensitive element of the light-measuring
244 WASHER September
device. The front of the lens faces the light source and the distance
separating the rear nodal point of the lens and the baffle covering the
light-sensitive element is adjusted to equality with the equivalent
focal length / of the lens. The opening in the baffle does not usually
exceed 1 mm except for some lenses of very long focal length in which
cases it is kept under 0.01 /. All parts of the equipment are shielded
so that only light from the source that passes through the lens can
reach the light-sensitive element.
Readings of the light meter are taken at each of the marked stop
openings. To minimize error arising from backlash, readings are
taken both for the condition of the setting at the marked / number
being made with the diaphragm ring of the lens moving in the closing
direction and with the diaphragm ring moving in the opening direc-
tion.* The readings from these two sets of observations are averaged
and this value is taken as the accepted reading of the light meter at a
given marked stop opening.
2. Procedure for the Standard Diaphragms
The lens is replaced by one of the series of standard diaphragms
which have centrally located circular openings with known diameters.
The reading of the light meter is taken and the distance D, from the
diaphragm to the baffle covering the light-sensitive element, is
measured. This operation is repeated for several of the standard dia-
phragms so selected that readings of the light meter are obtained
throughout the same range of readings that were observed for the
various marked apertures of the lens. The brightness of the source
and the sensitivity of the light meter are kept unchanged throughout
both parts of the experiment. To ensure constancy of brightness of
the source, a constant-voltage transformer is used to maintain a con-
stant voltage for the lamps that illuminate the broad uniform source.
To minimize error, two sets of data are taken for both the lens and the
series of standard diaphragms so intermingled that random fluctua-
tions in the brightness of the light source and in the sensitivity of the
light meter can be neglected.
Ideally the diameters of the standard diaphragm openings should be
so chosen that the same series of / numbers are present in both phases
* Ten lenses (Nos. 10, and 12 to 20, inclusive) were calibrated in this manner.
The remaining ten lenses were calibrated with the diaphragm ring moving in the
closing direction only in accordance with the recommendation contained in Report
No. 6 of the Subcommittee on Lens Calibration of the Society of Motion Picture
Engineers on November 6, 1947.
1948
CALIBRATION OF / NUMBER
245
of the experiment. Too, the distance D should equal the equivalent
focal length / of the lens. In practice, however, it has proved to be
more convenient to let D differ from/ and to place more reliance upon
the ratio D/A, where A is the diameter of the circular opening in a
standard diaphragm. When a wide variety of lenses is being cali-
brated, as is the case in this experiment, it is simpler to compute the
/ number of the standard diaphragm from the ratio D/A and to deter-
mine the performance for the conventional series of / numbers from
the curve of light-meter reading versus / number than to attempt to
reproduce the conventional set of / numbers by appropriate selection
of values of D and A .
D/
Fig. 1 — Calibration curve for computing /number
of standard diaphragms when the value of D/A is
known.
The / number for a lens is defined by the equation
1
/ number
2 sin
(D
where a is the angle between the axis and the extreme ray of the cir-
cular conical bundle transmitted by the lens. In the case of the stand-
ard diaphragm, the relation connecting the measured quantities D
and A is
D = 1_
'A 2 tan a
(2)
Accordingly the values of the / numbers for the standard diaphragms
can readily be computed from the known values of D/A. A suffi-
ciently accurate determination of the / number can be made with the
aid of a curve such as is shown in Fig. 1. To produce this curve, the
246
WASHER
September
values of the quantity, / number, D/A, are plotted as a function of
D/A. Hence, for a given value of D/A, the increment that must be
added thereto to yield the /number can be read easily from the graph.
For values of D/A greater than 15, the values of D/A and / number
are equal for all practical purposes since their difference is less than
0.1 per cent.
III. RESULTS OF MEASUREMENT
When the values of the scale deflections of the light meter are plot-
ted against the / numbers of the standard diaphragms on logarithmic
i
I
§
I10
\\
Fig. 2 — Scale deflection on light meter versus /
number. Curve No. 1 is for the standard diaphragms.
Curve No. 2 is for the lens under test.
paper, the resulting curve is a straight line with a slope nearly equal
to 2. The fact that the slope is not exactly 2 may be attributed to a
slight departure from linearity of the response of the light meter to
varying amounts of light indicated on the receiver. This curve, shown
as curve 1 in Fig. 2, shows the relation between the scale deflections of
the light meter and the /numbers of an ideal lens.
In a like manner, the values of the scale deflection of the light meter
are plotted against the / numbers of the actual lens on the same curve
1948 CALIBRATION OF / NUMBER 247
sheet. The resulting, curve, designated curve 2 in Fig. 2, is a straight
line parallel to curve 1 but displaced laterally therefrom. This dis-
placement shows in a striking manner the effect of light losses in the
actual lens. A fairly close approximation of the relative light trans-
mission of the actual lens at a given /number can be made at once, as
it is simply the ratio of the ordinates of curve 1 and curve 2 for the
given /number.
It must be mentioned that while curve 1 is always a straight line,
this is a consequence of its accurately determined / numbers. On the
other hand, the / numbers for curve 2 are read directly from the lens
markings and are subject to a variety of errors that will be discussed
later in the paper. As a result of these random and systematic errors
the points for curve 2 sometimes do not fall as close to the straight
line drawn as could be desired. This is especially noticeable at the
small apertures associated with the large / numbers. However, these
variations in no way interfere with validity of the final results but are
in fact helpful in tracking down errors in the /numbers.
The values of the calibrated / numbers for the actual lens may be
obtained readily from these curves. The calibrated / number is a
term used to designate the/ number of an ideal lens (i.e., a lens having
100 per cent transmittance) transmitting the same amount of light
that is transmitted by the actual lens at a given marked / number.
The terms T-aperture ratio or T stop 3- 6- 7 and equivalent-aperture
ratio 1 are other designations of this same quantity. To determine
the calibrated / number, the value of the scale deflection for a given
marked / number of the actual lens is noted and the value of the /
number of the ideal lens for which the same scale deflection is ob-
tained is read from curve 1. This has been done for twenty lenses
covering a wide range of focal lengths and / numbers. The results are
listed in Table I.
The unusual values of marked / numbers which are listed in the
first column result from assigning a calibrated / number to the maxi-
mum stop opening for each lens. The maximum stop opening of a
lens quite frequently does not fall in the commonly accepted series of
marked / numbers although the remaining marked / numbers of the
lens usually do. The calibrated / numbers, in most instances, are
larger than the marked / numbers. This is as expected because it is
known that some of the light incident on the front surface of a lens is
lost as a result of reflection back in the object space or by absorption
in the glass. The considerable differences in the calibrated / numbers
248
WASHER
September
for a given marked / number indicate appreciable differences in the
light-transmitting qualities of the various lenses. This is illustrated
in Fig. 3 where the calibrated / numbers are plotted on semilogarith-
mic paper for ten lenses. The values are given for the marked / num-
bers. 4, 8, and 16. Departures as great as 1/9 stop opening are indicated
in many instances. Since the departures may be in either direction
from the marked stop opening, it is possible to select two lenses such
that, on using each for the same scene at the same marked stop open-
ing, the effective difference in exposure is equal to that produced by a
change in excess of one full-stop opening. The fact that some lenses
have calibrated / numbers less than the marked stop opening may
seem anomalous in that it indicates a transmittance greater than
unity. This is, however, for the most part, an indication of errors
in the marked stop opening and will be discussed in more detail in a
later section.
'
1
"
LENS NUMBER
Fig. 3 — Departure of the calibrated / number
from the marked / number at //4, //8, and //1 6
for 10 lenses. The line separations shown are
equal to one stop opening.
Lens No. 7 is of especial interest in that the indicated stop openings
are marked in T stops, consequently the values of the calibrated /
numbers are quite close to the marked / numbers. Lenses Nos. 2, 3,
7, 9, 11, and 20 have coated surfaces to reduce reflection losses. The
gain in transmittance is definitely present but is somewhat obscured
in Table I because the marked aperture ratios frequently differ from
the true geometric-aperture ratio.
The fact that the calibrated / number varies so much from lens to
lens for the same nominal / number gives support to the proposition
MEASURED VALUE OF THE CALII
LENGTHS THAT RANGE FROM
1
2
3
17
18 19
20
Marked
f
Number
0.5
0.5
1.0 |
19.0
24.0
30.0
47.5
1.9
2.40
2.09
2.2
2.3
2.5
2.82
2.7
2.8
3.25
3.13
2.86
3.0
3.5
4.0
4.42
4.45
3.92
4.5
5.6
5.32
6.32
5.52
6.8
7.5
8.0
7.80
8.78
7.50
§
9.5
3
11.0
11.0
11.8
9.94
jl
13.6
14.3
12.5
15.6
15.0
16.0
16.0
15.4
17.2
13.6
5
19.2
19.8
19.3
17.3
22.0
21.3
!
1
25.8
28.2
26.7
23.3
32
3
37.6
40.9
39.2
34.7
45
3
50.8
59.9
53.4
48.6
64
3
69.6
86.8
79.0
71.1
90
3
98.0
117.0
99.0
97.6
128
143.0
SETTINGS OF THE STOP OPENINGS IN llO PER CENT TRANSMITTANCE FOR
Cali-
1 | 2
3 16
17 | 18 | 19
20
brated
f
number
0.5
0.5
1.0 £.5
19.0
24.0
30.0
47.5
2.8
2.33
2.48
2.74
4.0
3.58
3.60
4.08
5.7
5.92
5.10
5.80
8.0
8.10
7.24
8.57
11.3
11.3
10.5
12.7
*.60
16
16.4
14.9
19.1
5.1
13.2
12.4
12.8
14.6
22.6
23.2
&.6
19.2
18.2
18.8
20.6
32
i.9
27.4
25.2
26.2
29.0
45
a A
5.2
R Q
39.3
KQ ft
35.1
/IT Q
37.3
KQ A
41.0
CO O
194S CALIBRATION OF / NUMBER 249
that all lenses should be so marked that differences in light-transmit-
ting properties are negligible for a given / number. This can be done
from the curves shown in Fig. 2, by reversing the procedure used in
deriving the information reported in Table I. The deflection of the
light meter for a given / number of the ideal lens is noted on curve 1
and the / number of the actual lens which will yield the same deflec-
tion is read from curve 2. This can also be done by plotting the cali-
brated / number for a lens listed in Table I against the marked / num-
ber on logarithmic paper. The marked / number for a given cali-
brated / number can then be read directly from the graph. This has
been done for the same 20 lenses and the results are listed in Table
II. This table shows the proper settings in terms of the marked /
number so that each of these lenses will yield uniform performance
for each of a series of calibrated / numbers.
IV. SOURCES OF ERROR IN THE NOMINAL / NUMBER
In addition to the light losses in the lens arising from absorption and
reflection, there are several sources of error that affect the reproduci-
bility in the amount of light reaching the focal plane at a given stop
opening. The first of these is backlash in the iris-diaphragm stop and
results in differences in light transmission dependent upon the manner
in which the diaphragm is set at a given stop opening. The second
error is an actual error in the markings themselves and may arise from
errors in aperture, errors in equivalent focal length, or errors in both
at the same time. The backlash error varies for each lens while the
error in / markings contributes to variations in performance when
several different lenses are in use for the same type of work.
6 . Error in Setting the Lens at a Given f Number
When the diaphragm is set at a given / number, there is an appreci-
[ able difference in the amount of light passed by the lens dependent
upon the direction of movement of the diaphragm control. The error
arising from this source has been investigated and the results are
listed in Table III for several lenses. This backlash error is deter-
mined by two methods. In the first method, the lens is mounted on a
stand and the edges of the diaphragm are illuminated from tjie rear of
the lens by a fixed source. Photographs of the stop opening are made
with an auxiliary camera placed in front of the lens. Each stop open-
ing is photographed for the condition of the setting being made with
250
WASHER
September
the diaphragm closing and with the diaphragm opening. Prints are
made of these negatives and the area of each image is measured with
a planimeter. Let the area of the image, taken for the condition when
the setting is made by closing the diaphragm, be Ac; and the area of
TABLE III
RATIOS OF RELATIVE ILLUMINATION IN THE AXIAL REGION OF THE FOCAL PLANE
FOR LENSES USED UNDER IDENTICAL LIGHTING CONDITIONS, SETTINGS BEING
MADE WITH THE DIAPHRAGM CONTROL MOVING TO CLOSE AND WITH THE DIA-
PHRAGM CONTROL MOVING TO OPEN THE LENS
Nominal
Focal
Length,
Inches
Nominal
/ Number
Ratio of Light Transmissions
Diaphragm Closing to Diaphragm
Opening
Planimeter,
Ac/Ao
Light Meter,
Lc/Lo
Weighted
Average
16.5
9.5
1.01
1.04
1.03
11
1.01
1.02
1.02
16
1.02
1.04
1.03
22
1.02
1.07
1.06
32
1.05
1.11
1.10
45
1.13
1.08
1.09
64
1.11
1,08
1.09
19.0
11
1.00
1.00
1.00
16
1.06
1.02
1.03
22
1.05
1.04
1.04
32
1.07
1.06
1.06
45
1.10
1.09
1.10
64
1.24
1.26
1.26
24
11
1.00
1.00
1.00
16
1.00
1.03
.1.02
22
1.05
1.05
1.05
•
32
1.02
1.11
1.09
45
1.09
1.14
1.13
64
1.06
1.18
1.16
30
12.5
0.99
1.01
1.00
16
1.04
1.03
1.03
22
1.02
1.02
1.02
32
1.04
1.06
1.05
45
1.08
1.02
1.03
64
1.08
1.07
1.07
the image for the same stop opening, taken for the condition when the
setting is made by opening the diaphragm, be Ao. Then the ratio
Ac/Ao ist accepted as the ratio of the relative illuminations in the
axial region of the focal plane when the lens is used under identical
lighting conditions for these two processes of setting the lens at a
given / number.
1948
CALIBRATION OF / NUMBER
251
In the second method, the data taken in Section II are treated in
such manner as to separate the light-meter readings Lc, taken for the
condition of the setting being made with the diaphragm closing, and
the light-meter readings for the same stop opening Lo, taken for the
condition of the setting being made with the diaphragm opening.
Then the rate Lc/Lo is accepted as the ratio of the amounts of light
TABLE IV
COMPARISON OF NOMINAL AND MEASURED VALUES OF EQUIVALENT OF FOCAL
LENGTH AND EFFECTIVE APERTURE FOR A REPRESENTATIVE GROUP OF LENSES
Differ-
ence
in
Effective
Differ-
Equivalent
Focal Length
Equiva-
lent
Focal
Aperture
ence
in
Aper-
M!eas-
Lens
Nominal,
Measured,
Length,
Nominal,
ured,
ture,
No.
Mm
Mm
Per Cent
Mm
Mm
Per Cent
1
12.5
12.35
-1.2
6.58
7.07
+7.4
2
. 12.5
12.99 .
+3.5
5.00
5.07
+ 1.4
3
25.4
25.56
+ 1.0
13.37
13.65
+2.1
4
35.0
37.50
+7.1
12.96
14.06
+8.5
5
40.0
42.08
+5.2
14.81
14.94
+0.9
6
50.0
51.39
+2.8
18-. 52
19.62
+5.9
7
50.8
50.62
-0.4
25.40
24.40
-3.9
8
75.0
75.31
+0.4
26.78
27.36
+2.2
9
75.0
75.02
0.0
32.61
32.58
-0.1
10
76.2
74.71
-2.0
25.40
24.60
-3.2
11
101.6
99.42
-2.1
39.53
40.64
+2.8
12
177.8
180.81
+ 1.8
26.15
26.15
0.0
13
190.5
190.53
0.0
42.34
40.17
-5.1
14
279.4
284.85
+2.0
34.92
35.74
+2.3
15
342.9
351.60
+2.5
45.72
42.21
-7.7
16
419.1
418.14
-0.2
44.12
41.30
-6.4
17
482.6
481.97
-0.1
43.87
43.29
-1.3
18
609.6
605.55
-0.7
55.42
51.40
-7.2
19
762.0
756.54
-0.7
60.96
59.14
-3.0
20
1206.5
1207.60
+0.1
passing through the lens for these two conditions and is comparable to
Ac/Ao obtained by the first method.
The values of these ratios are tabulated in Table III for a series of
stop openings for four lenses. The differences by the two methods re-
sult mainly from the fact that a greater number of sets of data is used
in the determination of Lc/Lo, The third column gives the weighted
252
WASHER
September
average with a weight of 4 given to Lc/Lo and a weight of 1 given to
Ac/Ao. It is noteworthy that this error arising from backlash varies
from 1 to 2 per cent at the larger stop openings to as high as 10 to 26
per cent for the smaller stop openings. It is clear that error from this
cause can be avoided by always making the diaphragm setting in the
same manner, and preferably in the direction of closing the diaphragm.
TABLE V
NOMINAL AND MEASURED VALUES OF THE / NUMBER FOR A REPRESENTATIVE
GROUP OF LENSES
Lens
Number
Nominal
Focal
Length,
Mm
/ Number
Error in
/ Number,
Per Cent
Relative
Transmit-
tance
Nominal | Measured
1
12.5
1.9
1.77
-6.8
1.15
2
12.5
2.5
2.62
4.8
0.91
3
25.4
1.9
1.87
-1.6
1.03
4
35.0
2.7
2.67
-1.1
1.02
5
40.0
2.7
2.82
4.4
0.92
6
50.0
2.7
2'. 62
-3.0
1.06
7
50.8
2.2
2.07
-5.9
1.13
8
75.0
2.8
2.75
-1.8
1.04
9
75.0
2.3
2.30
0.0
1.00
10
76.2
3.0
3.04
1.3
0.97
11
101.6
2.5
2.51
0.4
0.99
12
177.8
6.8
6.91
1.6
0.97
13
190.5
4.5
4.74
5.3
0.90
14
279.4
8.0
7.97
-0.4
1.01
15
342.9
7.5
8.33
11.1
0.81
16
419.1
9.5
10.12
6.5
0.88
17
482.6
11.0
11.13
1.2
0.98
18
609.6
11.0
11.78
7.1
0.87
19
762.0
12.5
12.79
2.3
0.96
There still remains the random error of making the setting, even if
care is taken to move the control always in the same direction. This
error is, however, small in comparison to backlash error, and it is be-
lieved that it should be negligible for the careful worker at the larger
stop openings and perhaps rising to approximately one fourth of the
backlash error for the smaller stop openings.
2. Errors in the Existing Geometrical f Number
(a) At full aperture- — The true geometrical / number is obtained
by dividing the equivalent focal length of the lens by the diameter of
1948
CALIBRATION OF / NUMBER
253
the effective aperture. It is therefore obvious that errors in the value
of the equivalent focal length and the effective aperture will be re-
flected by errors in the / number. Table IV lists the nominal and
o
1
LENS NOI
/
x7
LE
JS N0.2
5 M2
iTEO
i
/''
f/l
» f02J
MM
/
,/
f/2
CO
5MM
x
^
/
/
^
/
"*/
/
* /
/'''
/
/*
-060 T
RANSMI
TTANCE
jf
v>
- 0.83
TRANSN
ITTANC
•
/
' /
¥
2
0 Z
8 4
2
l.c,
LE
JS N0.3
9 (=1 1
XTEO
/'
«K£0 AND CAL
CO
^
/
5 "^
57
40
28
2.0
/
/I
/
T^-
087 1
RANSM
TTANC
E
//
/'"
LEf
S NQ6
7 f-:
V/
//
1/2
CM
/
//
4
^7
4
//
>
//x
X
X-
078 T
RANSM
TTANC
-
/
//
/
//
113 16.0 22.6
TRUE F- NUMBER
Fig. 4 — Marked and calibrated values of / number versus true geometric/
number. The circles indicate the marked/ numbers and the crosses indicate
the calibrated / numbers. The circles would fall upon the dotted diagonal
line if marked and true / numbers were equal. The crosses would fall upon
the dotted line if the transmittance were 100 per cent. The separation of the
dotted and solid-line curve gives a measure of the transmittance of the lens.
The steps in the net equal one stop opening for ready appraisal of differences
in fractions of a stop opening.
measured values of equivalent focal length and effective aperture.
In those instances, where the nominal focal length was given in inches,
conversion has -been made to millimeters. The nominal values of
effective aperture are computed from the values of nominal focal
254
WASHER
September
length and nominal / number. Examination of this table shows that
the measured value of the equivalent focal length is within ±2 per
cent of the nominal focal length for 15 of the 20 lenses. The average
departure for the entire 20 lenses is ±1.7 per cent. The errors in
226
\jo
IS NO 8
e f-7.:
X
//
LE
MS NOl
5 f = 4
»TED
/
/
1/2
CM
y
^
CO
/<''
X
X
2
/
//
fl
F-NUMBER
•J c
x
^
/^
/
/,*
-073 1
RANSM
TTANCt
*
^
085 T
RANSMI
TTANCE
£
| 2.8
5
3 20
>
4
JX
^
2.8 40 57 80 1.3 160 226 320 2D 2.8 4.0 57 80 1
3 16.0 22.6 32 (
NO CALIBRATE:
i X i
LE
4S NO 10
(•JIN
^
/
4 LENS NO 13
f/45 (=7
/
/
f/3
/"y"
5 IN.
X
//
/
^
/
113
/
/X/
j
<'X/
X
X/
/>
/
^
-069
RANSM
TTANCE
y
//
-0727
RANSMI
TTANCE
V
S/
/
/ /
F
.
//
*-8 40 • 5.7 8.0 11.3 160 22.6 32O 450
_8 40 5.7 80 11.3 16.0 226 32fl 45
TRUE F-NUMBEF
Fig. 5 — Marked and calibrated values of / number versus true geometric /
number. The circles indicate the marked/ numbers. The circles would fall
upon the dotted diagonal line if marked and true/ numbers were equal. The
crosses would fall upon the dotted line if the transmittance were 100 per cent.
The separation of the dotted and solid-line curve gives a measure of the
transmittance of the lens. The steps in the net equal one stop opening for
ready appraisal of differences in fractions of a stop opening.
effective aperture are as high as ±8 per cent with an average for 19;
lenses of ±4 per cent. Nine of the nineteen lenses show errors in
effective aperture in excess of ±3 per cent. It is doubtful if the errors j
in focal length can be brought below ±2 per cent during the process j
1948
CALIBRATION OF / NUMBER
255
of manufacture but it does seem that the error in aperture at the maxi-
mum aperture could also be reduced to d=2 per cent.
As a result of these departures of the measured values of the
equivalent focal length and effective aperture from their nominal
640
450
320
22.6
LEF
f/6
IS NO 12
8 f«7
/
//
N.
?
~z_
2
//
/
//
/
77
*
11.3
60
«J7
/
/
075 TRANSMITTANCE
./
//'
fl
LENS NO.I5
'f/7.5 f- 13.5 IN.
076 TRANSMITTANCE
6O 113 16.0 22:6 320 450 640 90.0 57 80 M3 160 226 32.0 450 64X) 900
UJ
? 90.0
<
|64.0
8
< 450
320
226
16.0
11.3
e.o
LENS NO. 17
y
y
LEr
IS NO. IS
/
/
f/ll f«l9»
1
X
77
f/12
IN
}
X
/
</
/
/
>
/x^
/
X
^
//
y
' /
^
X/o
71 TRA
NSMIT7
ANCE.
0
y
,X°
67 TRA
NSMIT1
ANCE
;
y
/
xx>
2
/
ii.3 16.0 226 320 45.0 64O 900 I2i
).0 I
3 16.0 22.6 32.0 45.0 64.0 900 12
TRUE f- NUMBER
Fig. 6. — Marked and calibrated values of/ number versus true geometric/
number. The circles indicate the marked/ numbers and the crosses indicate
the calibrated / numbers. The circles would fall upon the dotted diagonal
line if marked and true / numbers were equal. The crosses would fall upon
the dotted line if the transmittance were 100 per cent. The separation of
the dotted and solid-line curve gives a measure of the transmittance of the
lens. The steps in the net equal one stop opening for ready appraisal of
differences in fractions of a stop opening.
values, appreciable errors in the / number are produced. This is
| shown in Table V, which lists the nominal and measured / numbers
for the same group of lenses. The errors in the / numbers range from
—6. 8 to +11.1 per cent. The effect of these errors in terms of relative
256
WASHER
September
transmittance is shown in the last column. These values of relative
transmittance show that, neglecting losses in the lens, the difference
between nominal / number and true geometric / number may alone
produce deviations of as much as 19 per cent between the expected!
and actual values of the amount of light passed by the lens. It must
be emphasized that these differences are present at maximum stop
TABLE VI
NOMINAL AND ACTUAL VALUES OF THE TRANSMITTANCE AT FULL APERTURE FOR A
REPRESENTATIVE GROUP OF LENSES
Lens
Num-
ber
Equiva-
lent
Focal
Length,
Inches
/ Number
Transmittance
Marked
True
Cali-
brated
Nominal
Actual
1
0.5
1.9
1.77
2.40
0.63
0.54
2
0.5
2.5
2.62
2.82
0.79
0.86
3
1.0
1.9
1.87
2.09
0.83
0.80
4
1.4
2.7
2.67
3.14
0.74
0.72
5
1.6
2.7
2.82
3.14
0.74
0.81
6
2.0
2.7
2.62
3.09
0.76
0.72
7
2.0
2.2
2.07
2.23
0.97
0.86
8
3.0
2.8
2.75
3.20
0.77
0.74
9
3.0
2.3
2.30
2.45
0.88
0.88
10
3.0
3.0
3.04
3.68
0.67
0.68
11
4.0
2.5
2.51
2.79
0.80
0.81
12
7.0
6.8
6.91
8.00
0.72
0.75
13
7.5
4.5
4.74
5.60
0.65
0.72
14
11.0
8.0
7.97
10.1
0.63
0.62
15
13.5
7.5
8.33
9.72
0.59
0.73
16
16.5
9.5
10.12
12.30
0.60
0.68
17
19.0
11.0
11.13
13.60
0.65
0.67
18
24.0
11.0
11.78
14.30
0.59
0.68
19
30.0
12.5
12.79
15.60
0.64
0.67
opening where the effective aperture is that of a true circular opening
and not that of a many-sided opening which is operative when the
aperture is determined by the iris diaphragm. In 6 out of 19 cases,
the relative transmittance deviates from unity by 10 per cent or* more,
which may produce significant differences in exposure time in someji
instances of use.
(6) Errors in the marked/ numbers at reduced apertures — It is clean
that errors of the type described in the preceding section are also pres-|
ent for all of the marked / numbers. Because the aperture formed byj
1948 CALIBRATION OF / NUMBER 257
the usual many-leaved iris diaphragm is a polygon, the accuracy of
determining the diameter of the effective aperture is somewhat less
than that for the full aperture where the limiting opening is circular.
Where the number of leaves is greater than six, two diameters at
right angles to one another are measured and the average is considered
to be the diameter of a circular opening of the same area. For those
diaphragms having four to six leaves, the area is computed from two
or three diameters, and the diameter of the equivalent circle is used in
computing the / number. It is believed that the / number obtained
in this manner is correct within ±2 per cent for the small / numbers
and rising to ±5 per cent on the average for / numbers greater than 22.
The errors in the /-number markings for twelve lenses are shown
graphically in Figs. 4, 5, and 6, where the marked / numbers are plot-
ted as ordinates and the true (measured) / numbers are plotted as
abscissas. The dotted line with slope of unity passing through the
origin is the line upon which the marked / numbers would lie if there
were no error in the markings. The points are plotted on logarithmic
paper so that one may see at a glance what the magnitude of the
error is in terms of fractions of a stop opening. For example, in the
case of lens No. 3, Fig. 4, the true / number corresponding to the /
number marked 16 is 12.9. This error of marking is clearly shown on
the graph to exceed one-half stop. For lens No. 10, Fig. 5, at //16,
the true / number is 18.4, or more than one-half stop in the opposite
direction. For lens No. 12, Fig. 6, the values of marked and true /
number are very close together throughout the range of the markings.
V. MEASUREMENT OF TRANSMITTANCE
1. Transmittance at Full Aperture
It is possible on the basis of the information obtained in the course
of this experiment to determine the light transmittance of the lens
itself. It must be emphasized, however, that the transmittance so
determined is the ratio of the amount of light passing through the lens
to amount of light incident on the front surface of the lens, and does
not differentiate between image-forming and nonimage-forming light.
There are two ways of making this determination. The first method
yields the nominal transmittance, and is simply the square of the
ratio of the nominal / number and the ideal / number that gives the
same deflection on the light meter. Values obtained by this method
258 WASHER September
are listed in Table VI, under the heading of nominal transmit tance.
Since no cognizance is taken of the errors in the nominal /number, the
nominal transmittance is affected by the error in / number as well as
by reflection and absorption losses in the lens.
The second method yields the actual transmittance, and is the
square of the ratio of the measured and calibrated / numbers. Since
this method rules out the error in / number, the actual transmittance
is affected only by reflection and absorption losses in the lens.
It is interesting to consider lenses Nos. 16, 17, 18, and 19. These
are all of the same type, having 8 glass-air surfaces but ranging in
focal length from 16.5 to 30 inches. The nominal transmittance for
these four lenses varies from 0.59 to 0.65, while the actual transmit-
tance is almost invariant, changing from 0.67 to 0.68.
The effect of antireflecting coatings on the lens surfaces can be seen
in this table. Lenses Nos. 2, 3, 7, 9, and 11 are coated and all have
transmittances which exceed 80 per cent. Only one, No. 5, of the un-
coated lenses has a transmittance above 80 per cent and the remaining
13 lenses have transmittances ranging from 62 to 75 per cent with one
lens (No. 1) falling as low as 54 per cent. The antireflecting coatings
increase the transmittance by 25 per cent or more. Even so, con-
sideration of the actual values of the transmittance shows that 10 per
cent or more of the incident light is still lost by the coated lens. This
is not surprising when it is remembered that antireflecting films
usually yield close to 100 per cent transmittance for only one wave-
length of light. Accordingly, when a broad region of the spectrum is
covered, as is the case for white light, the transmittance measured is
the average for the whole region.
The fact that the values of transmittance obtained by this pro-
cedure are affected in some small amount by the presence of non-
image-forming or scattered light cannot be considered as important.
It is improbable that markedly different values would be obtained by
the use of collimated light incident on the front surface of the lens
during the experiment. In any comparison between the broad source
method of measuring transmittance or calibrating a lens and the col-
limated light method, it is unlikely that light scattered by the lens
will produce appreciable difference in the end result. The broad
source fills the lens with light giving rise to a greater amount of scat-
tered light. However, the diaphragm in the focal plane rigidly re-
'stricts the measured scattered light to that falling within a small area.
The collimator system, at least for the larger aperture, illuminates the
1948 CALIBRATION OF / NUMBER 259
inner surface of the barrel with light at small angles of incidence
favorable for reflection. All the light that is scattered and emerges
from the lens is evaluated by the detector. A priori it is difficult to
say which will give the most weight to scattered light. Certainly for a
well-constructed lens the differences in results obtained by the two
methods will be small. For a lens purposely made to reflect the light
from the mount, the result is open to question. However such lenses
do not constitute a threat, because they would not make satisfactory
photographs. The extended source does give a measure of the light
(some of which is scattered) which will be incident on a central area
of the film when a subject is photographed with a reasonably average
illumination over the entire field. The collimator method gives a
measure of the light available over a central area of the film, plus all
scattered light, when a relatively small bright source is photographed
on a dark ground.
2. Average Transmittance for all Apertures
The value of transmittance obtained in the preceding section is a re-
liable one for full aperture, but since a lens is frequently used at a re-
duced stop opening it is advantageous to consider a method of deter-
mining average transmittance throughout the entire range of stops.
This is done by plotting the calibrated / number against the true /
numbers as has been done for 12 lenses in Figs. 4, 5, and 6. The crosses
show the relation thus obtained. It is clear that these crosses lie
on a straight line, shown as a solid line, parallel to the dotted diagonal
line. If the crosses should fall on the dotted line it would indicate a
transmittance of 100 per cent. As it is, the displacement of the solid
line from the dotted line gives at once a measure of the average
transmittance for all apertures. This has been computed from the
•curves and the value of the average transmittance for all apertures is
shown for each of the 12 lenses in the proper figure.
It is worthy of mention that this method of plotting the results of
measurement serves the dual purpose of showing the consistency of
the method of calibration and reliability of the measured values of
the true / number. Errors in either operation would cause the crosses
to fall away from the solid-line curve. The fact that these deviations
-are small indicates that both calibrated and true / numbers have been
quite accurately assigned.
260 WASHER
REFERENCES
(1) I. C. Gardner, "Compensation of the aperture ratio markings of a photo-
graphic lens for absorption, reflection, and vignetting losses," /. Soc. Mot. Pict.
Eng., vol. 49, pp. 96-111; August, 1947; /. Res. Nat. Bur. Stand., vol. 38, p.
643; June, 1947, RP 1803.
(2) M. G. Townsley, "An instrument for photometric calibration of lens iris
scales," /. Soc. Mot. Pict. Eng., vol. 49, pp. 111-122; August, 1947.
(3) F. G. Back, "A simplified method for precision calibration of effective
/ stops," J. Soc. Mot. Pict. Eng., vol. 49, pp. 122-130; August, 1947.
(4) L. T. Sachtleben, "Method of Calibrating Lenses," United States Patent
No. 2,419,421, issued April 22, 1947, and assigned to Radio Corporation of
America.
(5) A. E. Murray, "The photometric calibration of lens apertures," /. Soc.
Mot. Pict. Eng., vol. 47, pp. 142-152; August, 1946.
(6) C. R. Daily, "A lens calibrating system," /. Soc. Mot. Pict. Eng., vol.
46, pp. 343-357;- May, 1946.
(7) E. Berlant, "A system of lens stop calibration by transmission," J. Soc.
Mot. Pict. Eng., vol. 46, pp. 17-26; January, 1946.
(8) D. B. McRae, "Measurement of transmission and contrast in optical in-
struments," /. Opt. Soc. Amer., vol. 33, p. 229; April, 1943.
(9) E. W. Silvertooth, "Stop calibration of photographic objectives," /.
Soc. Mot. Pict. Eng., vol. 39, pp. 119-123; August, 1942.
(10) D. B. Clarke and G. Laube, "Lens calibration," /. Soc. Mot. Pict. Eng.,
vol. 36, pp. 50-65; January, 1941.
(11) D. B. Clarke and G. Laube, "Method and Means for Rating the Light
Speed of Lenses," United States Patent No. 2,334,906, issued November 23, 1943,
and assigned to Twentieth Century-Fox Film Corporation.
Projection Equipment for
Screening Rooms*
BY H. J. BENHAM
BRENKERT LIGHT PROJECTION COMPANY, DETROIT, MICHIGAN
Summary — Motion picture screening rooms have many and varied uses
such as for motion picture studios, film laboratories, recording studios, film
exchanges, and many other applications. The material to be presented
here, however, will be concerned with screening rooms which are used in
motion picture studios, and those in film laboratories. In many respects,
considerably more is required of the projection equipment used in such
screening rooms than is required when used in other types of screening
rooms, or in regular theaters.
UNTIL RECENTLY, standard motion picture equipment has been
supplied for most types of screening rooms. A study of the con-
ditions encountered, together with the experience gleaned from the
many installations of projection equipment made in screening rooms
during the past few years, has taught us that some modifications of
standard equipment are desirable in order to obtain best results. It
is important that the projection equipment used in these types of
screening rooms perform so as not to cause any undesirable screen
effects, which could be mistaken for errors made by the cameraman,
or poor work on the part of the laboratory.
Before discussing projection equipment for screening rooms, we
should first consider the purpose of these screening rooms so that it
will be easier to understand the reasons for some of the requirements
demanded of the equipment, and why it is desirable to modify slightly
some of the components used.
The main functions of a screening room in a motion picture studio
or a film laboratory are to check the action, direction, artist make-up,
sequence of scenes for editing purposes, set lighting, photography,
sound, and the laboratory processing of the film. In order to deter-
mine how a motion picture will appear on the screen of the average
theater, it is important that some of the conditions in the screening
room approach closely those which are encountered in a regular
theater. These conditions are such things as the intensity of light on
the screen, ratio of viewing distance to picture size, and amount of
* Presented October 16, 1945, at the SMPE Convention in New York.
SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 261
262
BENHAM
September
ambient light. Otherwise, a picture that looks well in the screening
room may look bad in a theater, and the opposite also can be true.
Consideration must be given, moreover, to the difference in the mag-
nification of the picture to determine its quality when projected on
very large screens such as that used at Radio City Music Hall, and
those used at drive-in theaters.
.
us.
Checking Photography
In checking the cameraman's photography, a close examinati
must be made of the objects in the picture for steadiness and focus.
It is important, therefore, that
the projection equipment be free
from vibration, be adjusted to
prevent any lateral or vertical film
weave, and that the projection
lens be held firmly in focus.
Whenever an out-of-focus con-
dition is noticed on the screen,
the projectionist usually is re-
quested to check the adjustment
of the projection lens. In order
to check the focus of the lens
quickly, the adjusting knob
should be located where it is ac-
cessible readily from either side
of the projector mechanism.
(Fig. 1)
Checking Laboratory Work
The projector must be in correct adjustment when checking the
work done by the laboratory, because improper adjustment of the
film guides, gate, and intermittent movement may easily result in a
poor picture motion on the screen, which may be taken for improper
printer registration or motion. These adjustments are important
because a lateral movement of two thousandths of an inch of the film
in the film trap will result in a picture movement of approximately
three eights of an inch on a twelve-foot screen. Unless the film gate
and guides are known to be correctly adjusted, it will be difficult to
determine whether poor picture motion is due to the projector or to
laboratory work.
The main causes for flicker inherent in the film have been explained
Fig. 1— Brenkert BX-80 projector
mechanism showing accessibility of
projection-lens processing knob and
framing knob from both sides of pro-
jector.
1948 SCREENING-ROOM PROJECTION EQUIPMENT 263
by Grignon,1 as being due to one or more of three things : the original
photography, printing, or to background projection. To determine if
flicker is present in the picture it is important that the flicker from the
projection equipment be negligible. Power for the arc lamp should
have a very low ripple content, and the intensity of light on the screen
should be kept within the limits recommended, as flicker due to the
projector shutters becomes more objectionable as the light intensity
is increased.
Checking Special Effects and Relative Density
The intensity, and the quality of the light on the screen in the
screening room plays an important part when checking night scenes,
subdued lighting scenes, special effects, background photography, the
lighting of the set where the picture was photographed, the artists'
make-up, and when judging by visual observation the optimum rela-
tive density of each scene on the film. To check these effects correctly,
the lighting conditions on the screen should coincide with those on the
screen in the average theater; the intensity of light should be kept
within the limits2 of 9 to 14 foot-lamberts, and it should be of daylight
quality, such as is obtained from a high-intensity arc lamp.
CONDITIONS ENCOUNTERED IN SCREENING ROOMS
Green Film and Film Splices
Much of the film projected in screening rooms at film studios and
laboratories is "green." Oftentimes difficulty is experienced when
this film is being projected unless several precautions are taken to
prevent the deposit of emulsion in the film trap. When emulsion does
collect in the film trap, it usually results in difficulty in keeping the
picture in focus, lateral and vertical picture motion on the screen, andr
because of the increased friction of the film in the gate, torn
sprocket holes. In order to avoid these difficulties, projectionists
have used many expedients such as decreasing the tension on the film
pads in the film gate, and dropping oil on the film as it passes into the
film trap. In some cases, the emulsion deposit has been so thick and
caused so much trouble that it has been necessary to stop the pro-
jector before running all of the film and clean the emulsion from the
film trap.
In many cases the film which is projected in these screening rooms
consists of short sequences spliced together. No modifications need be
264 BENHAM September
made to a well-designed projector mechanism in order to run a film
with a large number of patches. It is important, however, that the
gate be adjusted properly in order to minimize the picture jump when
a patch goes through the film trap.
Size of Screening Rooms
Many of the screening rooms in use are very small in size, necessitat-
ing the use of a small screen and a short "throw." Although a def-
inite relationship should be maintained between screen width and
viewing distance, it has been found difficult to maintain such a re-
lationship in many screening rooms.
In the past, low-intensity arc lamps were used almost exclusively
for screening-room projection. Because of the small screen used in
most cases, sufficient light could be obtained from this type of arc-
lamp to meet the ASA recommendation of the American Standards
Association of 9 to 14 foot-lamberts. Today, however, a large per-
centage of pictures made are in color and the color quality of the light
is equal in importance to the quantity of light. As a result, most of
the screening rooms use high-intensity arc lamps and copper-coated
Suprex carbons which produce light that has a snow-white color
characteristic. The current range of Suprex carbons is 40 to 50 am-
peres for the 7-mm positive and 6-mm negative, and 60 to 70 amperes
for the 8-mm positive and 7-mm negative. The arc lamps must be
operated- within the above ranges in order to obtain good operating
stability. The quantity of light, however, even when the carbons are
operated at the low end of the range, is usually more than is required
for the small screens used at these screening rooms. Therefore, in
order that the intensity of the light on the screen will fall within the
recommended limits of 9 to 14 foot-lamberts, it is sometimes necessary
to reduce the amount of light transmitted to the screen.
EQUIPMENT FOR SCREENING ROOMS
Projectors
The double-shutter type of projector has been found preferable ii
most of these types of screening rooms. However, in some cases a
few minor modifications are desirable. The problem of emulsion
collecting on the film shoes has been alleviated in some cases by re-
placing in the film trap the steel film shoes with highly polished
chrome-plated shoes. Although the steel shoes ordinarily supplied
are highly polished and work satisfactorily with film which has been
1948
SCREENING-ROOM PROJECTION EQUIPMENT
265
properly waxed, in some cases emulsion from "green" film adheres
more readily to polished steel shoes than to chrome-plated shoes.
Fig. 2 shows the location of the film shoes on the film trap used on
the Brenkert BX-80 projector. Also shown are the adjustable Holly-
wood film guides. Accurate adjustment of these guides allows correct
passage of the film through the film trap with negligible lateral motion
of the film. These guides may be easily and accurately aligned by
means of a gauge which may be purchased from the manufacturer.
Ftfm
If
Fig. 2 — Film trap and gate used with Brenkert BX-80 projector. All
three sets of film- tension pads are adjusted easily and simultaneously by one
adjusting knob.
It can be noted in Fig. 2 that the film gate has been designed for
three sets of film-tension pads. An equal amount of tension is applied
to each of the two upper sets of pads, but the tension applied to the
lower set of pads is somewhat greater, caused by the use of a heavier
spring on the lower set of pads. The design and construction of this
film gate aids greatly in holding the film steady in the film trap during
the time the picture is being projected. This is especially true when a
patch is being fed through the trap. One adjusting screw controls the
pressure of all pads at the correct ratio.
It was stated earlier that the use of high-intensity lamps for some
266
BENHAM
September
of the small screening rooms results in excessive screen brightness.
One way to reduce light on the screen, and at the same time increase
the threshold of nicker is to use three-bladed shutters on the projector
mechanisms. Fig. 3, which is reproduced from a paper by Engstrom,3
shows the relationship between screen brightness and nicker rate
when the screen is viewed at a distance of four and one half times its
width. It also shows that nicker decreases with an increase in the
percentage of time the image is illuminated during one frame cycle,
and decreases with an increase in the light impulse-frequency. Inas-
much as a three-bladed shutter increases the light impulse-frequency
to 72 cycles, the threshold of flicker is increased considerably and the
amount of flicker seen on a screen with a brightness of from 9 to 14
foot-lamberts is negligible.
One Frame-
•V 'Cycle 360'
(fl) Double Three-Bladsd
Shutte
One Frame-
Cycle 360'
34-5 7 /<?
Screen I// um /nation in ft Candles
(B)Doubfe T^o-S laded
Shutter Sy s fern
Fig. 3 — Chart showing effect of light-impulse frequency on threshold of
flicker. Screen viewed from a distance equal to six times screen width; vision
concentrated at center of screen. Data taken from: Proc. I.R.E., April,
1935 (E. W. Engstrom); Soc. Mot. Pict. Eng., October, 1942 (E. W. Kellogg).
Arc Lamps
Where the screens are exceptionally small, the screen brightness
will be excessive even though three-bladed shutters are used. In such
cases, additional steps must be taken to reduce the transmission of
light to the screen.
One of the best methods of obtaining an additional reduction in
light intensity is to step down the speed of the optical system in the
arc lamp by reducing the effective area of the reflector through the
use of a dull-black metal shield. Fig. 4 shows such a shield installed
on a reflector from a modern arc lamp. The size of the shield depends
on the screen brightness required. This method of reducing the
amount of transmitted light has the advantages of protecting the
1948
SCREENING-ROOM PROJECTION EQUIPMENT
267
mirror, reducing the radiant energy on the film aperture, and improv-
ing the depth of focus.
Power Source
A motor generator is preferred as a power source because of its low-
voltage ripple content, and because it is not critical to sudden voltage
changes. The capacity and regulation of the generator should be such
that no change in the light intensity will be noticed on the screen when
the second arc lamp is struck. Full- wave three-phase rectifiers are
satisfactory, however, when used in conjunction with a projector em-
ploying either the two- or three-bladed shutters. Four-tube rectifiers,
designed for. operation from . a
three-phase power source, usually
employ a Scott-connected trans-
; former which actually results in
a full-wave, two-phase rectifier.
This type of rectifier has been
found to give satisfactory results
when two-bladed shutters are
used on the projectors, but an-
noying flicker may develop when
used in conjunction with a pro-
jector employing three-bladed
shutters. The reason for this is
that a two-phase, full-wave rec-
tifier has a voltage-ripple fre-
Fig. 4 — The effective area of the arc-
lamp reflector can be reduced by a
metal shield whenever a reduction of
I quency of 240 cycles, which is an
iexact multiple of 48 cycles but
„ ,_rt light on the screen is required.
! not of 72 cycles. The intensity
of the ripple voltage from a two-phase, full- wave rectifier is also
; greater than that of the three-phase, full-wave rectifier. When using
any kind of a rectifier it is important that the alternating voltage
across each phase be substantially the same.
Screen
In all cases a seamless white screen is recommended. Such a screen
is obtainable in the sizes most frequently used.
Although most screens have black borders, it has been found that
light-colored borders are much more pleasing and comfortable for the
eyes. It has been pointed out by Luckiesh and Moss,4 that the screen
268 BENHAM
border should not be extremely dark because of bad physiologica
effects such as eye fatigue. They have proved that certain eye muscles
suffer more fatigue under conditions of dark surroundings than when
some general lighting is available.
The surrounding border should not be brighter than a dark area or
the screen. This would tend to make the observer more aware of th(
screen border than of the screen. Various shades of gray border have
been found to be. desirable. It has been recommended that the bordei
should be at least one thousandth to one five hundredth as bright as
the screen high lights.5 The contrast of a black- velvet border agains
the screen is estimated to be from one to ten thousandths as bright as
the screen high lights.6
CONCLUSIONS
Care should be given to the selection of projection equipment fol
screening rooms in film studios and film laboratories.
Because of the many tests made in these screening rooms, the equipl
ment must be kept in good adjustment at all times.
It may be necessary to reduce the amount of light transmitted ti
the screen when small screens and high -intensity arc lamps are used.
Screen brightness should be checked carefully so as to determine thl
approximate results which may be expected in the average theater.
The motor generator is the preferred power source for both two- and
three-bladed shutters. However, the full-wave, three-phase rectifie
was satisfactory when used w^ith either two- or three-bladed shut t era;
REFERENCES
(1) Lorin D. Grignon, "Flicker in motion pictures," J. Soc. Mot. Pict. Eng\\
vol. 33, pp. 235-248; September, 1939.
(2) "Standards Committee Report," J. Soc. Mot. Pict. Eng., vol. 35, p. 523
November, 1940.
(3) E. W. Engstrom, "A study of television image characteristics," Part (
Proc. I.R.E., vol. 21, pp. 1631-1652, December, 1933; Part II, Proc. I.R.Eti
vol. 23, pp. 295-310; April, 1935.
(4) M. Luckiesh and F. K. Moss, "The motion picture screen as a lightml
problem," /. Soc. Mot. Pict. Eng., vol. 26, pp. 578-592; . May, 1936.
(5) S. K. Wolfe, "An analysis of theater and screen illumination data," Ji
Soc. Mot. Pict. Eng., vol. 26, pp. 532-548; May, 1936.
(6) L. A. Jones, "The interior illumination of the motion picture theater, j
Trans., Soc. Mot. Pict. Eng., No. 10, pp. 83-97; October, 1920.
The Gaumont-Kalee
Model 21 Projector*
BY L. AUDIGIER
BRITISH ACOUSTIC FILMS, LTD., LONDON, ENGLAND
AND
R. ROBERTSON
A. KERSHAW AND SONS, LTD., LEEDS, ENGLAND
Summary — The main features of the design of this 35-mm model are a
completely enclosed projector, for silence, safety, and cleanliness. The
mechanism operates in a totally enclosed oil bath, and the equipment has
built-in accessories such as automatic change-over and fire-quenching
levices.
INTRODUCTION
HE PURPOSE of this paper is to indicate the tendency in design of
35-mm sound-film projectors in Europe, and so that members of
the Society could study the projector in detail, and compare it with
current practice in the United States, the Gaumont-Kalee Company
of Toronto brought a model to the Convention Exhibition.
When, nearly twenty years ago, the talking film passed from the
experimental to the commercial stage, sound-film equipment was
aaturally designed for use with existing picture-projection equipment.
'Sound was only an addition to the basic thing, the picture. For a
i.ong time after the arrival of sound films there was a clear-cut line
dividing the sound equipment from the picture equipment. The
complete picture and sound equipment was a mating dictated by ex-
bediency of the products of a number of different manufacturers.
More than one designer, in Europe and America, made a logical bid
:o end piecemeal design by producing a combined picture and sound-
lead, but although technically such a concept was attractive, com-
nercially it did not secure acceptance. The user's preference was for a
|nore flexible design that permitted the retention of existing projector
nechanisms, or of existing soundheads.
The design of complete equipment which is to satisfy expressed
^references both in Europe and America must take into account
* Presented October 22, 1947, by A. G. D. West, at the SMPE Convention in
S'ew York.
SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 269
270 AUDIGIER AND ROBERTSON September
established differences between equipments originating in the two hemi-
spheres. Thus, in Europe, for the past fifteen years, projector mech-
anisms in the medium- and high-price groups have had oil sumps
and automatic-pump lubrication. On the other hand, enclosure of
the operating side has been the exception. There has, in fact, been a
preference for the open machine which leaves the film path exposed.
In the event of a film fire, the burning film can more easily be re-
moved. The addition of automatically operated fire extinguishers
of the carbon-dioxide type, which quench a fire and simultaneously
cut the motor and the arc lamp, has been common, and in some dis-
tricts compulsory.
In England particularly, owing to population density, really large
cinema theaters and large screens are relatively more frequent than in
most parts of the world, and as smoking is universally permitted,
illumination requirements are high.
In the realm of vacuum tubes, there has been a tendency for each
European country to develop types dissimilar to those of its neighbor,
and dissimilar to North American types. Only the octal-base type
has secured any measure of international acceptance.
The Gaumont-Kalee 21 equipment was designed for the world mar-|
ket, as a complete picture and sound reproducer. Its designers set
up the following table of requirements:
1. Picture and sound performance to satisfy recommendations of
internationally accepted authorities.
2. Over-all reliability to be greatest that straightforward design
and high-grade components could attain.
3. Accessibility, including replacement of worn or defective me-
chanical or electrical components, to be such that unskilled personnel
could undertake necessary maintenance work in remote locations.
4. Complete sound channels to be built up from a minimum num-
ber of basic panel units, and component layout to be such as to facili-
tate comprehension of circuit function.
5. Mechanical assembly of stand, projector, soundhead, and arc
lamp to be conceived as a whole, and to incorporate such ancillaries as
carbon-dioxide fire-quenching equipment, picture change-over con-f
trol, and arc switches and meters, but major units, projector, sound-j
head, and arc lamp, to be capable of use with other equipment.
6. Projection lenses, coated, to have //1. 9 aperture over complete;
range of focal lengths up to 7 inches, and design of arc lamp and pro-j
jector to permit full use of this aperture.
1948 GAUMONT-KALEE PROJECTOR 271
7. Projector mechanism to have oil-bath lubrication, high-effi-
ciency flicker shutter, and enclosure of operating side.
8. Projector drive from soundhead to conform with American
practice.
PROJECTOR STAND
As the base upon which the mechanical assemblies are erected, the
description of the equipment commences with the stand (Fig. 1).
Fig. 1 — Operating side of assembly, all covers closed
for operation.
This incorporates platforms for soundhead and arc lamp, and the
bottom spool box is an integral part. Switches for motor, exciter
lamp, and picture change-over are grouped on a panel. A second
panel carries arc-control switches and meters, but these may be
omitted if not required. A door at the rear (Fig. 2) corresponding to
the spool box door in front, gives access to the chain-driven take-up,
and to the motor and arc switches controlled by the fire-extinguisher
equipment. Provision is made on the front end of the stand for all
cable entries, the internal wiring being run in the factory and terminated
272
AUDIGIER AND ROBERTSON
September
at a distribution board at the cable-entry point. Wiring arrange-
ments are, however, sufficiently flexible to suit other installation
requirements which may arise in practice. The stand is adjustable
for height in 3-inch steps by insertion of distance pieces, and has a tilt
adjustment by a concealed jackscrew, accessible through the door on
the nonoperating side. The possible tilt varies from 10 degrees up-
wards and from 20 to 30 degrees downwards, depending upon the
Fig. 2 — Rear view of assembly of stand, pro-
jector, soundhead, and arc lamp.
height of the stand. The fixed foot of the stand is a heavy iron cast-
ing; the tilting parts, "including the spool box and doors, are all sub-
stantial aluminum castings.
THE SOUNDHEAD
The soundhead bolts directly upon the horizontal upper surface of
the stand, which also supports the driving motor, thus making a very
rigid construction (Fig. 3). The soundhead is also arranged for the
L948
GAUMONT-KALEE PROJECTOR
273
more usual type of mounting on the back of a pedestal stand, when the
motor is then supported by the soundhead.
The aim motivating the soundhead design has been to secure a
high-grade performance that will remain stable over long periods of
time, and long life because of robust construction of all wearing rjarts.
From the maintenance point of view the soundhead is one that can be
kept in service for twenty years without being sent back to the fac-
tory. Accurate jigging and dimensional uniformity of component
; part
Fig. 3 — Rear view of assembly of stand, projector, soundhead,
and arc lamp. Doors open and covers removed.
s ensures that replacements will fit without requiring any tools
other than a screwdriver and spanner.
The soundtrack is scanned on the periphery of a rotary drum, and
stabilization of film speed past the scanning point is maintained by a
fluid flywheel mounted on the drum shaft (Fig. 4). The flywheel it-
self is a light aluminum shell containing a heavy viscous fluid, a de-
sign which eliminates the necessity for internal bearings to locate an
inner member in respect to the outer shell.
274
AUDIGIER AND ROBERTSON
September
The optical system is of the back-scanning or visible-image type.
Immediately in front of the exciter lamp is a large condenser which
projects the light horizontally forward to a prism mounted partly
within the scanning drum. The prism reverses the light path and
directs it back through the sound track, through the objective lens, and
on to a window carrying a mechanical slit. The window is in a hous- 1
ing containing a prism, which directs the received light vertically
downward on to the cathode of the phototube. As the optical mag-
nification is six times, an enlarged image, six times that of the
Fig. 4 — Operating side of soundhead.
actual soundtrack, is impressed on the window. With the film!
stationary it is possible to check whether the focus is approximately j
correct, and with the film running it is evident if either sprocket holes j
or the edge of the picture is being projected on to the slit. The win-
dow has fixed masks to accept the internationally accepted scanned ]
width of soundtrack of 0.084 inch. The adjustable tracking of the|!
lay-on roller, centers the scanned soundtrack on the window. The|
slit is correctly adjusted for azimuth at the factory and locked, j1
Various types of slits can be used with the 83 soundhead, depending j
1948
GAUMONT-KALEE PROJECTOR
275
upon the purpose for which the head is used. For re-recording, a
very fine slit is used so that a straight-line frequency response may be
obtained from the phototube. For all normal reproduction purposes a
comparatively wide slit is used, because the over-all frequency-response
curve recommended by the Academy of Motion Picture Arts and
Sciences entails curtailment above 2000 cycles. The standard repro-
ducing slit is 0.0108 inch wide, and taking into account the six-times
magnification of the optical system, corresponds to a slit dimension at
film of 0.0018 inch. This dimension results in an increased amount of
light being passed to the phototube, with a gain in output and an in-
._,
Fig. 5 — Rear view of soundhead; flywheel, and driving pulley removed.
creased signal voltage on the grid of the first tube. Its effect on the
frequency-response curve is progressively to attenuate the response
above 2000 cycles, giving a loss of —12 decibels at 8000 cycles.
The efficiency of the optical system is high by reason of the large
effective aperture of all optical components. All lenses and prisms
are hard-coated.
All the components of the scanning system, exciter lamp, optical
system, scanning drum, and phototube, are carried on a plate which is
resiliency mounted within the soundhead body proper.
There are three rotating shafts in the soundhead, the one carrying
the fluid flywheel and scanning drum, and two which carry a film
sprocket at one end and a gear wheel on the other (Fig. 5). These
276
AUDIGIER AND ROBERTSON
Septembe
three shafts are not carried in bearings located in the soundhead cast-
ing, but each shaft, with its bearings, is contained in a long, flanged
housing of circular cross section which in turn fits a machined bore in
the soundhead casting. The flywheel shaft runs on precision ball
bearings as it must impose a minimum load on the film. The two
sprocket shafts run on oilite bearings as they are driven by the motor.
When, after long service, it is necessary to renew a shaft and its bear-
ings, the complete housing can be withdrawn by taking out three
screws, and a new shaft and bearings, complete in a housing, replaces
the worn components. The two assemblies of sprocket shaft, bear-
ings, and housing are identical and interchangeable.
Fig. 6 — Soundhead dismantled.
The whole gear train of the soundhead is carried on the two sprocket
shafts plus one stationary layshaft. This layshaft is hardened and
ground, and held in a machined bore in the soundhead casting. All
the gearing is accessible when the soundhead cover is removed, and
the complete train can be taken down in a few minutes (Fig. 6).
Gears which rotate with their shafts are held thereto by key washers
and end screws.
Every component part of the soundhead, electrical, optical, and
mechanical, down to such items as small thrust washers, carries its
part number engraved on it.
The complete soundhead is rustproof. The soundhead body, the
scanning plate, and the doors are light alloy castings. The bearing
housings, the mounts for exciter lamp, condenser and prisms, the
1948- GAUMONT-KALEE PROJECTOR 277
phototube housing, the slit unit plate, the brackets for lay-on and pad
rollers, and the strippers are light alloy die castings. Small rollers
and retaining screws are either stainless steel or chrome-plated.
The resiliency mounted motor is carried in front of the soundhead
with its shaft horizontal, and parallel with the sprocket shafts of the
soundhead. The drive from motor to soundhead is by twin short can-
vas and rubber vee belts. The motor and the belt drive are protected
by a quickly detachable louvered cover, through which an inching
handle projects on the operating side. Motors are available for 25-,
30-, 40-, 50-, and 60-cycle supplies.
For studio requirements, a three-phase synchronous or an interlock
motor is used, and as truly synchronous speed must be maintained on
the film sprockets, gear drive takes the place of belt drive.
THE PROJECTOR
The projector is not bolted directly to the top of the soundhead, but
is mounted on a detachable base which in turn attaches to the sound-
head; this gives flexibility to suit other soundheads and avoids the
inconvenience which sometimes would arise, were it always necessary
to attach the projector by bolts from beneath the soundhead into
tapped holes in the projector base.
American practice employs a small 17-tooth pinion which meshes
with the projector gear train. This is an inconveniently small size,
which, used with an oil-bath mechanism, involves an external gear
train in order to keep the projector drive shaft at sufficient height
above the oil level to avoid danger of oil leakage, or the employment
of a stuffing box or equivalent expedient. -This difficulty has been
avoided by substituting for the 17-tooth pinion one of 34 teeth, run-
ning at the same speed and at the same relative center distance,
thus maintaining interchangeability of soundhead drives. This 34-
toothed pinion meshes directly with a drive gear mounted on the bot-
tom sprocket shaft which is carried through the frame on both sides.
This is the lowest bearing of the machine and is thus the limit to the
amount of oil which the mechanism can contain without overflow.
The projector body is a substantial box casting, the bottom of which
is the oil sump. The mechanism gear train runs in an oil bath, with
oil circulated by a gear pump and distributed after passing through a
filter, readily detachable for cleaning. The rear cover of the machine
has a large clear window for viewing the mechanism and the working
of the oil distribution; a "sight" window is provided at the operating
278 AUDIGIER AND ROBERTSON September
side to show the correct oil level. This is marked with a series of lines
to indicate the correct level corresponding to different angles of tilt.
Optional positions are provided for an oil-drain plug in the front end,
and on the nonoperating side of the mechanism, to suit different
soundheads. The floor of the box casting is sloped internally so that
the oil can be drained from the front, even in the case of a positive rake.
The projector gear train comprises throughout, cast-iron pinions
and fiber gears in pairs. All have helical teeth for quiet running and
their ratios have been worked out to secure a "hunting-tooth" con-
dition in each pair, conducive to quiet running. The drive to the
shutter shaft, which is at right angles to the main train, is by 45-degree
spiral gears. Racking or framing is effected by rotation of the inter-
mittent unit about the sprocket axis, timing compensation being
obtained by sliding in synchronization the spiral-driven gear on the
shutter shaft.
The intermittent unit has a large-size cross and cam of 2-inch
nominal diameter. All working parts are of heat-treated-steel pre-
cision-ground. The roller is rigidly supported on a fixed pin carried
between cheeks on both sides. The flywheel is mounted directly
upon the cam shaft, there being no gearing inside the unit. The
mechanism operates inside an oil box which is constantly flooded in
all working positions. Adequate oil-return arrangements prevent
leakage of oil. The unit is rigidly supported in the projector in a long
fixed quill in which it rotates for masking adjustment. The inter-
mittent sprocket, as all the projector sprockets, is hardened and
ground.
The top and bottom sprocket assemblies are constructed as units
which can be detached without dismantling. The intermittent unit
and also the pump are similarly removable as units. .The shutter
shaft which is supported in bearings in the frame can also be with-
drawn without dismantling. The rest of the gear train, including the
housing which receives the intermittent unit, is removable in the form
of two complete subassemblies.
Much attention has-been given to the achievement of maximum
shutter efficiency. The light must be cut off from the screen while
the film is moving from frame to frame, and again for an equal balanc-
ing period in order to obtain a sufficiently high frequency of obtura-
tion to avoid an objectionable flicker effect; hence the maximum effi-
ciency consistent with avoidance of "travel ghost" and marked flicker
is about 50 per cent.
1948 GAUMONT-KALEE PROJECTOR 279
In practice, efficiency can be increased to some extent by encroach-
ing into the period of film movement, which also enables a correspond-
ing reduction in the balancing cover period. This is possible by
talking advantage of the fact that there is a small but appreciable
period at both the beginning and end of the film movement when its
displacement is relatively small. The amount of encroachment toler-
able can only be determined by trial, since it is to some extent depend-
ent upon the intensity of illumination and also upon the rate of in-
crease in illumination which depends upon the characteristics of the
shutter.
Offsetting this the shutter does not cut a single ray of light, but a
beam of sensible diameter, hence its operation cannot be instantaneous
and the corresponding intervening twilight periods between full
illumination and full cutoff involve some loss of potentially useful
light, and the only effective way of increasing shutter efficiency lies in
shortening this period.
One method of attack, adopted in several projectors of recent design
has been the employment of twin shutter blades rotating in opposite
directions, thus making a scissors-type shutter which, by cutting the
beam simultaneously from opposite sides, cuts it in one half the time
required by the conventional single shutter. Experiments made with
this type of shutter showed a gain in average illumination of about 15
per cent. The current 21 projector employs a single-bladed shutter
running at twice the normal speed. This achieves the same efficiency
as the double shutter since it, too, cuts the light beam in one half the
time of that taken by a normal single shutter.
The principle has been employed in 16-mm projectors and was in
fact used in the first Kalee projector made 37 years ago. As applied
to the 21 projector it affords a very straightforward construction with
avoidance of external gears and oiling points. The high shutter
speed, 2880 revolutions per minute, demands adequate lubrication to
ensure quiet, troublefree running. A pipe furnishes a constant
supply of oil from the pump, and the shutter shaft has a spiral oil
groove which pumps oil continuously through the bearing, after which
it is again returned to the sump by a return passage, a "flinger"
assuring that no oil escapes into the shutter casing. Since the shutter
runs at twice normal speed it requires a single blade of approximately
180 degrees cover instead of the usual pair of 90-degree blades. The
blade is counterbalanced by steel plates riveted to the blade, which is
of light-gauge aluminum. This makes a very stiff construction in
280
AlJDlGIER AND ROBERTSON
September
perfect balance. The shutter casing houses the transformer for the
threading lamps and the magnet of the change-over unit, with asso-
ciated wiring and fuses.
An advantage of the open-sided mechanism is the freedom from
restrictions imposed upon the lens holder by enclosure, which makes it
easy to use large-diameter, big-aperture lenses. It has, however,
Fig. 7 — Operating side of projector.
been possible to retain much of this advantage by arranging the lensj
holder outside the enclosure (Fig. 7). The bore of the holder isjj
standard 2.781-inch diameter and it is furnished with a removable!
liner to take 2.062-inch diameter. The gripping length, while ade-j
quate, is kept short and close to the film plane, and permits the use of i
large-diameter lenses with still larger stepped-front elements. It isj
194S GAUMONT-KALEE PROJECTOR 281
thus possible to use //I. 9 lenses throughout the whole range of focal
lengths up to 7 inches.
The gate opens with a parallel action, is self-sustaining when open,
and operated by a conveniently located handle. The front part of the
gate assembly, which carries the spring-loaded film-guide rollers and
pressure pads, is carried in a box-shaped casting which also receives
the rear end of the projection lens. The gate simply hooks into a
location in the face of this box. This construction gives utmost
rigidity combined with accessibility as the gate assembly is instantly
removable for cleaning.
The gate has twin apertures; the lower one is the projection aper-
ture, and the upper one is for verification that the film is in frame.
When the gate is closed the stray end of a broken film cannot intrude
into the light path.
The whole gate assembly is detachable as a unit. A polished reflec-
tor is provided just behind the aperture to reject the heat of overspill
illumination and this, together with massive construction and ample
radiating surface, assures cool operation. The mask plate is of
hardened steel and retained in slots in the gate bracket from which it
is quickly detachable.
The framing aperture and the working side of the projector are
illuminated by a pair of small low-voltage lamps. The lamps are fed
from a transformer tapped to suit both 100/115- and 200/240-volt
supplies.
The safety shutter is housed in the rear of the gate unit, and is
actuated by a centrifugal governor on the shutter shaft of the projector.
PICTURE CHANGE-OVER
The electrically operated picture-change-over device operates on
the safety shutter, but in such a way that no derangement of the
change-over system can prevent the shutter falling when the force
exerted by the centrifugal governor fails when the machine slows down
or stops.
The safety shutter is raised by a floating lever acted upon inde-
pendently by both the governor and a change-over magnet. Neither
act ing alone can open the shutter, which can only open a.nd remain
open so long as both exert a pull. The change-over operating mech-
anism proper is very simple, consisting merely of a tractive magnet
arranged to pull down an armature connected to the floating lever
operating the shutter. Magnet core and armature are laminated and
282 AUDIGIER AND ROBERTSON September
fitted with slug rings, and wound for operating on alternating or
direct current supply at mains' voltage. The magnet is in circuit
the whole time that the picture is on the screen; change-over is
effected by a throw-over switch which breaks the magnet circuit of
the outgoing machine, closing the shutter, and simultaneously ener-
gizes the magnet of the incoming machine, the shutter of which opens
because its actuating lever is being pulled both by centrifugal and
magnetic force.
A two-station switch circuit is employed which allows operation
from either machine. This can be extended for three-machine opera-
tion. The picture change-over readily could be coupled to the sound
change-over, but there is divergence of opinion among operators as to
the merits of such a provision.
FIRE CONTROL
The Pyrene fire-extinguisher equipment comprises a sealed cylinder
of compressed carbon-dioxide gas and a spring-loaded piercer, which
punctures the seal and releases the gas. This piercer is held back by a
celluloid loop. A quick-burning gun-cotton fuse instantly transmits
a fire at any of several points along the film path to the loop which ig-
nites and releases the piercer. Pipes conduct the gas to various
points along the film path, effectively quenching any fire. The gas is
also led into both top and bottom spool boxes and to pistons which
knock off switches, cutting the power supply to both motor and lamp,
thus shutting' down the equipment. In practice, in the event of a
fire, not more than two frames are lost.
THE LAMP
The arc lamp employs a 16-inch diameter elliptical mirror with
focuses at 6 and 36 inches. This mirror works at a larger collective
angle than the more usual 14-inch mirror, and hence transmits more
light. Experience has also shown that important practical advan-
tages of its larger dimensions are that arc focus is less critical and the
greater crater distance results in substantial freedom from pitting and
reduced risk of mirror breakage. It has been found quite safe to
operate this lamp at rakes of as much as 30 degrees.
It has been possible to maintain the generally accepted optical-
center height with this large-diameter mirror by keeping the positive-
carbon drive to the rear of the lamphouse. This has resulted in a
clear unobstructed floor in front of the mirror. The lamp mechanism
is of straightforward orthodox type. The positive carbon is driven
1948 GAUMONT-KALEE PROJECTOR 283
directly by a variable-speed motor connected across the arc gap.
The negative carbon is driven from the same motor through a variable-
ratio drive comprising a cam-operated variable-stroke roller clutch.
The complete drive unit can be withdrawn through the rear of the
lamp. The whole of the mechanism and the mirror holder is mounted
on a stiff cast tray which forms the base of the lamp. The lamp-
house itself is constructed throughout of sheet steel fabricated and
welded into a stiff one-piece shell with flush fitting doors, similarly
fabricated.
Knobs on the operating side of the lamp, below the door line, give
independent manual control of positive and negative carbons. These
have quick releases for resetting and can be clutched together by
pressing in a push button on the rear control panel to focus the
crater, keeping the gap constant. A periscope system contained in-
side the lamphouse forms an image of the crater on a screen in the top
of the lamp. A push-button strike is provided.
A wedge-operated quick-release positive-carbon grip safeguards
against excessive clamping force and instantly dismantles for clean-
ing, and a tachometer shows the actual speed of the feed motor.
SOUND AMPLIFICATION
The complete amplifier system for a single or dual channel is built
from basic units, all of which are mounted on cadmium-plated, rust-
proof panels of uniform width (Fig. 8). Panels are mounted in the
vertical plane. All components, including tubes, are on the front
face, with terminals projecting through to the back, on which side, in
one plane, is all the wiring (Fig. 9). Every component is rated for
continual tropical use, and in the design, care has been taken to oper-
ate tubes and rectifiers at less than the rating permitted by the manu-
facturers for continuous service. Tubes used are exclusively of the
internationally accepted and available octal-base type.
The basic working panels comprise a three-stage preamplifier, a
volume control, a 30-watt power amplifier, a power-supply unit, a
meter panel, a dividing network, and exciter-lamp supply units.
From these units a complete single-channel system can be evolved.
A dual-channel equipment requires the addition of a switch-control
panel, and for technical reasons, a separate unit to provide heater
current for the tubes in the preamplifier. An optional unit, which
can be added to either a singie-or dual-channel equipment, is a panel-
mounted monitor and deaf-aid amplifier.
284
AUDIGIER AND ROBERTSON
September
The standard equipment is suitable for operation from 50- or 60-
c,ycle mains of any voltage from 90 to 130 volts and from 190 to 260
volts. Alternative power-supply and exciter-supply panels are avail-
able which permit of direct operation from 25 cycles. These 25-
Fig. 8 — Front view of cabinet rack.
cycle units will operate equally well from 30-, 40-, 50-, or 60-cycle sup-
plies. Such small items as the monitor and deaf-aid panel, and the
unit which provides heater current for the tubes in the preamplifiers,
are made in one model only suitable for connection to mains of any
periodicity between 25 and 60 cycles.
A complete single-channel amplifier is housed in two main units, a
1948
GAUMONT-KALEE PROJECTOR
285
wall-mounting case for the preamplifier, and a cabinet-type rack
for the power-amplifier — power-supply unit, dividing-network, and
exciter-supply units. Two exciter-supply units are provided so that
any soundhead test or adjust-
ment requiring an input to the
exciter lamp of the idle machine
can be made during program
hours.
Two small monitor speakers,
for suspension directly over the
two operating positions, are sup-
plied. The level from the moni-
tors is adjusted at the time of in-
stallation to suit the conditions
prevailing and the operator's
preference, and remains there-
after in direct relationship to
the volume of sound emitted
from the stage speakers. A
monitor on-ofT switch is provided,
and is fixed in a position adjacent
to the telephone in the operat-
ing enclosure.
The equipment is completed
by a switch-fuse distribution
unit which is fixed to the Avail,
alongside the main switch fuse
terminating the alternating cur-
rent run to the operating enclo-
sure. This distribution unit car-
ries a voltage-adjusting trans-
former so that, irrespective of the
incoming mains' voltage, the cor-
rect voltage may be fed to the untapped primary windings of the
various mains' transformers on the working panels.
A dual-channel equipment duplicates both the preamplifier and the
power amplifier. The two preamplifiers are housed in similar wall-
mounting cases and two cabinet-type racks house the remaining
panels. Normally each preamplifier serves one soundhead, but in
emergency either amplifier can be switched to serve two soundheads.
9 — Cabinet rack; rear cover re-
moved.
286 AUDIGIER AND ROBERTSON September
A control panel, housed in one of the racks, selects which of the two
30- watt channels is to be used, or in a third position of the switches,
links the two power amplifiers to give a total power output of 60 watts.
Correct matching of the power amplifiers, both in respect of input
impedance and output load is preserved irrespective of whether either
channel alone or the two linked are in use. The control panel also
carries a meter calibrated in decibels, and two gain-control switches,
which permit the two channels to be balanced accurately for sensi-
tivity. The meter also permits over-all frequency-response curves of
the two channels to be measured.
The preamplifier has three stages, resistance-capacitance-coupled
with negative feedback over the last two stages. Adjustment of fre-
quency response is by a unit, mounted on the amplifier panel, which
gives independent control of bass and treble response, permitting an
over-all curve to Academy or any other recommendation to be
obtained.
The main volume control is a 22-stud, click-action network with a
stationary scale and a rotating pointer. Sound change-over is by in-
stantaneous switch. Both volume and change-over can be remotely
controlled from positions adjacent to the two projectors.
The power amplifier has three stages, with negative feedback over-
all. The power stage comprises four 6L6G tubes in parallel push-pull,
and at the rated output of 30 watts there is less than lJ/2 per cent of
total harmonic distortion at the frequency where distortion is at a
maximum.
The power-supply unit is on a separate panel, both to reduce the
weight and complexity of the power amplifier, and to permit of its
ready interchange with the 25-cycle model. It utilizes two hard therm-
ionic rectifiers of the 5U4G type.
EXCITER SUPPLY
The exciter-supply units employ selenium-type metal rectifiers, and
the two-section, choke-input filter gives a smoothed output compar-
able to that obtained from accumulator batteries. The direct-cur-
rent output of each unit is controlled by the exciter-lamp switch on
the stand carrying the projector and soundhead. The alternating-
current input may be left connected to an unloaded unit for an in-
definite period without harm to the rectifier or any other component.
Throughout the design of the amplifier channel and associated
1948
GAUMONT-KALEE PROJECTOR
287
equipment, great attention has been paid to accessibility. The pre-
amplifiers are carried on hinged frames that permit immediate access
to the back of the panels. The cabinet-type racks have full-length
doors in front to give access to tubes and components, and quickly
detachable backs to give access to the wiring. Any component on any
Fig. 10 — Complete duosonic speaker assembly, small
size.
panel can be removed and replaced without disturbing any other
component. All high-tension smoothing capacitors are of the steel-
cased paper-dielectric type.
Components which are not immediately identifiable as to type and
value by manufacturers' markings are part numbered. All panels
carry on their front faces a metal label giving the type number of the
panel.
288
AUDIGIER AND ROBERTSON
LOUDSPEAKERS
September
The loudspeaker assemblies (Fig. 10) which are of the usual two-
way type, are supplied in four different sizes. High-frequency and
low-frequency units have permanent magnets of Alnico 5.
The dividing network has a crossover frequency of 500 cycles, with
TABLE I
PERFORMANCE OF GATJMONT-KALEE 21 PROJECTOR IN VARIOUS THEATERS IN
GREAT BRITAIN
O2
3
^
§
C3 c3
be r "}
0)
3
1
s
i— i O>
3 g
2 &
S -^
§ 1
3 "5
Theater
II
1
II
3 a
> 0
< fe
M ^
h-l CO
11
Astoria,
London, W.
40
62
2400
7,098
2.95
26
3.25
//2.2
18
Gaumont,
Princes Park,
Liverpool
38
60
2280
6,615
2.90
25
4.25
//1.9
19
Trocadero,
Elephant,
•
and Castle,
London
40
58
2320
10,887
4.69
25
5.00
//1.9
24
Gaumont,
Plymouth
39
59
2301
7,706
3.34
16
4.25
//1.9
26
Dominion,
London, W.
40
58
2320
6,425
2.64
18
5.75
//1.9
23
Classic, Belfast
38
58
2204
6,542
3.42
17
5.00
//1.9
23
Gaumont,
Birmingham
38
60
2280
8,658
3.79
16
4.25
//1.9
27
Gaumont,
Coventry
41
58
2378
7,581
3.14
16
4.50
//1.9
24
Picture House,
Glasgow
42
61
2562
8,114
3.20
20
5.00
jf/1.9
23
a 12-decibel-per-octave loss on each side of that frequency. An at-
tenuator with five 2-decibel steps is provided to equalize the acous-
tic output of high- and low-frequency horns.
The high-frequency horns are of all-metal construction, with from
eight to eighteen cells, according to the horizontal and vertical angles |
to be covered.
The low-frequency horn is of the direct-radiation pattern with no
HMS GAUMONT-KALEE PROJECTOR 289
hack emanation. In the larger sizes of horn, where the number of
low-frequency units employed would result in a departure from opti-
mum-load conditions, an impedance-matching transformer is used.
The design of the low-frequency horn is such that, in addition to the
normal back access to the low-frequency units, side access is provided
as well. This permits the loudspeaker assembly, where backstage
room is limited, to be positioned close against a rear wall without im-
pairing accessibility of the units.
THEATER PERFORMANCE
Some measurements of light efficiency, taken at various theaters in
London and the provinces, are shown in Table I. These indicate an
average performance as follows:
Arc watts 2400
Total lumens of light output ' 8000
Lumens per watt 3.3
Average illumination on a 23-foot-wide screen (projector running
with no film in the gate), foot-candles 20
Average brightness on a 23-foot-wide screen, foot-lamberts 14
Screen efficiency, per cent 70
NOTE BY G. T. LORANCE
When I was informed that the Chairman of the Board of Editors
had suggested that "perhaps it would be interesting to publish with
the paper a short note by Mr. Lorance discussing points wherein
British practice differs from that in this country," I felt quite compli-
mented and indicated my willingness to write such a note. It was not
until I received a copy of the paper by Messrs. Audigier and Robert-
son and began to study it that I got into trouble because it was clear
that the authors have already well covered the points which I was
supposed to emphasize.
At the risk of appearing to have developed a philosophical outlook,
let me first say that this is one of those cases, when, in comparing the
Gaumont-Kalee machine with American machines, an introvert would
find many, many similarities and an extrovert would find many,
many differences. To the introvert this is still a projector, a sound-
head, a motor, magazines, and an arc lamp mounted on a pedestal
much as they have always been designed and mounted. To the ex-
trovert this new assembly of equipment with the design of each piece
correlated to the design of its associated pieces and with certain new
details and with emphasis on them, this is a new projector and sound
290 AUDIGIER AND ROBERTSON September
system. Neither viewpoint, of course, is the correct one and in what
follows I shall endeavor to find a reasonable middle ground.
Audigier and Robertson refer to attempts to produce a combined
sound and picture head, indicating that such a concept was technically
attractive but commercially unacceptable and that the user preferred
a more flexible design that allowed combinations of equipment from
various sources and at various times. I cannot help but agree with
the authors regarding the attractiveness, from a technical standpoint,
of a combined design. I think they must have had it very much in
mind because the design which they describe in this paper represents
to me a very determined and worth-while endeavor to have your cake
and eat it too, in that they have striven for a unified and co-ordinated
over-all assembly of major components but have retained much of the
familiar breakdown into major components. Incidentally, it seems
reasonable to state at this time that, in portable and semiportable
equipment, unified designs of a combination of projector, soundhead,
and arc have been commercially successful for a number of years and
that the performance of some of these unified equipments compares
quite favorably with the performance of equipment of the regular
type as designed for permanent installation in projection rooms.
By and large this new Gaumont-Kalee equipment appears to follow
pretty generally accepted American practice although it has been, of
course, modernized in detail. A few points. are enumerated and dis-
cussed below. An attempt has been made to restrict them to points
of difference or to points of interest about which more knowledge
would be of value.
SOUNDHEAD
Rear scanning, similar in principle to that employed in the Gau-
mont-Kalee soundhead, has been used in American equipment. It is
possible, of course, to do good work with either front or rear scanning
and it is necessary to know more than is disclosed in this paper before
we would know if this system represents an improvement over Ameri-
can systems.
The use of a slit image 0.0018 inch high at the film represents a de-
parture from American practice that may be open to question. The
desire to utilize the dropping high-frequency response of such a slit
image as part of the desired over-all reproducing-system response char-
acteristic is understandable. It may, however, result in other unde-
sirable effects, such as irregular, rather than steady reproduction of
1948 GAUMONT-KALEE PROJECTOR 291
high frequencies. In my opinion, a slit image height of 0.0012 inch, as
used in much American equipment, is a little larger than is desirable if
quality of reproduction is to be stressed.
To the best of my knowledge, commercially available American
equipment has not used a fluid flywheel as described in this paper.
While realizing that the fluid flywheel eliminates the necessity for in-
ternal bearings, the quality of which is extremely critical, American
designers apparently have not yet given favorable consideration to the
fluid flywheel. This probably is because of a belief that to maintain
or improve -on performance, a much larger fluid flywheel would be
needed. Quantitative information on the performance of this
soundhead with this size of fluid flywheel would be of interest.
THE PROJECTOR
The description of the use of a 34-tooth pinion for the 17-tooth
pinion is not entirely clear to me . If, however, it provides for a higher
oil level in the projector without decreasing the speed of the pro-
jection drive shaft it may well represent a new and improved solution
to this general problem.
There is much more to shutter efficiency than the authors of this
paper have taken time and space to present. The use of a single-
bladed shutter, such as described, rotating at 2880 revolutions per
minute is one method of attack which has, to the best of my knowl-
edge, not yet appeared commercially in American machines. Fac-
tors affecting shutter efficiency involve such things as the number of
blades on the shutter, whether it is double or single, speed of rotation,
diameter of the shutter, and size of the light beam where the shutter
cuts it. Circumstances do not permit enlargement upon this thought
in this note.
While the lens holder is, dimensionally, in line with current Ameri-
can practice, there are indications that American designs may adopt a
larger lens diameter. At least one projector has been shown with
facilities for clamping a lens having a diameter of approximately 4
inches.
PICTURE CHANGE-OVER
From a technical standpoint it seems quite feasible to use the safety
shutter as the change-over as well and such a design probably does
simplify the complete system.
292 AUDIGIER AND ROBERTSON September
•FIRE CONTROL
To me, the use of built-in fire extinguisher equipment is novel in so
far as American equipment is concerned. The subject in general is of
such a controversial nature and exact and reproducible data are so
difficult to obtain that I do not feel competent to comment further on
this point.
THE LAMP
More information of a definite nature would be appreciated regard-
ing the lamp. We are not informed regarding the optical speeds in-
volved nor are we informed regarding the size of the carbons, conse-
quently we are unable to comment accurately on the performance of
the lamp.
SOUND AMPLIFICATION
While the available output power is stated, there does not appear to
be any statement regarding the noise level.
LOUDSPEAKERS
An unusual point is noted in regard to loudspeakers in that the de-
sign of the low-frequency horn is such that access to the low-fre-
quency units is provided through the side as well as through the back
of the horn. This is a detail which should be noted by American de-
signers.
THEATER PERFORMANCE
Figures are quoted regarding the amount of light put on the screen.
Information regarding the distribution of light on the screen would
also have been of interest. Reference should be made to the recently
published results of a survey by the Screen Brightness Committee,*
which gives in considerable detail the performance of a variety of
equipments in some eighteen American theaters.
In conclusion, may I state the hope that the above remarks will be
taken as being well intentioned. Messrs. Audigier and Robertson
have developed and designed some very interesting equipment which,
in my opinion, is well worthy of study by American engineers.
* /. Soc. Mot. Pict. Eng., vol. 50, pp. 260-274; March, 1948.
REPLIES BY AUTHORS TO QUERIES RAISED
ARC LAMP
The optical speed is //1. 9, to match the aperture of the projection
lens. Carbon sizes used for results given in Table I were 8-mm
positive and 7-mm negative, of "POOL" high-intensity type. (In
1948 GAUMONT-KALEE PROJECTOR 293
Great Britain during the war the number of types of carbons was
greatly reduced, and the product of all manufacturers pooled.)
INTERMITTENT SPROCKET
The diameter is 0.945 inch.
SOUND AMPLIFICATION
This paper has failed to give a clear impression of the physical
layout of the complete amplifier channel. The three-stage pre-
amplifier is a separate unit, mounted on the wall centrally between
the two machines, so that the circuit from the phototubes to the input
of the preamplifier is not unduly lengthy.
Low-capacity concentric cable is employed, and the ratio of photo-
tube-lead capacitance to total amplifier input capacitance is such that
a difference in length of several feet between the right-hand and left-
hand cell lead will result in a negligible disparity in the over-all re-
sponse curves of the two machines. The change in frequency charac-
teristic is less than 1 decibel at 8000 cycles for a 6-foot difference in
phototube-lead length.
The total amplifier noise level under normal operating conditions
is between 30 and 35 decibels below 0.006 watt.
SCANNING SLIT
Before adopting the wide slit, experiments were conducted with
slit widths of from 0.0004 to 0.0018 inch (equivalent at film), and
shaping the response curve by means of a wide slit was not found to
impair reproduction as compared with the alternative method of
shaping the amplifier's frequency response.
The wide slit has been used consistently since 1939.
FIRE CONTROL
Accumulated practical experience of the carbon-dioxide fire con-
trol, which has been an in-built feature of Gaumont-Kalee design
since 1939, shows that a film fire originating anywhere in the picture
mechanism is extinguished within less than one second of its incep-
tion, with a loss not exceeding two frames of film, and that the action
of the device is certain if charged and set.
Zoomar Lens for 35-Mm Film*
BY F. G. BACK
RESEARCH AND DEVELOPMENT LABORATORY, 381 FOURTH AVE.,
NEW YORK 16, NEW YORK
Summary— The 35-mm Zoomar is at present mainly used for newsreel
work where it has proved itself a valuable tool especially in the field of
sports shots. A studio Zoomar of more rigid construction and higher opti-
cal correction which can also be used for color work is in preparation.
THREE ARTICLES1 ~3 dealing with the basic principle of the Zoomar
lens have already been published in this JOURNAL, so it is hardly
necessary to go into theoretical details again. The 35-mm Zoomar has
already established its place as a valuable tool in newsreel photog-
raphy one year after appearance of its smaller brother, the 16-mm
Zoomar.
Numerous newsreel companies are using it regularly for their sport
shots. Fast sports, especially, change their center of interest rapidly
from one single player to a large group or even to the entire field. This
demands the use of a Zoomar lens because the standard turret does not
give satisfactory results. While changing lenses breaks the continuity
and makes the picture jumpy and confusing, the Zoomar gives a con-
tinuous transition which makes it easier to understand the game. It
is even possible to follow the players or participants over the whole
field and nevertheless keep them always the same size on the screen
regardless of their position. This is of great importance in certain
sports such as racing, football, and baseball. This striking effect
cannot be obtained by any other means.
The 35-mm Zoomar was primarily developed for use of newsreel
photography with emphasis on sports, because it was almost impos-
sible up to then to catch the high lights of a game except by accident.
It is self-evident that the 35-mm Zoomar lens for studio work has to
fulfill other requirements than the newsreel Zoomar. Each one will
have to be designed differently. The lens for news photography has to
be comparatively light in weight. It has to be capable of rapid transi-
tion but it works only for black-and-white. The studio lens on the
other hand may be much heavier but has to be more rigidly con-
structed because it must be suitable for taking unconscious zooms
* Presented October 21, 1947, at the SMPE Convention in New York.
294 SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
ZOOMAR LENS
295
which are never demanded of the newsreel Zoomar. The picture
quality of the studio lens has to be superior to that of the newsreel
lens, and above all the studio lens must be suitable for both black-and-
white and color.
Fig. 1 shows the 35-mm newsreel lens mounted on an Akeley
camera. The coupled Zoom viewfinder is of the same basic design as
the one in the 16-mm sport Zoomar. The cameraman is not confined
to a fixed position to view the finder image. He sees the finder frame
comparatively large and free from any ground-glass diffusion.
1
Fig. 1
A parallax adjustment of the viewfinder was not provided, for a very
good reason. Practical experience has shown that better results are
obtained when the cameraman is trained to make allowances for the
parallax instead of operating a parallax adjustment, because the tar-
get changes its distance from the camera and therefore a parallax
adjustment will have to be reset continuously throughout the take.
Actually, this is never done, and therefore the framing of a viewfinder
with parallax adjustment is not only inaccurate most of the time but
this inaccuracy changes continuously. The Zoomar finder has a fixed
parallax of five inches which is always maintained regardless of the
Zoom position and of the object distance. Once the cameraman has
learned to take these five inches into consideration he will never fail.
296 BACK September
The specifications for the 35-mm newsreel Zoomar are :
Aperture range : //4.5 to //22
Zoom range: Interchangeable wide-angle front lens, 2 to 7 inches.
Zoom range: Interchangeable telephoto front lens, 3.5 to 13 inches.
Field coverage: Difference in field area in any one continuous shot, 15 times.
Object distance: From 12 feet to infinity.
Height: 3 inches
Length: 26 inches
Width: 8 inches, including viewfinder
Weight: 11 pounds, including viewfinder
The design for the 35-mm Zoomar for studio work has not yet been
completed. There is also a special Zoomar for animation work in prep-
aration.
Without doubt the 16- and 35-mm Zoomar lens have proved their
merits and have stimulated enthusiastic cameramen to develop many
new ideas for their uses in a variety of fields. These pioneers will
eventually develop a new technique which will prove a valuable con-
tribution to the art of motion picture production.
REFERENCES
(1) Frank G. Back, "A positive vari-focal viewfinder for motion picture
cameras," /. Soc. Mot. Pict. Eng., vol. 45, pp. 466-472; December, 1945.
(2) Frank G. Back, "Zoom lens for motion picture cameras with single-barrel
linear movement," /. Soc. Mot. Pict. Eng., vol. 47, pp. 464^69; December, 1946.
(3) Frank G. Back, "The physical properties and the practical application of
the Zoomar lens," J. Soc. Mot. Pict. Eng., vol. 49, pp. 57-64; July, 1947.
DISCUSSION
DR. K. PESTRECOV: What is the speed of the lens?
DR. FRANK G. BACK: The geometrical speed of the lens is//4.5.
DR. PESTRECOV: Does the speed change while zooming?
DR. BACK: No. The speed is independent of the zoom. If you set your lens
for a certain speed this speed remains constant over the entire zoom.
MR. R. E. LEWIS: What is the resolving power of the lens in lines per milli-
meter?
DR. BACK : The 35-mm lens has a resolution of approximately forty lines in the
center and a little less toward the edges. We are working now on increasing this
resolution.
CHAIRMAN WILLIAM H. RIVERS: What approximately is the weight of the
35-mm lens and is there any additional support needed for field use of this lens?
DR. BACK: The weight of this lens is not more than the weight of one of the
large telephoto lenses. For instance, if you take a 25-inch telephoto lens it is
heavier than this one. Just as it is, it weighs approximately 11 pounds, and we
have found that for general use it is not absolutely necessary to put any support
underneath. Of course, a support can be used with the lens very easily.
1948 ZOOMAR LENS 297
DR. PESTBECOV: What is the range of focal length covered?
DR. BACK: The range of this particular lens with the telephoto front lens is
from three and a half inches up to thirteen inches. We are working now on in-
creasing the range as well as the focal length. If we use the wide-angle front lens
by exchanging the front lenses the range goes from approximately two inches up
to seven inches.
MR. WILLARD W. JONES: What is the depth of field?
DR. BACK: The depth of field is exactly the same as on any other lens of com-
parative focal length. For instance, if you take the lens in the telephoto position
the depth of field becomes rather shallow as it would be, for instance, on a thirteen-
inch lens at//4.5. If the zoom lever is in the wide-angle position the depth of field
increases accordingly. In this case you would have the same depth of field as a
lens of three and a half inches focal length at//4.5.
This is the reason why it is not necessary to refocus on those follow shots where
the subject moves toward or away from the camera and still remains the same size
on the screen. These follow shots have been focused in the telephoto position
where the subject is at the greatest distance from the camera because in this posi-
tion the depth of field is very small. If everything is sharp in this position even
when the subject comes closer and we go into our wide-angle position by moving
the zoom lever everything still remains in focus, due to the greater depth of field
in this position.
FORTY YEARS AGO
Moving Pictures from a Balloon
Berlin, April 25. — Photographs for the cinematograph have just been
taken from a balloon successfully by Herr Ernemann, a Dresden engi-
neer. As the exciting aerial voyage was ending he passed over the
Sensteberg coal mine. Here, too, Ernemann succeeded in taking fine
photographs. But just then the balloon shot down so suddenly that
even the cinematograph apparatus had to be thrown from the basket.
Luckily the pictures were afterward found intact. — New York World.
—The Moving Picture World, May 2, 1908
Parabolic Sound Concentrators
BY R. C. COILE
WASHINGTON, D. C.
Summary — Parabolic sound concentrators have long been investigated
for application to military antiaircraft location, radio broadcasting, and
motion picture recording. Olson and Wolff, of the Radio Corporation of
America, developed a combination horn-reflector concentrator in 1929.
Obata and Yosida, of Tokyo University, published measurements of ampli-
fication in 1930. Hanson, of the National Broadcasting Company, described
the use of parabolic reflectors in broadcasting in 1931. Dreher reported
in 1931 on the use of microphone concentrators in motion picture produc-
tion. Sato and Sasao published the results of tests on the sound fields of
deep parabolic reflectors in 1932.
Rocard published an analysis of the theory of the amplification of the
reflector-type parabola in 1932. Schneider of the Moscow Radio Center
made amplification and directivity measurements in 1935 while studying
the application of parabolic concentrators to Russian broadcasting and
checked his amplification data with Rocard's theory. Gutin, in Leningrad,
independently derived the theory of amplification and went on to analyze
directivity in 1935.
This paper presents the pertinent historical background and reports on
an experimental verification of the theoretical acoustical directivity of
parabolic concentrators as well as further checks of the amplification theory.
The sound fields inside parabolic reflectors have also been investigated
experimentally with an agreement found with theoretical fields calculated
by principles of geometrical optics.
HISTORICAL BACKGROUND
THE IDEA OF USING a parabolic mirror as a concentrator of sound
by placing one's ear or a microphone at the,focus was a subject of
research in World War I. Waetzmann1 has described German parab-
olas and Tucker2 has reported on English and French development
of plaster and concrete parabolic reflectors. The only quantitative
data given in these reports is an estimate by Waetzmann that for a
parabola having an opening diameter of 3.2 meters and a depth of 0.8
meter the magnification was about five times compared with unaided
ears for whispers and less for lower notes.
The first quantitative work published on sound concentrators was a
report by Olson and Wolff3 in 1929 of their development of a combi-
nation horn and reflector. The theory behind this was that the
amplification of a reflector-type sound concentrator depends on the
298 SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
PARABOLIC SOUND CONCENTRATORS 299
wavelength of the impinging sound being less than the dimensions
of the reflector. Hence the low frequencies whose wavelengths are
larger than the dimensions are amplified very little. But by build-
ing a horn on the parabolic reflector, the amplification of the horn
raised the low-frequency response. This design worked fairly well
and microphone concentrators of this type have been used in Holly-
wood for recording motion pictures.
Obata and Yosida,4 engineers of the Tokyo Imperial University's
Aeronautical Research Institute, made a'study of acoustical proper-
ties of some sound collectors for the aircraft sound locator in
1930. They made measurements of the amplification and directivity
of two different horns and two 200-centimeter diameter open-bowl
parabolic reflectors of different focal distance.
Dreher5 reported on the use of microphone concentrators in
motion picture production in this JOURNAL in 1931. Military
searchlights with a microphone at the focus were used in outdoor re-
cording, and other types of parabolic bowls were also used.
The developments of the National Broadcasting Company were
announced in 1931 by Hanson,6 chief engineer. Measurements of the
; amplification, directivity, and effect of microphone position on the
' axis were reported on a design of an open-bowl parabolic reflector built
by NBC engineers.
Engineers of the Aeronautical Research Institute, Sato, Sasao,
Kubo, and Nisiyama published several papers7' 8 on the sound fields
i of parabolas in 1932. Their measurements were performed on deep
parabolic reflectors and hence the results are rather complicated look-
ing. The measurements were taken in the region beyond the focus,
for the most part. These writers did not explain these results but
merely said, "The experiment was very laborious and troublesome
and therefore was carried out with only two pitches of sound, C2 and
C4. . . . For C4, the sound field becomes very complex and many
maxima and minima due to interference fill up the space in front of
the mirror."
In 1932, the first theoretical treatment was published in Rocard's
paper9 on "Les Paraboloides Acoustique" in the Revue d'Acoustique,
where Rocard derived an expression for the amplification of a para-
bolic reflector.
In 1935, Rocard 's theoretical predictions were experimentally veri-
fied by Schneider,10 an engineer of the Moscow Radio Center.
Schneider's paper in the Zhurnal Teknicheskoi Fiziki examined all
300 COILE September
previously published work and reported on measurements of am-
plification which checked Rocard's predictions.
Neither Rocard nor Schneider had been able to cope with the
theory of the directivity of a parabolic reflector. In 1935, Gutin,11 a
physicist in Leningrad, knowing nothing of the work of either Rocard
or Schneider, derived independently the expression for amplification
and went on to work out the theory of directivity which he published
in the Izvestia Elektropromishlennosti Slabova Toka.
EXPERIMENTAL STUDY OF PARABOLIC CONCENTRATORS
An experimental study has been made of the following characteris-
tics of the parabolic reflector: (1) frequency response, (2) amplifica-
tion, (3) directivity, and (4) sound fields. The published experimen-
tal work on reflector-type concentrators has been very meager as out-
lined above. Most of the published papers show the results of experi-
ments completed prior to 1932. The microphones used were not
always of the highest quality or of small size — a desirable feature of a
sound-field measuring device. Some of the work by Obata and
Yosida,4 for example, was done using a large homemade condensei
microphone with most of the experimental work performed indoors
with the sound source rather close to the parabola. What work had
been done outside is open to considerable question because of groum
effects, as the parabola was simply placed upright about a foot off the
ground.
Other experimenters have used parabolas with opening diameters
ranging from 40 to 300 centimeters. A 130-centimeter copper para-
bolic reflector was used in this experimental setup to simplify measure-
ment of the sound fields inside the reflector for we might expect acous
tical reflection similar to optical reflection when the sound wave-
lengths are small compared to the dimensions of the parabola am
diffraction effects when the wavelengths are comparable to the dimen-
sions. The large size of the parabola indicated outdoor measurements
to avoid errors from reflected sound although outdoor measurements
present difficulties of wind and extraneous noises.
Kellogg12 described five methods for minimizing echo errors in &
paper in the Journal of the Acoustical Society some years ago. Fig. 1
depicts these schemes. In A both the loudspeaker and the microph* >i i <
are well above the ground. If the distance is large compared with
the wavelength of the lowest frequency employed the image sourcea
1948
PARABOLIC SOUND CONCENTRATORS
301
are negligible. In B image sources are taken into consideration by
placing both the loudspeaker and the microphone on the ground
i so that the difference between path length r from loudspeaker to
I microphone and the path length r' from image to microphone is less
than a quarter wavelength of the sound. In C the speaker is sup-
ported in the air with the microphone on the ground. The sound re-
flected from the ground, if the ground is not a good absorbent, is some-
times strong enough to cause some back pressure on the loudspeaker.
This can be fixed by putting the microphone on a slope as shown in D,
so that the sound is reflected off at such an angle that it has little
i effect on the loudspeaker.
I One more method is to get
j the microphone out on a
| boom as far from any build-
[ing as possible and to have
I the sound source at the cor-
j ner of the building as illus-
trated in E.
The most convenient
j method for this particular
LOUDSPEAKER AND
MICROPHONE WELL
ABOVE GROUND
TAKING ACCOUNT
OF IMAGE
MICBOPHONE ON
GROUND
REACTION OF ECHO
ON SPEAKER. AVOIDED
WORKING AT COR.NER. OF BUILDING
Fig. 1 — Arrangements for minimizing echo
errors.
I experimental setup was a
i variation of D as illustrated
in Fig. 2. The sound source
[was a General Radio beat-
I frequency oscillator which
| excited a Western Electric
| loudspeaker unit in a 6-foot
exponential horn suspended
out of a window of one of
j the sound laboratories at the Massachusetts Institute of Technology.
j The parabolic reflector was placed about 100 feet from the side of the
! building and was pointed toward the sound source. A Western
Electric Type 630-A moving-coil microphone, step-up transformer,
General Radio amplifiers, and General Radio output meter were
! used in the frequency response-amplification measurements, and a
General Radio sound-level meter with a Brush sound-cell crystal
microphone for sound-field measurements.
FREQUENCY RESPONSE- AMPLIFICATION
An expression for the theoretical frequency response and
302
COILE
September
amplification of parabolic reflectors was derived independently by
Rocard and Gutin. This expression is as follows :
I = depth of parabola
R = radius of opening
3RD FLOOR. WINDOW OF THE
SOUND LABORATORY
where
P/s = pressure with concentrator
Pa = pressure without concentrator
a = focal distance
For the parabola under test,
R the radius of opening was
65 centimeters; a the focal
distance was 30 centimeters;
and I the depth was 35 centi-
meters. The expression for
the amplification of this pa-
rabola reduces to
Fig. 2 — Experimental setup.
Equipment: 1. Parabolic bowl reflec-
tor: diameter, 130 centimeters; depth, 35
centimeters; focal distance, 30 centimeters.
2. Microphone, Western Electric 630-
A moving-coil microphone.
3. Step-up transformer, 30 to 100,000
ohms.
4. General Radio battery-operated
amplifier.
5. Step-down transformer.
6. Transmission line.
7. Step-up transformer, 30 to 100,000
ohms.
8. General Radio battery-operated
amplifier.
9. General Radio output-level meter.
10. General Radio beat-frequency oscil-
lator, 20 to 20,000 cycles per second.
11. Six-foot exponential horn with
Western Electric 555 unit.
12. General Radio sound-level meter
with Brush crystal microphone.
This theoretical amplification
at the focus is an inverse func-
tion of the wavelength and is
plotted as the straight line in
Fig. 3.
The frequency response and
amplification of the parabolic
concentrator were determined
first by measuring the response
of the microphone alone in free
space, and then measuring the
response of the microphone in
the concentrator. The micro-
phone was placed at the focus
of the paraboloid. The meas-
ured frequency response and
amplification characteristic are!
shown in Fig. 3. This agreement between measured gain of the con-
centrator and the computed values of amplification is as good as that j
reported by Gutin,11 for the work of Obata and Yosida,4 and the
comparisons reported by Schneider.10 The measured amplification j
differs from the computed amplification by about 10 decibels at the
1948
PAKABOLIC SOUND CONCENTRATORS
303
higher frequencies (7000 cycles) for the parabola investigated.
Obata and Yosida's4 measurements for frequencies from 475 to 188
cycles show the same trend, a divergence between theory and meas-
urement for the higher frequencies. At their highest frequency of
475 cycles, the difference between theoretical and measured amplifica-
tion was on the order of 20 per cent for one parabola 200 centimeters
in diameter and 72.5 centimeters in focal length; and about 80 per
cent difference for another parabola of the same diameter with a 54.5
40
335
UJ
cfl
S30
Q
225
_1
LJ20
>
U
J.5
1-
|,0
g5
0
2(
^X
COMPUT
(J.) ,
ED AMPLIFICATION
4TT 0, ^ Cl+R2 )
^x
vRc
A ' 4al y
/
/
/
/ /
^x
*^~ AMPLIFICATION CP
CONCENT!^
ATOR.
/
x^^
D=I3O cm, f -io
(e=65CYn a =
Cm, d-35
30cm)
CWl
,/X
7^' ' ~
)0 ' 500 1000 MOO 10000
IN CpS
Fig. 3 — Frequency-response amplification of the parabolic
sound concentrator.
centimeter focal length. The curves of Schneider show this same
trend. Schneider does not draw theoretical curves for frequencies
higher than 4000 cycles because the disagreement is so large.
The size of the probe microphone used in the measurements affects
the accuracy of the results. Schneider used three microphones in his
tests, a large Reisz (carbon) with a diaphragm area of 70 square centi-
meters, a small Reisz with a diaphragm area of 40 square centimeters,
and a condenser microphone with an area of 20 square centimeters.
Examination of his data shows that the smaller the diaphragm area,
the better the agreement between theoretical and measured results.
Schneider did not attempt to explain this phenomenon. It may
possibly be attributed to phase-cancellation effects, the diaphragm
being so large that the higher-frequency sounds which behave more or
less as geometrical optics predict, arriving in pencils of rays, hit the
diaphragm in different phase thus reducing the output. Instead of
304 COILE September
having an infinitely small collector of the sound arriving at the focus in
phase, we have a large sound-receiving surface that can pick up sound
of different phase which will tend to reduce the output. The Western
Electric Type 630-A microphone used in the frequency response-am-
plification tests has a diaphragm area estimated at 10 square centi-
meters, and the Brush sound-cell crystal used as a probe microphone
in tracing out the sound fields has an area estimated at 2.5 square
centimeters.
DIRECTIVITY
The directivity characteristic is important in many applications of
the parabolic reflector. Sato and Sasao have published experimental
directivity curves and Schneider published some curves. The first
published paper on an analysis of the theory of the directivity of the
parabolic reflector is that of Gutin, who derived an expression for the
coefficient of amplification at the focus of a paraboloid for an arriving
sound wave whose normal made an angle a with the axis as follows
/•
JO
O 1 +
R
,_ ,
Gutin derived a simpler expression using the theorem of reciprocity
that is essentially the same at higher frequencies. This expression for
the coefficient of amplification at the focus is
Neglecting the incoming wave with respect to the reflected wave,
the directivity characteristic expressed as a fraction of the maximum
amplification (i.e., a = 0) is
.R
f2a h(2akt siii «)
Jo
(I + t2) In (1 -
Gutin has given a table showing the position of the first minimum of,
the directivity for values of R/2a.
The measured directivity characteristic is shown in Fig. 4. The
directivity was determined first by lining up the axis of the parabola |
1948
PARABOLIC SOUND CONCENTRATORS
TABLE I
FIRST MINIMUM OF DIRECTIVITY
R/2a
sin a
1.0
0.683X/#
0.7
0.66 \/R
0.6
0.64 \/R
0.4
0.62 \/R
0
0.61 X/fl
305
150'
Fig. 4 — Directivity characteristics: Massachusetts
Institute of Technology parabolic concentrator.
with the axis of the 6-foot exponential-horn sound source, and then
tilting the parabola and measuring the response at different angles.
The lobes other than the fundamental were negligible and could not
be distinguished from background noise. The main features of in-
terest are the angles at which the response falls to its first minimum.
These angles can be calculated by the method developed by Gutin.
The procedure is as follows : The angle at which the directional char-
acteristic goes through its first minimum is given by the expression
a = sm~1K \/R
Where X = wavelength of sound being received
R = radius of opening of parabola
K = constant depending on 72/2a (see Table I).
306 COILE September
The constant K is determined by calculation of R/2a and then use of
Table I. For the parabola investigated R = 65 centimeters; a = 30
centimeters; R/2a = 1.08; K = 0.69. Hence we can compute the
angle of the first minimum.
TABLE II
CHECK OF THEORY AND MEASUREMENTS OF DIRECTIVITY
/
500 cycles
2000 cycles
5000 cycles
X
69 centimeters
17.2 centimeters
6.9 centimeters
\/R
1.06
0.264
0.106
0.69 \/R
0.725
0.180
0.0725
a (computed)
46 . 5 degrees
10. 2 degrees
4 . 2 degrees
a (measured)
45 degrees
15 degrees
5 degrees
This agreement of theoretical and measured angles for the first
minimum was a reasonably good check of Gutin's directivity theory.
It is of interest to note that the Izvestia Elektropromishlennosti
Slabovo Toka is available at so few libraries in the world that even
Schneider, another Russian, publishing his paper in the Zhurnal Tek-
nicheskoi Fiziki, also in 1935, stated "The story of the concentrator
is very complicated. The amplification has an approximate solu-
tion. . . . The directivity characteristics are without theory. ..."
The experimental directivity curves of Schneider measured with a
condenser microphone have also been compared with theoretical pre-
dictions. The calculations have been carried through in a manner
similar to those for the 'parabola at the Massachusetts Institute of
Technology: R = 47 centimeters; a = 27 A centimeters; R/2a =
0.86; K = 0.675.
TABLE III
DIRECTIVITY CHECK OF SCHNEIDER'S DATA
/ 700 cycles 1600 cycles 3000 cycles 5500 cycles
X 49.2 centi- 21.5 centi- 11.4 centi- 6.25 centi-
meters meters meters meters
\/R 1.04 0.456 0.242 0.133
0.675 \/R 0.705- 0.308 0.163 0.0896
a (computed) 45 degrees 18 degrees 9. 4 degrees 5. 2 degrees
a (measured) 50 degrees 42 degrees 11 degrees 5 degrees
The results check reasonably well .with the exception of the 1600-
cycle data. However, Schneider's measurements were made indoorsi
1948
PARABOLIC SOUND CONCENTRATORS
307
so that there is a greater possiblity for a freak measurement than if the
measurements had been made outside with less chance of reflections
introducing errors.
SOUND FIELDS
There has been very little published on the sound fields of parabolic
reflectors. Sato and Sasao8 have reported measurements on fields in
a deep parabola. These previous experiments studied complex sound
fields in regions beyond the focus. It was thought of interest to ex-
amine the region between the focus and the vertex.
DIRECTRIX. B
Fig. 5 — Reflection from a parabolic mirror.*
* Wood, R. W.: "Physical Optics," Macrnillan Company, New York, N. Y.,
1934, p. 47.
According to the principles of geometrical optics a source placed at
the focus emitting spherical waves will have them reflected at the
walls of the parabola and sent out as plane wave fronts. And, con-
versely, plane waves arriving at the parabola will be reflected as
spherical waves converging on the focus. When the incident and re-
flected waves meet there can be either constructive or destructive in-
terference. If the difference in path length is m A, where m = 0, 1, 2,
3, 4, ... there will be constructive interference. If the difference in
path Itngth is (2m + 1) (A/2) where m = 0, 1, 2, 3, 4, . . .there will be
destructive interference.
Hence, in attacking this portion of the problem, the contours of
constructive and destructive interference were first determined by
COILE
September
geometrical optical construction and then measured by the acous-
tical setup described.
The construction of the reflected wave fronts is a simple matter.
The fundamental definition of a parabola is that it is the locus of
points equidistant from a fixed line called the directrix and a point
called the focus. Reflected wave fronts may be constructed in a
graphical manner similar to that outlined by Wood.13
In Fig. 5 let 0 be the focus of the parabola and line BD the direc-
trix. Let the unreflected wave front be represented by line HG.
+ MAXIMA- POINTS OF CONSTRUCTIVE INTERFERENCE
- MINIMA-POINTS OF DESTRUCTIVE INTERFERENCE
UNCEFLECTED
WAVE FRONTS
REFLECTED
WAVE FRONTS
CONTOURS
OF MAXIMA'
CONTOURjS
OF MINIMA^
Fig. 6 — Calculation of sound-field contours by
geometric optical interference phenomena.
/ = 1720 cycles per second
X = 20 centimeters
Through any two points on the parabola A and C draw lines from 0,
the focus. Construct circles about points A and C of radius equal to
the distance from these points on the parabola to the unreflected wave
front. A circle drawn about 0 with radius OE will pass through point
F and will be the reflected wave front. This may be proved as fol-
lows : every point on the parabola is equidistant from focus and di-
rectrix, OA = AB and OC = CD; the small circles constructed about
A and C had radii of AE = AH and CG = CF; but now OE = BH
and OF — DG adding the two parts of each line. But since DG = BH,
for the unreflected wave front is parallel to the directrix, hence OE =
1948
PARABOLIC SOUND CONCENTRATORS
309
OF and a circle of the reflected wave front has been determined.
Now we can see that as the unreflected wave front moves into
the parabola, the reflected wave fronts become smaller and smaller
circles converging on the focus.
SCALE : I cm = 5 cm.
COMPUTED
CONTOURS OF
MAXIMA.
MEASURED POINTS
OF MAXIMA
FOCUS
PARABOLIC BOWL
BEFLECTOB.
DIAMETER. 13O Cm
DEPTH 35 cm
FOGAL DISTANCE
30 Cin
^. MEASURED POINTS
OF MINIMA
COMPUTED
^CONTOUR.6
OF MINIMA
Fig. ±7 — Sound-field contpurs^off constructive and
t A •*£* destructive interference.
/ = 3440 cycles per second
X = 10 centimeters
Brush sound-cell microphone; Federal Radio
sound-level meter.
A useful short cut in drawing these wave fronts is apparent on ex-
amination. The reflected wave fronts, circles about the focus, inter-
sect the parabola at the same points as the unreflected plane wave
front. Therefore, it is an easy matter to draw a circle with center at
the focus and radius equal to the distance from the focus to the points
of intersection of the parabola and the plane wave front.
310 COILE September
Using this simple method of constructing reflected wave fronts, the
contours of points of constructive interference (maxima) and points of
destructive interference (minima) may be traced out after finding
these points by checking path lengths. This construction is illus-
trated in Fig. 6. A series of plane unreflected wave fronts approach-
ing the parabola has been drawn spaced a half wavelength apart.
The frequency of 1720 cycles per second has been chosen to give a con-
venient wavelength of 20 centimeters (X = c/f = 34400/1720). A
series of concentric circles converging on the focus has been drawn
corresponding to the c/f approaching plane wave fronts. There are
numerous points of intersection. For each of these points we trace
out the difference in path length between the incident and reflected
wave. For example, at the surface of the parabola the path-length
difference is zero, hence constructive interference; but moving out
from the parabola along any circle of a reflected wave front there are
points whose path-length difference is X/2 designated by a minus
sign; X designated as plus, 3X/2 minus, etc. The points of construc-
tive interference marked "plus" have been joined and in a similar
fashion lines have been drawn through the "minus" points. These
lines are the contours of maxima and minima.
These contours were traced out by a crystal probe microphone and
a General Radio sound-level meter. Predicted and measured data
have been plotted in Fig. 7. Examination of the figure shows the
theoretical contours of constructive interference plotted as solid lines
in the upper half of the parabola and the theoretical contours of de-
structive interference plotted as' dotted lines in the lower half. Of
course, both contours occur in three dimensions as a series of confocal
shells of paraboloids of revolution, but for our purposes the contours
are cross-section pictures of the parabola with the contours of minima
made invisible in the upper half and the contours of maxima made in-
visible in the lower half.
The experimental points have been plotted as little circles for
maxima and X's for minima. There are two places where the points
do not check so well, points a great distance from the axis, and points
on the axis. The explanation for the discrepancy of the points quite
distant from the axis may be attributed to two things : distortion of
the sound field by ground effects, and/or lack of rigidity of the
microphone probe equipment. Measurements could not be made with
sufficient precision to determine quantitatively the magnitude of the
ground effect. For the purpose of checking the theoretical contours
1948 PARABOLIC SOUND CONCENTRATORS 311
of constructive and destructive interference the experimental points
nearer the axis must suffice and it is felt that, on the whole, a reasonable
agreement is found within the magnitude of the errors of measurement.
REFERENCES
(1) E. Waetzmann, "Parabolic reflectors" (In German), Zeit.filr Tech. Phys.,
vol. 2, p. 191; 1921.
(2) W. S. Tucker, "Sound reception," Royal Aero. Soc. Proc., vol. 28, p. 504;
1924.
(3) H. F. Olson and I. Wolff, "Sound concentrator for microphones," J.
Aeons. Soc. Amer., vol. 1, pp. 410-417; March, 1929.
(4) Juichi Obata and Yekio Yosida, "Acoustical properties of some sound
collectors for the aircraft sound locator" (In English), Aero. Res. Inst., Tokyo Im-
perial University, vol. 5, pp. 231-249; July, 1930.
(5) Carl Dreher, "Microphone concentrators in picture production," J. Soc.
Mot. Pict. Eng., vol. 16, pp. 23-31 ; January, 1931.
(6) O. B. Hanson, "Microphone technique in broadcasting," J. Acous. Soc.
Amer., vol. 3, pp. 81-93; 1931.
(7) Kozi Sato, Masaki Sasao, Keiiti Kubo, and Masao Nisiyama, "On the
acoustical properties of parabolic reflectors" (In Japanese), Aero. Res. Inst.,
Tokyo Imperial University, vol. 8, pp. 18-64; 1932; pp. 339-356; 1933.
(8) Kozi Sato and Masaki Sasao, "On the sound field of parabolic reflectors"
(In English), Proc.*Physics-Math. Soc. (Japan), vol. 14, pp. 363-372; 1932.
(9) M. Y. Rocard, "Les paraboloides acoustique" (In French), Rev. d'Acous-
tique, vol. 1, pp. 222-231 ; 1932.
(10) J. I. Schneider, "A microphone concentrator" (In Russian), Zhurnal Tek-
nicheskoi Fiziki (Jour. Tech. Phys.}, vol. 5, pp. 855-867; 1935.
(11) L. J. Gutin, "On the theory of the parabolic sound reflector" (In Russian),
Izvestia Elektropromishlennosti Slabovo Toka (Leningrad), vol. 9, pp. 9-25, 75-76;
1935.
(12) E. W. Kellogg, "Loud speaker sound pressure measurements," /. Acous.
Soc. Amer., vol. 2, p. 157; 1930.
Committees of the Society
(CORRECT TO AUGUST 10, 1948)
ADMISSIONS
To pass upon all applications for membership, applications for transfer,
and to review the Student and Associate membership list periodically
for possible transfer to the Associate and Active grades, respectively.
The duties of each committee are limited to applications and transfers
originating in the geographic area covered.
D. B. JOT, Chairman, East
30 E. 42d St.
New York 17, N. Y.
E. A. BERTRAM JAMES FRANK, JR. G. T. LORANCE
M. R. BOYER L. E. JONES PIERRE MERTZ
HERBERT GRIFFIN, Chairman, West
133 E. Santa Anita Ave.
Bui-bank, Calif.
J. W. BOYLE P. E. BRIGANDI H. W. MOYSE
C. R. DAILY
•
BOARD OF EDITORS
To pass upon the suitability of all material submitted for publication, or
for presentation at conventions, and publish the JOURNAL.
A. C. DOWNES, Chairman
2181 Niagara Dr.
Lakewood 7, Ohio
A. W. COOK C. W. HANDLEY P. J. LARSEN
J. G. FRAYNE A. C. HARDY G. E. MATTHEWS}
A. M. GUNDELFINGER PlERRE MERTZ
CINEMATOGRAPHY
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of motion picture cameras, accessory equip-
ment, studio and outdoor set lighting arrangements, camera technique,
and the varied uses of motion picture negative films for general photog-
raphy.
C. G. CLARKE, Chairman
328 S. Bedford Dr.
Beverly Hills, Calif.
J. W. BOYLE ARTHUR MILLER ARTHUR REEVES
KARL FREUND JOSEPH RUTTENBERGI
312 SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
COMMITTEES OF THE SOCIETY 313
COLOR
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of color motion picture processes, accessory
equipment, studio lighting, selection of studio set colors, color cameras,
color motion picture films, and general color photography.
HERMAN H. DUEBR, Chairman
Ansco
Binghamton, N. Y.
J. A. BALL L. E. CLARK J. G. FRAYNE
R. H. BINGHAM R. O. DREW L. T. GOLDSMITH
M. R. BOYER ALBERT DURYEA A. M. GUNDELFINGER
H. E. BRAGG R. M. EVANS A. J. MILLER
0. 0. CECCARINI G. F. RACKETT
CONVENTION
To assist the Convention Vice-President in the responsibilities pertain-
ing to arrangements and details of the Society's technical conventions.
W. C. KTJNZMANN, Chairman
Box 6087
Cleveland 1, Ohio
JAMES FRANK, JR. H. F. HEIDEGGER A. SHAPIRO
E. R. GEIB O. F. NEU L. H. WALTERS
EXCHANGE PRACTICE
To make recommendations and prepare specifications on the engineer-
ing or technical methods and equipment that contribute to efficiency in
handling and storage of motion picture prints, so far as can be obtained
by proper design, construction, and operation of film-handling equip-
ment, air-conditioning systems, and exchange office buildings.
(Under Organization)
FELLOW AWARD
To consider qualifications of Active members as candidates for elevation
to Fellow, and to submit such nominations to the Board of Governors.
D. E. HYNDMAN, Chairman
342 Madison Ave.
New York 17, N. Y.
E. A. BERTRAM W. C. KTTNZMANN . L. L. RYDER
M. R. BOYER G. T. LORANCE A. SHAPIRO
JAMES FRANK, JR. J. A. MATJRER E. I. SPONABLE
C. R. KEITH W. V. WOLFE
314
COMMITTEES OF THE SOCIETY
September
FILM DIMENSIONS
To make recommendations and prepare specifications on those film
dimensions which affect performance and interchangeability, and to
investigate new methods of cutting and perforating motion picture film
in addition to the study of its physical properties.
(Under Organization)
FILM-PROJECTION PRACTICE
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of motion picture projection equipment, pro-
jection rooms, film-storage facilities, stage arrangement, screen dimen-
sions and placement, and maintenance of loudspeakers to improve the
quality of reproduced sound and the quality of the projected picture in the
theater.
(Under Organization)
HIGH-SPEED PHOTOGRAPHY
To make recommendations and prepare specifications for the con-
struction, installation, operation, and servicing of equipment for photo-
graphing and projecting pictures taken at high repetition rates or with
extremely short exposure times.
JOHN H. WADDELL, Chairman
463 West St.
New York 14, N. Y.
R. E. FARNHAM
COMMANDER FRAZER
H. M. LESTER*
L. MARTIN
J. J. McDEVITT
E. ANDRES
K. M. BAIRD
M. BEARD
B. BRIXNER
HAROLD EDGERTON
MAJOR ELLIOTT
A. NEYHART
BRIAN O'BRIEN
EDWARD SCHMIDT
C. SLACK
M. L. SANDELL
E. W. WIGHTMAN**
* Representing Photographic Society of America.
** Representing Photographic Engineering Society.
HISTORICAL AND MUSEUM
To collect facts and assemble data relating to the historical development
of the motion picture industry to encourage pioneers to place their work
on record in the form of papers for publication in the JOURNAL, and to
place in suitable depositories equipment pertaining to the indus-
try.
(Under Organization)
HONORARY MEMBERSHIP
To search diligently for candidates who through their basic inventions or
outstanding accomplishments have contributed to the advancement of
the motion picture industry and are thus worthy of becoming Honorary
members of the Society.
1948 COMMITTEES OF THE SOCIETY 315
HONORARY MEMBERSHIP (continued)
H. W. REMERSCHEID, Chairman
716 N. La Brea Ave.
Hollywood, Calif.
F. E. CAHILL, JR. C. W. HANDLEY
WALTER CLARK REEVE O. STROCK
JOURNAL AWARD
To recommend to the Board of Governors the author or authors of the
most outstanding paper originally published in the JOURNAL during the
preceding calendar year to receive the Society's JOURNAL Award.
J. I. CRABTREE, Chairman
Eastman Kodak Company
Rochester 4, N. Y.
C. R. DAILY R. E. LEWIS
L. T. GOLDSMITH D. R. WHITE
LABORATORY PRACTICE
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of motion picture printers, processing ma-
chines, inspection projectors, splicing machines, film-cleaning and treat-
ing equipment, rewinding equipment, any type of film-handling acces-
sories, methods, and processes which offer increased efficiency and im-
provements in the photographic quality of the final print.
(Under Organization)
A. A. DURYEA, Chairman
90 W. Horton St.
City Island, N. Y.
MEMBERSHIP AND SUBSCRIPTION
To solicit new members, obtain nonmember subscriptions for the JOUR-
NAL, and to arouse general interest in the activities of the Society and its
publications.
L. E. JONES, Chairman, East
427 W. 42d St.
New York 18, N. Y.
Atlantic Coast
BERTIL CARLSON C. F. HORSTMAN C. W. SEAGER
A. R. GALLO W. C. KUNZMANN A. G. SMITH
T. J. GASKI O. F. NEU C. M. WEBER
N. D. GOLDEN P. D. RIES C. R. WOOD, SR.
316 COMMITTEES OF THE SOCIETY
MEMBERSHIP AND SUBSCRIPTION
H. H. WILSON, Chairman, Midwest
240 E. Ontario St.
Chicago, 111.
JOHN POWERS
T. I. RBSS
JOHN SPINNEWEBER
B. W. DsPEU
R. E. FARNHAM
C. E. HEPPEBERGER
September
(continued)
LLOYD THOMPSON
ELMER VOLTZ
JOHN ZUBER
G. C. MISENER, Chairman, Pacific Coast
6424 Santa Monica Blvd.
Hollywood 38, Calif.
L. W. CHASE, JR. L. T. GOLDSMITH W. A. MUELLER
J. P. CORCORAN HERBERT GRIFFIN G. E. SAWYER
C. R. DAILY EMERY HUSE W. L. THAYER
J. G. FRAYNE K. F. MORGAN W. V. WOLFE
H. W. MOYSE
16-Mm
G. A. CHAMBERS C. L. LOOTENS A. G. PETRASEK
A. W. COOK L. R. MARTIN F. B. ROGERS
R. C. HOLSLAG W. H. OFFENHAUSER, JR. R. J. SHERRY
W. F. KRUSE LLOYD THOMPSON
Foreign
R. J. ENGLER B. J. KROGER H. S. WALKER
R. O. STROCK
Television
R. B. AUSTRIAN, Chairman
247 Park Ave.
New York 17, N. Y.
NOMINATIONS
To recommend nominations to the Board of Governors for annual elec-
tion of officers and governors.
E. A. WILLIFORD, Chairman
40 Charles St.
Binghamton, N. Y.
JAMES FRANK, JR.
L. T. GOLDSMITH
J. K. MILLIARD
EMERY HUSE
E. W. KELLOGG
P. J. LARSEN
K. F. MORGAN
M. G. TOWNSLEY
1948 COMMITTEES OF THE SOCIETY 317
PAPERS
To solicit papers and provide the program for semiannual conventions,
and make available to local sections for their meetings papers pre-
sented at national conventions.
G. A. CHAMBERS, Chairman
343 State St.
Rochester 4, N. Y.
JOSEPH E. AIKEN, Vice-Chairman R. T. VAN NIMAN, Vice-Chairman
225 Orange St., S. E. 4331 W. Lake St.
Washington 20, D. C. Chicago 24, 111.
E. S. SEELEY, V T ice-Chairman N. L. SIMMONS, Vice- Chairman
161 Sixth Ave. 6706 Santa Monica Blvd.
New York 13, N. Y. Hollywood 38, Calif.
H. S. WALKER, Vice- Chairman
1620 Notre Dame St., W.
Montreal, Que., Canada
G. M. BEST C. R. DAILY G. E. MATTHEWS
P. E. BRIGANDI A. R. DAVIS J. H. MAYNARD
J. P. CORCORAN W. P. DUTTON P. A. McGuiRE
G. R. CRANE J. L. FORREST EDWARD SCHMIDT
PRESERVATION OF FILM
To make recommendations and prepare specifications on methods of
treating and storage of motion picture film for active, archival, and
permanent record purposes, so far as can be prepared within both the
economic and historical value of the films.
CHARLES G. WEBER, Chairman
National Bureau of Standards
Washington 25, D. C.
J, G. BRADLEY J. I. CRABTREE TERRY RAMSAYE
H. T. COWLING J. L. FORREST V. B. SEASE
J. E. GIBSON
PROCESS PHOTOGRAPHY
To make recommendations and prepare specifications on motion picture
optical printers, process projectors (background process), matte proc-
esses, special process lighting technique, special processing machines,
miniature-set requirements, special-effects devices, and the like, that
will lead to improvement hi this phase of the production art.
(Under Organization)
LIN WOOD DUNN, Chairman
RKO-Radio Pictures
780 Gower St.,
Los Angeles 3, Calif.
318 COMMITTEES OF THE SOCIETY September
PROGRESS
To prepare an annual report on progress in the motion picture industry.
C. R. SAWYER, Chairman
233 Broadway
New York 7, N. Y.
J. E. AIKEN R. E. LEWIS J. W. THATCHER
C. W. -HANDLEY W. A. MUELLER W. V. WOLFE
W. L. TESCH
PROGRESS MEDAL AWARD
To recommend to the Board of Governors a candidate who by his in-
ventions, research, or development has contributed in a significant man-
ner to the advancement of motion picture technology, and is deemed
worthy of receiving the Progress Medal Award of the Society.
F. E. CARLSON, Chairman
Nela Park
Cleveland 12, Ohio
HARRY BRAUN R. M. CORBIN
J. W. BOYLE W. C. MILLER
PUBLICITY
To assist the Convention Vice-President in the release of publicity ma-
terial concerning the Society's semiannual technical conventions.
*HAROLD DESFOR, Chairman
RCA Victor Division
Radio Corporation of America
Camden, N. J.
*LEONARD BIDWELL P. A. McGuiRE
HARRY SHERMAN
* Advisory Member.
SCREEN BRIGHTNESS
To make recommendations, prepare specifications, and test methods
for determining and standardizing the brightness of the motion picture
screen image at various parts of the screen, and for special means or
devices in the projection room adapted to the control or improvement of
screen brightness.
E. R. GEIB, Chairman
Box 6087
Cleveland 1, Ohio
HERBERT BARNETT W. W. LOZIER C. W. TUTTLE
F. E. CARLSON G. M. RENTOTTMIS C. R. UNDERBILL
ARTHUR HATCH ALLEN STIMSON H. E. WHITE
W. F. LITTLE A. T. WILLIAMS
1948 COMMITTEES OP THE SOCIETY 319
16-MM AND 8-MM MOTION PICTURES
(Formerly Nontheatrical Equipment)
To make recommendations and prepare specifications for 16-mm and 8-
mm cameras, 16-mm sound recorders and sound-recording practices,
16-mm and 8-mm printers and other film laboratory equipment and
practices, 16-mm and 8-mm projectors, splicing machines, screen di-
mensions and placement, loudspeaker output and placement, preview
or theater arrangements, test films, and the like, which will improve the
quality of 16-mm and 8-mm motion pictures.
H. J. HOOD, Chairman
333 State St.
Rochester 4, N. Y.
W. C. BOWEN R. C. HOLSLAG D. A. PBITCHABD
F. L. BRETHAUER R. KINGSLAKE L. T. SACHTLEBEN
F. E. BROOKER G. T. LORANCE R. SPOTTISWOOD
S. L. CHERTOK W. W. LOZIER H. H. STRONG
E. W. D'ARCY D. F. LYMAN A. L. TERLOUW
J. W. EVANS W. C. MILLER L. THOMPSON
C. R. FORDYCE W. H. OFFENHAUSER, JR. M. G. TOWNSLEY
A. G. PETRASEK
SOUND
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of motion picture film, sound recorders,
re-recorders, and reproducing equipment, methods of recording sound,
sound-film processing, and the like, to obtain means of standardizing
procedures that will result in the production of better uniform quality
sound hi the theater.
L. T. GOLDSMITH, Chairman
Warner Brothers Pictures, Inc.
Burbank, Calif.
G. L. DIMMICK, V 'ice-Chairman
RCA Victor Division
Camden, N. J.
A. C. BLANEY L. D. GRIGNON OTTO SANDVIK
D. J. BLOOMBERG ROBERT HERR G. E. SAWYER
F. E. CAHILL, JR. J. K. MILLIARD R. R. SCOVILLE
E. W. D'ARCY L. B. ISAAC W. L. THAYER
R. J. ENGLER E. W. KELLOGG M. G. TOWNSLEY
J. G. FRAYNE J. P. LIVADARY R. T. VAN NIMAN
W. C. MILLER
320
COMMITTEES OF THE SOCIETY
Septembei
STANDARDS
To survey constantly all engineering phases of motion picture produc-
tion, distribution, and exhibition, to make recommendations and pre-
pare specifications that may become proposals for American Standards.
This Committee should follow carefully the work of all other commit-
tees on engineering and may request any committee to investigate and
prepare a report on the phase of motion picture engineering to which it
is assigned. .
F. E. CARLSON, Chairman
Nela Park
Cleveland 12, Ohio
F. S. BERMAN
CHARLES CLARKE
H. H. DUERR
LINWOOD DUNN
A. A. DURYEA
F. T. BOWDITCH
E. K. CARVER
GORDON EDWARDS
L. A. JONES
C. R. KEITH
Chairmen of Engineering Committees
E. R. GEIB
L. T. GOLDSMITH
M. A. HANKINS
H. J. HOOD
M ember s-at'Large
E. W. KELLOGG
R. KlNGSLAKE
Members Ex-OJjicio
V. O. KNUDSEN
STUDIO LIGHTING
D. E. HYNDMAI
LEONARD SATZ
JOHN WADDEL
C. G. WEBER
D. R. WHITE
G. T. LORANCE
D. LYMAN
OTTO SANDVIK
GEORGE Nixoi
F. W. SEARS
To make recommendations and prepare specifications for the operation,
maintenance, and servicing of all types of studio and outdoor auxiliary
lighting equipment, tungsten light and carbon-arc sources, lighting-effect
devices, diffusers, special light screens, etc., to increase the general en-
gineering knowledge of the art.
M. A. HANKINS, Chairman
937 N. Sycamore Ave.
Hollywood 38, Calif.
J. W. BOYLE
W. E. BLACKBURN
RICHARD BLOUNT
KARL FRETJND
C. W. HANDLEY
TELEVISION
C. R. LONG
W. W. LOZIER
D. W. PRIDEAI
To study television art with special reference to the technical interre-
lationships of the television and motion picture industries, and to make
recommendations and prepare specifications for equipment, methods,
and nomenclature designed to meet the special problems encountered at
the junction of the two industries.
1948
COMMITTEES OF THE SOCIETY
321
TELEVISION (continued)
D. R. WHITE, Chairman
E. I. du Pont de Nemours and Company
Parlin, N. J.
R. B. AUSTRIAN A. N. GOLDSMITH
F. T. BOWDITCH T. T. GOLDSMITH, JR.
F. E. CAHILL HERBERT GRIFFIN
A. W. COOK C. F. HORSTMAN
E. D. COOK L. B. ISAAC
C. E. DEAN P. J. LARSEN
BERNARD ERDE C. LARSON
R. L. GARMAN NATHAN LEVINSON
FRANK GOLDBACH J. P. LIVADARY
P. C. GOLDMARK H. B. LUBCKE
THEATER TELEVISION
PIERRE MERTZ
H. C. MlLHOLLAND
W. C. MILLER
J. R. POPPELE
PAUL RAIBOURN
OTTO SANDVIK
G. E. SAWYER
R. E. SHELBY
E. I. SPONABLE
H. E. WHITE
To make recommendations and prepare specifications for the construc-
tion, installation, operation, maintenance, and servicing of equipment
for projecting television pictures in the motion picture theater, as well as
projection-room arrangements necessary for such equipment, and such
picture-dimensional and screen-characteristic matters as may be in-
volved in high-quality theater-television presentations.
D. E. HYNDMAN, Chairman
342 Madison Ave.
New York 17, N. Y.
P. J. LARSEN, Vice-Chairman
508 S. Tulane St.
Albuquerque, N. M.
NATHAN LEVINSON
OTTO SANDVIK
EDWARD SCHMIDT
E. I. SPONABLE
J. E. VOLKMAN
G. L. BEERS E. P. GENOCK
F. CAHILL A. N. GOLDSMITH
A. W. COOK T. T. GOLDSMITH, JR.
JAMES FRANK, JR. C. F. HORSTMAN
R, L. GARMAN L. B. ISAAC
A. G. JENSEN
TEST-FILM QUALITY
To supervise, inspect, and approve all print quality control of sound and
picture test films prepared by any committee on engineering before the
prints are released by the Society for general practical use.
F. S. BERMAN, Chairman
111-14 76th Ave.
Forest Hills, L. L, N. Y.
F. R. WILSON
C. F. HORSTMAN
322
COMMITTEES OF THE SOCIETY
THEATER ENGINEERING, CONSTRUCTION, AND
OPERATION
To make recommendations and prepare specifications of engineering
methods and equipment of motion picture theaters in relation to their
contribution to the physical comfort and safety of patrons, so far as can
be enhanced by correct theater design, construction, and operation of
equipment.
LEONARD SATZ, Chairman
132 W. 43rd St.
New York 18, N. Y.
FELIX ALEXA
HENRY ANDERSON
A. G. ASHCROFT
CHARLES BACHMAN
F. E. CARLSON
E. J. CONTENT
JAMES FRANK, Jr.
EVANS PERKINS
BEN SCHLANGER
SEYMOUR SEIDER
EMIL WANDELMAIER
SMPE REPRESENTATIVES TO OTHER ORGANIZATIONS
AMERICAN STANDARDS ASSOCIATION
Sectional Committee on:
Standardization of Letter Symbols
and Abbreviations for Science and
Engineering, Z10
S. L. CHERTOK
Motion Pictures, Z22
C. R. KEITH, Chairman
A. N. GOLDSMITH, Honorary Chairman
F. E. CARLSON E. K. CARVER
D. F. LYMAN
Sectional Committee on:
Acoustical Measurements and Termi-
nology, Z24
F. C. SCHMID
Photography, Z38
J. I. CRABTREE
Standards Council, ASA Member-
Bodies
D. E. HYNDMAN
R. M. EVANS, Chairman
INTER-SOCIETY COLOR COUNCIL RADIO TECHNICAL PLANNING
BOARD
P. J. LARSEN E. I. SPONABLE^
t Alternate.
AMERICAN DOCUMENTATION
INSTITUTE
J. E. ABBOTT
J. A. BALL L. E. CLARK
F. T. BOWDITCH A. M. GijNDELFINGER
M. R. BOYER H. C. HARSH
64th Semiannual Convention
SOCIETY OF MOTION PICTURE ENGINEERS
Hotel Statler • October 25-29 • Washington 6, D. C.
OFFICERS OF THE SOCIETY
LOREN L. RYDER ............................... President
DONALD E. HYNDMAN ........................... Past-President
EARL I. SPONABLE ........... .' .................. Executive Vice-President
JOHN A. MAURER ............................... Engineering Vice-President
CLYDE R. KEITH ............................... Editorial Vice-President
JAMES FRANK, JR ............................... Financial Vice-President
WILLIAM C. KUNZMANN ......................... Convention Vice-President
( 1 . TOEL LORANCE .............................. Secretary
RALPH B. AUSTRIAN ............................ Treasurer
New York, General Office
BOYCE NEMEC .................................. Executive Secretary
HELEN M. STOTE ............................... Journal Editor
SIGMUND M. MUSKAT ........................... Office Manager
DIRECTORY OF COMMITTEE CHAIRMEN
LOCAL ARRANGEMENTS
N. D. Golden, Chairman
REGISTRATION AND INFORMATION
W. C. Kunzmann, Chairman
Assisted by E. R. Geib and J. C. Greenfield
PAPERS PUBLICITY
G. A. Chambers, Chairman Harold Desfor, Chairman
J. E. Aiken, Vice-Chairman, Washington, D. C. Assisted by Leonard Bidwell
LUNCHEON AND BANQUET LADIES' RECEPTION
J. G. Bradley, Chairman Mrs. N. D. Golden, Chairman
HOTEL AND INFORMATION MEMBERSHIP AND SUBSCRIPTION
J. C. Greenfield, Chairman Lee Jones, Chairman
PUBLIC-ADDRESS EQUIPMENT PROJECTION PROGRAM — 16 MM
W. P. Button, Chairman R. B. Dame, Chairman
PROJECTION PROGRAM — 35 MM
H. F. Heidegger, Chairman
A. Pratt, Vice-Chairman
Assisted by officers and members of Washington Projectionists Local 224
SEPTEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 323
324 SEMIANNUAL CONVENTION September
GENERAL INFORMATION
Hotel Reservations and Rates
The Hotel Statler, Washington, D. C., will be the Convention Headquarters.
Room-reservation cards were mailed to the membership in August. These should
be checked to indicate the accommodations desired for the 64th Semiannual Con-
vention and returned to the hotel promptly, so that the hotel can book and confirm
room reservations.
Reservations are subject to change of arrival date or cancellation prior to Oc-
tober 10.
The following daily rates (European Plan) are extended SMPE members and
guests.
Single room, with tub and shower, $4.50 to $7.50
Double room, with tub and shower, $8.00 to $10.00
Twin beds, with tub and shower, $9.00 to $13.00
Parlor suites, with connecting bedroom, $17.50 to $26.50
Rail, Pullman, and Plane Travel
The Convention Committee suggests arranging travel accommodations at
least a month prior to the Convention, since travel conditions still remain acute,
especially into Washington, D. C.
Convention Registration and Papers Program
The Papers Committee can only function successfully in the early assembly,
scheduling, and release of the tentative and final Convention programs by re-
ceiving the title of paper to be presented, name of the author, and a complet •
manuscript mailed to one of the following vice-chairmen of this committee :
J. E. AIKEN N. L. SIMMONS
225 Orange St., S. E. 6708 Santa Monica Blvd.
Washington 20, D. C. Hollywood 38, Calif.
E. S. SEELEY H. L. WALKER R. T. VAN NIMAN
250 W. 57th St. P. O. Drawer 279 4431 W. Lake St.
New York 19, N. Y. Montreal 3, Que., Canada Chicago 24, 111.
The Convention business and technical sessions will be held in the Presidential
Ballroom of the hotel. Registration and Information Headquarters will be set up
in the Capitol Terrace, adjacent to the Presidential Ballroom. All persons at-
tending the Convention should register and receive their Convention badges, also
identification cards, which will admit them to all sessions held at, and away from,
the hotel. These cards also will be honored at the de luxe motion picture theaters
in Washington. Only through your registration can the Society derive the
revenue needed to defray the Convention expenses. Please co-operate. Conven-
tion Press Headquarters and headquarters of Harold Desfor, SMPK Publicity
Committee Chairman, will be located in the Continental Room.
1948 SEMIANNUAL CONVENTION 325
Special Meeting
It is expected that the Thursday evening meeting will be held at a naval station.
Tickets will be required for admission to this meeting, which can be obtained at
the time of registration. Noncitizens will be required to register on Monday if
they wish to attend this meeting. Busses will be available for transportation and,
due to naval security regulations, all those attending will be required to go on such
busses for which there will be a small charge.
Convention Get-Together Luncheon
The 64th Semiannual Convention Get-Together Luncheon will be held in the
hotel's Congressional Room at 12:30 P.M. on Monday, October 25. Although
there will be no technical session scheduled for that morning, Registration Head-
quarters will be open from 9:30 A.M. to noon in the hotel's Capitol Terrace, so
that you may register and purchase Luncheon and Banquet tickets.
Seating at the luncheon will be assured only if tickets have been purchased from
W. C. Kunzmann, who will be at the hotel several days before the Convention,
or at the Registration Headquarters prior to noon on October 25. Only through
your co-operation can the committee and hotel provide satisfactory accommoda-
tions for this function. Checks or money orders made payable to W. C. Kunz-
mann, Convention Vice-President, may be mailed to W. C. Kunzmann, c/o Hotel
Staffer, Washington, D. C., from October 18-25 for Luncheon and Banquet tick-
ets. Advance reservations should be picked up at Registration Headquarters.
Tickets for the Luncheon must be purchased in advance and there will be no re-
fund for tickets not used.
Luncheon and Banquet fees will be announced in the Convention bulletin, pub-
| lished in the October Journal.
Convention Social Cocktail Hour
The Convention Cocktail Hour for holders of Banquet tickets will be held in
the Hotel Statler on October 27 in the Congressional room.
Informal Banquet
The Convention informal banquet (dress optional) will be held in the Presiden-
tial Ballroom on the evening of October 27. At this time the annual Awards will
be presented, and there will be dancing and entertainment.
Table reservations should be made at the Registration Headquarters. No
tables for the Banquet will be reserved except for holders of tickets that have been
purchased before noon of October 27, and there will be no refunds for tickets not
Used.
Ladies' Headquarters and Registration
The Ladies' Reception and Registration Headquarters will be located in the
Potomac Room in the hotel, and open daily during the Convention dates, from
10:00 A.M. to 5:00 P.M. Mrs. Nathan D. Golden will serve the Convention as
Hostess to the visiting and local ladies attending the 64th Semiannual Convention.
The ladies' entertainment program will be announced in later released convention
bulletins.
326 SEMIANNUAL CONVENTION
Motion Pictures and Recreation
The identification cards issued to registered members and guests will be hon-
ored at the following motion picture theaters in downtown Washington: Loew's
Capitol, Loew's Palace, Metropolitan, RKO Keith, and Warner.
Literature and information will be available at the Registration Headquarters
on the many places of historic interest in Washington and vicinity. The Con-
vention recreational features will be released later by the local arrangements com-
mittee.
TENTATIVE PROGRAM
Monday, October 25, 1948
9:30 A.M. Registration, Capitol Terrace Room. Advance Sale of Luncheon
and Banquet Tickets
12 : 30 P.M. Get-Together Luncheon, Congressional Room
3 : 00 P.M. Technical Session, Presidential Ballroom
8:00 P.M. Technical Session, Presidential Ballroom
Tuesday, October 26, 1948
9:30 A.M. Registration, Capitol Terrace Room. Advance Sale of Banquet
Tickets
10:00 A.M. Technical Session, Presidential Ballroom
2 : 00 P.M. Technical Session, Presidential Ballroom
3:00 P.M. Business Session of the Society, Presidential Ballroom
3 : 30 P.M. Resumption of Technical Session, Presidential Ballroom
Open Evening
Wednesday, October 27, 1948
9:30 A.M. Registration, Capitol Terrace Room
10:00 A.M. Advance Sale of Banquet tickets
10:00 A.M. Technical Session, Presidential Ballroom
Open Afternoon
6:45 P.M. Cocktail Hour, Congressional Room
8:00 P.M. 64th Semiannual Banquet, Presidential Ballroom
Thursday, October 28, 1948
Open Morning
2:00 P.M. Technical Session, Presidential Ballroom
8:00 P.M. Technical Session, location to be announced later
Friday, October 29, 1948
10:00 A.M. Technical Session, Congressional Ballroom
2:00 P.M. Technical Session, Congressional Ballroom
5:00 P.M. Adjournment of the 64th Semiannual Convention
Section Meeting
Cleveland Meeting
The June 18, 1948, meeting of the Midwest Section was held in Cleveland, Ohio,
at the General Electric Lighting Institute at Nela Park. Seventy-five guests and
members attended this all-day affair and about twenty-five ladies attended the
dinner and a special evening program.
The doors were open at 9 : 30 A.M. and coffee was served during the registration
period.
At 10:00 A.M. the group was divided into smaller sections to visit the Sundeck
Gallery, Horizon House, store, office, and school. Expertly handled demon-
strations illustrated the methods used in creating lighting combinations which
were truly dramatic and functional as well.
At 11:30 A.M., R. T. Van Niman, chairman, called the meeting to order in the
Auditorium. F. T. Bowditch gave a resume of the Report of the Standards com-
mittee as presented in Santa Monica. Frank Carlson was called upon as the
representative of our host, the General Electric Lighting Institute, and welcomed
all to Nela Park.
Gordon Chambers reported on the Santa Monica convention. He began with a
series of Kodachrome slides which were projected on a large screen showing the
Del Mar Beach Club, the surrounding territory, and some prominent members of
the Society of Motion Picture Engineers. Mr. Chambers then summarized a few
of the papers presented at the 63rd Semiannual Convention. Also he reported
the high spots in the group of color papers which were presented on the coast.
"The Engineering Aspects of Drive-In Theaters," by George M. Peterson,
Cleveland, Ohio, was next presented by the author. It was revealed in this paper
that there are approximately 800 drive-in theaters in this country with average
capacity for 500 cars each. Mr. Peterson stated that many operators "build
their theaters- without any engineering assistance," a fact which he greatly de-
plored. Subjects covered in this paper were: traffic problems, grading, ramps,
sound circuits, surfacing, and screen building.
At 1:00 o'clock one half of the group visited the Automotive Lighting Labora-
tory, and the other half visited the Optical and Photographic Laboratories. After
' luncheon, which was served in the Cafeteria, the groups were reversed for visits
to the Optical and Photographic Laboratories and the Automotive Lighting
Laboratory.
The meeting was resumed in the Auditorium, with R. T. Van Niman presiding,
at 3:15 P.M.
"Practical Applications of New Photographic Techniques," by John Campbell,
vice-president, Jam Handy Organization, was supplemented by an 800-foot reel
of 16-mm pictures showing samples of the various techniques described in the
paper. "Light Sources for Television Studio Lighting," was given by Richard
Blount of the General Electric Company. M. D. Temple of the Brush Develop-
ment Company presented his talk, "Some Applications of Magnetic Recording
in the Motion Picture Field," from a reel of tape which was recorded a few days
previously in his living room and edited to match the series of slides which were
simultaneously projected. Boyce Nemec, executive secretary of the SMPE, gave
327
328
SECTION MEETING
September
R.T.VanNiman,
chief sound engi-
neer for Motio-
graph, chairman of
the Midwest Sec-
tion, and in charge
of the program, and
Frank E. Carlson,
General Electric
Lamp Department
illuminating engi-
neer and host to
the group.
Typical scenes at the General Electric Lighting Institute, Nela Park, where
the Midwest Section of the Society of Motion Picture Engineers held its June
meeting at Cleveland. Nearly seventy members were in attendance for the
day portion of the program.
Frank E. Carlson
lights a tiny grain-
of- wheat lamp with
a huge 50-kilowatt
lamp for R. T. Van
Niman and G. W.
Colburn, president
of the G. W. Col-
burn Laboratories
and secretary-treas-
urer of the Mid-
west Section.
1948 SECTION MEETING 329
a rather complete report on "Flicker in Motion Pictures; Further Studies," by
L. D. Grignon, Twentieth Century-Fox Film Corporation. This paper was pre-
sented at Santa Monica in May and was considered an important contribution to
tho art for design of future equipment.
At 5:15 P.M. the meeting adjourned for refreshments at the Coffee Bar for
members, guests, and their ladies. This was followed by dinner, which was served
hi the Managers' dining room. The only speech was by Mr. Van Niman leading
a rising vote of thanks to our hosts, the General Electric Lighting Institute.
At 7:00 P.M. a popular lecture by Alston Rodgers of the General Electric
Company called, "New Horizons in Lamp Research," was given. This was a
combination magic and vaudeville show with amazing stage props which was
highly entertaining and enlightening to engineer and layman alike.
"A Gearless, Sprocketless 8-Mm Projector," by Otto R. Nemeth, included a
demonstration of this new 8-mm projector following a discussion of engineering
features. This projector without gears or sprockets is driven directly from motor
to shutter and cam shaft with a belt. The lamp is 750 watts, lens //1. 6, one inch,
coated, and the mechanism is built into a self-contained carrying case with total
weight 12y2 pounds.
Mr. Nemeth then gave a brief description of "A Professional Wire Recorder
for Studio Use." The complete paper was presented at the Santa Monica Con-
vention. This machine features a magazine for handling the wire and automatic
threading.
"The Optimum Performance of High-Brightness Carbon Arcs," was next pre-
sented by F. T. Bowditch and M. T. Jones of the National Carbon Company.
The arc trim described is applicable to studio lighting. The 16-mm positive and
11-mm negative carbon holders are water-cooled jaws. Current at the crater is
about 450 amperes. The light output is in excess of 40,000 lumens.
"Tungsten-Filament Sources for Picture Projection," by D. A. Pritchard, of
the General Electric Company, dealt with photometric measurements of various
places in the optical system of a group of competitive projectors. The measured
results indicate output performance as a percentage of light output. The report
clearly indicated that peak performance may be obtained from standard equip-
ment if there is proper alignment of the tungsten filament.
"A Photometric Analysis of Picture-Projection Systems," by Edward E. Bickel
of the Simpson Optical Manufacturing Company, was comprised of a mathemati-
cal and geometrical analysis of factors limiting the light output of motion picture
projection systems. Based on mathematical values, performance results were
computed that compared with actual laboratory test results. The formulas given
establish limits beyond which it is physically impossible to go.
While the foregoing papers were presented, the ladies attended special demon-
strations at "Horizon House" by Aileen Page and "Color and Indoor Sunshine"
by Alston Rodgers.
Book Review
The Preparation and Use of Visual Aids, by Kenneth B. Haas and
Harry G. Packer
Published (1946) by Prentice-Hall, Inc., 70 Fifth Avenue, New York 11,.
N. Y. 218 pages + XII pages + 6-page index. 167 illustrations. 6x/4 X 9x/4
inches. Price, $4.00.
Unique in a long procession of recent publications in the field of the preparation
and use of audio-visual materials, this book provides a truly how-to-do-it ap-
proach. Tempered with enough of the philosophical to point out clearly the
strengths and advantages, in terms of the learning process inherent to the use of
the several mechanical divisions within the broad medium of audio-visual pres-
entation, detailed explanation continues to show how in the local training
situation the preparation of valuable teaching materials may be undertaken.
Designed primarily for use in personnel training, sales demonstrations, adult-
education programs, and advertising, the book should find a use or place in the
school professional library as well.
Stress is continually made that visual materials are to be regarded as necessary
supplementing experiences to good training programs. Too often the impression
is given that here is a "new broom." Rather, this book stresses the idea that
visual materials are not intended to displace but rather to improve and to supplement.
The authors, Packer and Haas, have very methodically organized the discussion
of the several audio- visual materials: motion pictures, filmstrips, slides, opaque
projection, flash cards, maps, charts, posters, manuals, photographs, the black-
board, the bulletin board, the field trip, objects and specimens, and television. In
each case they have ended the chapter considering the several materials with a
detailed "how to arrange it," "how to do it," set of instructions regarding pro-
jection equipment and the production of the materials to be projected.
The book is well illustrated and includes numerous sketches and photographic
examples to help the interested person to follow out the thinking and helpful ideas
stressed in the book — indeed a valuable addition to the growing literature in this
field.
W. A. WlTTICH
Bureau of Visual Instruction
University of Wisconsin
Madison 6, Wis.
FORTY YEARS AGO
Moving Pictures for Medical Students
In one of the New .York hospitals moving pictures have been made
of epileptic patients, as well as of persons affected with locomotor ataxia.
This is following the example set in Vienna, where moving pictures have
been made of celebrated surgeons performing critical operations. THe
purpose in both cases is, of course, to enable students and practitioners
to study the peculiarities of diseases and the methods of distinguished
operators.
— The Moving Picture World, April 18, 1908
330
Journal of the
Society of Motion Picture Engineers
VOLUME 51 OCTOBER 1948 NUMBER 4
PAGE
Improved Safety Motion Picture Film Support
CHARLES R. FORDYCE 331
Color-Television Film Scanner BERNARD ERDE 351
35-Mm Process Projector
HAROLD MILLER AND E. C. MANDERFELD 373
New Theater Loudspeaker System
H. F. HOPKINS AND C. R. KEITH 385
Modern Film Re-Recording Equipment
WESLEY C. MILLER AND G. R. CRANE 399
Motion Picture Research Council W. F. KELLEY 418
Use of 16-Mm Motion Pictures for Educational Reconditioning
EDWIN W. SCHULTZ 424
Report of Studio-Lighting Committee 431
Proposed 16-Mm and 8-Mm Sprocket Standards .' 437
Thomas Armat 441
Louis Lumiere 442
Thad C. Barrows 442
Book Reviews:
"Enlarging — Technique of the Positive," by C. I. Jacobson
Reviewed by Joseph S. Friedman 443
" Camera and Lens," by Ansel Adams
Reviewed by Lloyd E. Varden 443
"Informational Film Year Book 1947"
Reviewed by Glenn E. Matthews 444
Current Literature 445
Journal Exchanges 446
ARTHUR C. DOWNES HELEN M. STOTE GORDON A. CHAMBERS
Chairman Editor Chairman
Board of Editors Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930,
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOURNAL must be obtained in writing from the General Office of the Society.
Copyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC EXECUTIVE SECRETARY
OFFICERS
1947-1948
PRESIDENT
Loren L. Ryder
5451 Marathon St.
Hollywood 38, Calif.
PAST-PRESIDENT
Donald E. Hyndman
342 Madison Ave.
New York 17, N. Y.
EXECUTIVE VICE-PRESIDENT
Earl I. Sponable
460 West 54 St.
New York 19, N. Y.
EDITORIAL VICE-PRESIDENT
Clyde R. Keith
233 Broadway
New York 7, N. Y.
CONVENTION VICE-PRESIDENT
William C. Kunzmann
Box 6087
Cleveland, Ohio
SECRETARY
G. T. Lorance
55 La France Ave.
Bloomfield, N. J.
1948-1949
ENGINEERING VICE-PRESIDENT
John A. Maurer
37-01—31 St.
Long Island City 1, N. Y.
TREASURER
Ralph B. Austrian
25 W. 54 St.
New York, N. Y.
Governors
FINANCIAL VICE-PRESIDENT
James Frank, Jr.
426 Luckie St., N. W.
Atlanta, Ga.
John W. Boyle
1207 N. Mansfield Ave.
Hollywood 38, Calif.
David B. Joy
30 E. 42 St.
New York 17, N. Y.
1947-1948
Robert M. Corbin Charles R. Daily
343 State St. 5451 Marathon St.
Rochester 4, N. Y. Hollywood 38, Calif.
Hollis W. Moyse
6656 Santa Monica Blvd.
Hollywood, Calif.
William H. Rivers
342 Madison Ave.
New York 17, N. Y.
1948
S. P. Solow
959 Seward St.
Hollywood, Calif.
1948-1949
Alan W. Cook
4 Druid PI.
Binghampton, N. Y.
Lloyd T. Goldsmith
Burbank, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Section Officers and Office Staff listed on page 447.
R. T. Van Niman
4431 W. Lake St.
Chicago, 111.
Gordon E. Sawyer
857 N. Martel St.
Hollywood, Calif.
Improved Safety
Motion Picture .Film Support5
BY CHARLES R. FORDYCE
EASTMAN KODAK COMPANY, ROCHESTER, NEW YORK
Summary — Extensive experimental work on safety cine film support has
resulted in an improved product which offers possibilities for professional
motion picture use.
This product is a highly acetylated cellulose acetate with physical proper-
ties which are considerably different from those of ordinary commercial
cellulose acetate previously used. Certain improved physical characteristics
and improved aging properties of this base material are described in detail.
As a cine positive film support the high-acetyl cellulose acetate is shown to
give satisfactory behavior in printing, processing, and projection operations
and compares favorably with present standard release positive film.
Experimental studies on the use of the high-acetyl base for 35-mm nega-
tive film are described showing that this base will lend itself to use for nega-
tive materials. Particularly important is the fact that this base offers a very
low degree of shrinkage on long time keeping.
THE MOTION PICTURE industry has for many years employed two
types of film stock; one on cellulose nitrate base for professional
use, and the other on cellulose acetate base mostly in widths of less
than 35 mm. The requirement of safety for amateur film has made
the use of acetate necessary in this field, regardless of its comparative
qualities in other respects. An improved safety-base stock, made of
a cellulose acetate propionate was adopted by the Eastman Kodak
Company in 1937, and afforded physical properties midway between
those of cellulose nitrate and the former acetate. The character-
istics of these films were discussed in detail by Calhoun1 in 1944.
The cellulose acetate propionate base was an improvement over
cellulose acetate in many respects. It was less subject to brittleness
at low humidities, and more resistant to dimensional change by
moisture under varying conditions. During the war years of 1941 to
1945 this safety film gave very satisfactory service for many purposes,
including theater use for short subjects. . For rigorous professional
motion picture use, however, this product fell somewhat short of
requirements. Its comparatively low strength provided insufficient
* Presented May 17, 1948, at the SMPE. Convention in Santa Monica.
OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51 331
332 FORDYCE October
wearing qualities and it lacked necessary rigidity for screen steadiness
in projection with high-intensity lamps. For these reasons still
further improvements, particularly in strength and rigidity of the
base, were desirable.
In continued experimental work toward further improvements in
safety base it has been found that of all plastic materials which have
offered potential possibilities of better quality film support, the
product most promising was a cellulose acetate selected in the range
of higher degrees of acetylation than the product commonly used.
In the manufacture of cellulose acetate a hydrolysis step is usually
employed to remove part of the acetyl groups and provide a material
soluble in acetone. To gain this advantage in- solubility the product
is transformed from a strong, more rigid, very neat-resistant acetate
to one more plastic. This is adverse to the basic requirements of
good film support.
Cellulose triacetate, the product of complete acetylation of cellulose
is soluble in only a limited number of organic solvents, and would be
of doubtful success for motion picture film base because of the diffi-
culty of splicing. Furthermore, casting procedures are difficult with
this material, tending to give brittle film. By selecting an inter-
mediate chemical composition, within the range of 42.5 to 44.0 per cent
acetyl content, it has been found possible to retain the advantages of
high physical strength and at the same time eliminate the problem of
proper manufacturing quality and splicing behavior.
The chemical nature of these safety base materials is shown graph-
ically in Fig. 1. The trilinear chart is used to show the relative
chemical compositions of cellulose acetates and cellulose acetate pro-
pionates.2 Cellulose acetates of different acetyl contents lie along the
left boundary line of the triangle. Cellulose propionates are iden-
tified along the right boundary line. Within the area of the triangle
are mixed esters of acetic and propionic acid. Cellulose triacetate
(44.8 per cent acetyl) and cellulose tripropionate (51.8 per cent pro-
pionyl) are identified at points which mark complete esterification of
the cellulose. Point A identifies the commercial acetone-soluble cel-
lulose acetate used in safety base before 1937. The range B designates
the material used in the present new safety-base development. With-
in the area of the triangle the point C identifies the cellulose acetate
propionate used as Eastman Safety Base since 1937.
The new safety base has proved to be a useful improvement for
both 35-mm and narrow-width motion picture films and is being used
1948
SAFETY FILM SUPPORT
333
at the present time for some Eastman films. These include safety
release positive film, in both 16-mm and 35-mm widths, as well as the
32-mm width film, later to be converted to 16-mm product; fine-grain
duplicating positive film, type 5365, in both 35-mm and 16-mm
widths; sound-recording film, type 5373, in both 35-mm and 16-
mm widths; and high-contrast positive film, type 5363, in 16-mm
width. In addition, the new base is being used for certain profes-
sional 35-mm duplitized-color positive films.
CELLULOSE
CELLULOSE
TRIACETATE
ACETYL
CELLULOSE
- TRIPROPIONATE
PROPIONYL
Fig. 1 — Chemical composition of cellulose esters used in safety film base.
A, acetone-soluble cellulose acetate; B, high-acetyl cellulose acetate;
C, cellulose acetate propionate.
Careful evaluation of this base for 35-mm film indicates it should
be suitable for professional motion picture positive and negative
stock, for which cellulose nitrate base is now employed. The results
of these tests are summarized in the following experimental sections.
PROPERTIES OF FILM BASE
Consideration of the physical properties of the film base will serve
as a comparison of the high-acetyl cellulose acetate support with
nitrate and safety products which have been in standard use (Table
I) . In these measurements both lengthwise and width wise directions
of the support have been included because of the slight difference in
these two directions. In certain cases such differences may be of
importance.
334
FORDYCE
October
B
g.tf
<1 o
H ^
S^
£ M
QQ g
3-6
2*
0 0
o
Machin
Directio
' OOOOOO
•<tf T^ iO 10 CO «O
.i
TO jo
I | i
lit
>^ -2 *S
•g a
llf
S a J
O) OJ bC
< ^ s
<N HO
«x iJ t^
b£
.ss
"61
a) x
a
03 COOO^
0000
oooo ^ ooo
fc I + I
o OOO
QOQOo:
X
o
rH I
§8g
ood
888 888
^^OJQJ aT^aJaj
O03-^^j bfiw-*^-^
^^ 3 g _^x5 2
bCO-3 f-±bca)-S
02
1948
SAFETY FILM SUPPORT
335
The tensile strength of the high-acetyl base is considerably im-
proved over the former safety base and approaches that of the nitrate,
indicating a corresponding increase in the mechanical wearing
quality of the material. Flexibility is of importance in general
handling behavior, and is in the same range for all three products.
Tear values of the high-acetyl base are somewhat below those of
cellulose nitrate and acetate propionate, and may be the cause of
some concern if this property should prove to be critical. Young's
modulus is a measure of the stiffness and rigidity of the support and is
100
80
2 60
40
20
B\.
Q)
s*<
f ^HIGH ACETYL SAFETY
^» ACETATE PROPIONATE SAFETY
10
20
25
30
TIME. DAYS
Fig. 2 — Viscosity retention of film base at 100 degrees centi-
grade.
of importance in resisting temporary or permanent deformation. The
high-acetyl base here shows appreciable improvement over the former
safety base, and is somewhat inferior to the nitrate. Cold flow
characteristics represent the tendency of the material to undergo
permanent deformation under stress. Here again, the high-acetyl
base lies midway between the other two products.
A further evaluation of the film base may be made by testing its
permanence under accelerated aging. These tests involve heating
samples for periods of time at temperatures above those which
ordinarily would be met in standard use with the assumption that
this severe condition will predict their behavior for much longer
periods of time in normal use. No quantitative relation between
336
FORDYCE
October
accelerated and normal keeping times can be given. It may be
pointed out, however, that a National Bureau of Standards test for
Archive films8 employs an incubation time of three days at a temper-
ature of 100 degrees centigrade.
The chemical stability of cellulose derivatives is best measured by
their resistance to viscosity degradation. Samples of film support
may be incubated at elevated temperatures for periods of time after
which they may be dissolved in suitable solvents and the viscosities
compared with those of the same material before heating. Chemical
deterioration results in loss of viscosity which is proportional to the
degree of degradation. Viscosity curves of safety and nitrate film
0 5 10 15 20
TIME. DAYS
Fig. 3 — Retention of flexibility of film base at 100 degrees
centigrade.
support upon heating at 100 degrees centigrade for increasing periods
of time are given in Fig. 2. It will be noted that the safety bases
undergo this treatment with no appreciable chemical degradation,
while cellulose nitrate rapidly and progressively decreases in vis-
cosity. This comparative behavior is well known, and illustrates the
possibilities of distinctly superior keeping qualities of cellulose ace-
tate safety base.
Flexibility retention in accelerated aging tests at 100 degrees centi-
grade is shown in Fig. 3. Very little loss in flexibility up to 30 days
at this temperature has resulted in either safety base, from which it
may be pre:licte 1 that very long times, under standard storage
conditions, should be possible without difficulty. Cellulose nitrate
1948
SAFETY FILM SUPPORT
337
support dropped rapidly in flexibility at this temperature, and became
completely brittle within ten days.
As an indication of the retention of tear strength the curves of Fig.
4 give measurements of tear values of the three film bases after in-
creasing periods of incubation. Although the high-acetyl safety
base has somewhat lower initial tear values than the other products,
the fact that there is little or no loss in tear strength under this very
severe incubation indicates probable satisfactory behavior in film use.
Cellulose nitrate support again deteriorates rapidly under this
treatment.
80
I
ACETATE PROPION/
TE SAFET1
\
D
0
T~r
\ NITRATE
\
\
HIGH
ACETYL
SAFETY
(SHATTERS)
10
15
TIME. DAYS
20
25
30
Fig. 4— Retention of tear strength of film base at 100 de-
grees centigrade.
An important property of motion picture film is its permanence of
dimension upon aging. Results of accelerated shrinkage tests on the
base at 100 degrees centigrade are shown in the curves of Fig. 5. It
will be noted that the high-acetyl base here exhibits a lower order of
shrinkage than either the safety or nitrate standard materials. Be-
cause of the severe temperature used in this test, a second series of
shrinkage measurements were carried out at 71 degrees centigrade
(160 degrees Fahrenheit) to confirm this shrinkage behavior (Fig. 6).
Here again, the high-acetyl base exhibited a very low order of shrink-
age as compared with the other materials. From these character-
istics it may be predicted that the experimental base should give film
of excellent aging shrinkage properties.
To summarize the physical properties of the base materials, the
high-acetyl cellulose acetate is an improvement over former safety
338
FORDYCE
October
TIME. DAYS
Fig. 5 — Rate of shrinkage (lengthwise) of motion picture posi-
tive film base at 100 degrees centigrade.
base in most of its properties, and particularly in tensile strength and
rigidity, which are most needed. Compared with cellulose nitrate,
most properties are somewhat lower in original measurements, but
permanence tests show that there is very little change in quality even
under severe aging tests. Perhaps the most unique property of the
U£3
1.00
0.75
0.50
0.25
n
A
CETATE PRC
>PIONATE S*
FETY
D<
^v
,^--*D"
^A'
NITRATE
. A-
— o '
HIGH ACETYL SAFETY
. -o •
L
10 15 20
TIME. DAYS
25
30
Fig. 6 — Rate of shrinkage (lengthwise) of motion picture
positive film base at 71 degrees centigrade (160 degrees
Fahrenheit).
1948 SAFETY FILM SUPPORT 339
high-acetyl acetate is its exceptionally low shrinkage as compared with
materials previously in use.
PROPERTIES OF CINE FILM
Proper evaluation of the experimental film base for cine products
requires consideration of its use as both positive and negative stock.
In the case of positive films, processing, printing, and projection be-
havior should be considered in detail. Negative products introduce
the factors of camera behavior and proper shrinkage characteristics
as additional important qualities.
Positive 85-Mm Film
Testing of positive 35-mm films was done by comparison with stand-
ard safety release print stock (5302) on cellulose acetate propionate
base and nitrate release print stock (1302).
Processing Tests — An important factor in processing motion
picture film is the degree of swelling which takes place in the develop-
ing step. If the longitudinal swell is too rapid or too great some
processing machines encounter trouble from excessive slackness which
allows the film to become displaced on the bottom rollers. Likewise,
excessive swelling during development may result in correspondingly
excessive shrinkage at the beginning of the drying operation. This
has been a difficulty with previous safety films, not only because of the
magnitude of swelling, but also because of the very rapid rate of
shrinkage of the safety film upon drying, causing rapid building up of
tension in a critical area. Of importance also is any tendency which
the film may have to curl too highly negative (away from the emul-
sion) at the beginning of drying, or too highly positive (toward the
emulsion) when completely dry.
Results of preliminary processing tests are summarized in Table II.
The curl characteristics of the high-acetyl film will be seen to be of the
same order as the standard check materials. A slight positive curl
takes place after fixing which changes to a slight negative curl at the
beginning of the drying operation. This changes again to a positive
curl when leaving the drying cabinet, which returns to a very slight
positive curl at the time of rewinding. The development shrinkage
is also of small magnitude, and characteristic of present product.
The lengthwise swelling of the experimental film during processing is
0.28 per cent. This is slightly greater than the type 1302 nitrate film,
but less than the acetate propionate safety 5302 material.
The swelling and shrinkage behavior of these films may be seen in
340
FORDYCE
October
more detail in Fig. 7. Upon immersion in water the lengthwise
swelling takes place quite rapidly for about 30 minutes then ap-
proaches a maximum value. The rates of shrinkage upon drying
after 30 minutes' swelling are shown by descending curves. It will
be seen that during the first ten minutes of drying the high-acetyl
film shrinks 0.14 per cent, the nitrate 0.07 per cent and the acetate
propionate safety 0.31 per cent. Thus, while the high-acetyl film
exhibits an appreciable amount of swelling its comparatively slow
TIME. MINUTES
Fig. 7 — Rates of swell in water at 68 degrees Fahrenheit (heavy
curves) and rates of shrinkage upon drying at 70 degrees Fahrenheit,
50 per cent relative humidity (light curves) for positive films.
rate of shrinkage upon drying tends to reduce the possibility of
excessive operating tensions.
Following these preliminary measurements extensive processing
tests were carried out in two commercial East Coast Laboratories
each test involving several thousand feet of film, to insure in so far as
possible that the test represented stable continuous processing
behavior. In these tests the experimental product proved to be
satisfactory, giving no "indication on any of the machines of greater
tension than normal (Table III). Likewise, in none of the tests did
the swelling during development cause difficulty from slackness.
Projection Tests — Evaluation of the projection quality of 35-mm
positive film is probably the most important factor in the testing
program. Many characteristics must be considered, and can only be
1948 SAFETY FILM SUPPORT 341
TABLE III
PROCESSING OF HIGH-ACETYL 5302 (35-Mai) FILM
Tests in Commercial Laboratories
Machine
Film
Threading
Approximate Curl at
Length,
Drying
Feet of
Rewind,
Laboratory
Feet
Conditions
Test
Inch
Difficulties
A
1750
77 °F.— 42% R.H.
21,000
+0.09
None
B
2120
63 °F.— 60% R.H.
12,000
+0.14
None
determined by experimental projection under conditions of actual use.
Preliminary tests of a laboratory nature were made in this investiga-
tion, and were followed by trade tests involving prints of several
commercial feature pictures issued through selected film exchanges in
different parts of this country.
Preliminary laboratory tests for physical behavior of the film in-
volved continuous projection of short lengths of film for increasing
lengths of time, followed by examination of the film for perforation
damage and general appearance. A summary of the results is given
in Table IV. It will be noted that slight perforation damage began
to take place after about 200 projections for both the high-acetyl and
the nitrate films, as compared with 100 projections for the acetate-
propionate safety film. This became progressively more severe on all
products until failure by complete perforation breakdown at 520 runs
for the high-acetyl film as compared with 380 runs for the acetate-
propionate safety and 644 for the nitrate. It should be emphasized
that the numerical values of runs before failure are of significance only
for comparative purposes, and do not necessarily indicate the number
of runs to be expected in trade use.
Another type of laboratory test involved projection of rolls of the
three types of film on a Simplex E-7 projector, with a projection
throw of 157 feet to a screen 30 feet by 40 feet. Initial tests with non-
rotating positive high-intensity arcs up to 65 amperes in mirror
optical system lamps resulted in entirely satisfactory performance of
all three types of film. A more severe test was then undertaken,
using a rotating positive high-intensity arc (13.6-mm positive carbon)
at 175 amperes in a condenser optical system lamp and employing
an Aklo No. 3966 heat-absorbing glass filter. Certain charac-
teristic differences in the films became evident in this test, as recorded
in Table V. The acetate propionate safety 5302 film here showed
342
FORDYCE
October
TABLE IV
WEARING QUALITY OF MOTION PICTURE POSITIVE FILM
Times
Projected
High-Acetyl
Acetate 5302
Acetate Propionate
5302
Nitrate 1302
100
A
B
A
200
B
C
B
300
C
D
B
400
D
Failure (380)
C
500
D
D
600
700
Failure (520)
••
D
Failure (644)
Condition of Film
A — No perforation damage
B — Damage in one perforation in a frame
C — Damage in two perforations in a frame
D — Damage in three perforations in a frame
considerable unsteadiness in focus,4 and rather severe embossing
after projection. The high-acetyl safety film and the nitrate 1302
were satisfactory and nearly identical in behavior. Upon examina-
tion after projection the high-acetyl film showed somewhat less
embossing effect than the nitrate.
On the basis of the above background, indicating satisfactory
behavior of the high-acetyl film for commercial use in regard to both
TABLE V
LABORATORY PROJECTION QUALITY OF CINE POSITIVE FILM
Arc Intensity: 175 Amperes
High-
Acetyl
5302
Acetate
Propionate
5302
Nitrate
1302
Screen Quality
Original sharpness and definition O.K. O.K. O.K.
Focus drift Normal Excessive Normal
Tendency to image flutter Slight Slight Slight
Tendency to in-and-out of focus Slight Excessive Slight
Film Appearance
Frame embossing Very slight Appreciable Slight
Image embossing Very slight Appreciable Very slight
1948 SAFETY FILM SUPPORT 343
processing and projection behavior, it was decided to undertake
trade tests with prints released in the regular manner for theater use.
In these tests, which included four different features and a total of 22
experimental prints, each print was assembled with approximately
half of the reels on the experimental stock and half on standard
nitrate. The first two reels in each case were of one type of stock and
the next two of the other, and so forth, to insure that each material
would be used on both projectors in any theater.
Throughout these tests no difference was noted between any of the
experimental and standard reels as regards condition of focus,
steadiness on the screen, or general quality of either picture or sound.
Likewise no detectable difference was noted in the tendency toward
scratching. The conditions of the prints after completion of their
trade use are summarized in Table VI. All films were comparable
throughout in curl and in brittleness at low humidity. Shrinkage of
the experimental film was consistently less than that of the nitrate
stock. Likewise the tendency of the film to become embossed or buckled
after long use was noted to be less in the experimental film. In per-
foration damage, no marked differences were evident, although the
experimental films showed somewhat greater damage in areas of
severe wear. This is in agreement with the preliminary projection
tests, which indicated a slight advantage in mechanical wearing
quality for the nitrate film, but is believed to be due in part also to
the higher shrinkage characteristics of the nitrate, which give that
material the advantage of more nearly fitting the projector sprockets.
The lower shrinkage values noted for the experimental film in these
practical use tests are in agreement with the predicted behavior ob-
served in the accelerated shrinkage tests (Figs. 5 and 6) and have been
further confirmed by laboratory keeping tests of processed film under
normal conditions (Fig. 8). Here the high-acetyl film will be noted
to undergo a shrinkage of 0.20 per cent in one year as compared with
0.29 per cent for standard nitrate and a higher value of 0.46 per cent
for film on cellulose-acetate propionate safety base.
It may be well to point out in connection with these trade tests that
they were carried out while all theaters were using 0.935-inch-diam-
eter intermittent sprockets. Because of the recently adopted
change in standard to the 0.943-inch-diameter sprocket it may be
expected that the new sprockets will soon replace the former in most
theaters. This will be an advantage to films with low shrinkage
characteristics, such as this experimental material, and should offer
344
FORDYCE
Oc1
an improvement in wearing quality for high-acetyl cellulose acetat
film even greater than that anticipated for film on nitrate base.
Splicing — The question of proper splicing behavior is one of im-
portance for motion picture films. It was pointed out that the chem-
ical composition of the cellulose acetate used in this film base is in a
range of very limited solubility in organic solvents. This fact limits
the formulation of effective cement mixtures to carefully chosen
solvents, properly balanced to give good results with this specific
product. For this reason it should not be expected that film cements
designed for products previously in use should give good performance.
0.6
0.4
TIME, MONTHS
Fig. 8 — Rate of shrinkage (lengthwise) of motion picture posi-
tive films. Individual developed strips were stored, freely ex-
posed to circulating air, and shrinkage values calculated from
the initial dimension of the raw stock.
It has been demonstrated, however, that properly formulated
cements can be made without great difficulty, and suitable cements
are now available for the purpose. With these the cementing prop-
erties of the new film are quite similar to those of other types of film
with cements commonly "used for them.
Proper splicing behavior must in all cases be qualified with the re-
quirement that the emulsion be scraped properly from the film support
in the area to be cemented. It is essential here to remove the bonding
sublayer beneath the emulsion so that the cement solvents will have
sufficient opportunity to attack the film base. This removal of sub-
layer is somewhat more critical on safety than on nitrate base, and
SAFETY FILM SUPPORT
345
c •»—
1:1
O T3
II
QJ Q^
*!
*C o
pq ^
££
•42 -3
^ '3
o 'C
^ PH
O O
0 0
o q
d d
I I
o
1 I
0 0
PQ
«
CO CO
d d
o o
I I
1 I
•
PQ
O pq
33
GO CO
1-1 CO
CO 00
<N <N
d d
o o
I I
1 1
0 0
CD 43
O
O pq
33
pq <!
§§
d d
o o
I I
1!
g >
r-3 §
-a
O §
'o O
o £
||
0) O
Jl p
03 "S
"
<U rj
^ O
pq PH
346 FORDYCE October
should be understood properly by the operator. Duplitized cine films
which carry emulsions on both front and back surfaces of the film,
should be recognized as far more difficult to splice than single-
coated films, and should receive special consideration in removal of
sublayer coatings before cement is applied.
Negative 35-Mm Film
In general, mechanical properties which are necessary for positive
film are also advantageous for negative. Its use on continuous
printers, however, demands proper shrinkage characteristics to
provide the necessary range of perforation pitch to give good printing
quality. It is well understood that standard processing laboratory
practice requires that negative films which are manufactured with
the standard perforation pitch as raw stock undergo a shrinkage of at
least 0.20 per cent for use as a negative on continuous drum printers.5'6
A satisfactory range for good printing quality is usually considered to
be a shrinkage of between 0.2 and 0.4 per cent. If shrinkage should
exceed this range, however, the pitch becomes too short and again
results in unsatisfactory prints.
These shrinkage characteristics of negative film have been controlled
during manufacturing by allowing a small but controlled amount
of solvent to remain in the support. As this escapes from the film
during processing and subsequent storage of one to three months a
corresponding shrinkage takes place.
A better way of meeting the requirements of negative film might be
to employ a base of very low shrinkage properties and to change the
standard of perforation so that the pitch will be optimum for printing
throughout the life of the negative. For this to be successful tjie
shrinkage upon aging would have to be exceptionally low, to main-
tain good printing quality on long keeping.
The low shrinkage characteristics of the experimental high-acetyl
acetate base presented possibilities for such a product. To test this
a set of experimental films was specially prepared, including both
normal shrinkage and .low shrinkage base, perforated to both stand-
ard and optimum pitch dimensions. These products are tabulated
in Table VII. Sample A was made to correspond to standard nitrate
negative (1231) in its shrinkage characteristics while samples B and C
were made to represent low shrinkage films. This is shown by the
"accelerated aging" measurements, which represent the degree of
shrinkage which would normally take place over a considerable
SAFETY FILM SUPPORT
347
9.
00 00 00 O
^1 i-H ^H CO
dodo
CO CO CO (N
dodo
12 o o 2
dodo
o o
oo
ll
tbC M ^bfl -ij
<1 PQ O P
1
s
I
s
•8
period of time. Sample B was per-
forated with a pitch of 0.1869 inch
and sample C with a shorter pitch
of 0.1866 inch to correspond to a
shrinkage of 0.20 per cent.
Test rolls of each sample of nega-
tive film were exposed in a Mitchell
camera and were processed by
standard procedure. These were
used to make prints on standard
fine-grain release positive (1302)
stock shortly after processing and
at repeated intervals of time over a
period of several months. Results
of printing quality were found to
be in agreement with pitch values
at the time of printing. The re-
sults of shrinkage behavior are
shown in Fig. 9. It will be seen
that sample A reproduces in general
the shrinkage behavior of the stand-
ard (1231) negative, sample D.
Both of these negatives gave some-
what unsteady prints when printed
soon after processing, but were
satisfactory on later tests, by
which time shrinkage had resulted
in shorter pitch measurements.
Sample C, on the other hand, cor-
responded to a shrinkage of 0.20
per cent when freshly processed
because of perforation adjustment,
and as a result gave immediate
satisfactory printing quality. Also,
because of its low shrinkage, it
continued to give good printing
quality after long keeping. As
would be expected, the sample B
with standard pitch perforations
on low shrinkage base gave
348
FORDYCE
October
unsteady prints the first three months of the test period.
These results indicate the possibility of obtaining a satisfactory
safety cine negative film either by standard perforation of stock
duplicating present negative film in shrinkage characteristics, or by a
properly adjusted perforation of stock employing low shrinkage base.
The latter product may prove to be attractive to the industry be-
cause of its improved permanence characteristics and because of
possible improved printing quality immediately after processing as
well as after long aging.
-0.6
TIME, MONTHS
Fig. 9 — Change in longitudinal pitch of developed motion picture negative
films stored in rolls in untaped cans. Film rewound daily for first 30 days.
A, high-acetyl safety base, normal negative shrinkage, initial pitch 0.1869
inch; B, high-acetyl safety base, low shrinkage, initial pitch 0.1869 inch;
C, high-acetyl safety base, low shrinkage, initial pitch 0.1866 inch; D, ni-
trate base, normal negative shrinkage, initial pitch 0.1869 inch.
CONCLUSIONS
1. An evaluation of an improved safety motion picture film sup-
port, made from a high-acetyl type of cellulose acetate, has been
presented.
2. The general physical characteristics of the high-acetyl acetate
base are superior to those of former safety base materials, and are in
the range of base from cellulose nitrate.
3. The properties of the new base have been confirmed by com-
mercial tests of positive film in which satisfactory quality for pro-
fessional motion picture use was obtained.
1948 SAFETY FILM SUPPORT 349
4. Laboratory tests on negative film have been carried out, which
indicate probable satisfactory behavior. Low shrinkage character-
; istics of the high-acetyl acetate base offer the possibility of improved
i printing characteristics under proper conditions.
ACKNOWLEDGMENT
The author wishes to express his appreciation for the assistance and
I helpful suggestions received from many members of the Department
of Manufacturing Experiments of the Eastman Kodak Company in
the preparation of this paper.
-
REFERENCES
(1) J. M. Calhoun, "The physical properties and dimensional behavior of
motion picture film," J. Soc. Mot. Pict. Eng., vol. 43, pp. 227-267; October, 1944.
(2) C. J. Malm, C. R. Fordyce, and H. A. Tanner, "Properties of cellulose
esters of acetic, propionic, and butyric acids," Ind. Eng. Ghent., vol. 34, p. 430;
April, 1942.
(3) J. R. Hill and C. G. Weber, "Stability of motion picture films as deter-
mined by accelerated aging," Research Paper RP 950, Jour. Res. Nat. Bur.
Stand., vol. 17, p. 871; December, 1936.
(4) E. K. Carver, R. H. Talbot, and H. A. Loomis, "Effect of high-intensity
arcs upon 35-mm film projection," J. Soc. Mot. Pict. Eng., vol. 41, pp. 69-88;
July, 1943.
(5) R. H. Talbot, "Some relationships between the physical properties and the
behavior of motion picture film," J. Soc. Mot. Pict. Eng., vol. 45, pp. 209-218;
September, 1945.
(6) J. Crabtree, "Sound film printing," /. Soc. Mot. Pict. Eng., vol. 21, pp. 294-
323; October, 1933.
DISCUSSION
CHAIRMAN W. V. WOLFE: Dr. Fordyce, I understand that your company is
now putting out a universal film cement which is good for the old acetate and the
new safety-base film. Is that correct, and do I also understand that you are no
longer making available the old nitrate-base film cement?
DR. CHARLES R. FORDYCE: The first part of that statement is correct. We
are putting out a film cement which we call Universal Cement for all types of film.
I think I am right in stating that you are also correct in the second statement that
we no longer supply the other, but I might be wrong.
CHAIRMAN WOLFE: I particularly ask that question because there has been
some comment about the possibility that the Universal Cement was not so good
for a nitrate-base film as the old nitrate film cement.
DR. FORDYCE: Yes, that is a question. Of course, as you know, in testing
film cements, it is quite difficult to get more than two people to test them in the
same way, and I do not really know what a majority opinion would be on the
basis of these two cements. We like the newer cement in the tests we have run,
and in our use, but maybe some of the trade would rather have the other cement.
350 FORDYCE
DR. C. R. DAILY: Some years ago, we ran a series of pitch checks on release
negative, starting with the first and printing from that negative, and then measur-
ing successively 25, 50, and up to 300 prints during the release life of that negative.
During that time the negative was aerated through the printer and rewound that
many times, and pitch checks were made throughout the entire run to determine
properties posted in the last slide that you showed, and to help determine the
matter of initial pitch. Have you made live tests through a release print cycle for
the aeration and pull on the negative to determine how the acetate film compares
with nitrate?
DR. FORDYCE: We have given you all the shrinkage data that we have; in
other words, keeping tests, but not tests made at intervals during actual use of
that negative. Our data are only laboratory incubation tests. I think quali-
tatively that actual use data would have the same trend as our laboratory tests.
MR. K. B. LAMBERT: We have used some of this film recently, both with the
long perforation and some specially perforated. We have had very successful
results from the shorter perforations. With the longer perforations, we con-
stantly encountered unsteadiness on all of the early prints. In fact, we never got
away from it until we short-perforated the film.
The Research Council is at this time considering the possibility of recommend-
ing, in conjunction with Eastman, the shorter perforation of this type of film, be-
cause we are faced here with the problem of making the highest quality prints
first, immediately after the negative is processed; then perhaps the negative lies
idle for quite a while and some more prints are made, but if you can have a nega-
tive which can be printed over a very long period of time, and have both the first
and last prints all good, it would seem to be very advantageous.
MR. G. J. BADGELY: As Eastman people are pretty well aware, we are inter-
ested in high-temperature development. What is the action of this film when it
is subject to developing temperature of 125 degrees?
DR. FORDYCE: It is going to be a problem connected with the emulsion. We
can say that this base is more resistant than former safety base but the degree of
swelling in water is more than nitrate; if you increase the temperature, you have
a higher amount of swelling. So far as I know, it is not high enough to interfere.
QUESTION: Would there be any elasticity added to the film as a result of the
higher temperature? Has it a tendency to stretch under strain?
DR. FORDYCE: Yes, the higher the temperature, the softer it will be. In other
words, it will be more easily distorted at the higher temperature but not so much
so as prior safety films.
DR. J. G. FRAYNE: It seems a shame that we are still worrying about printing
from antediluvian sprocket- type printers. Instead of spoiling a fine job, why does
not the whole industry do something about the printer situation?
CHAIRMAN WOLFE : Dr. Fordyce cannot very well answer that question, nor
for that matter, can anyone else answer for the industry. We agree with you, Dr.
Frayne, that it would be a good idea to improve our printers as they stand today.
Color-Television Film Scanner*
BY BERNARD ERDE
COLUMBIA BROADCASTING SYSTEM, NEW YORK 22, NEW YORK
Summary — The transformation of moving color-film images into video
signals is accomplished with the most faithful rendition when the pickup tube
is of the continuous-cathode, or nonstorage type. This, however, imposes the
limitation that the film motion, in the associated film scanner, be constant in
velocity, rather than intermittent. In the color-film scanner of the
Columbia Broadcasting System, the pickup tube is the Farnsworth daylight
image dissector. An optical-electronic method, requiring no moving optical
parts, is used to compensate for the continuous motion of the film.
THE COLUMBIA Broadcasting System postwar system of color tele-
vision was put into operation in January of 1946. At first, the
color-television pictures had their origin in 16-mm color film and 2-X
2-inch color slides. In the spring of the same year, the live pickup
camera and equipment were completed and put into use.
Commencing with a brief review of the basic characteristics of the
entire system, the remainder of this discussion will concern itself with
a description of the methods involved and problems encountered in
scanning the color film and slides. Particular attention will be paid
to the design and function of the optical, mechanical, and electronic
equipment involved in the process of transforming moving color-film
images into video signals.
The fundamental property of the system is one of sequential,
additive-color scanning, in which the subject matter is analyzed into
three primary-color impulses of varying amplitudes following each
other in sufficiently rapid succession to be integrated by the observer's
eyes. Rotating color disks, one in front of the pickup device and one
in front of the receiving tube, properly synchronized and phased,
produce the color analysis at the transmitter and the color synthesis
at the receiver. The color images are scanned horizontally in 525
lines, interlaced 2 to 1, and interlaced fields are scanned vertically at
the rate of 144 per second. Each field, of Vi44-second duration, is
scanned and reproduced in succession through a different primary-
color filter, so that the three color fields are presented to the viewer
in y48 of a second, a sufficiently short interval of time to allow the
* Presented May 10, 1946, at the SMPE Convention in New York.
OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51 351
352 ERDE October
eye's property of persistence of vision to give an apparent fusion of
the separate colors into their resultant additive mixture.
The film pickup tube is the Farnsworth image dissector with a day-
light photoelectric characteristic. Its property of nonstorage photo-
emissivity makes the dissector particularly suitable for color use,
since there is no stored charge on the unscanned interlaced lines to be
carried over from one color field to the next to produce spurious color
values. Another distinct asset is its inherent freedom from shading
effects. These two important characteristics of the daylight dis-
sector account, to an enormous degree, for its fidelity of color
rendition.
The light source is a Peerless Hy-Candescent high-intensity carbon
arc operated at 175 amperes. This is necessitated by the low sen-
sitivity of the dissector and by the transmission loss in the color
filters. Although the daylight dissector has been especially adapted
for color work by improved response in the visible spectrum, it still
has appreciable sensitivity in the near infrared. Since the red, blue,
and green filters transmit freely in the infrared, this unwanted
radiation must be removed by suitable filters if color contamination
is to be avoided. A water cell containing a disk of heat-absorbing
glass of the desired characteristic is inserted in the carbon-arc beam
between condenser and gate aperture and is effective in transmitting
a high ratio of visible light to total radiant energy. For valid color
reproduction it has been found that the proportion of infrared
response must not exceed more than 5 per cent of the maximum signal
in the dissector output. However, since dissector tubes vary some-
what in their spectral response, provision has been made to insert
elsewhere in a cooler part of the light beam an additional heat filter,
if necessary, to bring the infrared content below the permissible
maximum level.
OPTICAL SYSTEM
Of greatest interest, perhaps, will be the optical system. Before
describing this, it may be well to discuss first the factors instrumental
in determining the choice of system to be used.
The image dissector, as has been mentioned before, is a tube with an
instantaneous photoemissive, or nonstorage, type of cathode. As such
it imposes one basic limitation upon the type of projection equipment
to be used with it; that is, the film cannot be intermittently projected
upon the cathode. This restriction is caused by the shortness
1948 COLOR-TELEVISION FILM SCANNER 353
of the vertical blanking interval; Vi44o of a second. Whereas, with a
storage type of tube the film can be projected during this short
interval of time and then at relative leisure be pulled down in the film-
gate aperture during the remaining 9/i44o of a^second, with a dissector
tube the film would have to be treated in the converse manner,
i.e., projected during the 9/i44o of a second scanning period and then
pulled down during the 1/i44o of a second blanking period. This,
considering the masses and acceleration of the film and pulldown
mechanism involved, is an impracticably short period of time and
obviously rules out all consideration of intermittent projection.
There are several well-known methods of continuous projection
some of which have undoubted merit. l> 2 All involve the use of one or
more optical elements, i.e., lens, mirror, or prism, actuated or driven
by cam, wheel, or drum, to compensate for the displacement of the
continuously moving film. It was felt desirable to avoid the use of a
number of precisely adjusted optical elements driven by precisely
machined mechanical members. It was also felt that most of the
optical errors inherent to these types of optical displacement com-
pensation, or inherent to the means employed to actuate them could
be avoided. A method of optical compensation was employed which
had earlier been used successfully at CBS in the transmission of
black-and-white film.3 The optical elements are six in number and
consist of segments of simple achromatic doublets. They are entirely
stationary, easily adjusted, and remain permanently fixed in position.
The only rotating component is purely mechanical in function — a
rotating slotted selector disk which exposes the lens segments one at a
time. Since the electronic scanning process, as will be pointed out, is
also instrumental in offsetting the movement of the film, this may be
termed an optical-electronic method of film-movement compensation.
Since the film is moving at the rate of 24 frames per second and must
be scanned at the rate of 144 per second, it is apparent that each frame
must be scanned six times. This means that each frame will be com-
pletely scanned as it moves a distance equal to one sixth of the per-
foration pitch. ' Fig. 1 is a simplification of the action that takes place.
The heavy arrows represent frame A occupying successive positions
after each YJW of a second interval as it moves down through the gate
at constant linear velocity. (For the sake of clarity the arrows are
shown slightly displaced to the right in each position.) At the start,
frame A is in position 1 in the- top of the gate, its lower five sixths (the
heavier portion of the arrow) exposed. After l/u of a second it has
354
ERDE
October
1948 COLOR-TELEVISION FILM SCANNER 355
moved to position 7 in the lower part of the gate, and has been scanned
six times in the interim; frame B now stands in position 1 ready to
repeat the cycle. It will be seen from the diagram that the height of
the gate aperture must be equal to !2/3 times the perforation pitch.
Corresponding to each of these six positions of the frame in the gate
is one of six lens-segment projection elements. These lens segments are
arranged in a common plane normal to the main optical axis, with
their centers in a straight line parallel to the direction of travel of the
film. Each lens is so adjusted that it lies on the straight line joining
the center of the scanned area on the cathode with the center of the
lower five sixths of the frame in each position in the gate. In effect
then, these lens segments are projecting upon the cathode sucessive
overlapping images of the gate aperture, each displaced by an amount
equal to one sixth of the perforation pitch. If, now, each lens segment
is exposed alone for the time during which the frame is moving across
the corresponding part of the gate, it will project upon the common
scanned area of the cathode an image of the frame moving upward
one sixth of the perforation pitch in l/m of a second. This moving
optical image can then be scanned by deflecting its corresponding
electronic image in an opposite direction over the physical scanning
aperture at the other end of the dissector tube.
It should be borne in mind that the optical image on the cathode is
not stationary, but is moving upward for one sixth of the perforation
pitch, repeating this for every vertical scanning period over the identi-
cal portion of the cathode. The effectual immobilization of the image
i is brought about by the action of the vertical scanning in a direction
opposite to the image movement, i.e., the over-all vertical scan is the
; resultant of one sixth upward motion of the image and five sixths (or
i less) downward motion of the electronic scanning. The expression
"or less" is used here because it is obvious that a resultant vertical
scan of one perforation pitch is a little more than is usually desired
I and would allow the frame line to be seen. In practice, the verti-
; cal-scanning amplitude is adjusted to a value slightly less than five
sixths of the magnified perforation pitch (more exactly, five sixths
of the perforation pitch less the difference between the perforation
pitch and the standard projector gate aperture height) to give a
scanned picture height equivalent to the standard projector gate
aperture height.
Because of the geometry of the optics wherein six points spaced
0.050 inch apart (this is equal to one sixth of the perforation pitch of
356 ERDE October
0.300 inch) are focused to a common point, it is obvious that the six
projection-lens elements would have to be extremely small to avoid
mutual physical interference. In practice, a standard projection lens
is employed to form an enlarged virtual image of the gate aperture, so
that the center-to-center spacing of the projection elements now be-
comes of practicable dimensions. This is shown in Fig. 2, where for
clarity, a simple plano-convex lens replaces the standard projection
lens. In this diagram, a frame is shown in the No. 2 position in the
gate with its enlarged virtual image projected through the No. 2 lens
segment upon the scanned cathode area. The No. 2 lens segment is
exposed through one of six slits in a rotating selector disk. These
slits are concentric, adjacent arcs of 60 degrees and of differing radii,
so that each slit exposes its associated lens segment at the proper time.
The selector disk is driven at 24 revolutions per second by a syn-
chronous motor, thus exposing the six lens segments in order every
x/24 of a second. Since the selector disk must be synchronized with
the film movement in order to have each lens segment exposed at the
appropriate time, the motor frame is mounted so that it may be
phased manually.
The size of the selector disk is determined by the ratio of blanking-
to -total-vertical-scan period; in this case, 1 to 10. Therefore, if the
radius of the innermost slit is made such a value as to give the slit a
length equal to 10 times the lens segment effective aperture length,
the optical change-over will occur wholly within the blanking period.
Allowing for the radial increments of the five other slits plus a small
guard rim, the selector disk is 29 inches in over-all diameter. It is
completely enclosed within a housing, and driven by a V4-horsepower,
3-phase, 1800-revolution-per-minute synchronous motor, through
a 5-to-4 reduction gear box, with complete absence of vibration. This
latter point is of importance where extremely fine optical registration
is to be maintained.
LENS SEGMENTS
The derivation of the lens segments may be understood by refer-
ence to Fig. 3. Each segment is originally a fully formed, cemented
achromatic doublet of 14-inch focus and l3/4-inch diameter. From
each, the limb on each side of the optical centerline is cut or ground
away until a segment of the desired thickness remains. It is evident
that these lens segments of rectangular aperture will transmit more
light than equivalent circular lenses of small-enough diameter to
COLOR-TELEVISION FILM SCANNER
357
358
ERDE
October
maintain the same spacing. The maximum lens segment center-to-
center spacing for the optical constants involved is about 0. 140 inch.
Allowing 0.010 inch between segments for adjustment purposes,
this would leave a narrow, fragile piece of glass only 0.130 inch thick,
if the lens were cut symmetrically with respect to the optical center-
line. However, by cutting the lenses unequally on each side of the
optical centerline, it is possible to maintain the same center-to-center
spacing of the segments while at the same time materially increasing
their thickness. This is shown on the right of Fig. 3 where the degree
of asymmetry increases progressively with the distance of the lens
segment from the common centerline. By this means the lens thick-
ness has been increased to 0.160 inch, thus imparting a sturdier quality
VERTICAL CL
'OPTICAL CL-
6 NARROW SYMMETRICAL
SEGMENTS SPACED FOR
GIVEN M" BETWEEN
OPTICAL CENTER LINES
6 THICKER ASYMMETRICAL
SEGMENTS SPACED FOR
SAME "d"
LENS SEGMENT CUT FROM ENTIRE
ACHROMATIC DOUBLET
Fig. 3 — Derivation of lens segments.
to the segments, and more important still, increasing their light trans-
mission by about 20 per cent. In addition to their strongly asym-
metrical cut, as shown, the top and bottom segments, not being as con-
fined as the interior ones, actually are about 50 per cent thicker in
order to allow them to transmit still more light and thus improve the
average signal-to-noise ratio.
Since the six lens segments are to form six superimposed congruent
images, it is necessary, for the sake of good resolution, that their real
magnifications match very closely. Since the real magnification is
closely dependent upon the focal length, this parameter must be
controlled carefully in manufacture if satisfactory coincidence of
images is to be obtained.
1948 COLOR-TELEVISION FILM SCANNER 359
If it is assumed that the images are to agree in size within ± 1/z of a
picture-line pitch, and if the centers of the images are made to coincide
exactly, then the greatest disagreement will be at the extreme top (or
bottom) where the picture lines will fall within ± */4 of a picture-line
pitch, which is quite satisfactory. For a 500-line picture, a tolerance
of ±*/2 of a line pitch is equivalent to a tolerance in the magni-
fication of ±0.1 per cent. In this instance, the focal length is 14
inches and the real magnification is 0.43 times. From the funda-
mental relationship among focal length, magnification, and image
distance, the permissible variation in focal length turns out to be
±0.010 inch, or a percentage variation of ±0.08 per cent. This can
be achieved in good optical practice.
Three other properties required of the lens segments for satisfactory
coincidence of images are freedom from curvature of field, negligible
distortion, and good color focus. Curvature of field is reduped to a
negligible minimum by good design aided by the long focal length and
narrow angle of projection (a maximum of 2 degrees for the top of a
frame entering the gate aperture, or the bottom of an exiting one).
Distortion has been found to be almost entirely a function of the
projection lens forming the virtual image. A high-quality projection
lens will introduce no noticeable distortion. If the projection lens is
also well corrected for color and if the lens segments have been achro-
matized for the C and F lines, the color focus of the combination is
found to be entirely satisfactory. In addition, the projection lens
should have a wide aperture to avoid vignetting of the upper and
lower frame images. The Bausch and Lomb //2, 41/2-inch focus Super-
Cinephor lens has been found satisfactory in all the above respects.
In order to realize fully the accuracy with which the lens segments
are fashioned, it is necessary that they be mounted in such a manner
that they may be carefully adjusted for accurate optical alignment
and then rigidly and permanently fastened in place. A mounting relief,
Vie of an inch deep, ground into the ends of the segments, is utilized
for fastening. Fig. 4A is a view of the lens-segment mount in posi-
tion, with the selector disk shown behind. A sturdy brass aperture
plate forms the basis of the mount and an arrangement of small
metallic holding members and screws permits the individual segments
to be adjusted vertically and horizontally and then clamped rigidly
in place.
The actual alignment of the lens segments is accomplished in the
following manner. A reel, or loop of film, of a suitable geometric
360 ERDE October
resolution pattern is run through the scanner with the selector disk
properly phased, and the resolution pattern is reproduced upon the
picture monitor. In front of the lens segments, a slotted aperture
frame (shown in Fig. 4B) receives masks containing rectangular
apertures of different sizes and combinations so that the lens segments
may be exposed singly or in combination. First, the No. 4 lens,
lying just below the optical axis of symmetry, is exposed and the
dissector electronic equipment is adjusted for normal picture-
scanning amplitudes and optimum electronic focus. Then the dis-
sector optical focus is adjusted by longitudinal racking of the dis-
sector-tube mount until the image seen upon the picture monitor is
Fig. 4 A — Lens segment mount in posi- Fig. 4B — Aperture frame and fixed,
tion. ' segmented, tricolor filter.
in best focus. The dissector tube is left in this position during the
remainder of the alignment while each lens segment is adjusted in
turn with respect to the No. 4 lens segment so that the images
exactly coincide. By this method, coincidence of the six images can
quickly be obtained to within a fraction of a line pitch in both the
horizontal and vertical directions.
The light efficiency pf the optical system is unavoidably low, since
the//2, 41/2-inch focus projection lens is in effect stopped down by the
six lens segments to an average rectangular aperture of 0.180 X 1.375
inches. The diameter of the equivalent circular aperture is 0.560 inch.
Since the front element of the projection lens is almost completely
filled with light and since the light beam has diverged only slightly
when it falls upon the lens segments, the effective relative aperture of
1948 COLOR-TELEVISION FILM SCANNER 361
the complete system may be considered to be the ratio of 4.5 inches to
0.56 or about //8.
Nevertheless, despite the light lost in heat and color filtering as well
as through the restrictive lens segments, the light flux incident upon
the scanned portion of the cathode, an area equal to I1/ 2 X 7/s inches,
is in the order of 4 lumens. This is sufficient to give a signal having
an acceptable signal-to-noise ratio.
Fig. 5 is a schematic view of the entire optical system. Here are
shown the carbon-arc source, condenser, water cell, and heat-absorbing
filter. Following these in order are the components of the film-
scanning portion — negative field lens (to be explained later) film gate,
projection lens, optically polished auxiliary heat filter, selector disk,
lens segments, a fixed segmented tricolor filter, and the dissector tube.
It will be noted that there is no rotating color disk in this portion of
the film-scanning optical system. In its place and performing the
identical function of inserting sequentially in the light beam the three
primary-color filters, the fixed segmented tricolor filter is used. This
can be done because the lens segments are exposed sequentially ,and
are six in number while the primary colors are three in number. Thus
each lens segment is associated with a given primary color w^hich it is
called upon to transmit to the exclusion of the other two primary
colors. The color order of the lens segments from top to bottom is
green, red, blue, green, red, blue. The fixed tricolor filter consists of
six rectangular strips of green, red, and blue Wratten gelatin filter
arranged in the corresponding color order and cemented between two
squares of optically polished glass. This composite filter is held in
place by means of a fixed aperture plate, close to the lens segments, so
ithat each segment transmits its own associated color. The three
primary color filters are green No. 58, red No. 25, and blue No. 47.
In Fig. 4B, the tricolor filter is shown in place over the lens segments,
with the individual colors somewhat indistinguishable in a black-
and-white photograph.
For color-slide projection a 45-degree plane, chromium-plated
mirror is swung into the arc beam in front of the water cell to deflect
the light through the slide-projection components. An auxiliary
condenser lens, another 45-degree mirror, and a field lens serve to relay
the arc beam and illuminate the aperture of the 2- X 2-inch slide
carrier. From this point the light passes through another optically
polished heat filter and then through the color filters of a synchronously
rotating tricolor disk. The filters are six in number and arranged to
362
ERDE
1948 COLOR-TELEVISION FILM SCANNER 363
give exposure in the red, blue, green order. The disk rotates at 1440
revolutions per minute so that each filter exposes the beam for Vi44 of
j. a second. A Kollmorgen 8-inch-focus projection lens projects the
slide image onto the cathode of the dissector tube which can be moved
over to line up with this new axis of projection. Whenever a black-
and-white image is needed for test purposes, an auxiliary light source,
using a 500-watt incandescent lamp and condenser, can be cut in by
swinging out the second mirror. Since the color disk is not needed
for this latter purpose, a hinged mounting on the disk-and-drive
structure permits it to be swung aside out of the light path.
FILM-SHRINKAGE COMPENSATION
As with all types of continuously moving film projectors, some
; method must be employed to compensate for film shrinkage. The
I necessity for this is evident upon consideration that the distances
; between the centers of the six lens segments in the vertical direction
have been permanently fixed and correspond to certain definite
distances between the centers of successive frame positions in the
gate. For another film, of less or greater shrinkage, these centers of
successive frame positions will no longer correspond with the fixed lens
segments, center-to-center distances, with the result that the super-
; imposed images, although still remaining in sharp focus, will no longer
coincide. To restore this coincidence of images, it is necessary only
to refocus the //2 projection lens slightly in or out, and thereby
diminish or enlarge the virtual image of the frame just sufficiently to
realign the successive virtual frame position centers with their corre-
sponding lens-segment centers. This adjustment, while restoring co-
1 incidence, also alters the focus, which, however, is regained by shift-
) ing the real image plane, i.e., the dissector tube, longitudinally. The
!> result is exactly coincident and sharply focused images for any degree
[• of film shrinkage encountered.
Published data4 and our own experiences have indicated that a
range of 0 to 1.5 per cent shrinkage should be accommodated. The
; projection-lens barrel is calibrated directly in film shrinkage over this
range, for both the standard and the nonstandard emulsion positions.
| Calibration practice consists in making the initial lens-segment-
spacing adjustment, which has already been described, with a test
film of measured shrinkage in the gate. This shrinkage is then
marked on the lens barrel opposite a fixed reference point. Several
other test films of different shrinkages are then run through the
364
ERDE
October
scanner, and for each the projection lens is focused and the dissector
mount readjusted to give an image on the picture monitor tube hav-
ing the sharpest focus and the best coincidence. Each shrinkage set-
ting is marked on the lens barrel and the determination of four such
points is sufficient to allow a smooth curve to be drawn for the inter-
polation of additional shrinkage settings.
In operation, the shrinkage of the film to be run is measured before
Fig. 6 — Side view of film scanner.
threading and the projection-lens barrel is preset to the corresponding
shrinkage. Then, while the film is being scanned, the dissector-tube
mount is adjusted for best resolution of the monitor image, whereupon
optimum image coincidence and focus are simultaneously achieved.
Because the depth of focus of the projection lens combined with the
lens segments (the real magnification is less than unity and thej
effective aperture is //8) is large compared with the permissible shift |
of the projection lens as far as image coincidence is concerned, small j
changes of shrinkage in the body of the running film can be accommo-
dated by a slight refocusing of the projection lens. This always brings
1948 COLOR-TELEVISION FILM SCANNER 365
the images back to exact coincidence leaving the focus substantially
unchanged.
Film shrinkage is most easily, and with sufficient accuracy, de-
termined by measuring the length of a given number of sprocket
holes with a scale and dividing by the number of frames to obtain the
average perforation pitch. It is convenient to use 39 sprocket holes
and a 30-centimeter scale. This makes actual counting of the sprocket
Fig. 7 — Film scanner and dissector electronic scanning equipment.
holes unnecessary, since the 39th hole will always fall between 293 and
297 mm for a 0 to 1.5 per cent shrinkage range. The measurement
can then be left in millimeters per 39 frames, or transformed to per cent
shrinkage by subtracting from and dividing by 297 mm, the nominal'
length of 39 frames of unshrunk film.
FILM-DRIVE MECHANISM
Figs. 6, 7, 8, and 9 are views of the complete scanning equipment.
On the left of Fig. 6 is the front of the arc lamp containing the water
cell. In the center is the film-drive mechanism, and associated with
366
ERDE
October
it the upper and lower reel holders, projection lens, selector disk and
lens-segment housing, selector-disk drive and manual phaser, and the
control panel. On the right is the dissector-tube housing containing
the tube and the scanning and focusing coils. The small knob
beneath the dissector mount in Fig. 7 is for optical focusing and moves
the tube mount longitudinally. The table-type rack beneath the dis-
sector mount contains the electronic equipment for operating the dis-
sector, namely, power supplies, scanning generators, and associated
equipment.
Fig. 8 — Front view of film scanner.
Fig. 9 — Slide-scanning components.
Fig. 6 is of special interest in that it reveals the origin of the mech-
anism for pulling down the film at a constant rate of speed. When
design of the film scanner was first begun, serious consideration was
given to the possibility of adapting some existing commercially
available 16Tmm sound-film projector to television-film scanning. Of
several models examined with this end in view, the Ampro Premier No.
10 projector was selected as being the most conveniently adaptable
from the constructional standpoint. The soundhead assembly was
retained in its original form while the rest of the projector was modi-
fied by removing the intermittent pull-down mechanism, the shutter
drum, and the motor drive and blower. Only that portion contain-
ing the pull-down and take-up sprockets and the associated gear train
1948
COLOR-TELEVISION FILM SCANNER
367
was retained. Around this as a nucleus, the rest of the film drive
was designed.
As shown in Fig. 6, a Maurer precision sprocket for constant-speed
film drive is built in just underneath the film gate. This sprocket is
driven through a mechanical filter consisting of a 10-inch flywheel,
spring coupling, and suitable damping. These are shown in Fig. 10,
Fig. 10 — Mechanical filter for constant-speed
sprocket.
which is a top view of the scanner with covers removed. Although
the speed of the sprocket is low, being only 3 revolutions per second,
the mechanical filter is extremely efficient in smoothing out any fluc-
tuations in the drive that might tend to be imparted to the film
motion, and the images are gratifyingly steady. The inherent
24-cycle-per-second sprocket modulation normally introduced by the
sprocket teeth has been minimized by suitable sprocket design, and
368 ERDE October
there is no observable impairment of optical resolution from this
source.
The motive power is furnished by an 1800-revolution-per-minute
synchronous motor coupled to the constant-speed sprocket assembly
through a 1 : 1 right-angled spiral gear drive and a 10 : 1 worm and
worm gear. A motor-shaft extension engages directly with the Ampro
gear train to drive the pull-down and take-up sprockets. This
arrangement effectively isolates the constant-speed sprocket behind
its mechanical filter and frees it from any motional irregularities intro-
duced by the remainder of the mechanism. For picture framing, the
motor shell can be rotated manually in its support by means of a long
shaft and a knob located conveniently on the front of the scanner.
Also shown in Fig. 10 are some of the components of the slide
projection system, the auxiliary condensers, slide holder, color disk
and drive, and the stand-by incandescent-lamp source. An air blast
for cooling the film and slide gates and the dissector tube is obtained
from a central blower beneath the scanner and linked to those points
through manifold and air-hose connections.
FLICKER
The nicker to be discussed here at length is that which has its
origin in the optical characteristics of the film-scanning system and
not the flicker arising as a function of the field-repetition rate and
the brightness level at the receiver.
Regarding the latter, however, this much will be mentioned as of
general interest. The CBS prewar system of color television was
based on a field repetition rate of 120 per second. This permitted a
high-light brightness at the receiver of 2 foot-lamberts for just per-
ceptible flicker. Postwar investigations indicated the desirability of a
higher field-repetition rate, and when this was raised to the present
144 per second, a high-light brightness at the receiver of 9 foot-
lamberts was obtained. Subsequent changes in receiver-filter
characteristics have enabled the receiver-flicker threshold to be
raised to a value of 20. foot-lamberts.
It might seem, from a review of the basic optical design of this type
of film scanner, that flicker would be rather a vexing problem. This
conclusion might, quite naturally, be drawn upon consideration of two
innate characteristics of the optical system. First, successive images
of a frame in successive portions of the film-gate aperture are, in effect,
superimposed in projection on the photocathode. Should the gate
1948 COLOR-TELEVISION FILM SCANNER 369
aperture be nonuniformly illuminated, corresponding areas of the
image may go through a cyclic variation in brightness and cause a 24-
cycle-per-second regional flicker. Second, in performing their
function of consecutively projecting a frame as it occupies different
portions of the gate aperture, the lens segments must suc-
cessively select different portions of the cross-sectional area of the
light beam. If the light beam is nonhomogeneous, or if one of a pair
of lens segments transmits light unequally, again there may be a
periodic variation in the brightnesses of the superimposed cathode
images, evidencing itself this time as an over-all 24-cycle-per-second
flicker. (The lens segments are paired in the sense that the first and
fourth are filtered to transmit only green light; the second and fifth,
red light; and the third and sixth, blue light.)
It follows then, that the solution of the flicker problem depends
upon two requirements. First, that of securing adequately uniform
distribution of illumination in the gate aperture over its full height
(equal to approximately 10/6 X perforation pitch of 16-mm film, or
0.500 inch) and second, that of obtaining equivalent transmissions of
light through both lens segments in a pair.
The first requirement, that of adequate uniformity of gate-
aperture illumination, is fairly easily met by the fact that the 16-mm
gate-aperture dimensions (even though extended of necessity to 0.500
inch in the vertical direction) are less than those of a 3o-mm film
aperture, for use with which the arc-lamp condenser was designed.
In addition, a negative field lens (whose function is of greater impor-
tance in enabling the second requirement to be met) behind the gate
aperture permits a more enlarged crater image, so that by proper re-
focusing of the arc condenser a compromise between light intensity
and light distribution can be effected in which the gradient of illumi-
nation from the center of the aperture upward and downward is not
large enough to introduce any regional flicker as the images of a
frame in the different gate positions coregister on the cathode.
The second requirement, that of obtaining equivalent transmissions
of light through both lens segments in a pair, too, has not offered
great difficulty. The negative field lens inserted behind the gate
aperture has a focal length of 145 mm. The effective focal length of
the combination of this lens1 with the 41/2-inch-focus projection lens
is such as to project into the plane of the six lens segments a reduced'
image of the front surface of the forward arc-lamp condenser lens, and
of a circumference closely circumscribing the total rectangular area of
370 ERDE October
the six lens segments. Thus, while also insuring the maximum re-
laying of light through the lens segments, the homogeneity of the cross
section of the light beam in this plane is considerably improved.
Although this would give substantially equal transmissions of light
through lens segments of equal areas, it should be recalled that the
top and bottom lens segments have been intentionally designed to have
appreciably greater area than the other segments. As has been
pointed out earlier, the top and bottom lens segments do not have the
spacing restrictions that are of necessity imposed on the interior ones.
Consequently, these outer segments have been made with a 50 per cent
greater area in order to increase the average light transmission and
thereby improve* the average signal-to-noise ratio. To exploit this
situation fully, it is necessary that some means be employed to convert
the repetitive light pulses of unequal level into resultant signal pulses
of equal level. The mechanism to be described accomplishes this by a
form of automatic gain compensation synchronized with the sequential
exposure of the lens segments that permits the equalization of the
magnitude of any pair of lens segments' light-to-signal conversion
simply by adjusting the corresponding knobs on a control plane
while the actual film-scanning process is under way.
A rotating switch arm attached to a shaft rotating at the syn-
chronous speed of 1440 revolutions per minute sweeps over six con-
tacts fixed at the periphery of a manually adjustable disk. Each
contact is connected to its own flicker-control potentiometer and the
six potentiometers are shunted across a common stage of the elec-
tron multiplier in the dissector tube. With the equipment in oper-
ation but with no film in the gate, with the selector disk running and
light falling through the lens segments to focus the gate-aperture image
on the cathode, the contact disk is phased by hand until the switching
change-over falls wholly within a vertical blanking period (as shown
upon either the wave-form or the picture monitor) and is then clamped
permanently in place. It can be seen that the gain of the multiplier
stage during a given color field now will depend upon the resistance of
only that potentiometer which is connected across it for the du-
ration of that color field.
If the wave-form monitor sweep frequency has been adjusted to give
three color fields, it will be noted that each color field consists of two
fields superimposed (two green, two red, and two blue). Since each
of the six flicker-control potentiometers affects the level of its
corresponding color field only, it is then a comparatively simple
1948 COLOR-TELEVISION FILM SCANNER 371
matter, while observing the wave-form monitor, to adjust the paired
color-field amplitudes until the two components of each pair are
equal and at a maximum, whereupon equal and maximum signal am-
plitudes for the paired lens segments of each color will have been
derived. For slide transmission, the flicker-control potentiometers
are switched off since they need not then, of course, be used. Once
this adjustment has been made, no further flicker adjustment is
thenceforth necessary, barring excessive maladjustments in the
carbon-arc trim or condenser alignment.
Color mixing is accomplished in a somewhat similar manner, with
the difference that the response through both of a pair of lens segments
(instead of through each individual lens segment) is varied simul-
taneously and automatically. Color mixing means, simply, vary-
ing the ratios of red-to-blue-to-green signal levels. Although the
color-filter chromaticity and transmission values have so been
selected that normal color reproduction (with also the widest pos-
sible range of colors and color saturation) will be obtained when the
red, blue, and green signal levels are equal, there often arises the need
for altering these ratios either to achieve a more pleasing effect or to
compensate for color-balance deficiencies in the film. A second
rotating switch arm attached to the flicker-control switch arm
shaft and space-phased to it sweeps in like manner over six contacts.
These six contacts are connected diametrically in pairs and each pair
is connected to its own color-level-control potentiometer.
The three color-level-control potentiometers are shunted across a
second common stage of the dissector-tube electron multiplier. When
proper phasing has been obtained the levels of the red, blue, and green
signals can be varied independently from zero to a maximum simply
by turning the corresponding knob on the color-control panel.
Other controls on the color-mixing panel include the usual bright-
ness-level control, a master gain control which varies the red, blue,
and green signal levels simultaneously without altering their ratios,
and a gamma-correction control for varying the contrast distribution
of the entire system.
An interesting feature of the film scanner is that it is not uniquely
a color-television standards device. Provision, in fact, has been made
for a quick change-over in a matter of minutes from color to the Radio
Manufacturers Association standard black-and-white transmission.
Should that be desired, it is only necessary to replace the six lens
segments holder with one containing five lens segments adjusted for a
372 ERDE
VB perforation-pitch interval of 0.060 inch of a center-to-center spacing
of 0.168 inch. The six-slit selector disk is also replaced with one
containing five slits arranged in a 1, 3, 5, 2, 4 order (instead of the con-
secutive order as in color transmission) . Then with the gear-box trans-
mission shifted to 720 revolutions per minute the selector disk rotates
at the proper speed to allow a frame to be projected and scanned, as it
moves 2/o of a perforation pitch, in 1/M of a second, giving 5 scans of 2
frames in 1/i2 of a second, or 60 fields per second.
Although the foregoing description may have created an impression
of delicacy and complexity in the function of the CBS color-tele-
vision film scanner, it must be emphasized, in concluding, that in
two and one half years of constant use, this film scanner has
given quite definite proof of the practical nature of its design.
During this time it has given dependable and consistently satis-
factory results as a transformer of moving color-film images into
video signals, with no more than the normal amount of operating
adjustment and maintenance required of any piece of studio equip-
ment. Several similar scanners have since been built, and in every
case the requisite optical, mechanical, and electronic precision and
dependability have been easily reproduced.
REFERENCES
(1) Fordyce Tuttle and Charles D. Reid, "The problem of motion picture pro-
jection from continuously moving film," J. Soc. Mot. Pict. Eng., vol. 20, pp. 3-31 ;
January, 1933.
(2) F. Ehrenhaft and F. G. Back, "A non-intermittent motion picture projec-
tor," /. Soc. Mot. Pict. Eng., vol. 34, pp. 223-232; February, 1940.
(3) Peter C. Goldmark, "A continuous type television film scanner," /. Soc.
Mot. Pict. Eng., vol. 33, pp. 18-26; July, 1939.
(4) J. A. Maurer and W. Bach, "The shrinkage of acetate-base motion picture
films," /. Soc. Mot. Pict. Eng., vol. 31, pp. 15-28; July, 1938.
FORTY YEARS AGO
The Future
The future of the moving picture machine is a theatrical problem.
Some theatrical men believe that it will prove a serious competitor of
the vaudeville. They suggest the time when the phonograph will work
with it, and the best act of the newest New York comic opera will be
flashed on the screen and sung out of the phonograph.
Others, and probably these are right, say that the picture machines
have hit their highest notch.
— The Moving Picture World, January 4, 1908
35-Mm Process Projector*
BY HAROLD MILLER AND E. C. MANDERFELD
MITCHELL CAMERA CORPORATION, GLENDALE, CALIFORNIA
Summary—A studio type of process projector, designed and built to meet
the specifications as set forth by the Motion Picture Research Council Com-
mittee, is described. Both the single- and the triple-head projectors are
discussed.
THE MITCHELL background projector as supplied to most of the
major studios is an outgrowth of a development originally
started about 1934. Previous to this time, as well as during the period
up to now, background projectors have been for the most part semi-
experimental laboratory-type machines built up by studio technicians
from various odds and ends available from the studio camera shop.
The present projector design is based on the recommendations as
stated in the Academy specifications entitled "Recommendations on
Process Projection Equipment/7 published in February, 1939, by the
Process Projection Equipment Committee of the Research Council of
the Academy of Motion Picture Arts and Sciences. The Research
Council report showed the need for process projection equipment that
could be used on the sound stage without the use of blimps or portable
projection rooms to contain the noise of the equipment, that would be
portable on a suitable dolly and movable to various stages, that could
deliver the maximum light possible with a modern optical system, and
that could project an absolute steady picture.
The following is a description of the various components that consti-
tute the complete process projector which meets these requirements.
PROJECTOR HEAD
The projector head is composed of a vertical drive shaft and four
driven cross shafts coupled by helical gears and held in accurate align-
ment by oilite bearings. The cross shafts drive two 32-tooth sprock-
ets, the movement, and the 180-degree rear shutter. This mecha-
nism is enclosed in an invar steel housing, and is lubricated by grease
of 2500 to 3000 viscosity through three Zerk fittings. The unit is
* Presented May 18, 1948, at the SMPE Convention in Santa Monica.
OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51 373
374 MILLER AND MANDERFELD October
mounted in an aluminum case which is fitted with the necessary idler
rollers to guide the film to and from the sprockets and the magazines.
The Mitchell compensating link camera movement, Fig. 1, is used!
in the head to provide pilot registration pins necessary for "rock-
steady" projection. The movement is modified for process projection
by placing the claw and operating mechanism below the aperture away
from the light beam and heat. The removable back plate and register
plate are open between the film tracks to release condensation of mois-
ture from the film and to eliminate the possibility of scratching. The
movement is provided with an adjusting screw to adjust for shrinkage,
and also to adjust for minimum
film noise while running both
forward and backward.
The movement is coupled to
the head by a key coupling
which can be engaged in only
one position thereby insuring
that the movement is in time
with the shutter. The movement
position is set by a dowel pin
to keep the necessary alignment
between aperture and optical ele-
ments, and is locked in place by
w" • t -D.-' • thumbscrew clamps. A dummy
Fig. 1 — Projector movement.
aperture is used in place of the
movement to align the optical system and check screen illumination.
Both movement and dummy apertures accommodate mattes of
Academy and Standard apertures.
UPPER AND LOWER MAGAZINES
As shown in Fig. 2, the upper magazine is mounted on top of the
head and the lower magazine bolts to the back of the head. The
magazines are 13 inches inside diameter, giving ample hand room
when using 1000-foot -reels or spools. They are lined with corduroy-
velvet for protection to the film. The magazine doors are fitted with
8-inch windows, thereby providing a full view of the film passing
through the head. Both magazines are equipped with adjustable
overriding clutch, felt friction disk take-up drives. Also there are
adjustable pull-down brakes inside of the magazines to provide the
proper film tension for running forward or backward.
1 1948 PROCESS PROJECTOR 375
LENS MOUNTING
The lens mount bolts on to the front of the projector head. It has a
} 5V2-inch diameter opening to accommodate //2.0 lenses from 4 to 8
inches focal length.
The lens mount is equipped with jackets to hold various focal-length
lenses. The jackets are slipped in the lens mount and are held in place
I by a retaining pin. The lenses are easily changed by lifting the knob
I on the retaining pin, removing the jacket, and inserting another.
The lens mount has a manual focusing knob at the operating side
and a Selsyn receiver motor for remote control from the camera
position.
Fig. 2 — Projector head and light tube.
REMOTE-FOCUSING CONTROL
The remote focusing of the objective lens on the background screen
is practically standard procedure in most studios for several reasons.
First, the correct focus position is that whereby the projected image
looks best as viewed from the camera position. Second, while the set is
being prepared for the next "take" it is quite customary for the light-
ing crew to set up and test the various lights. This light, when falling
on the background screen, for the most part, makes it difficult and at
times practically impossible to see the image on the screen from the
projector position.
Several methods or means for remote focusing were considered.
The method adopted was the use of Selsyn motors. Essentially it con-
sists in interlocking two single-phase Selsyn motors, one of which is
376 MILLER AND MANDERFELD October
mounted as a receiver in a manner to actuate the focus of the projector
objective lens, and the other one at the remote focusing point which is
used as the transmitter. The latter motor shaft is provided with a
small handle so it can be rotated manually in either direction and thus
through the electrical interlock rotate the lens-actuating motor.
The remote-focusing motor is encased in a housing having a con-
venient handle for carrying the motor about on the stage. Because of
its general appearance it is commonly referred to as the "beer mug."
A push button is incorporated in the carrying handle which first must
be pressed down in order to excite the two motors electrically. This is
a safety device to prevent accidental movement of the projector lens
by inadvertent movement of the "beer-mug" rotating handle.
HEAD MOUNTING
The complete projector head, magazines, and motor are coupled by
means of a ring coupling to an aluminum casting, Fig. 2, commonly
known as a light tube. This permits the projector to be revolved
about its optical axis. An adjustable worm gear controls the rotation
15 degrees each side of vertical. However, by rotating farther, there-
by disengaging the worm from the gear, the projector head can be re-
volved to any position within 125 degrees each side of vertical. The
head is locked in place by a clamping screw on the coupling ring.
The relay lens (described in more detail later), Fig. 3, is located in
the light-tube casting behind the projector head. This unit is
mounted with the necessary controls so that it can be adjusted hori-
zontally, vertically, and axially for focus. The condenser lenses in the
arc house have a similar set of controls.
The relay lens holder is a hard, chrome-plated, aluminum casting
and is provided with a 2-inch space between the two elements. This
space is filled with boiled distilled water for the purpose of removing
much of the heat in the condenser-light beam.
To remove the heat from the distilled water, ordinary tap water is
circulated in a hollow space around the periphery of the relay lens
holder. This circulating water also passes through the cooling jacket
in the Mole-Richardson lamp, then to a vertical flow-type radiator
located under the lamphouse. The radiator unit includes a four-
bladed fan driven by a 1/4-horsepower, direct-current motor. This
motor circuit is interlocked with the lamp control in a manner which
makes it impossible to operate the arc unless the cooling fan is running.
1948 PROCESS PROJECTOR 377
The fan operates normally at 1700 revolutions per minute, but drops
to a reduced speed when the projector driving motor switch is on.
This procedure reduces the noise level during a "take."
Also included in the radiator unit is a circulating-water pump direct-
connected to a small alternating-current motor. This motor circuit is
also interconnected in a manner to prevent the lamp from being oper-
ated unless the circulating motor is operating.
Fig. 3 — Relay lens.
CONDENSER OPTICAL SYSTEM
For background projection work it is very essential that the screen
illumination be adequate, uniform across the screen, and of the proper
color quality. To achieve all three of these characteristics requires a
somewhat more complicated optical system than normally used in
theater projectors.
In the development of our optical arrangement, the previous art on
condenser systems as developed by Bausch and Lomb, Technicolor,
Paramount, and others was investigated. From the literature cover-
ing this art, it appears that the basic ideas underlying such optical sys-
tems are not too new, and it seems that most of the recent improve-
ment is due to more careful designing and the use of more precisely
made optical parts. In Fig. 4, it may be seen that the Mitchell system
makes use of a relay lens (already mentioned) used in conjunction with
378
MILLER AND MANDERFELD
October
a standard condenser unit with a collecting angle of nearly 90 degrees.
An image of the arc crater is projected by the two-element condenser
to an image point located at A . This enlarged image is again focused
by means of the relay lens at a point B which in general is close to the
rear surface of the projector objective lens.
In transferring the crater image from point A to the objective lens,
an interesting intersection of light rays takes place which is of prime
importance to the quantity as well as the quality of the screen illumi-
nation. Starting at point 3 of image A, three rays can be traced
through the relay lens, through the film aperture and to point 3 of
image B. It should be particularly noted that these three rays pass
Fig. 4 — Condenser-lens optical schematic.
through the film aperture at the top, center, and bottom. In other
words, the light emanating from point 3 of image A covers the entire
area of the film aperture. If now a similar bundle or cone of light rays
is traced from point 2 of image A to point 2 of image B, it will be found
that as before this point radiates a light cone which covers the
entire film aperture. Thus every point of the image plane A will indi-
vidually cover every portion of the entire film aperture. The resultant
effect is a uniform light intensity over the entire area of the film aper-
ture regardless of the light uniformity of the image source at A.
In tracing the light rays back from the image plane A to the arc
crater, another interesting feature may be observed. Starting at point
3 of image A it should be noted that the lower ray intercepts the larger
condenser element very nearly at the rim and progresses through the
other condenser element to the upper edge of the arc crater. The cen-
tral ray from point 3 passes through the condenser element at a point
approximately in the center, whereas the upper ray passes through the
PROCESS PROJECTOR
379
denser element close to the outer rim. Thus point 3 of image A is
for practical purposes illuminated from rays emanating from the entire
surface of the condenser element.
In a similar fashion, if one traces a series of rays from point 2 of
image A to the arc crater one will find that the illumination of the cen-
ter point of the image at A also depends on the entire surface of the
condenser element. All this can be summarized by saying that this
optical system is so designed that it will provide very uniform illumi-
nation over the aperture area and in addition makes effective use of all
the light that can be picked up by the condenser system.
FIRE SHUTTER
An automatic fire shutter, Fig.
5, is located behind the relay lens
in the light-tube casting, and is
mounted from the top of the
casting. The shutter unit is com-
posed of a cast-Inconel dowser
blade to withstand the heat of
the light beam. A direct-current
solenoid holds the shutter open
when energized through a circuit
controlled by the fire-shutter
governor. The governor unit
includes the necessary adjust-
ments to operate the fire-shutter
solenoid when the projector motor
reaches 700-revolutions-per-friin-
ute speed. It releases the shutter when the motor speed drops to
600 revolutions per minute, thus protecting the film from the exces-
sive heat of the arc. A handle is also provided for manually operat-
ing the fire shutter, thereby enabling the projectionist to flash the
light on the screen without the necessity of running the projector.
A variable-light aperture of cast Inconel is mounted in the light-
tube casting immediately behind the governor-controlled fire shutter.
This aperture is controlled by an eccentric coupled to an operating
knob on the outside of the light tube. An indicator dial is provided to
allow accurate setting of the light aperture.
Variation of the light aperture does not affect the uniformity of the
light intensity on the film and therefore the arc carbons can be burned
Fig.
5 — Variable aperture and fire
shutter.
380
MILLER AND MANDERFELD
October
at the maximum amperage consistent with good operating efficiency
and steadiness of light.
ARC LIGHT
The light source is a Mole-Richardson Type 250 arc lamp designed
and built for the Mitchell Camera Company. The details of the con-
struction and operation of this lamp have been covered in this JOUR-
NAL for July, 1947, in a paper entitled "Recent developments of super-
Fig. 6 — Electrical control box.
high-intensity carbon-arc lamps," by M. A. Hankins. As described
in this paper, the lamp is used in conjunction with a ballast grid and
automatically maintained at the proper operating position by a
photoelectric control system, thereby insuring very constant operat-
ing characteristics without the necessity of constant vigilance on the
part of the projectionist.
ELECTRICAL CONTROLS
The electrical-control-circuit layout for the background projector is
based on the general over-all requirements as requested by several
1948
PROCESS PROJECTOR
381
studios. In general, the circuit arrangement allows the projector to
run either forward or backward in interlock, or "wild." In an emer-
gency all the leads to the driving motor can be broken simultaneously.
The actual switching is done by Leach relays which in turn are under
control of heavy-duty snap switches.
All electrical connections are brought into or out of the background
projector unit by means of Cannon plugs. Consequently it is a rela-
tively simple matter to replace complete assemblies or to disconnect
power supplies for routine maintenance. All important circuits are
supplied with fuses to eliminate
the danger of equipment over-
load.
Mechanically, the various com-
ponents of this electrical system
are disposed as follows:
All the electrical relays, fuses,
rectifier, and the intercommuni-
cation amplifier are located in a
large metal box, Fig. 6, located
just below the arc house and on
the nonoperating side of the pro-
jector. The snap switches used
for controlling the various cir-
cuits are located on the arc-light
control panel on the operating side of the projector, Fig. 7. The 3-
phase rheostat used for "wild" operation of the driving motor is also
located on this pajiel.
Fig. 7 — Single-head projector.
720-REVOLUTION-PER-MlNUTE OPERATION
The photographing by means of a standard motion picture camera
of an image projected from a background projector necessitates that
the shutter on the camera be properly synchronized with the shutter
on the projector both as to rotational speed as well as phase or instan-
taneous angular position.
Assume the use of 2-pole, 3-phase interlock driving motors on both
the camera as well as the projector, both electrically coupled to a com-
mon distributor. If the stators of the motors and the distributor are
excited from a common 60-cycle, 3-phase power supply and all the
rotor circuits are properly interconnected, each rotor will align itself
382 . MILLER AND MANDERFELD October
in some mechanical position so that electrically there is no transfer of
power from any one rotor to either of the others. If one rotor is
mechanically rotated or displaced to some new position, the other
rotors will follow mechanically, in order to rebalance themselves
electrically.
If now the shutters of the camera and projector are mechanically
aligned, while all the interlocked rotors are in electrical balance, it can
readily be seen that if the distributor (assumed to be 4-pole) is rotated
by means of a mechanically coupled driving motor, say at 720 revolu-
tions per minute, the 2-pole interlocked motors will operate at 1440
revolutions per minute, and both the shutters will rotate in synchro-
nism as well as maintain their relative angular phase while rotating.
Now it so happens that with a 2-pole motor there is only one
mechanical alignment position for an electrical power balance of the
rotor. With 4-pole motors there are two mechanical alignment posi-
tions 180 degrees apart where the rotor can be electrically in power
balance with the distributor. However, if the shutter shaft is coupled
to a 4-pole driving motor by means of a 2-to-l gear unit, it will act as
though it were coupled directly to a 2-pole motor.
Practically, there is an advantage in using a 4-pole interlock driving
motor for the background projector, namely, that of availability. In
general, 2-pole interlock motors are available only for camera drives
and, because of the small frame size and the low resultant power out-
put, this type of motor was not considered adequate for driving the
background projector and thus a 4-pole motor was used instead.
Since the camera and projector shutters must both run at 1440 revolu-
tions per minute, a 720-revolution-per-minute, 4-pole, or a 1440-revolu-
tion-per-minute, 2-pole distributor has to be used.
From the foregoing discussion it becomes rather apparent that the
use of a 720-revolution-per-minute interlock driving motor electrically
coupled to a 4-pole, 720-revolution-per-minute driven distributor
always insures the proper shutter alignment after the initial adjust-
ment. If some other .driving speed such as 1200 revolutions per min-
ute were used, this would not be the case.
It should be mentioned that the use of the 720-revolution-per-min-
ute interlock motor-drive system for background projection use was
not original with the Mitchell company. Several studios, notably
Paramount and Warner Brothers, have used this method for interlock-
ing for some years and therefore should be given due credit.
1948
PROCESS PROJECTOR
SINGLE- AND TRIPLE-HEAD BASE
383
The projector head and associated lamp unit are so designed that
they are interchangeable for either single- or triple-head assembly.
The triple-head unit (Fig. 8) of course requires two reflecting mirrors
for two of the three heads. This facility of being able to disassemble
a triple head readily into three separate projectors greatly increases
the range of activity for the process department.
Fig. 8 — Triple-head projector.
The "dolly" base of the single-head projector provides "rocklike"
stability when "locked off" for operation, and yet it can be easily
moved about the stage by two men. The panning and tilting mecha-
nism operates with smoothness and precision.
Four solid-rubber-tired wheels are attached to the fabricated steel
base of the single-head projector, Fig. 7. The rear wheels are
mounted on swivels and hinged to the frame. They are connected by
a tie-rod, enabling the base to remain level when moved over an
uneven surface.
384 MILLER AND MANDERFELD
By means of a handwheel on the center column, the optical axis can
be raised from 4 feet 9 inches to a height of 6 feet 3 inches. The col-
umn is also equipped with a handwheel to pan over an arc of 180 de-
grees. A third handwheel tilts the projector 12 degrees up or down
from horizontal.
In order to provide levelness and stability during the operation of
the projector, the base is equipped with three screw jacks.
The T-shaped base of the triple-head projector, Fig. 8, is constructed
of fabricated steel, and equipped with solid-rubber-tired swivel wheels
and screw jacks at each end of the T base. It is equipped with two
handwheels to pan 22 degrees and tilt 11 degrees up or down from
horizontal. The height of the optical axis of the three projectors on
this base is 5 feet 6 inches when parallel to the floor. Both bases are
equipped with removable tow bars.
ACKNOWLEDGMENT
In conclusion, the Mitchell Camera Corporation wishes to acknowl-
edge the valuable help and suggestions made by the process depart-
ment personnel of the various studios. In particular, we wish to
thank Mr. Farciot Edouart and Mr. Hal Corl of Paramount Studios
for their efforts in obtaining data of various sorts and for their techni-
cal suggestions during the development of this process projector.
FORTY YEARS AGO
It is alleged that many of the moving picture theaters in this city are
still having their machinery operated by boys under sixteen years of
age, especially on the latter East Side. The scheme is said to be to have
some matured operator go before the authorities and pass an examina-
tion and then turn the license over to the youngster. The Board of
Fire Underwriters and the Fire Department had better look into this
and if found to be correct to lock the offending manager up, send the
person who took the examination to Blackwell's Island and send the
"kid" operator to the Reformatory. No punishment is too severe for
people who conspire to do things that put human life in jeopardy.
— The Moving Picture World, May 80, 1908
New Theater Loudspeaker System*
BY H. F. HOPKINS
BELL TELEPHONE LABORATORIES, MURRAY HILL, NEW JERSEY
AND
C. R. KEITH
WESTERN ELECTRIC COMPANY, NEW YORK, NEW YORK
Summary— The new system employs sectoral high-frequency horns and
a crossover frequency of 800 cycles. Improvement is obtained in uniformity
of distribution and reduction of size and weight.
4 LOUDSPEAKER SYSTEM of the excellence required for motion pic-
JL\_ ture applications must be founded on sound fundamental prin-
ciples, many of which are understood fully only as a result of a con-
siderable background of experience in the field. Such a background
must be established before the application of technical skill can be
successfully applied to the development effort. Since the interpreta-
tion of the basic principles is so important to the success of the ven-
ture, it seems appropriate to discuss briefly the considerations upon
which the new Western Electric loudspeaker designs were based.
The description of the physical embodiment which follows later will
then have greater significance.
Before setting the course for a specific development of this type,
many diverse and complex phenomena must be weighed, and the
various merits and demerits of certain opposing characteristics must
be reconciled. The following paragraphs will attempt to rationalize
such a procedure. Where possible, data illustrating the effects dis-
cussed as well as the performance of the final loudspeaker system will
be presented. In order to simplify the discussion, the various general
attributes of a loudspeaker will be considered individually, and their
relation to practice defined.
FREQUENCY RESPONSE
It is quite generally agreed that one of the more important yard-
sticks in the determination of loudspeaker performance is the fre-
quency-response characteristic. In the development of speakers for
* Presented October 24, 1947, at the SMPE Convention in New York.
OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51 385
386
HOPKINS AND KEITH
Octob<
any specific use, therefore, it becomes necessary to determine what
response characteristic is needed to fulfill- the requirements of the situa-
tion best. The employment of shaped, or "distorted" response to
provide certain desirable characteristics such as high intelligibility in
the presence of high ambient noise is well known. The commonly
Fig. 1 — Arrangement of horn and loudspeaker for outdoor
response measurements.
accepted criterion for high-quality or "natural" reproduction such as
is required for motion picture applications is flatness of response over
the whole audible frequency range. This presumes, however, that
the conception of flat response, or uniform pressure-frequency re-
lationship is adequately understood.
The so-called free-field pressure measurement is the only type in
1948
THEATER LOUDSPEAKER SYSTEM
387
which complete uniformity of pressure with frequency is at all likely
to be found, and exact free-space conditions are very difficult to
realize. Fig. 1 shows a satisfactory arrangement of horn and micro-
phone for outdoor measurements. Furthermore, except in the ideal
case where the pressure distribution is identical at all frequencies,
uniform free-field response at a point does not indicate a uniform
power-radiation condition, which is probably nearer to the desired
characteristic when the loudspeaker is to be used in an enclosed
space. Since practical low-frequency speakers now in use are all less
directive at low frequencies, a flat free-field response which might be
excellent for outdoor situations is not at all desirable for indoor use.
Indoor measurements, on the other hand, are rather difficult to
500 1000
FREQUENCY IN CYCLES PER SECOND
Fig. 2 — Free-field response of loudspeaker phased for optimum
performance as judged by listening tests.
make. Rotating microphones, multiple microphones, and various
other devices are resorted to, but all have their shortcomings. The
use of rooms with staggered wall surfaces which tend to increase the
number of reflections, but to reduce the severity of the interference
effects appears to have some advantages. A rather ragged response
curve is obtained in such a room, but the peaks and dips are so close
together that a reasonable response trend may be inferred from the
data. Unfortunately, however, such measurements are useful only
for a rough evaluation, for the room in which the speaker is actually
to be used probably will have different characteristics. A great deal
can be learned about frequency balance from data of this type, but
the limitations of the measurement must be understood.
Ideally it would seem that a speaker should have a perfectly smooth
388
HOPKINS AND KEITH
October
free-field pressure response whether its trend be flat or otherwise. It
is recognized, however, that reflections and standing waves in rooms
cause raggedness in response that is far in excess of that exhibited in
the free-field response of high-grade loudspeakers, and that in many
cases the speaker having a smoother outdoor response does not appear
superior when measured under indoor conditions. Under listening
conditions where music or speech is involved, the transient nature of
the reproduced material makes the audible effect of reflections and
standing waves less evident than they are with the single frequencies
used in measurements. Reflected sound, however, constitutes the
greater part of the energy audible to the observer under indoor con-
ditions, so that a certain amount of "raggedness" must be present in
m
i
z
/*
k. — .
~\
^r
\
i]
*
I
/
<J
\/
^
>
V
TV
V
,
\
*i
11
%
LJ
A
A
/
fj
il
Ul
>
A
5
\
1
7
1
1
/
i
Ml
0
FREQUENCY IN CYCLES PER SECOND
Fig. 3 — Free-field response of loudspeaker of Fig. 2 phased for
smoothest free-field response.
what is heard. This is true, of course, whether the sound source is
"live" or whether it is radiated from a loudspeaker. In many in-
stances, where multiple- or dual-loudspeaker systems are used, free-
field response dips caused by out-of-phase radiation from the various
sources may be completely obscured under an indoor setup. It is
the opinion of most experienced observers, however, that while the
ear tolerates a certain degree of nonuniformity, a loudspeaker having
a smooth response will generally be more acceptable under all listening
conditions.
In connection with the free-field dips in response which may result
from the relative phasing of the sources in a multiple-unit speaker,
Figs. 2 and 3 are of particular interest. The free-field response of a
1948
THEATER LOUDSPEAKER SYSTEM
389
loudspeaker system phased for optimum performance as judged by
listening tests is shown on Fig. 2. The same speaker phased for the
smoothest free-field response is shown on Fig. 3.
In regard to the "flatness" of response and the frequency range,
present-day high-quality speaker systems do not, in general, follow
the pattern of idealized response. In practically all high-quality re-
producing systems, the high-frequency response is purposely
"drooped" to provide the most natural and pleasing quality from the
listener's viewpoint. In many instances "tailoring" is provided to
enhance "presence" or to compensate for some defect in the recording
>r reproducing medium. It is interesting to note, that while the ear
500 1000
FREQUENCY IN CYCLES PER SECOND
Fig. 4 — Superposed response curves taken every 5 degrees off
axis in horizontal direction.
irs a certain tolerance for sharp variations in sound pressure, it is
able to discern rather small changes in response trend. It has been
demonstrated that a "hump" 1 decibel in magnitude at about 2000
cycles tapering to zero at 1000 and 3000 cycles is easily detectable on
an A-B test (direct comparison of "with" and "without" conditions).
Many engineers are of the opinion that a "hump" in this frequency
range is desirable for the enhancement of "presence." This leads to
the conclusion that while measurement is an essential part of loud-
speaker development, the ear, as the final judge, must be resorted to
for the last adjustment to take care of subjective reactions under the
actual conditions of use.
On the other hand, the importance of the response curve to the de-
lopment engineer must not be minimized. For example, let us
390
HOPKINS AND KEITH
October
consider Fig. 3 showing the outdoor axial response of the new Western
Electric RA-450 60- watt system. The engineer, who is familiar with
the acoustic environment of the setup used, will recognize this as one
of the smoothest response characteristics that has been attained in
the highest-quality loudspeaker systems. He will appreciate from
experience that the relatively small irregularities which appear closer
together with increasing frequencies are caused by reflections which
he has been unable to avoid even in a very careful setup. The low-end
response is found to be rising at a rate of approximately 6 decibels
per octave which indicates a uniform radiation of power over this
range. At higher frequencies, where uniform distribution is obtained
the response is flat which indicates a uniform power radiation in this
.EDGE DAMPING
EDGE DAMPING -
POLE PIECES
MAGNETS
VOICE COIL
-ACOUSTIC RESISTANCE
Fig. 5 — Cross section of low-frequency unit.
range. In measuring this response curve, the diaphragms of both
low- and high-frequency units have been located so as to be the same
distance from the measuring microphone, and, therefore, no irregular-
ites around crossover are indicated. The engineer realizes, however,
that he could have separated the diaphragms by a distance equal to a
half wavelength at crossover frequency and obtain an outdoor curve
indicating irregularities in response around crossover. He also recog-
nizes that while the outdoor response of such a setup would look in-
ferior, the speaker would sound just as satisfactory under listening
conditions. His indoor response curves will not show these phasing
difficulties due to crossover.
DISTRIBUTION
The most frequently published loudspeaker frequency-response
curves are those measured on the axis of the speaker in free space.
Such data are a measure of a very minute portion of the total energy
1948
THEATER LOUDSPEAKER SYSTEM
391
radiated by the device, and were nothing more known about the
speaker, they would provide a very incomplete picture of the perform-
ance of the instrument. For most applications it is conceded that
the response of the speaker should be uniform over an angle encom-
passing the area in which listening is to take place. For outdoor in-
stallations this is usually a simple and logical requirement to set.
When the device is intended for use indoors, however, the situation
is not so clear-cut. No loudspeaker known at the present time will
provide uniform response over a desired area and zero response out-
side this area. Thus, in a room such as would be suitable for listening,
500 1000
FREQUENCY IN CYCLES PER SECOND
Fig. 6 — Impedance of low-frequency unit computed at high side
of 24- to 4-ohm transformer.
much of the sound reaching an observer's ears will be reflected energy,
and the effect of the directional characteristics minimized. In spite of
this situation, most experienced observers agree that a uniform direc-
tional characteristic is a desirable attribute even for indoor listening.
It is very difficult to design a practical loudspeaker which is capable
of producing the same directivity pattern for all frequencies in its
radiation spectrum. Multiple-unit devices approach this objective
by limiting the frequency range reproduced by the individual units.
As will be evident from an inspection of Fig. 4, two unit systems may
be made to produce a remarkable uniformity of distribution. Wide
variations of pressure are observed to occur at only one or two fre-
quencies throughout the range at the extremes of the coverage angle.
Sectoral horns, if carefully designed, are capable of producing uni-
form radiation patterns over wide frequency ranges. The desirable
392
HOPKINS AND KEITH
October
directional characteristics are obtained at frequencies where the wave-
length of the radiated sound is small compared to the width of the
sector measured at the horn mouth. For this reason they are ex-
cellent devices for use with high-frequency units. Because of the size
involved, it becomes economically impracticable to use sectoral horns
as low-frequency radiators. Flat-mouth, rectangular section, low-
frequency horns are commonly used because of their structural sim-
plicity, but they have the disadvantage that their directionality in-
creases with frequency. This may be compensated by proportioning
the mouth so that it provides the desired distribution of sound at the
\
500 1000
FREQUENCY IN CYCLES PER SECOND
Fig. 7 — Free-field response of low-frequency unit in box baffle.
crossover frequency. If the low-frequency houdspeaker is then de-
signed with a drooping low-end response under free-space conditions,
the energy radiated may be made to be approximately uniform.
Under indoor conditions, a satisfactorily uniform spread of energy
will be apparent when these design objectives are attained.
EFFICIENCY AND POWER CAPACITY
The acoustic output available from a loudspeaker at which accept-
able freedom from distortion exists and at which no mechanical failure
occurs, is, in general, limited by the mechanical design. An increase in
the efficiency of the device can only make it possible to achieve this
limiting output with lower-powered amplifiers. A compromise be-
tween the cost and weight of amplifier and loudspeaker must be
struck in the interests of over-all system economy. Obviously effi-
ciency, power capacity, and amplifier power must be considered in
determining the needs of a given installation.
1948
THEATER LOUDSPEAKER SYSTEM
393
A method of determining the loudness-efficiency rating of loud-
speakers, and of applying it to power requirements in enclosures has
been described at a meeting of this Society,1 and published in the
Proceedings of the I.R.E. When measured in accordance with this
method, a loudness efficiency rating of 20 per cent is indicated for the
loudspeakers described herein. This efficiency is based on the acoustic
power radiated over a 300- to 3000-cycle sweep-frequency band and is
ENCLOSURE FOR TERMINALS
DIAPHRAGM
POLE PI EC
MAGNET
Fig. 8 — Cross section of high-frequency unit.
higher than the figure for previous commercial Western Electric
systems. An accurate comparison with other systems is not possible
until they are measured on the basis given in the above paper.
While the efficiency of loudspeakers is generally regarded as an im-
portant consideration in determining their suitability, it is a term
which is frequently misinterpreted. Axial-response data are often
exhibited as an indication of efficiency, whereas, as has been pointed
out, this is a measure of a very small portion of the energy radiated.
The directivity must be taken into account in any determination of
efficiency based on pressure response.
DAMPING
Horn-type speakers, if used for the frequency range well above" the
designed cutoff frequency of the horn, usually have well-damped
mechanical systems. Because of size limitations, however, it is
394
HOPKINS AND KEITH
October
customary to use low-frequency horns at frequencies down to, and even
below, cutoff. At these extremely low frequencies, the air loading will
be small, and other means must be resorted to. Acoustical and
mechanical resistance elements have been built into the Western
Electric speakers to provide the required low-frequency damping.
Such features result in a ' 'firmer," less boomy, bass response. The use
of bass "booster" devices is, in general, inimical to well-damped
response.
Fig. 9 — Sectoral high-frequency horn.
DISTORTION
The problem of distortion ratings for loudspeakers is rather com
plex. Speakers do not, in general, exhibit uniform distortion-fre
quency characteristics, and, therefore, the choice of frequencies o
which to base distortion measurements is likely to be different fo
each type of instrument. Furthermore, loudspeakers may produc
relatively large amounts of distortion in one or two narrow-frequenc
bands, and it is difficult to evaluate the subjective effect due to such
condition- as compared to a smaller degree of distortion over a wide
frequency range. It would seem reasonable to base a distortioi
rating on total acoustic power output, which makes it necessary t
search wide-frequency bands over wide-dispersion areas, or to comput
the acoustic outputs at the various frequencies through the use o
1948 THEATER LOUDSPEAKER SYSTEM 395
a theoretically derived directivity index. The problem of distortion
measurements in loudspeaker systems is not insoluble, but up to the
present time only limited work in this field has been undertaken, and
no standardized procedure has been worked out. It must be pointed
out, too, that the results of steady-state measurements may not be a
proper indication of the performance of a device intended to reproduce
chiefly transient material.
Certain facts are obvious, however, and qualitatively, at least, the
distortion may be controlled. It is known, for instance, that non-
linearity must exist in a system in order for distortion to be present.
Consequently, control of stiffness and utilization of a linear flux field
over the amplitude range required will control the first-order effects.
Since various modes of diaphragm vibration may show up within the
frequency range, and since nonlinearity may exist in some of these
modes, distortion may result. Such modes, however, may be con-
trolled by the judicious use of damping material. Listening tests
usually will evaluate the efficacy of the measures taken. The effect
of such damping on both the steady-state and transient response plays
a large part in the clean performance of a high-quality loudspeaker
system.
DESIGN FEATURES OF NEW SYSTEMS
With the above considerations in mind a series of theater loud-
speakers has been designed to cover the range of power input and
angular distribution needed for theaters of various sizes and shapes.
In designing these horn systems, a crossover frequency of 800 cycles
was chosen as the result of listening tests on systems having various
crossover frequencies. It has been found an advantage to have any
effects due to out-of-phase conditions between the low- and high-
frequency loudspeakers come above the region of maximum energy
transmission rather than in the middle of this range. This relatively
high crossover frequency makes it possible to use smaller high-fre-
quency units since less power is transmitted in the high-frequency
horn system. It also makes possible the use of a smaller high-fre-
quency horn due to its higher cutoff frequency. However, such a
crossover is possible only through the use of a low-frequency unit and
horn designed to transmit adequately the wider low-frequency band.
A new type of low-frequency unit and a special low-frequency horn
make this possible.
The low-frequency units, Fig. 5, utilize a comparatively flat dia-
phragm in place of the usual cone in order to reduce phase differences
396
HOPKINS AND KEITH
October
in the sound radiated from different portions of the diaphragm. The
diaphragm surface is designed to provide high rigidity with light
weight and to reduce "breakup" of the diaphragm at higher frequen-
cies. This is accomplished by means of an approximately spherical
central dome and an outer portion in the form of a surface of revolu-
tion of a logarithmic curve. The voice coil is attached at the junction
of the central dome and the curved outer surface. The arrangement
Fig. 10 — Front view of 60-watt theater system.
of permanent magnets (Alnico 5) is shown in Fig. 5, consisting of two
cylindrical magnets, one inside of the other. This provides high flux
density with minimum weight and depth. Special acoustic and me-
chanical resistance elements provide relatively high damping. This
is illustrated by the impedance curve, Fig. 6. The unit has excellent
high-frequency response as shown in Fig. 7 and is used in many in-
stances as a full-range loudspeaker, but in this system it is utilized
only for the range between 50 and 800 cycles.
The low-frequency horn has an exponential taper with a sector-
shaped horizontal section. This makes possible the combination of
1948 THEATER LOUDSPEAKER SYSTEM 397
two low-frequency horns side by side for additional power output
without making the combination too directional. The low-frequency
units are mounted in an enclosed cavity back of the horn which ob-
viates the difficulties due to back radiation, such as an increase in
response at certain frequencies and a decrease at others with attendant
"hangover." Sound-absorbing material within the cavity prevents
Fig. 11 — Rear view of 60- watt theater system.
standing waves, which might react on the diaphragm and cause ir-
regularities of response. Flat baffle sections are provided to improve
the response at low frequencies without producing resonant effects.
The high-frequency units used in this system are similar to units
previously used except for increased power capacity and efficiency.
The former is accomplished by the use of a phenolic diaphragm and
the latter by an improved permanent magnet. As shown in Fig. 8,
sound waves created by motion of the diaphragm are conducted
through expanding channels to a throat extending through the central
pole. These units are capable of excellent reproduction up to con-
siderably above 10,000 cycles.
398 HOPKINS AND KEITH
High-frequency horns are designed as single units with exponential
taper and a horizontal section of uniformly increasing width. This
sector-type construction is simple and capable of giving smoother re-
sponse at high frequencies and better distribution than previous types.
The design is such that two horns, designed for a horizontal distribu-
tion of 50 degrees, Fig. 9, each with a driving unit, may be combined
to give one horn of 100-degree distribution. This combination avoids
the possibility of impedance irregularity which may occur when a
double throat is used on a single horn. An 80-degree horn of similar
design is used where this distribution angle is required.
The • dividing network is designed to operate from an amplifier
having an output rating impedance of 24 ohms although both low-
and high-frequency units have impedances of 4 ohms each. This re-
duces the size of the network and the wires leading to it from the pro-
jection booth. Step-down transformers are incorporated in both
low- and high-frequency circuits to provide a proper impedance match
whether one or two units are used in either low- or high-frequency
circuits. An adjustable attenuator having 1-decibel steps between 0
and 5 decibels is included in the high-frequency circuit.
Outdoor measurements on a typical system (60-watt, 100-degree)
are shown in Figs. 3 and 4. Fig. 3 shows response on the axis while
Fig. 4 consists of superposed curves taken every 5 degrees off the axis
in a horizontal direction. It will be noted that except in the neigh-
borhood of the crossover frequency, 800 cycles, the curves for the
various horizontal angles fall very close together. The departures
near crossover are only at the extreme angles and cover such narrow-
frequency bands that they can be neglected. The general construc-
tion of a typical system is shown in Figs. 10 and 11.
REFERENCE
(1) H. F. Hopkins and N. R. Stryker, "A proposed loudness-efficiency rating
for loudspeakers and the determination of system power requirements for en
closures," presented April 24, 1947, at the SMPE Convention in Chicago; Proa
I.R.E., vol. 36, pp. 307-315; March, 1948.
Modern Film
Re- Recording Equipment*
BY WESLEY C. MILLER
METRO-GOLDWYN-MAYER STUDIOS, CULVER ClTY, CALIFORNIA
AND
G. R. CRANE
WESTERN ELECTRIC COMPANY, HOLLYWOOD, CALIFORNIA
Summary — Here is described a recently installed modern cabinet-type re-
recording equipment, having a completely new approach in design and opera-
tion. Radically new, but proved features have been incorporated for ease
and economy in installation, operation, and maintenance. The over-all
functional design is based on original concepts by Metro-Goldwyn-Mayer
and includes the manufacturer's recent basic developments in film-pulling
mechanisms and optical systems.
THE RECENT completion of the new film re-recording installation
at the MGM Studio in Culver City and the progress of the
parallel installations in the Elstree -Studio in London and the various
MGM International Studios suggest it may be an opportune time
to describe the installation and apparatus components and to com-
ment on the underlying philosophy of the design concepts involved.
Some preliminary work was done several years ago resulting in the
completion of experimental models of apparatus units described at
that time1. As a result of the experience gained from operation with
these models, and also to take advantage of the improved film
motion developed by Western Electric, for which recognition was
recently given by the Academy of Motion Picture Arts and Sciences,
further design work has resulted in the present MGM units. This
paper primarily will be concerned with the re-recording machine, and
although a number of special features were custombuilt for the MGM
installations, the basic design principles are incorporated in the
standard Western Electric units.
Motion picture re-recording has many ramifications. It is a part
of the picture-making technique which reflects the sound engineer's
ingenuity in finding answers to the many problems, suggestions, and
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
OCTOBEK, 1948 JOURNAL OF THE SMPE VOLUME 51 399
400 MILLER AND CRANE October
inspirations presented by all of the other individuals and groups who
contribute to the finished product. It is also the place where the
soundman ceases to be an engineer and becomes a controlling and
creative factor hi the successful presentation of the product to the
public. To an increasing degree the interpretations of the producer,
the director, the musician, and the editor are dependent upon the re-
recording processes and upon the skill and understanding of the re-
recording mixer.
There was a time when a single domestic release negative would be
re-recorded from perhaps three or four sound tracks, made up of a
dialog track, .one or two music tracks, and one or two effects tracks.
Such simplicity sometimes would be welcome now but the require-
ments and demands of the modern product require a great deal more
complexity. Eight re-recording tracks are probably a fair average
requirement and the need to use ten or twelve tracks arises fre-
quently. In particularly complicated reels or sequences there are
many cases where twice this number of tracks may be used.
It is common practice to make sound effects and music negatives
for foreign synchronized versions at the same time that the domestic
release version is re-recorded. This involves additional recording
machines together with the amplifier channel extensions needed for
the multiple job. Moreover, a 16-mm version is now standard
practice. Added to these requirements is the need for several
productions to be in work at the same time, together with all of the
routine preparation of playback material for production stage use,
temporary re-recording for immediate editorial purposes, publicity
and broadcast material, and other irregular but continuing demands.
This variety of requirements obviously presumes a large amount of
equipment. For the MGM Culver City Studio this necessitates forty-
one film reproducers, or film dummies as they are commonly known,
and eight film recorders, of which some are for theater release 100-mil
variable-density track, others for 200-mil push-pull for various studio
and international uses, one is for 16-mm release and another for vari-
able-area release when required. These machines operate with any
one of four re-recording channels and from corresponding re-recording
auditoriums and projection rooms. To these film facilities is added
the necessary disk equipment for recording and reproducing which
are constantly in demand. Later there will be magnetic equipment
as well.
The artistic phase of re-recording and the technique of equalization,
1948 FILM RE-RECORDING EQUIPMENT 401
balance, and editorial construction are of such nature that it may well
be expected that they will be changing continually to conform to
current and future requirements. While there are certain elements
of uniformity, departures from the uniform pattern are not only ex-
pected but are to be desired in the attempt to produce improved
entertainment and technical quality.
On the other hand, the physical handling of film material from
which re-recording is done is a matter which should be undertaken on
a basis which is largely routine and which can be semiautomatic in
character. It is with this phase of the work that the present dis-
cussion is principally involved.
EQUIPMENT REQUIREMENTS
Some of the more important attributes of re-recording equipment
which will meet present requirements and which may be expected to
continue to meet these requirements for many years to come may
readily be listed as follows:
(1) Virtually absolute uniformity of film motion, regardless of
film or drive irregularities.
(2) Consistent and dependable operation of all electrical and op-
tical elements.
(3) Simplicity and rapidity of operation and manipulation. .
(4) Ease and economy of maintenance.
(5) Installation simplicity and economy.
The experience of the past few months with the installation at
Culver City has shown that the design concept is admirably suited to
the operating requirements. Similar reports have been received from
the smaller installations.
MGM INSTALLATION
Three kinds of equipment units are involved; a film recorder,
its associated control cabinet, and the film reproducers, or re-record-
ers. Fig. 1 shows the form which the equipment takes. Each unit
is housed in a rectangular sheet-metal cabinet containing all of the
equipment associated with the unit . The height and depth of all units
are the same, with the width varying with the nature of the unit.
This permits arranging the units in rows of any desired length and
in any pattern which floor plan and operational procedure suggest.
Inasmuch as all of the elements associated with a given unit are in
402
MlLLEK AND CRANE
October
the same cabinet, the complete unit can be completely assembled,
wired, and tested under shop manufacturing conditions. The in-
stallation thus becomes very simple as it requires a minimum of
external wiring to terminal blocks in each cabinet.
Past experience has shown that the installation costs are very high
when it is necessary to install and wire a number of individual
.
Fig. 1— Typical arrangement of control cabinet, re-
corder, and re-recorders as used at MGM Studio, showing
similarity of the units.
mechanical units, amplifiers, and all of the associated items when the
work has to be' done as a matter of individual installation. As com-
pared with this the installation of complete equipment units can be
done quickly and inexpensively.
In this connection, as will be shown in more detail later, one differ-
ence between the MGM re-recorders and the standard Western
Electric units lies in the arrangement of the panel above the film
1948
FILM RE-RECORDING EQUIPMENT
403
compartment of the re-recorder. An MGM requirement was the use of
a six-position rotary switch in this location, arranged to switch the re-
recorder interlock motor to any one of six separate distributor trunks.
With a suitable safety provision in this switch it then becomes prac-
ticable to shift a reproducer rapidly from use with one re-recording
system to any other. This facility of interchange is important in
saving time as assignments in various rooms are constantly changing.
To accommodate these switches some fifty-odd trunk lines of power
capacity are involved and with normal cabling and connection
Fig. 2 — Top view of equipment shown in Fig. 1, with covers removed to show
the distribution of wiring and switching arrangement used at MGM.
methods this would have been an expensive installation item. Fig. 2
shows the solution attained. A flat bakelite-supported form was
constructed before installation. At each junction point in the trunks
and also at each switch connection appropriate Stakon connections
were employed. When the equipment units were in place, the pre-
fabricated flat form was dropped in place and the Stakon connectors
plugged together. For a row of six reproducers and their associated
recorder and controls the number of connections is of the order of 1200
so that this substitution of Stakons and the prefabricated form for
soldered connections and standard cable forms results in an appre-
ciable saving in initial installation. The standard Western Electric
machine uses this space for drawer-type mounting of the phototube
404 MILLER AND CRANE October
amplifier available from the rear, and the front area is used for lamp
and other controls. In the MGM form the phototube amplifier is
placed in the bottom compartment, also accessible from the rear.
The floor plan for the MGM installation is shown by Fig. 3. In
this case, each row of units includes six film reproducers, one film
recorder, and one control cabinet, so that in the two opposing rows
there is the direct association of twice this number of units. This
choice of numbers and arrangement is based upon the particular
experience and requirements at MGM. It varies, of course, among
studios, and the use of existing building space is an important con-
sideration. In the case of the MGM installation all of the units are
finished in white enamel to accentuate the impression of cleanliness.
This feature has created favorable comment and is of course a depar-
ture from what has heretofore been thought of as standard practice
for this type of equipment.
The conventional overhead sprinkler system is used for fire pro-
tection. However, high-temperature heads are installed to mini-
mize the possibility of sprinkler operation at the wrong time. In
this connection it should be noted that there is no potential fire-
producing element in the machines themselves, barring the remote
possibility of a static condition.
RE-RECORDING MACHINE
The film motion and mechanical drive are identical in the repro-
ducer and recorder. Minor changes provide for the use of 17.5-
mm film in the reproducer and of 35-mm film in the recorder. Aside
from this there are, of course, the expected differences which permit
the use of the proper modulator and of unexposed film in the recorder,
• whereas the reproducer is adapted for positive track and for the re-
producing optical and phototube systems. A pair of the standard re-
recorders is shown by Fig. 4 and the most unique features are perhaps
the disappearing doors, the removable mechanism unit, the film path,
and the optical system. A two-section cabinet permits the choice
of a separate base cabinet unit, which may be either a plain unit or one
containing a series of fixed and adjustable rollers for handling a loop
up to 25 feet long. However, the loop may extend from one lower
loop rack area into adjacent machines if desired. The cabinet itself
is of rigid construction, well ribbed to avoid panel vibration, and has
removable rear doors for accessibility to all components.
In spite of the fact that the room in which this equipment is
1948
FILM RE-RECOKDING EQUIPMENT
405
d t
d t
I
m
I
I
406
MILLER AND CRANE
October
installed may be air-conditioned it was deemed desirable to have doors
which close the film compartments of the reproducers. The doors are
of heavy glass and are so hinged that they are flush with the front of
the cabinet in both the open and closed positions. When swung open
they disappear into the sides of the cabinet. These doors normally
*%!&*'
Fig. 4 — Two RA-1251-type re-recording ma-
chines with the standard control panel and loop-
rack cabinets. Left machine threaded for normal
operation; right machine for loop operation using
upper or lower racks or both.
are operated manually .with rubber bumpers provided for protection,
but in the case of the MGM machines, they are electrically operated
by a momentary contact push button located at a convenient height
on the side of the machine as shown in Fig. 1. Suitable precautions
have, of course, been taken so that the hand or any obstacle caught in
1948
FILM RE-RECORDING EQUIPMENT
407
the door will cause the door to stall without damage or excessive force.
These cabinets are intended to be anchored to the floor with wiring
coming up through the floor.
MECHANICAL DESIGN
Attention is particularly invited to the unusual arrangement where-
by the actual film-motion unit and its associated accessories for both
reproducer and recorder are readily removable from the cabinet. The
film-motion unit slides in and out from the rear and is arranged to be
kept in mechanical alignment when it is pushed into place, and to
V
lls
Fig. 5 — Rear view of removable mechanism unit of the
RA-1251-type re-recording machine.
make the necessary electrical connections, with the exception of a
plug to the phototube amplifier. In addition, each of the major
mechanism components is removable as a complete subassembly and
all such assemblies are interchangeable in case of emergency. This
complete mechanism unit is shown by Fig. 5, and facilities are pro-
vided for the removal and replacement of the various major subas-
semblies without loss of precison adjustments. For example, the motor
position is adjustable vertically and laterally for shaft alignment,
after which the adjustments are locked and the motor and its base
may be removed thereafter and replaced without losing the align-
ment. The same principle is applied to the flywheel and drum
408 MILLER AND CRANE October
assembly to retain optical adjustments related to the portion of the
optical system which is mounted within the scanning drum.
The scanner assembly, shown by Fig. 6, contains a scanning drum
driven by the film, and it is rigidly connected to a solid flywheel, the
shaft being supported on two very small ball bearings so that the
friction is held to a minimum and in addition, any moderate imper-
fections in the bearings are negligible, thereby avoiding a high degree
of bearing selection. Fig. 7 shows the gear-drive assembly which also
contains the sprockets and associated pad rollers. The gear re-
Fig. 6 — Complete scanning drum and flywheel assem-
bly, including optical subassembly under cap at end of
scanning drum.
duction is accomplished in two stages with the rear section contain-
ing a high-ratio, right-angle drive for which gears are available to
accommodate motor speeds of 720, 1000, or 1200 revolutions per
minute, or other ratios as might be required for 16-mm operation.
This assembly also contains a built-in clutch to disconnect the motor
from the recorder mechanism so that the latter alone may be turned
over by hand. It has been found to be a great convenience to be able
to thread rapidly and then register the start mark on the film without
disturbing the motor from its interlock position. The clutch is, of
course, of the positive type which does not permit any slip during
normal operation. The sprockets have flanges to facilitate threading
1948 FILM RE-RECORDING EQUIPMENT 409
and improve film guiding. The sprocket teeth are unusually large,
having a base of 74 mils in the direction of the film. With a base diam-
eter of 0.942 inch, this sprocket permits nearly optimum operation
over a shrinkage range of 0 to 0.8 per cent and considerably greater
shrinkage can be accommodated without significant damage to the
film. These sprockets reduce to negligible amount the so-called
"crossover" effect, which is the erratic motion of the film over the
sprocket within the limit of sprocket-tooth clearance in the sprocket
holes. The sprocket pad rollers are of uniquefdesign in which the pad-
Fig. 7 — Complete drive-unit assembly from motor coupling
to sprockets. Vertical lever releases built-in clutch for thread-
ing release.
roller assembly pivots in the same plane as the sprocket axis and the
finger pads are lo'cated so that the two pad-roller assemblies may be
opened simultaneously by the thumb and first finger. Fig. 8 shows
the filter-arm subassembly consisting of two rollers mounted on
pivoted arms controlled by springs and arranged so that only the
rollers appear through slots on the front of the panel. One spring pro-
vides for the tension in the film path between two sprockets. The
other spring compensates for the total accumulated friction of all
rotating components in the path between the two sprockets. This is
sometimes referred to as the "gravity spring" and is anchored to the
410
MILLER AND CRANE
Octobei
frame through a cam control appearing on the front of the panel)
This cam then becomes a vernier adjustment of synchronization whilj
the machine is running or stationary, and a range of approximately i
±2 sprocket holes is available. The filter rollers as well as. the fixed
rollers in the film path are, ol
course, ball bearing with pre ];
cautions taken to reduce friction]
to a low value. The lower arni
is provided with a fluid dashpolj
for proper damping and is arl
ranged to prevent spillage of thl
fluid at any angle.
Film rewinding is provided bjl
a motor and gear-reduction unit
located behind the upper pane
and connected to the feed-reel
4|1 shaft with an automatic cuton
assembty located in the uppei
right-hand corner of the angle
plate assembly. This facility
provides automatic rewind once
the film is threaded and rewind
ing started by the operator throw'
ing a small idler roller into con
tact with the film. The rewinc
time is adjustable between 3(
seconds and one minute for 10CK
feet of 35-mm film and the film
velocity is reasonably constanl
as provided by the usual char-
acteristics of a high-speed, series-type motor.
FILM MOTION
The film propulsion of this machine is essentially the same in prin-
ciple as that previously described in the JOURNAL. 2 As shown by Fig.
9, a taut film path between two 16-tooth sprockets drives a scanning
drum and flywheel by belt action, and passes over two compliance)
rollers, one to the left of each sprocket. One of these rollers id
provided with viscous damping and this film-propulsion system hasj
demonstrated its ability to suppress all mechanical disturbances genern
ated in the drive mechanism as well as those caused by film splices.
• Fig. 8 — Filter-roller assembly show-
ing damping device attached to lower
roller arm and locking lever which is
actuated by the lower sprocket pad
roller. Spring at right determines
film tension in filtered path and that
at left is anchored on cam for syn-
chronization adjustment.
1948
FILM RE-RECORDING EQUIPMENT
411
As an aid in threading, the two filter rollers are locked in the normal
operating position when the lower sprocket pad roller is open for
threading. Threading is therefore reduced to a very simple and fast
Dperation since no loops are required. As the film is placed over the
lower sprocket and the pad roller closed, the filter arms are thereby
Fig. 9 — Front view of mechanism unit showing film threaded
for operation.
released and ready for operation. As previously described, the clutch
button may then be used for setting the start mark.
OPTICAL SYSTEM
The optical system employed in this re-recorder is somewhat of a
departure from previous optical systems and was designed to meet the
requirements of convenience in operation, relatively high efficiency,
412
MILLER AND CRANE
Octoben
freedom from fire hazards, and versatility in scanning any type oij
track in current use, with a high degree of uniformity in light in-j
tensity and definition in the scanning beam. Years of practical
experience in optical systems have demonstrated the safety of son
called front-scanning systems, but these systems are not readily obJj
served for scanning performance and require frequent use of various^ j
types of alignment or test tracks. The convenience of rear-scanning
systems for visually determining scanning is well known, but manw
PHOTOTUBE
CYLINDRICAL LENS
SPHERICAL LENS
SPHERICAL LENS
SPHERICAL LENS
CYLINDRICAL LENS
CYLINDRICAL OBJECTIVE LENS
YLINDRICAL OBJECTIVE APERTURE MASK
CYLINDRICAL RELAY LENSES
-SLIT MASK
"CYLINDRICAL CONDENSER LENS (VERTICAL)
"CYLINDRICAL CONDENSER LENS (LATERAL)
"CURVED FILAMENT EXCITER LAMP
Fig. 10 — Optical schematic, with omission of prism and mirrors used only for I
turning optical axis by 40 degrees.
such systems have contained fire hazards because of the relatively]'
large amount of light placed upon the film by the condenser-lens i
system. In order to retain the advantages of each of these systems, II
the one used in this machine combines the convenience of rear scanning|
with the safety of front scanning. This is accomplished by using
the front-scanning method of placing a scanning line upon the film in]
which the width of the line determines the frequency characteristic!
in reproduction, but which is of sufficient length to more than cover I;
the area occupied by all sound tracks in current use. A rear-scanning i
type of system is employed whereby an enlarged image of the film and]
1948 FILM RE-RECORDING EQUIPMENT 413
this scanning line is produced upon a mask so that the limits of scan-
ning are readily observable and adjustable at any time. Fig. 10 shows
this system schematically, omitting a prism and mirrors which merely
turn the optical axis by 90 degrees for mechanical convenience.
The basic requirement of the front-scanning section was that of pro-
ducing a line in which the intensity and the definition were to be reason-
ably uniform throughout its length as well as being highly efficient.
This was accomplished by the use of cylindrical lenses only and the
scanning line is an image of a physical slit produced by a relatively
short focal-length cylinder. The height of this line is equivalent to an
ideal scanning slit of approximately 1.0 mil and its quality and defini-
tion are quite uniform throughout its length. The illumination is
uniform to within ±0.5 decibel. The light source is a 10-volt, 5-
ampere, curved-filament lamp and the optical constants are such that
its vertical position is not critical, thereby permitting prefocused
lamps and essentially eliminating microphonic noise generated by
lamp vibration. The stereopticon type of system is used and the
advantages of the curved-filament lamp have been described by
Carlson.3
The rear-scanning system consists of a combination of spherical
and cylindrical elements, the design of which is primarily dependent
upon the ability to collect and control all of the light coming from the
scanning assembly and to meet the physical requirements of the
machine design. It consists of three spherical lenses with the ad-
dition of two cylindrical elements between the film and the scanning
mask. The first cylinder is located just behind the film to collect the
relatively large vertical angle of light, and this cylinder in com-
bination with a negative cylinder located on the rear of the third
spherical lens produce a vertical enlargement of the scanning line of
the order of 100 : 1 to permit more convenient observation of the
scanning limits relative to the mask. Although not shown on Fig. 9,
there is a prism located between the first two spherical lenses for the
purpose of offsetting the beam to clear the scanning drum. Other
mirrors are used to bend the light depending upon the particular
application. The three spherical lenses produce an image at the mask
which is magnified laterally approximately 3 : 1 and a scanning mask
contains three sets of openings, one of which is registered with the
projected light beam to limit the scanning. These openings provide
for 200-mil push-pull in either the standard or offset positions and
for 100-mil single or push-pull. In the vertical direction all of the
414 MILLER AND CRANE October
projected light beam being scanned passes through the opening in the
mask and appropriate field lenses are located just behind these open-
ings to direct the light into the phototube. An RCA 920 tube is used
and the patterns on the cell cathodes represent filament images about
I/B by 3/s inch long so that they are readily accommodated by the
standard cathode construction. These images are essentially variable
in intensity only, regardless of the manner in which the light is atten-
uated at the film plane. Fig. 11 shows the mechanical embodiment
of the optical assembly between the lamp and the film. In this
Fig. 11 — Complete scanning optical system from
lamp to film, which places the lamp behind the
front panel and produces a line of light on the film
1 mil high by 230 mils long.
machine the lamp is located behind the front plate to eliminate all
heat and fire hazard. A mirror turns the axis 90 degrees and this
complete subassembly contains lateral and vertical lamp adjustments
as well as focus and azimuth adjustments. The latter is attained by
a rotation of the entire subassembly about its optical axis. The
scanning assembly contains the third spherical lens of the collector
system, the scanning masks, field lenses, and associated controls.
The left-hand knob controls the position of the mask for the type of
scanning desired. The right-hand knob moves the mask laterally for
proper scanning registration and this knob is calibrated to indicate
the normal scanning position or the departure from this normal with
1948
FILM RE-RECORDING EQUIPMENT
415
calibration in mils. A viewing window is provided on the front of
this assembly with a sliding cover. This system has met the design
objectives and gives an output signal at the phototube which results
in an unusually high signal-to-noise ratio relative to mechanical
disturbances which frequently have been so troublesome in the past.
This optical system has also been applied to theater-type sound-
heads and a typical example is shown by Fig. 12. The same facilities
Fig. 12— Theater- type reproducer using the same optical system as Fig. 10.
are available and the performance is essentially equivalent. In this
case the elements between lamp and film are contained in a circular
tube, and are designed to conform to the usual distance between lamp
and film, which is approximately 47/i6 inches.
PHOTOTUBE AMPLIFIER
The phototube amplifier employes a new feature heretofore not
used on this type of equipment. This consists of a special feedback
input circuit which permits several feet of shielded cable to be used
between the phototube mesh and the amplifier. By locating the
416
MILLER AND CRANE
October
amplifier away from the congested area around the machinery, greater
freedom of design is obtained both for the mechanical system and for
the amplifier. Also less microphonic noise and other noise disturbance
generally will be obtained with the separation. Ordinarily the capac-
itance associated with a cable connecting phototubes and amplifier
Fig. 13 — Phototube coupling unit showing front view with phototube
switch and balancing controls, and rear view with cover removed.
would cause a serious loss of high-frequency response. However, in
this application the feedback from plate to grid of the first tube
effectively cancels the capacitance of the cable so that an approximately
flat frequency characteristic is obtained up to about 8000 cycles per
second. No appreciable loss in signal-to-noise ratio is incurred over
that of more conventional methods. The amplifier has three
stages with feedback around the last two stages as well as the* first
stage. In the standard machine, it is mounted on a drawer
1948 FILM RE-RECORDING EQUIPMENT 417
arrangement which plugs in, thus providing for good accessibility
and quick change in case of emergency.
Fig. 13 shows the phototube-coupling unit consisting of a complete
assembly mounted on rubber. It contains all of the electronic com-
ponents necessary for coupling the phototube through a single un-
balanced, shielded line to the amplifier which is several feet away. A
push-pull to parallel switch and a balancing potentiometer for push-
pull balance adjustment are available through the front of the panel,
but normally are under a cover since they are merely routine main-
tenance adjustments and are not required during operation.
CONCLUSION
At present the re-recording operations throughout the industry are
almost entirely from film reproduction to film recording. However,
there are good prospects of the adoption of magnetic methods for all
original studio recording and for all editorial work. The adaptation
of this new equipment to magnetic methods w~as kept in mind during
its design so that the change can be made with a moderate amount of
difficulty and expense. It is anticipated that magnetic methods will
be useful for every phase of studio work up to, but excluding, the
actual release negative.
The experience thus far with this new equipment at MGM and
elsewhere has been remarkably good. Uniformity of operation from
machine to machine had heretofore been difficult to attain with a
large number of machines. At MGM the machines are now assigned
to any class of work by number and any departure from uniformity
is virtually unheard of. The arrangement of machines and the
rapidity of threading and rewinding have greatly reduced the time
between successive rehearsals and takes. This is of importance to
the mixers and the producing group since it increases the amount of
production material which can be completed in each room each day
and thereby reflects directly on the operating economy of the re-
recording work.
REFERENCES
(1) Wesley C. Miller, "The MGM recorder and reproducer units," J. Soc. Mot.
Pict. Eng., vol. 40, pp. 301-326; May, 1943.
(2) C. C. Davis, "An improved film drive filter mechanism," /. Soc. Mot. Pict.
Eng., vol. 46, pp. 454-464; June, 1946.
(3) F. E. Carlson, "Properties of lamps and optical systems for sound reproduc-
tion," J. Soc. Mot. Pict. Eng., vol. 33, pp. 80-97; July, 1939.
Motion Picture Research Council*
BY W. F. KELLEY
SECRETARY-TREASURER, MOTION PICTURE RESEARCH COUNCIL,
HOLLYWOOD, CALIFORNIA
Summary — The purpose of this paper is to explain the organization, func-
tions, and activities of the Motion Picture Research Council, Inc.. A brief
resume of the Council's history and the reasons for its reorganization will be
given as an introduction.
SHORTLY AFTER the organization of the Academy of Motion Pictim
Arts and Sciences in 1927, a technical bureau was formed withinl
the Academy. The technical bureau, under the chairmanship 01
Irving Thalberg, collected and published information on the use 01
incandescent lamps on sound sets, conducted courses in industrial
education, and contributed materially to the solution of many prob^
lems encountered in establishing sound in motion pictures.
In 1932, the Academy was reorganized. The Research Council reJ
placed the technical bureau and functioned under the Academy by-
laws, but was sponsored and financed by the Association of Motion
Picture Producers. The governing body of the Research Council
consisted of one technical representative from each of the ten studios,
plus an executive-producer chairman. These chairmen have been
S. J. Briskin, William Koenig, Darryl Zanuck, and Y. Frank Freeman.
Experience in the operation of the Academy Research Council
demonstrated the restrictions of its particular organizational struc-
ture. Its activities were limited by lack of funds and staff, and the
necessity of operating through committees of technicians volunteering
their tune. However, the work of the Academy Research Council
demonstrated unmistakably the need and the possibilities of a prop-
erly organized and adequately financed research and development
program.
Several proposals for a research program were presented to and
discussed by studio executives, both in Hollywood and in New York.
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
418 OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51
RESEARCH COUNCIL 419
A committee of Herbert Freston, Peter Rathvon, and Y. Frank Free-
man, was appointed to decide upon a plan for the producers. Two
recommendations were made and approved. First, it was decided to
appropriate sufficient funds to establish an expanded program with
proper organization and adequate staff. Second, it was decided to
use the existing Research Council as the nucleus of the new
organization.
As a first step, it was necessary to secure a director for the new pro-
gram. In August, 1947, the services of Wallace V. Wolfe were ob-
tained. His executive ability, engineering background, and broad
knowledge of the motion picture picture industry made him particu-
larly well qualified for this position.
The second step was the establishment of a proper organization.
Operation of the Academy Research Council under the Academy had
been entirely satisfactory and practical, but limited. The new
organization needed greater freedom of action. It had to be able to
negotiate contracts, obtain patents, grant licenses, buy and sell prop-
erty. It had to be responsible directly to its sponsors for the ex-
penditure of funds.
This was discussed with the Academy and the necessity for a change
in the organization of the Council was recognized and approved by
Jean Hersholt, president of the Academy, and the Academy Board of
Governors. As a result, Herbert Freston drew up articles for a new
corporation and the Motion Picture Research Council was estab-
lished as a nonprofit California corporation on October 14, 1947.
Under its bylaws, its purposes are to engage in engineering develop-
ment and research, to find solutions to common problems, to develop
and improve equipment and methods, to promote standardization
and the interchange of ideas and information, and to act as a liaison
between studios and suppliers.
The corporation has ten company members : Columbia, Goldwyn,
Loew's, Paramount, RKO, Republic, Roach, Twentieth Century-
Fox, Universal, and Warner Brothers. These company members elect
a board of twelve directors. As presently constituted, the board con-
sists of one representative from each member company, plus an execu-
tive-producer chairman and the president of the corporation.
The management of the corporation is vested in a board of
directors, the officers, and committees.
420 KELLEY October
OFFICERS
WALLACE V. WOLFE, President R. A. KLUNE, V ice-President
W. F. KELLEY, Secretary-Treasurer
BOARD OF DIRECTORS
Y. FRANK FREEMAN, Chairman
THOMAS MOULTON, Vice-Chairman
JOHN AALBERG GERALD F. RACKETT
DANIEL J. BLOOMBERG ELMER R. RAGUSE
FARCIOT EDOUART GORDON SAWYER
BERNARD HERZBRUN DOUGLAS SHEARER
NATHAN LEVINSON WALLACE V. WOLFE
Committees are of three types: permanent committees, special
committees, and task groups or subcommittees. Permanent com-
mittees, appointed by the board, represent each major technical
division of motion picture production. There are fifteen such per-
manent committees:
Art Direction Electrical Set Grip Equipment
Color Laboratory Sound
Directors of Cinematography Photographic Standards
Editorial Production Managers Subgauge Film
Effects Set Construction Television
These committees are charged with the responsibility of acting in a
technological advisory capacity to the board and staff on all activities
pertaining to that committee's particular phase of production.
Special committees, also appointed by the board, advise on par-
ticular projects not within the scope of any permanent committee.
Task groups, subcommittees of permanent committees, are ap-
pointed to act in an advisory capacity on a single phase of a project,
and to make recommendations directly to its permanent committee.
The next step in the program was to secure proper quarters and
staff personnel. Office and laboratory space was obtained at 1421
North Western Avenue in Hollywood. This location has the ad*
vantage of being adjacent to studio facilities where operating tests
can be made.
Personnel requirements were based on the recognition that produc-
tion problems embrace every phase of the engineering profession. We,
therefore, employed an engineer for each of the major phases of our
program, consistent with the budget limitations of a new organization.
At present, in addition to Mr. Wolfe and myself, we have on the staff
1948 RESEARCH COUNCIL 421
eight engineers with practical experience in chemistry, physics, con-
struction, standards, electricity, lighting, and mechanical design.
As the organization was being set up, specific activities were being
outlined. Our program is still in its formative stage, but our present
activities indicate our future program.
First, the staff is analyzing new ideas and new inventions submitted
for possible application to motion pictures. Since the first of the year
we have considered such items as three-dimensional systems, tele-
vision patents, special cameras, color systems, aerial and underwater
photography, and universal focus lenses.
Second, we are searching for and examining equipment and ma-
terials developed by other industries for adaptation to the motion
picture industry. Included are such items as magnetic recording,
"liquid-envelope" materials, static eliminators, wall coverings, shellac
substitutes, nylon products, and molded screens.
Third, we are disseminating information to our member companies
through reports and bulletins and by discussion in committee meet-
ings.
Fourth, we are promoting standardization through the establish-
ment of industry practices and with a recommendation for American
Standards where these industry practices apply. We are presently
concerned with the standardization of screen illumination, dimen-
sions and speed of magnetic-recording mediums, pitch of sound-re-
cording negative, laboratory procedures for 16-mm release of 35-mm
material, elimination of frame lines, Dubray-Howell perforation, and
fused plugs and cables.
Fifth, we are actively engaged in short- and long-range projects.
Short-range projects may cover a period of a few days to a year or two,
while long-range projects may cover from one to five years. We are
now carrying on two long-range projects: set construction and set
lighting.
The purpose of the set-lighting project is to provide improved
lighting methods and tools to enable production crews to accomplish
their job with greater flexibility and improved efficiency.
A review of the literature is in progress and a survey of present
methods and techniques has been undertaken. This survey, partially
completed, includes light sources, lamphouses, optical systems,
filters, control equipment, power supply and, most important, the
manner in which such equipment is used on the set. At the same
time, an investigation was initiated to determine the possibility of
422 KELLEY October
employing new light sources which might be available. We are
actively following the development of the mercury-cadmium compact
light source and the zirconium lamp, and we are presenting designers
and manufacturers with broad operating specifications for the motion
picture use of. these lamps. Since this project is still in the staff-in-
vestigation phase, results cannot be reported at this time.
The set-construction project is also in the survey stage. A survey
of present methods and the literature is being carried on simultane-
ously with an investigation of the possibility of adapting products of
other industries to the construction of motion picture sets.
We have a number of short-term projects and it will suffice to ex-
plain one of these in some detail and merely list the others.
One of our short-term projects is the design and construction of a
small camera crane. There has beeri greater and greater need for
camera flexibility as production methods have progressed. The
camera has advanced from the stationary tripod to the dolly and to
cranes or booms. The dolly is somewhat limited in its use and the
large crane unwieldy and expensive to operate. Therefore, there is a
need for a crane with the mobility and flexibility of a dolly, but with
the camera range and broader application of the large crane. Several
small cranes had been developed previously. Metro-Goldwyn-Mayer
built, and still has in use, several cranes of intermediate size. Sub-
sequently, Twentieth Century-Fox developed a small, motor-driven
crane. The Council's crane is similar, differing mainly in the type of
construction and accessories.
Acting under the direction and advice of our Camera Crane Com-
mittee, and taking advantage of the previous knowledge and experi-
ence gained in the use of the small cranes, a staff engineer designed a
crane which was acceptable to all of the studios.
Normally, we would provide manufacturing firms with performance
specifications or a complete design, and they would manufacture and
sell directly to the industry. This would complete the Council's proj-
ect. Equipment required by studios, however, is often so specialized
that manufacturers are reluctant to produce the item for sale on the
open market. Such was the case with the crane. We were unable to
find a firm willing to undertake the manufacture because of the risk
involved compared to the possible market for such specialized equip-
ment. The Council, therefore, found it necessary to correlate studio
orders and arranged for the manufacture of an initial order of twenty-
five.
, 1948 RESEARCH COUNCIL 423
As a part of the crane project, we are building a prototype of a dual
camera head and have in the design stage a location carriage for the
crane.
Some other short-term projects under consideration are polonium
for static elimination, red-sensitive photoconductive tubes, materials
and equipment for simulating fog, more efficient wind machines, du-
plication of color stills for stereopticons, and nonfadirig dye agents for
plaster.
In conclusion, I should like to point out first, that we are engaged
in applied research, rather than in the opposite extremes of pure re-
search or manufacture. We are working closely with research groups
in and outside the industry and bringing to their attention problems
of importance to our industry.
Second, we are acting as a liaison between studios and suppliers.
On one hand suppliers are using the Council to distribute directly to
those concerned in the studios, information on new products. This
informational service is set up and working. On the other hand we are
correlating studio needs and desires and presenting such information
to the suppliers. This procedure saves time, standardizes methods
and practices, and results in better and less expensive equipment.
Third, our effort is an industry effort. The Research Council has
been set up by this industry for this industry. To quote Mr. Freeman,
our chairman, "we have in the motion picture industry one of the
largest reservoirs of competent and experienced engineering and
technical personnel of any industry." The Council needs the benefit
of this experience and knowledge. We not only welcome, but request
your help, your suggestions, and your advice.
Use of 16- Mm Motion Pictures
for Educational Reconditioning*
BY EDWIN W. SCHULTZ
ARMY MEDICAL CENTER, WASHINGTON, D. C.
Summary — This paper will cover in a general way some of the things
which have been done at the Walter Reed General Hospital with 16-mm
films and at the same time will offer a few suggestions which, it is believed,
will help to improve the motion picture industry so far as the 16-mm non-
theatrical field is concerned.
REAT IMPETUS has been given to the development of visual educa-
tional methods as a result of the demand for accelerated training
programs in the Armed Services. Training films, documentary re-
ports, and general informational subjects to aid in the orientation of
the soldier have established the value of motion pictures in supple-
menting other media of learning; also research studies have shown
the motion picture to be one of the most popular forms of entertain-
ment and diversion among American servicemen.
Within each hospital there are wide opportunities for educational
reconditioning personnel to develop programs well implemented with
carefully selected screen subjects that will aid in the psychological re-
conditioning and contribute to the resocialization of individual
patients.
The film program must be planned in a way that will accomplish
the following four objectives:
(1) To contribute to the individual's personal adjustment by pro-
viding information and fostering understanding of the hospital pro-
gram and providing local orientation to restore confidence, establish
respect, and develop a "sense of belonging."
(2) To develop the concept that the struggle in which we were en-
gaged required the total and continued effort of all, not only to win in
the field but also to secure a society dedicated to the principles of
democratic living.
(3) To offer occupational information co-ordinated with the pro-
grams of counseling and vocational guidance to aid in the exploration
* Presented October 15, 1945, at the SMPE Convention in New York.
424 OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51
FILMS IN REHABILITATION 425
of job opportunities and benefits available to the prospective dis-
charged soldier, should he require further education or vocational
rehabilitation.
(4) To supplement the content of specific courses of instruction
offered in educational reconditioning or the convalescent training
program.
The building of a film program, that will adequately serve the ob-
jectives of reconditioning, demands skillful balancing of films with a
serious purpose and those of a diversional and entertaining nature.
In order to carry out the film program as outlined, the Surgeon
General of the Army established a Visual Aids Center at the Walter
Reed General Hospital in Washington, D. C. It is the responsibility
of the Center to (1) provide necessary films, projectors, screens, and
other film equipment; (2) train projectionists and maintain equip-
ment; (3) determine all film sources and secure films for the programs
from all fields, insuring proper screening technique; (4) assist in the
proper utilization of films in the reconditioning, information, and
educational programs; (5) assist hi the adoption and improvisation
• of equipment for all such purposes in the hospital; (6) advise military
personnel of the hospital on the most effective ways of using motion
pictures and other visual aids in the programs; and (7) recommend
•to the Surgeon General production of visual material which can be
utilized in all Army Service Forces Hospitals to further recondition-
ing activities.
So far has been outlined the plan for the use of the 16-mm film in
the reconditioning program. Now let us talk a little about how it
i is used at Walter Reed General Hospital and later some of the
problems which confront us.
First, the theme at Walter Reed Hospital has been "If the man
cannot come to the movies, take the movies to the man." In other
• words, we believe that, what we in the hospital know as the "class
four" or bedfast patient is just as entitled to see entertainment and
educational movies as the "class three" or ambulatory man, who can
go to the places where motion pictures are shown. In order to do this
portable carts equipped with 16-mm sound projectors and screens
have been provided. These can be wheeled into the wards and films
shown for the men confined to their beds. These units, in the hands
1 of trained projectionists, give remarkably near-professional shows.
1 The lighting is good and the sound adequate. Audiences number
from a dozen or so up to 50 or 60 depending on the number of men
426 SCHULTZ October
In the ward. The reconditioning service confines its shows to educa-
tional, documentary, and vocational-type films primarily, while the
Red Cross furnishes the full-length entertainment features.
Fig. 1 — Cabinet- type booth for the 16-mm projector
which is in use at the Walter Reed General Hospital. This
booth is on the balcony of the Auditorium of the Red
Cross building at the hospital. The amplifier of the 16-mm
projector is jacked into the permanent 35-mm sound sys-
tem on the stage which makes it possible to show 16-mm
sound pictures on the big screen on the stage and get maxi-
mum tone qualities. This provides a more or less per-
manent setup in the auditorium for showing 16-mm pic-
tures at any time to the ambulatory patients.
Second, a daily information-education program is conducted in a
large theater auditorium which is attended by wheel-chair, crutch,
and walking patients. These programs usually last an hour and are
package or unit programs, a week being assigned to a subject during
1948 FILMS IN REHABILITATION 427
which outstanding speakers are brought in to discuss the subjects
along with which films are shown dealing with the various phases of
the subject. The sessions usually draw large audiences and the
program is broadcast over the hospital radio system so that men in
bed can hear it over their individual bedside headsets.
Fig. 2 — A discarded wheeled stretcher has been utilized
to make a cart for carrying complete projection unit to the
wards for ward showings of films at the Walter Reed
General Hospital. The addition of a bottom shelf and
the wooden top converts the stretcher into an easily
handled, practically silent, rolling cart. There's suf-
ficient room for projector, amplifier, speaker, screen,
films, and extra reels and is easily handled by one man.
- The third place where films are used is in the various crafts and
vocational shops. Among these shops is included carpentry, music,
art, leather tooling, and typewriting and business machines. Films
covering many phases of work in these crafts are shown to the patients
in a specially provided "little theater." While no attempt is made to
go into actual vocational training, an effort is made toward doing
exploratory work to help the patient find his fitness for a certain
428
SCHULTZ
October
craft or trade, and at the same time occupy his mind while the|
medical officers are curing him of his physical ills. Here we can!
only touch on the immense problem which we faced, and which still ]
faces us, and the progress that has been made. It will require a I
visit to Walter Reed Hospital to see how the job is being done.
Films are being employed in the treatment of the neuropsychiatric I
patients with very good results. Men mentally disturbed react!
extremely well toward films which
are shown them, and they have!
motion pictures almost every day. I
As an indication of the use of I
16-mm films at Walter Reed Hos-
pital, it might be well to give
you the figures on utilization
during the 3-month period of
June, July, and August, 1945.
During those three months, 598
films were shown with a total
of .1470 showings and a total
attendance of 46,760.
Among the problems which
were faced in bringing a film
program to the patients, was
first, the lack of the right kind
of 16-mm films available. Of
the many film subjects that are
available today in 16-mm sound,
a majority of them are unsuitable,
the sequences are bad, the pho-
tography poor, and the sound inadequate. The films made by the
Signal Corps Army Pictorial Service, were all excellent but were
unsuited for use in the hospital as they were filmed for military
training and our patients are through with that phase of their career.
The Surgeon General's Office started work on the evaluation and
procurement of available 16-mm films suitable for hospital use, but
only a few were acceptable. That office also launched certain pro-
ductions. The Visual Aids Center at Walter Reed started a campaign
to borrow from any source such pictures as could be used in these
programs. We faced the rental-fee problem right off, but we were
not in a position to pay rentals on all the films we wanted to use. In
Fig. 3 — A class in auto mechanics
see a film on first echelon maintenance
in the auto shop of the Educational
Reconditioning Section of the Walter
Reed General Hospital. More than
30 films are used continually by this
course during classes.
1948 FILMS IN REHABILITATION 429
most cases the firms having 16-mm films were most co-operative and
we secured the use of literally hundreds of prints with no charge other
than transportation.
In speaking of the lack of quality subject in 16-mm, let me make
it clear that I am not speaking of the 16-mm Hollywood feature
I pictures which have been released, but those of an educational or
^documentary nature. It has been my experience that there is a
very definite need for more high-class, well-produced films of an
educational and documentary type. It offers a great field, and
educational institutions consistently will demand a better quality
product; which is where an organization such as the Society of
Motion Picture Engineers can play a very definite part. Regardless
i of what some of the 35-mm producers want to believe, the sub-
standard 16-mm film is here to stay and the sooner all men in the
L industry realize it and make an effort to see that only the best quality
16-mm films are produced, the better off the entire industry will be.
It may be worth while to look ahead somewhat and consider the
great need the Veterans' hospitals will have for 16-mm films to carry
on the work of rehabilitation, resocialization, and vocational training
which eventually will be their responsibility. The work that is
being done is small, compared to what faces the Veterans Adminis-
tration in the years to come. While only a very small percentage
of the total number of veterans of World War II are in our hospitals,
' later there will be tens of thousands seeking the help of the Veterans
Administration.
Should we not look ahead and be prepared, and not face a "Pearl
Harbor" situation with respect to this matter?
Should we not use the experience we have gained and the time we
i. now have to prepare, rather than procrastinate and have to make a
i mad dash later?
If pioneer groups such as the SMPE and other agencies, both
i private and governmental, will carry on while the momentum is
rapid, there will be much accomplished; certainly, the future outlook
for 16-mm educational films was never brighter.
The third and last phase is that of projection equipment and
the shooting of pictures in 16-mm sound rather than the 35-mm
to 16-mm reduction method.
Considerable difficulty has been experienced during the war
years with 16-mm projection equipment. Many of the available
projectors were made so rapidly and out of such poor material that
430 SCHULTZ
maintenance became a major problem because of rough handling
and lack of well-experienced operators. Again a scarcity of repair
parts and good repairmen was a problem. The remarkable fact
is that they stood up as well as they did.
Now that the war is over, there should be an over-all improvement
in the quality of projection production, servicing, and a lowering
of costs, not only in the purchase prices of new projectors but in
the cost of replacement parts and maintenance. These facts are
mentioned as it is believed that 16-mm projection equipment must
be priced at such a figure as to make projectors within the reach of
other than the wealthy individual, the school, or the industrial
concern with unlimited funds.
Sixteen-millimeter projection equipment also must be improved to
make maintenance a minor problem for schools and industrial con-
cerns and improvements made in the amplifier systems in order
to give better quality sound. If the quality of the equipment can
be improved and the cost reduced, the future of the 16-mm film
as an educational medium is assured. The next aim should be an
improvement in 16-mm productions for educational use. All 16-
mm educational films should be filmed in 16-mm with original
sound to get away from bad effects caused in reduction of prints from
35-mm film. This will not only mean a reduction in cost but will
mean better prints in my belief.
A great field seems open to enterprising business concerns willing
to sponsor good-quality educational films as an advertising medium,
provided they do not ruin the sustained interest by too much ad-
vertising. This is especially true with the outlook, of present-day
trends as regards television, since already a number of 16-mm films
have been produced and televised.
ACKNOWLEDGMENT
The photographs used to illustrate this article were furnished by
The Visual Aids Center, Army Medical Center, Washington, D. C.
Report of
Studio- Lighting Committee
THIS REPORT describes motion picture studio-lighting power
sources and completes a series of reports covering all phases of
studio-lighting equipment.1"4
Direct-current motion picture studio-lighting power sources may
be divided into three general types: (1) Permanent installations con-
Columbia Pictures Corporation
Fig. 1 — Main generator room. 500-kilowatt General Electric motor-generator
sets.
sisting of motor-generator sets usually installed in a centrally located
powerhouse and with suitable underground cable connecting them to
the stages. (2) Portable motor-generator sets mounted on trucks or
trailers which may be located outside stages for extra power where
heavy loads are used, or which may be sent out on location. (3) In-
ternal-combustion engine-driven generators which find their greatest
use on locations where power from electric lines is not available.
* Original manuscript received October 11, 1947.
OCTOBER, 1948 JOURNAL OF THE SMPE VOLUME 51
431
432
STUDIO-LIGHTING REPORT
October
oo
CQ
I
H
W
I
h- 1 O
gs
*!
Generato
§i|
•^.s.g-
.
lillil! Illiil
.,-
0>£oo -oO^o
lliPI
1948 STUDIO-LIGHTING REPORT 433
The types of studio-lighting power sources in present use are legion
and it is outside of the scope of this report to catalog all of them.
Table I gives the characteristics of typical units and includes such
factors as desired ripple characteristics which may not be present in
some equipment, but which are desirable to minimize the necessity of
choke coils and niters when using carbon-arc lamps.5
Metro-Goldwyn-Mayer Studios
Fig. 2 — Portable motor-generator set, 300 kilowatts as in No. 3, Table I.
Metro-Goldwyn-Mayer Studios
Fig. 3 — Same as Fig. 2 except unit in closed position.
Compound-wound generators are used because of their ability to
maintain voltage under widely fluctuating load conditions. Com-
pound-wound generators may be, and often are, paralleled by the use
of an equalizer, or low-resistance connection, between the machines
which places their series fields in parallel. However, in order to ac-
complish the foregoing the generators must be of identical, or similar,
electrical characteristics. Either similar or dissimilar compound
434
STUDIO-LIGHTING REPORT
October
Walt Disney Productions
Fig. 4 — Portable internal-combustion engine-driven generator. 150-kilo-
watt generator, 290-horsepower motor, 1400 amperes, Thiotron automatic
voltage control.
the use of automatic voltage
by
generators may be paralleled
regulators.
Permanent-installation motor-generator sets are mounted on con-
crete bases. Approximate weights for one known installation are
34,000 pounds for the 500-kilowatt set and 20,000 pounds for the
300-kilowatt set without bases.
Fig. 5 — Same as Fig. 4.
Walt Disney Productions
Control panel.
1948
STUDIO-LIGHTING REPORT
435
Drive motors on portable motor-generator sets are usually made
for operation on more than one voltage and those used in Hollywood
are capable of operation at both 50 and 60 cycles.
Internal-combustion engine-driven generator sets may be made up
of any combination of generator and gas or diesel engine provided the
generator characteristics conform to the speed-horsepower charac-
teristics of the engine. It is desirable that the generator be capable
Mole-Richardson Company
Fig. 6 — A part of fleet of rental internal-combustion engine-driven generators.
Rated at 1400 amperes, 120 volts, 230-horsepower engine.
of maintaining voltage with approximately 65 per cent load at around
65 per cent of its rated speed.
These sets usually have a capacity of between 750 and 1400 am-
peres at 125 volts. For design estimates a basis of 5 amperes of gen-
erator output per engine horsepower may be used. Engine horse-
power so estimated includes that necessary to drive the water pump,
fuel pump, and other usual engine accessories. As a protective
measure, engine horsepower should be below that capable of driving
the generator at an injurious load.
Incandescent lamps may be, and sometimes are, operated on alter-
nating current to relieve a heavy direct-current load. This practice
is not common, however, because of the difficulties encountered with
two types of power being delivered through similar distribution sys-
tems on the sets.
Mercury-arc rectifiers have been considered as a direct-current
source of power for motion picture studio lighting but no installations
have been made to date.6
436
STUDIO-LIGHTING REPORT
Westinghouse Electric Corporation
Fig. 7 — Warner Brothers Studios powerhouse, 500-kilowatt Westinghouse
motor generators.
REFERENCES
(1) R. G., Linderman, C. W. Handley, and A. Rodgers "Illumination in mo-
tion picture production," /. Soc. Mot. Pict. Eng., vol. 40, pp. 333-368; June,
1943.
(2) "Report of the Studio-Lighting Committee," J. Soc. Mot. Pict. Eng., vol
45, pp. 249-261; October, 1945.
(3) "Report of the Studio-Lighting Committee," /. Soc. Mot. Pict. Eng., vol.
47, pp. 113-118; August, 1946.
(4) "Report of the Studio-Lighting Committee," /. Soc. Mot. Pict. Eng.,
vol. 45, pp. 279-289; September, 1947.
(5) B. F. Miller, "A motion picture arc-lighting generator filter," J. Soc. Mot.
Pict. Eng., vol. 41, pp. 367-374; November, 1943.
(6) L. A. Umansky, "Power rectifiers for studio lighting," J. Soc. Mot. Pict.
Eng., vol. 45, pp. 414-441; December, 1945.
STUDIO-LIGHTING COMMITTEE
(1947)
J. W. BOYLE
R. E. FARNHAM
C. W. HANDLEY, Chairman
KARL FREUND
M. A. HANKINS
C. R. LONG
W. W. LOZIER
D. W. PRIDEAUX
,
roposed 16-Mm and 8-Mm
Sprocket Standards
THE PAPER entitled "Proposals for 16-Mm and 8-Mm .Sprocket
Standards/' by J. S. Chandler, D. F. Lyman, and L. R. Martin,
was published in the June, 1947, issue of the JOURNAL OF THE SMPE
for the purpose of inviting comment on the work that has been done
on the design of sprockets. Discussion by letter was received from
Mr. E. W. Kellogg of the Radio Corporation of America in which he
questioned the propriety of the sprocket proposals as American
Standards. Mr. Kellogg's letter was not received in time to be pub-
lished at the same time as the paper; however, a notice appeared
with the paper stating that Mr. Kellogg's discussion and a reply
from the authors would appear at a later date. These two letters
are published here.
JOHN A. MAURER
Engineering V ice-President
MR. E. W. KELLOGG: Much as I admire the work which the authors have done,
I think that the title given the paper is not well chosen. There are, of course,
many possible titles. One might be "A Method of Designing Sprocket Teeth for
Minimum Flutter." We do not see any reason for suggesting that the resulting
tooth shape be made a "Standard." In fact, the fundamental purpose of Stand-
ards is to make interchangeability possible. Obviously films must be inter-
changeable so that they will run on all machines, but if a manufacturer puts out a
machine which performs well with a standard film, and the film is not subjected
to undue wear, and his customers are happy, he has complied with all of the con-
ditions which are important. Subject to the above conditions we might say that
it is no one else's business what shape tooth he uses. The occasion for attempting
actually to standardize sprocket designs would come only when and if it becomes
the practice of projector and camera manufacturers to procure their sprockets
from certain sprocket manufacturers who specialize in stock designs which are
interchangeable and are to be used by various equipment manufacturers. We do
not foresee such a development.
There is, on the other hand, a definite objection to giving the Society's official
sanction to a specific design when sprocket teeth of other designs are in wide, suc-
cessful use. The argument against such standardization has been well stated in
correspondence between members of the present Subcommittee on "8-Mm and
16-Mm Projector Sprocket Standards." Messrs. Sachtleben and Isom of RCA
pointed out
that while the plan to extend the usefulness of existing sprocket-design infor-
mation, by putting it into the form of formulas including several of the vari-
ables of projector design, was highly commendable, standardization of design
OCTOBER. 1948 JOURNAL OF THE SMPE VOLUME 51 437
438 DISCUSSION October
on the basis of such formulas was not desirable. This was because the grow-
ing prestige of standardization automatically would put a projector employing
nonstandard sprockets at a competitive disadvantage; whereas it could be
possible that under actual test the nonstandard sprockets would exhibit
superior performance.
Mr. Isom emphasized
from his experience in the educational field, the prestige and authority stand-
ards enjoy among persons charged with the responsibility for institutional
purchases. These people make their choices on the principle of elimination ,
and seize upon published standards as their aid and authority in this process.
For a number of years the Society has published information on the basis of
which a designer can produce sprocket teeth which have been found to be
satisfactory.
These designs have been published under the caption "SMPE Standards."
While we hold to the position that such information should be designated as a
ommended design procedure rather than a standard, we would not hesitate for
moment to say that the information has been useful to the industry. If tl
Society wishes to continue to call any sprocket- tooth design "Standard," it w<
be justified in changing to a new standard only if the new design had been showi
after prolonged use, to be definitely superior. We think for the present tl
Society should go no farther than to include it among acceptable designs and
experience provide the ultimate answer as to superiority. Although RCA engi-
neers are not at the present time in a position to pass on the performance of the pr
posed tooth, we may say the following in regard to the tooth design which we havt
used practically without change for a good many years. This design corresponds
very closely to the old SMPE "Standard" as published in 1934.
Many thousands of projectors have been in service for years in the Army and
elsewhere where the service is severe. In all this experience practically no com-
plaints have been received of trouble with sprockets. Mr. del Valle has given me
the story of a series of tests of film life made, for the most part, during the summer
of 1944. Twelve-foot loops were run through our projectors (including the inter-
mittent) in a setup having a tensioning roller which put about 4 ounces tension
each on the supply and takeup lines. The average life of a number of loops of
film run in different machines, properly adjusted, was of the order of 25,000 pas-
sages, and one loop ran 86,000 passages before slight cracks appeared at the cor-
ners of the perforations. It cannot be said that this was a case of perfect fit be-
cause the test of this loop extended over many weeks during the summertime, and
there must have been some shrinkage.
If we compare a tooth designed for a typical case, according to the formula given
in this paper, with one designed according to the previously recommended prac-
tice, there is a striking difference in that the sides of the new tooth slope much
more near the bottom, whereas the sides of the old tooth start more nearly ver-
tical. The effect of this difference would be to give the film much more tendency
to climb up the tooth or to make it necessary to resort to special measures to hold
the film down against the sprocket body. It does not seem to me practical to
force the film all the way down. Shoes must offer sufficient clearance to permit
1948 PROPOSED SPROCKET STANDARDS 439
splices to pass. The film can still climb 5 or 6 mils up the tooth if it wishes to, and
even ride at that position throughout its engagement with the sprocket.
We might expect then, in spite of the very well worked-out theory, that the
new tooth, which under ideal conditions would minimize flutter, would give more
erratic results, or require greater niceties of design and C3nstruction in order to
realize the possible benefit. Although the old tooth undoubtedly would give more
nutter at the sprocket than the new tooth when working at its best, we depend on
mechanical filtering to take out the flutter, and it is a serious question whether
other desirable properties should be sacrificed to reduce flutter at the sprocket.
The actual 24-cycle or 96-cycle flutter is very effectively filtered. Much harder to
filter out would be variations in the manner in which the film rides on the sprocket
which might result in random phase shifts at the sprocket, having components of
much lower frequency.
While my remarks are based largely on theory and it may well be that the
authors of the paper have gotten excellent results from their sprockets, I think
that there is abundant reason for a very cautious attitude on the part of the
Society with regard to recommending the new design as superior, or of giving it
the special sanction of calling it "Standard," especially when the former designs
are giving such excellent satisfaction.
There are special cases where mechanical filtering is not employed, such as
sprocket-type printers and certain recorders, and for these applications the
authors' approach to the problem of minimizing flutter deserves careful
consideration.
J. S. CHANDLER, D. F. LYMAN, AND L. R. MARTIN: The authors are grateful
to Mr. Kellogg for the careful attention he has given this subject, and for his
comments, which serve to re-emphasize some of the important points in the
paper. We agree that the desirability of standardizing sprockets on. the basis
of insuring good performance is open to question, since the usual function of
standards is to provide for interchangeability of parts. But the present ASA
standard for 16-mm sprockets, Z22. 6-1941, deals with dimensional specifications
that are related not only to interchangeability but also to performance. We
agree also with Mr. Kellogg's statement that the information in the present
standard has been useful to the industry. It has been realized for some time,
however, that the existing standard is inadequate in that the same tooth profile
is recommended for all sprockets, regardless of the number of teeth on the sprocket.
Moreover, no account is taken of such related and dependent factors as the range
of film shrinkage for which accommodation is to be made and the shape of the
path followed by the film as it approaches and leaves the sprocket. If for these
reasons the current standard is allowed to lapse and no substitute is made avail-
able, the industry will soon be keenly aware that it lacks an authoritative source
of information on this important subject.
It was largely because of these latter considerations that a committee on
sprockets was formed and the project leading to the present proposals was insti-
gated. One point should be emphasized: the profile recommended for the
tooth is designed to give maximum film life, not necessarily minimum flutter.
The fact that the recommended tooth would, theoretically at least, result in less
flutter than would be obtained with the alternative shape shown in Fig. 9 of our
paper is incidental; it follows from the basic principle that if the acceleration
440 DISCUSSION
of the film is kept low as it strips off the driving tooth, the impact of the film
against the next tooth is less severe.
As Mr. Kellogg points out, the recommended tooth slants more at the base
than does the tooth in the existing ASA standard. This criticism was offered
early in the work on the proposals, and that is why the alternative profile was
included in Fig. 9. Experiments have demonstrated that the flexing of the
film in the region of the loaded edge of the perforation, rather than impact loading,
determines the life of the film. If further studies show that the alternative tooth
shape, which is steeper at the base but more sloping at the tip, affords longer
film life, it should be the one that is recommended and the formulas should be
changed accordingly.
The purpose of submitting this proposal is to encourage more trials by a number
of investigators, since sprocket specifications resulting in optimal performance
can be determined only after extensive tests. The form of the proposal is such
that the equations can be altered readily to make them comply with the results
of such tests, the importance of which cannot be overemphasized.
Incorporation of American Standards Association
rpiHE AMERICAN Standards Association became the American Standards
J- Association, Incorporated, August 2, 1948, through incorporation under the
laws of the State of New York. This is the third in a series of changes which have
consistently recognized the enlarging scope of the Association's work.
Organized in 1918 as the American Engineering Standards Committee, a co-
ordinating committee for the standardization work of five of the country's im-
portant technical societies, the scope and organizational setup were soon broad-
ened to include associations and government agencies. This led to the extension
of the work into the field of safety standards. In 1928, an entire reorganization
took place, changing the Committee into a full-fledged "American Standards
Association," the nation's clearing house for standards and the United States
medium for international contacts on standardization. The present change to an
incorporated Association again recognizes the enlarged activities and responsibili-
ties of the organization, giving it and its members the protection and benefits
which corporation law affords and which is considered essential in the light of the
scope of the Association's activities.
Bills seeking Federal incorporation are now before Congress, having been intro-
duced in the House by the Honorable Kenneth B. Keating of New York, and in
the Senate by the Honorable Ralph Flanders of Vermont.
The Association's co-ordinating functions now extend to standards in the
mechanical, electrical, building, photographic, mining, safety, and consumer-goods
fields, as well as such general work as that on office equipment and abbreviations
and symbols for use in engineering and scientific literature.
Frederick R. Lack, vice-president, Western Electric Company, is president of
the Incorporated Association. Vice-Admiral G. F. Hussey, Jr. (United States
Navy, retired), is secretary and adminstrative head, and Cyril Ainsworth is tech-
nical director.
THOMAS ARMAT
THOMAS ARMAT, pioneer inventor of the motion picture projec-
tor, who died on September 30, 1948, received a special award
this year from the Academy of Motion Picture Arts and Sciences
for his contributions to the development of the motion picture.
The statement of the Academy reads as follows:
"Academy Special Award to Thomas Armat, one of the small
group of pioneers whose belief in a new medium, and whose
contributions to its development, blazed the trail along which
the motion picture progressed, in their lifetime, from obscurity
to world-wide acclaim/'
In 1946, on the occasion of the Fiftieth Anniversary of the first
exhibition of motion pictures in a theater, Mr. Armat was awarded
a Citation by the Society of Motion Picture Engineers in recogni-
tion of his distinguished inventions which were outstanding features
of his first projecting machine.
Mr. Armat was born in Fredericksburg, Virginia, on October 26,
1866, and in February, 1896, he demonstrated a motion picture
projector of his own design to Thomas A. Edison at his laboratory
in West Orange, New Jersey. This projector, known then as the
Vitascope, was the first to incorporate a loop-forming means and
a longer period of rest and illumination than the time required to
move the film from one frame to the next. These features were a
major step in the development of modern motion picture projec-
tors and were incorporated subsequently in most commercially
successful projection machines.
Mr. Armat was elected an Honorary Member of the Society of
Motion Picture Engineers in October, 1935.
LOUIS LUMIERE
Louis LUMIERE, 83, foremost Frenchman of the cinema, died
on June 6, 1948, at Bandol on the Riviera.
M. Lumiere and his brother, Auguste, were among the pioneers
outside the United States who developed the possibilities of the
motion picture Kinetoscope. On December 28, 1895, they opened
an exhibition in the basement of the Grand Cafe in Paris. This
marked the beginning of commercial motion picture exhibition in
France.
The Lumieres were manufacturers of photographic materials at
Lyon, in France. They set to work, as did so many others, to join
the Kinetoscope's peep-show pictures with the magic lantern to
achieve projection. Lacking film base, which could only be had
from their American competitor, Eastman, they sent to New York
for a makeshift material, strips of celluloid from the American Cel-
luloid Company. Being economically minded, also, they cut the
rate of motion picture photography from Edison's 48 frames a second
to 16 frames.
After the demonstration of his motion picture camera Lumiere
experimented with color photography and developed a number
of photographic appliances. So outstanding were his contribu-
tions that on April 22, 1935, he received a tribute from the Motion
Picture Producers and Distributors.
M. Lumiere was, for a short time, a member of Marshal Petain's
advisory council of State. He was honorary president of the
French Chamber of Cinema, a member of the French
Academy of Sciences, a grand officer in the Legion of Honor, and an
Honorary member of the Society of Motion Picture Engineers.
TERRY RAMSAYE
THAD C. BARROWS
THAD C. BARROWS, 59, president of Boston Local 182 from 1918
until 1947, died as a result of a heart attack on June 2, 1948.
Until the day of his death he was actively engaged in his craft
in Boston's Metropolitan Theater, and his interest in technological
developments in his work was unflagging.
Recognition of his enthusiastic devotion to his field came in
1929 when he was unanimously elected the first president of the
Projection Advisory Council, a national organization which con-
tributed greatly to the industry during the difficult years of transi-
tion to sound motion pictures. He was an Active member of the
Society of Motion Picture Engineers for 20 years.
His sincerity, courage, and honesty won the affection and respect
of all who knew him.
Book Reviews
Enlarging — Technique of the Positive, by C. I. Jacobson
Published (1948) by the Focal Press, Inc., 381 Fourth Ave., New York 16, N. Y.
307 pages + xx pages + 9-page index. 77 illustrations. 5:/4 X 7*/2 inches.
Price, $3.50.
The culmination of the photographer's work is the print. Involved in its
preparation is a whole series of events including materials, techniques, and
equipment. These are the negative, the printing media, the enlarger, the proc-
essing technique, and the aftertreatment. Involved also are psychophysical
and physiological aspects such as perspective and other distortions, definition,
and visual acuity. All of these topics are treated in a chatty manner which makes
for easy reading. As with the companion volume (see the review on "Develop-
ing," by the same author, published in the July, 1948, issue of THE JOURNAL,
page 105), the emphasis has been laid on description without the use of techni-
cal language, and without the presumption of a technical background. And
yet an adequate panorama of the field is given.
The book will be valuable to all who desire to learn what is involved behind the
scenes, when a camera record is converted into a final print. After a discussion
of the negative material and its characteristics, as examplified by the negative
to be printed, it goes on to treat in detail the printing media upon which the
negative is to be copied. The first quarter of the book is therefore concerned
with the materials used. But equipment and techniques are also involved in the
cycle of events. The discussion of these incidentals is the subject matter of the
remainder of the book. One obtains a working knowledge of the intricacies of
enlarging equipment, of the various printing techniques, of tone separation
processes, montages, and other matters. All in all it is a darkroom man's ele-
mentary handbook on printing, and it will serve him as an excellent guide to help
him solve old problems, or indicate to him new ones.
JOSEPH S. FRIEDMAN
Ansco
Johnson City, New York
Camera and Lens, by Ansel Adams
Published (1948) by Morgan and Lester, 101 Park Ave., New York 17, N. Y.
117 pages + 3-page index + viii pages. 77 illustrations. 6V4 X 91/* inches.
Price, $3.00.
This book is the first in a series of six volumes on basic photography to be
written by Ansel Adams. It is intended to acquaint the aspiring still photographer
with those fundamentals of camera operation which the author considers essential
in creative photography. However, there are six short chapters pertaining to
darkroom layout and construction, darkroom equipment, the finishing room,
negative storage, print storage, and print-display devices, none of which has
much relation to the camera and lens.
Mr. Adams is a photographer of repute. Some of his pictures rank among
the best that have ever been produced, which attests his ability as a competent
judge of aesthetic and photographic quality. It is unfortunate that he did not
443
444 BOOK REVIEWS
choose to write a book in these fields, for he does not appear to be sufficiently
versed in the technical aspects of photography to discuss them authentically.
For a beginner's book too many terms are used before they are defined, and in
some instances the terms are nowhere properly defined. For example, on page 5
the term "parallax" is used without being defined, and not until page 15 is it stated
that lens speed is expressed as //8, //3.5, etc., although this designation is used
freely on previous pages. And in the chapter beginning on page 88, / number is
improperly defined.
Mr. Adam's discussion on composition is considerably better than one finds
usually in the photographic literature.
LLOYD E. VARDEN
Pavelle Color
New York 19, N. Y.
Informational Film Year Book 1947
Published (1947) by the Albyn Press, 42 Frederick St., Edinburgh 2, Scotland.
174 pages. 25 illustrations. 5XA X 83/4 inches. Price, 10s. Qd. net.
The rapid growth of the nontheatrical film in recent years is indicated clearly
in several of the articles in this Film Year Book. Twelve short articles by well-
known writers such as Paul Rotha, John Grierson, Andrew Buchanan, Forsythe
Hardy, and Basil Wright comprise about one half of the book. Subjects discussed
include documentary films, the conditions in nontheatrical film industry in
America, the services rendered by the film in industry, the classroom film and
films for children, and the use of films by the United Nations Educational, Scien-
tific, and Cultural Organization (UNESCO). Summarizing the place of the non-
theatrical film in the world today, Norman Wilson states, "It should be the aim of
everyone who believes in democracy to make the freedom of the screen as much a
reality as the freedom of the press."
The latter half of this interesting volume contains a "Buyers' Guide" on new
substandard apparatus; a group of stills from documentary films of the year; a
list of the informational films of the year; also lists of film-producing organiza-
tions, cine societies, studios, laboratories, libraries, manufacturers of cine appa-
ratus, specialist cinemas, and film periodicals.
GLENN E. MATTHEWS
Kodak Research Laboratories
Kodak Park, Rochester 4, New York
Current Literature
rpHE EDITORS present for convenient reference a list of articles dealing with
J- subjects cognate to motion picture engineering published in a number of se-
lected journals. Photostatic or microfilm copies of articles in magazines that are
available may be obtained from The Library of Congress, Washington, D. C., or
from the New York Public Library, New York, N. Y., at prevailing rates.
American Cinematographer Ideal Kinema
29, 6, June, 1948 14, June 10, 1948
The Application of Motion Picture Consistency and Colour of Screen
Technique to Television (p. 194)
R. B. AUSTRIAN
U. S. Navy Develops Super-Speed
Cameras (p. 207)
29, 7, July, 1948
Photography for Television (p. 229)
F. FOSTER
29, 8, August, 1948
Transition Lens for Television Cam-
eras (p. 266) F. FOSTER
The New "Spectra" Measures Color
Temperature (p. 267) F. GATELY
3000 Frames Per Second (p. 269)
British Kinematography
12, 5, May, 1948
Sound-on-Film Reproducing Equip-
ment:
I. The Sound Head (p. 155) A. T
SINCLAIR
II. Electrical, Electronic and
Acoustic Design (p. 159) H. J.
O'DELL
Television Production in Contrast to
Film Production (p. 164) P. H.
DORTE
Electronic Engineering
20, June, 1948
A New Television Film Scanner
(p. 174)
Improving Circuit Diagrams (p. 175)
L. H. BAINBRIDGE-BELL
Electronics
21, 7, July, 1948
Design Factors for Intercarrier Tele-
vision Sound (p. 72) S. W. SEELEY
Photometry in Television Engineer-
ing (p. 110) D. W. EPSTEIN
Illumination (p. 20) R. H. CRICKS
International Projectionist
23, 6, June, 1948
New Acetate Film for Release Prints
(p. 5) C. R. FORDYCE
Projection Factors of New Acetate
Film (p. 6)
23, 7, July, 1948
Control of Sound-Film Reproduction
(p. 5) R. A. MITCHELL
New Century Sound Systems Fea-
ture Fundamental Reproducer Ad-
vances (p. 8)
Television: How It Works (p. 10)
W. BOUIE
Theater Television: A General Sur-
vey (p. 19) A. N. GOLDSMITH
23, 8, August, 1948
Television: How It Works. Pt. 2
(p. 9) W. BOUIE
Projector Progress in Great Britain
(p. 18) H. HILL
Proceedings of the I.R.E.
36, 6, June, 1948
The Application of Projective Geom-
etry to the Theory of Color Mix-
ture (p. 709) F. J. BINGLEY
36, 7, July, 1948
Avenues of Improvement in Present-
Day Television (p. 896) D. G. FINK
Radio and Television News
40, 2, August, 1948
The Recording and Reproduction of
Sound. Pt. 18 (p. 49) O. READ
RCA Review
9, 2, June, 1948
Motion Picture Photography of Tele-
vision Images (p. 202) R. M.
FRASER
445
Journal Exchange
Dr. R. F. Nicholson wishes to dispose
of a complete bound set of JOURNALS
of the SMPE from January, 1928,
through December, 1947. Prospec-
tive purchasers should write to Dr.
Nicholson at 230 Albany St., Cam-
bridge 39, Mass., and mark the en-
velope "Personal."
The Society of Motion Picture Engi-
neers has for sale certain back copies of
the TRANSACTIONS and the JOURNAL, as
listed below.
TRANSACTIONS
Numbers 3, 4, 10, 11, 12, 13, 14, 15, 17,
19, 20, 21, 22, 23, 24, 25, 27, 28, 29,
and 32 at $1.25 each; 33, 34, 35, and
36 at $2.50 each; and 37 and 38 at
$3.00 each.
JOURNALS — $1.25 each
Entire Year: 1930, 1931, 1935, 1937,
1942, and 1943.
1932— no May
1933 — no May, June, October
1934 — no January, February, March,
April
1936— no April, May
1937— no April, May
1938 — no January, March, May
1939— no April
1940— no June
1941 — no March and July
1944 — no February, March, May, July
1945 — no January, February, April,
August
1946 — no January, February, March
1947 — no January
1948— all to date
Send your order with remittance to
the Society of Motion Picture Engi-
neers, 342 Madison Avenue, New
York 17, N. Y.
35-MM AND 16-MM TEST FILMS
— for the manufacturer —
As an important tool for the design engineer and as a
performance standard in quality control and inspection.
— for the dealer and serviceman —
To demonstrate new projectors and sound reproducers as
well as to adjust equipment in service or in the process of
being repaired.
— for 16- and 35-mm exhibitors —
To insure peak performance for those who maintain their
own equipment.
Twenty-nine different test films, both 16- and 35-mm are
available for all purposes from the Society of Motion Picture
Engineers and the Motion Picture Research Council. A catalog
describing each in detail is available at no cost from the Society
of Motion Picture Engineers, 342 Madison Avenue, New York 17,
N. Y.
446
SECTION OFFICERS
Atlantic Coast
Chairman
William H. Rivers
Eastman Kodak Co.
342 Madison Ave.
New York 17, N. Y.
Secretary -Treasurer
Edward Schmidt
E. I. du Pont de Nemours & Co
350 Fifth Ave.
New York 1, N. Y.
Chairman
R. T. Van Niman
Motiograph
4431 W. Lake St.
Chicago 24, 111.
Midwest
Secretary-Treasurer
George W. Colburn
George W. Colburn Laboratory
164 N. Wacker Dr.
Chicago 6, 111.
Pacific Coast
Chairman
S. P. Solow
Consolidated Film Industries
959 Seward St.
Hollywood, Calif.
Secre tary-Trea surer
G. R. Crane
212—24 St.
Santa Monica, Calif.
Student Chapter
University of Southern Calfornia
Chairman
Thomas Gavey
1046 N. Ridgewood PI.
Hollywood 38, Calif.
Secretary-Treasurer
John Barn well
University of Southern California
Los Angeles, Calif.
Office Staff— New York
EXECUTIVE SECRETARY
Boyce Nemec
STAFF ENGINEER
William H. Deacy, Jr.
OFFICE MANAGER
Sigmund M. Muskat
JOURNAL EDITOR
Helen M. Stote
Helen Goodwyn
Dorothy Johnson
Beatrice Melican
Thelma Klinow
Ethel Lewis
447
Statement of the Ownership, Management, Circulation, Etc., Required by the
Acts of Congress of August 24, 1912, as Amended by the Acts of March 3, 1933
and July 2, 1946, of Journal of the Society of Motion Picture Engineers, published
monthly at Easton, Pa., for October 1, 1948.
State of New York )
County of New York j Sw
Before me, a Notary Public in and for the State and county aforesaid, person-
ally appeared Boyce Nemec, who, having been duly sworn according to law, de-
poses and says that he is the Executive Secretary 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,
weekly, semiweekly, or triweekly newspaper, the circulation), etc., of the aforesaid
publication for the date shown in the above caption, required by the Act of August
24, 1912, as amended by the Acts of March 3, 1933 and July 2, 1946 embodied
in section 537, 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 o/— Post Office Address —
Publisher, Society of Motion Picture Engineers, Inc., 342 Madison Ave., New
York 17, N. Y.
Editor, Helen M. Stote, 342 Madison Ave., New York 17, N. Y.
Managing Editor, None.
Business Manager, Boyce Nemec, 342 Madison Ave., New York 17, 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, Inc., 342 Madison Ave., New York 17, N. Y.
Loren L. Ryder, President, 5451 Marathon St., Hollywood, Calif.
G. T. Lorance, Secretary, 63 Bedford Rd., Pleasantville, N. Y.
Ralph B. Austrian, Treasurer, 247 Park Ave., New York 17, N. Y.
No stockholders.
3. That the known bondholders, mortgagees, and other security holders
owning or holding one 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 aifiant
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
twelve months preceding the date shown above is : (This information is required
from daily, weekly, semiweekly, and triweekly newspapers only.)
BOYCE NEMEC, Exec. Secy., Business Manager.
Sworn to and subscribed before me this 18th day of August, 1948.
(Seal) Elisabeth J. Rubino
Notary Public, Clerk's No. 2986,
Queens County. Reg. No. 86-R-9
(My commission expires March 30, 1949)
Journal of the
Society of Motion Picture Engineers
VOLUME 51 NOVEMBER 1948 NUMBER 5
PAGE
Proposed Standards for the Measurement of Distortion in
Sound Recording 449
Magnetic Recording for the Technician DOROTHY O'DEA 468
35-Mm Magnetic-Recording System EARL MASTERSON 481
Optimum High-Frequency Bias in Magnetic Recording
G. L. DIMMICK AND S. W. JOHNSON 489
Variable-Area Recording with the Light Valve
JOHN G. FRAYNE 501
Variable-Area Light- Valve Modulator LEWIS B. BROWDER 521
Nine Recent American Standards 534
Section Meetings 549
Book Reviews:
"The Diary and Sundry Observations of Thomas Alva
Edison"
Edited by Dagobert D. Runes
Reviewed by Terry Ramsaye 550
"L'Annuaire du Cinema 1948 (Motion Picture Yearbook
for 1948)"
Published by Editions Bellefaye 551
Current Literature : 552
New Products . . 553
ARTHUR C. DOWNES HELEN M. STOTE GORDON A. CHAMBERS
Chairman Editor Chairman
Board of Editors Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930,
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOURNAL must be obtained in writing from the General Office of the Society.
Copyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC . . . EXECUTIVE SECRETARY
OFFICERS
1947-1948
PRESIDENT EDITORIAL VICE-PRESIDENT
Loren L. Ryder Clyde R. Keith
5451 Marathon St. 233 Broadway
Hollywood 38, Calif. New York 7, N. Y.
PAST-PRESIDENT CONVENTION VICE-PRESIDENT
Donald E. Hyndman William C. Kunzmann
342 Madison Ave. Box 6087
New York 17, N. Y. Cleveland, Ohio
EXECUTIVE VICE-PRESIDENT SECRETARY
Earl I. Sponable G. T. Lorance
460 West 54 St. 55 La France Ave.
New York 19, N. Y. Bloomfield, N. J.
1948-1949
ENGINEERING VICE-PRESIDENT FINANCIAL VICE-PRESIDENT
John A. Maurer James Frank, Jr.
37-01—31 St. 426 Luckie St., N. W.
Long Island City 1, N. Y. Atlanta, Ga.
TREASURER
Ralph B. Austrian
25 W. 54 St.
New York, N. Y.
Governors
1947-1948
John W. Boyle Robert M. Corbin Charles R. Daily
1207 N. Mansfield Ave. 343 State St. 5451 Marathon St.
Hollywood 38, Calif. Rochester 4, N. Y. Hollywood 38, Calif.
David B. Joy Hollis W. Moyse
30 E. 42 St. 6656 Santa Monica Blvd.
New York 17, N. Y. Hollywood, Calif.
1948
William H. Rivers S. P. Solow R. T. Van Niman
342 Madison Ave. 959 Seward St. 4431 W. Lake St.
New York 17, N. Y. Hollywood, Calif. Chicago, 111.
1948-1949
Alan W. Cook Gordon E. Sawyer
4 Druid PI. Lloyd T. Goldsmith 857 N. Martel St.
Binghampton, N. Y. Burbank, Calif. Hollywood, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Section Officers and Office Staff listed on page 554.
Proposed Standards
for the Measurement of
Distortion in Sound Recording*
ORGANIZATION AND PLAN OF WORK
ON OCTOBER 25, 1947, a meeting was held of the American Stand-
ards Association Committee on Standards for Sound Re-
cording, under the chairmanship of George Nixon of the National
Broadcasting Company. On this committee the Society of Motion
Picture Engineers was represented by J. A. Maurer, C. R. Keith,
Otto Sandvik, and E. W. Kellogg, who was asked to assume the chair-
manship of a subcommittee to recommend standards with respect to
the measurement of performance characteristics and distortion in
sound recording and reproducing systems. The membership of the
subcommittee is as follows :
EDWARD W. KELLOGG, Chairman
RCA Victor Division
Camden, N. J.
W. S. BACHMAN G. T. LORANCE
Columbia Recording Corporation General Precision Laboratories
New York, N. Y. Pleasantville, N. Y.
PETER CHRZANOWSKI R. A. LYNN
National Bureau of Standards National Broadcasting Company
Washington, D. C. New York, N. Y.
L. C. HOLMES OTTO SANDVIK
Stromberg-Carlson Company Eastman Kodak Company
Rochester, N. Y. Rochester, N. Y.
(r. B. MEYER HARRY SCHECTER
Westinghouse Electric Corporation Air Materiel Command
Baltimore, Md. Cambridge, Mass.
K. B. LAMBERT R. R. SCOVILLE
Metro-Goldwyn-Mayer Pictures Electrical Research Products
Culver City, Calif. Los Angeles, Calif.
W. H. OFFENHAUSER, JR.
New Canaan, Conn.
* Presented May 18, 1948, at the SMPE Convention in Santa Monica.
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 449
450 DISTORTION MEASUREMENT November
The work had been carried on largely by correspondence, with one
meeting held at ASA headquarters on January 21, 1948. The
method for the most part has been for the chairman, with the help of
the members, to attempt to ascertain what standards already exist,
or what procedures are current, and then to formulate proposals for
the subcommittee members to criticize or correct. The circular let-
ters with proposals, discussions, and questions have also been sent to
numerous qualified persons who are not members of the subcommittee,
and many valuable letters have been received. It is appropriate to
acknowledge here the thoughtful comments plus information re-
ceived from Howard Chinn of the Columbia Broadcasting System,
New York; R. C. Moyer of the RCA Victor Division, In-
dianapolis; W. R. Furst, of Furst Electronics, Chicago; J. W.
Bayliss and Kurt Singer of the RCA Victor Division, Holly-
wood; John T. Mullin of W. A. Palmer Co., San Francisco (rep-
resenting also views of Ampex Corporation, San Carlos, Calif.);
J. K. Hilliard of the Altec Lansing Company, Hollywood; Captain
R. Bennett of the Navy Electronics Laboratory, San Diego; H. H.
Scott of H. H. Scott, Inc., Cambridge: S. J. Begun, of Brush De-
velopment Company, Cleveland; and A. R. Morgan of the RCA
Laboratories, Princeton. Few of the proposals here outlined have
had final approval of the subcommittee membership for submission
to the main committee, but it seems desirable to take advantage of
the opportunity offered by the Spring Convention of this Society
to obtain a wider consideration of the questions on whose answers
any standards must depend. This method is similar to what the
SMPE Sound Committee did last year in submitting their propo-
sals on Flutter Standards in a report, for consideration by the entire
Society membership. This afforded them the benefit of wide dis-
cussion and many viewpoints before the proposals were taken up by
the Standards Committee. We hope to have a similar experience
with respect to the proposals submitted here.
TYPES OF DISTORTION
Distortion in reproduced sound may be broadly divided into four
types.
(1) Inaccuracies of pitch or frequency, especially fluctuations in
pitch relative to the original, which are generally called "flutter" or
"wow."
1948 DISTORTION MEASUREMENT 451
(2) Inequalities in the amplification of sounds
(a) depending on their frequency (lack of flat frequency re-
sponse) ;
(b) depending on their amplitude (volume expansion or com-
pression);
(c) variations with time (fluctuations in level relative to original) .
(3) Phase distortion. In the reproduced sound certain compo-
nents of complex sounds are delayed with respect to others. Transient
distortion due to equipment resonances is one type of phase distortion,
in that energy is absorbed from the sound for a short period of time
and then released. Irregularities in frequency-response characteris-
tics (as, for example, peaks due to resonance) are nearly always ac-
companied by phase distortion.*
(4) Production of spurious sounds, not in the original
(a) overtones, due in general to nonlinearity in the instantaneous
(or dynamic) output-to-input relationship in one or more of
the elements of the system;
(b) beat tones, also due in general to nonlinearity, resulting in
rectification, but also caused by proximity effects in the rec-
ord as in high-frequency variable-area recording;
(c) noise, generally divided into that due to hum, mechanical
vibration, and microphonic elements, and that due to the
granular character of the record, plus foreign particles and
minor injuries or abrasions. There are also noises of ther-
mal and tube origin.
TOPICS COVERED
The assignment of our subcommittee does not include attempting
to set standards of performance, but only to recommend such stand-
ards as can be agreed upon for the measurement of performance char-
acteristics and distortion. This may, however, involve specifica-
tions for the measuring equipment itself.
Standardization in measurements has, for its main object, the en-
abling of persons in different laboratories to check and compare meas-
urements; but a second and no less important object is to prevent
•* If phase distortions of this kind produce any audible quality changes, they are
usually so overshadowed by the variations in response as to be of secondary
importance. Moreover, measures which correct the response irregularities usu-
ally also correct the phase distortion.
452 DISTORTION MEASUREMENT November
the misunderstandings which result from use of different systems,
and especially from use of the same term with different understand-
ings of its meaning. In many cases it may not be possible to get
people in various organizations to adopt a single standard method of
making tests and, when this situation is encountered, it still should
be possible to prevent misunderstandings, by establishing definitions
of terms and by calling for adequate information to accompany re-
ports of tests, so that the reader may know when certain figures are
not directly comparable. Hence, two important parts of the stand-
ards toward which we are working are definitions of terms and stipu-
lations about reporting results.
Although, to the listener, distortion is anything which makes the
sounds reaching his ears different in quality from those produced
by the original source, the present work is limited to the quality
changes which are caused by the operations of recording and playing
back or, in other words, the differences between direct and transcribed
sound.
The principal topics so far discussed are :
1. Flutter or wow.
2. Frequency-response characteristics.
3. Distortions of the type which cause changes in wave shape.
4. Noise and signal-to-noise ratio.
STANDARDS FOR FLUTTER MEASUREMENT
The Sound Committee of the SMPE, under the able chairmanshipj
of J. G. Frayne, worked out a series of standards proposals with
respect to "flutter" or "wow," and these were published in the
JOURNAL for August, 1947. A great deal of thought and work had
gone into these proposals, and comments by many individuals repre-
sentative of different organizations had been invited and considered.
There was thus available to our ASA subcommittee a draft ofl
specifications dealing with this important problem, which could weal
be made the basis of the ASA specification, provided they were like-j
wise acceptable to makers and users of sound-recording equipment in
other fields, especially equipment for use with disk records. Broad-j
cast stations and disk-record manufacturers are represented on our!
subcommittee and opinions have been invited from others. We be-j
lieve that the proposals submitted by your Sound committee will
1948 DISTORTION MEASUREMENT 453
find general acceptance by all groups interested in sound recording
perhaps with some additions and changes of wording, but with no
essential changes of meanings.
On one item, however, our subcommittee took the position at the
February, 1948, meeting, that recommendation as an ASA standard
would be premature, namely, the quantitative part of the definition
of the term "Flutter Index." While it is admitted by all that the
information given in the formulation is interesting and valuable,
several of our members questioned whether it is directly applicable
to recorded sound in general, the perception threshold for continuous
tones in a live room being much lower than for ordinary music under
average listening conditions. In view of the divergent viewpoints,
recommendation of a standard for Flutter Index may be delayed as
compared with the remainder of the flutter standards, but it is not
our intention to postpone action any longer than is needed to arrive
at an all-around understanding.
Although none of the subcommittee members has objected to
specifying flutter in terms of root-mean-square magnitude, a number
of people have expressed the idea that the peak values of flutter are
more significant with respect to quality damage than the root-mean-
square values. The chairman of this committee wishes to take ad-
vantage of this opportunity to present his personal discussion of this
and one or two other questions. For better continuity of the report
of the ASA work, this discussion is put in an Appendix.
PARAGRAPH NUMBERING
In what follows, the paragraphs which are numbered, as 3.1, 3.2,
etc., and the "Notes" directly under them, are the proposed definitions
or specifications, and notes of explanation which would, if approved,
'be incorporated in the standards with substantially their present
wording. Paragraphs not numbered in this manner, or in other words
the remainder of this paper, are the writer's explanations or discus-
sions, which have not been put into the form of standards. The num-
bering employed is part of a topical system, and is of no concern to
the reader except as he may find some cross references.
FREQUENCY-RESPONSE CHARACTERISTIC
It has seemed to the chairman unnecessary to go into details of
measuring methods for this determination, since the main requirements
&re rather obvious. However, if any are of the opinion that mate-
rial should be added to what is here suggested, it is hoped that we
454 DISTORTION MEASUREMENT November
shall hear from them. A definition is in order, and the requirements
to avoid errors due to measuring harmonics or noise instead of the
true useful output, and it is necessary to specify something about the
levels at which the tests are to be run.
3.1 Definition — Frequency-Response Characteristic.
The frequency-response characteristic (sometimes shortened to "fre-
quency characteristic") of a sound system, or any portion of a sound
system, is the output level (or "response") as a function of frequency
for constant sine-wave input, the output being usually expressed
relative to some arbitrary level, for example, the level at 1000 cycles.
3.2 All distortion components (harmonics or noise) must be
excluded from the output measurement.
3.3 The input level chosen shall be high enough so that at no
frequency within the range covered is the true output less than 10
decibels above noise, and low enough so that at no frequency, within
the range covered, will overloading occur in sufficient degree to af-
fect the reading appreciably.
3.3a Another way of stating the second requirement is that the
input level must be sufficiently below that which would cause over-
load, so that a characteristic taken at a slightly (say 2 db) lower level
will, if plotted on a decibel scale, have identical shape.
3.4 It is permissible, provided the above conditions are met, to
take part of the measurements with a different input level than the
remainder. This may be advantageous in testing systems in which a
large amount of pre-emphasis of high frequencies is employed.
3.5 The output-input relationship shall be that which occurs
under steady-state conditions. Thus, if the practice is followed
of employing a continuously variable input frequency and changing
this rapidly, the rate of sweep shall be slow enough to give results
identical with those obtained with a slower sweep.
NOTE 1 : It is customary to express input and output levels in
amplitude terms (not power) or else in decibels above or below a I
chosen reference.
NOTE 2: The frequency-response characteristic may include in-
tentional departures from uniformity as, for example, a rising re-|
cording characteristic by way of pre-emphasis of high frequencies.
FREQUENCY CHARACTERISTIC OF A RECORD
For testing the characteristics of reproducing systems, it is necessary:
to employ test records whose characteristics can be definitely
1948 DISTORTION MEASUREMENT 455
specified. It is not always possible to make a logical distinction be-
tween the losses or distortions which are to be attributed to the record
and those which should be charged against the reproducing system.
Whenever it is possible to state the characteristics of the record with-
out specifying anything about the reproducing system, this is desir-
able. Hence, it is proposed that
3.6 The recorded level on a disk record is the velocity (maximum
of cycle) corresponding to the slope of the centerline of the groove.
Or, as an alternative definition,
3.6a The recorded level on a disk record is 2irfa, in which / is
the frequency and a is the amplitude of the recorded wave.
NOTE 1 : In the case of records cut in wax and properly processed,
the velocity of the recording stylus is the recorded level. In the case
of a lacquer disk, recorded with a cutter having a burnishing surface,
there is some springback in the record material, and the cutting
stylus velocity is not a safe guide to the recorded level.
The amplitude of the recorded waves can be measured microscopi-
cally. The light pattern method of checking a record calibration is
believed to be, when carefully carried out, a reliable indication of the
recorded level.
3.7 The recorded level on a film is the amplitude of the fundamen-
tal (sine- wave) component of the sound-track transmission.
NOTE 1 : A photographic record may be calibrated by measuring
the reproduced level, using a reproducing system whose performance
and characteristics are known. The scanning beam should form on
the film a rectangular image of uniform intensity, having a length
equal to the standard for the type of track under test and a height
which is small in comparison with the wavelength, and not more than
10 per cent of the total light should fall outside the boundaries of the
rectangle. Correction is then made for the finite width of the scan-
ning image, by multiplying the output by x/sin x where x = Trw/2\,
w being the width (or height) of the scanning image and X the length
of the recorded waves. It is recommended that this correction be
made small by making w as small as will afford adequate light and
satisfactory stray-light ratio.
Specification of recorded level in terms of the cyclic peak is, to the
best of the writer's information, more widely current than the use of
root-mean-square figures. The probable reason is that it is more
simply related to the overload point. The fact that electrical levels
are specified in root-mean-square terms might be thought to lead to
456 DISTORTION MEASUREMENT November
possible confusion, but that danger is lessened by the fact that the re-
corded level is not an electrical quantity. However, statements of
recorder or reproducer response must be so worded as to avoid possible
misunderstanding.
3.8 In the case of a magnetic recording no way of specifying re-
corded level seems feasible at present, except as the record is tested
with a specified reproducing system.
WAVE-SHAPE DISTORTION
Since the problem of measuring distortion in recording and re-
producing systems is in general the same as in any audio transmission
system (for example, in an amplifier), existing standards are appli-
cable. Test systems tend to crystallize around developed and avail-
able equipment. Four types of distortion-measuring equipment have
found wide use.
(1) Wave Analyzers, by which the amplitude of each overtone or
harmonic, relative to the fundamental, can be measured,
(2) Distortion-Factor Meters, which suppress the fundamental
and measure the sum total of what is left (overtones, rumble, hum,
and surface noise), expressing the root-mean-square magnitude of
this residue, relative to that of the fundamental. At the higher
levels, and with reasonable control of rumble and hum, distortion-
factor meters serve to measure total harmonic distortion.
(3) Intermodulation Analyzers, which measure the fluctuations in
level of a low-amplitude, relatively high-frequency tone when super-
imposed on a high-level, low-frequency tone. Levels 20 and 80 per
cent, respectively, of normal full sound-track amplitude have been
widely employed. Intermodulation is a more sensitive test (higher
readings for the same distortion) than total harmonic distortion.
It has been especially useful in variable-density photographic record-
ing, and has been employed in a limited way in studying distortion
in disk recording systems. Equipment now in use gives the choice of |
several frequencies for the low- and high-frequency components.
(4) Cross-Modulation Analyzers — A high-frequency tone modu-
lated at a relatively low frequency is recorded. The high-frequency \
tone is suppressed in reproduction, and only the output (if any) at the
modulation frequency is measured. This is essentially a test for rec-
tification. It has been especially useful in variable-area photographic
recording.
There does not appear to be any serious danger of confusion or j
1948 DISTORTION MEASUREMENT 457
misunderstanding of the results of the test methods listed above, pro-
vided the practice is followed of stating the results as " total harmonic
distortion," "intermodulation distortion," and so forth. Hence,
our committee may not be called upon to recommend any modifica-
tion or amplification of existing standards. It is altogether likely
that in the application of intermodulation testing to disk recording
(and perhaps to magnetic recording) results may prove to be more
informative when other frequencies are employed than those adopted
for variable-density sound tracks. Thus, it would be inadvisable to
recommend present standardization of frequencies. However, it is
desirable that intermodulation measurement figures be accompanied
by statements of the component frequencies.
Up to the present, trouble has been experienced in making wave-
analyzer measurements of reproduced tones, because in the available
meters the filters are so sharp that speed imperfections in the recording
or reproducing machines have prevented the proper functioning of the
wave analyzer. We understand that instruments with broader
filters will, in the near future, be available. The same problem can
occur in distortion-factor meters, but several models have been on
the market which have been entirely satisfactory in this respect.
NOISE AND SlGNAL-TO-NoiSE RATIO
In the measurement of noise there has been considerable variation
of practice, and still more divergent practices are followed in specify-
ing the signal level with reference to which the noise is to be stated.
In some cases the noise is measured with a "flat" reproducing system,
and in others with a reproducing system which has purposely been
given a drooping characteristic to lessen the noise. Filters to elimi-
nate hum and rumble are sometimes employed, especially when the
purpose is purely a study of record materials. In stating signal level
it has been customary in the film industry to use practically the maxi-
mum permissible recording level. The same practice will probably
be followed in magnetic recording. On the other hand, in the field of
mechanical or disk recording it has long been customary to employ a
reference level or 1000-cycle standard signal which on a volume indi-
cator gives a reading that safely may be equaled when recording regu-
lar program material. In this there is an allowance of about 10
decibels for peaks. The standard reference signal generally adopted
is t\vo inches or five centimeters per second, maximum velocity. The
recordist notes the reading on his volume indicator produced by the
458 DISTORTION MEASUREMENT November
standard signal, and then knows that if he does not let his volume in-
dicator (which is somewhat sluggish) go above that reading, he is rea-
sonably safe with respect to the sudden peaks. Such a reference
signal has unquestioned utility, but it does not have to be identical
with the reference for specifying the signal-to-noise ratio of which a given
recording system is capable. The difference in practice may readily
lead one not familiar with the situation to think that a mechanical
system is about 10 decibels worse than it is actually is, in comparison
with other systems.
Fully conscious of the difficulties of inducing groups of people to
change any of their practices, we have nevertheless thought it worth
while to raise the question whether unified practice and terminology
are attainable and, in order that those who are interested may judge
better what might be involved, an attempt has been made to draw up
some specific proposals which, it is hoped, might be acceptable as not
upsetting established practices, while reducing the likelihood of mis-
interpretations When it is not feasible to have all people follow
identical procedures, it should at least help prevent misunderstand-
ings if certain information is given when reporting results, a require-
ment to which scarcely anyone could object In the first place, it is
proposed that the present 5-centimeter-per-second tone be called the
"recording reference signal," and that another signal, to be called
"maximum signal," be employed for determining the signal-to-noise
ratio of a system. Since the recording reference signal would not be
directly employed in signal-to-noise determinations, its definition
does not belong in the present specification, but a note is included to
point out the distinction.
Proposed definitions and specifications for signal-to-noise deter-
minations are as follows :
5.1 Noise — The term noise, as applied to a sound-reproducing
system, means any output power which is not of the same frequency
as the input, except that distortion products (harmonics or rectifica-
tion terms ) are not usually regarded technically as noise.
NOTE 1: Noise is commonly comprised of hum, rumble, or the
effect of mechanical vibrations, microphonics, thermal noise (in low-
level input circuits), tube noise, phototube hiss, and, in recorded
sound, surface noise in the case of mechanical records or graininess
plus scratches and dirt in photographic records.
Noises such as thermal noise, phototube hiss, record surface noise,
and graininess in film sound tracks have an energy distribution such
1948 DISTORTION MEASUREMENT 459
that the power within any frequency band is approximately propor-
tional to the width of the band in cycles.
NOTE 2: The usual method of measuring noise in any type of
sound record, is to reproduce from an unmodulated record and meas-
ure the noise in the reproducing system.
5.2 Modulation Noise — The presence in a record of the factor
which produces output (light transmission in a film, slope of a groove,
or a magnetization in a wire or tape) results in an increase in noise as
compared with the condition of no modulation. This extra noise,
which is modulated with the signal (generally at double signal fre-
quency) is called "modulation noise." The increase in noise due to
removal of ground-noise reduction bias in photographic sound tracks
is not true modulation noise.
NOTE 1 : Modulation noise may be measured by recording a tone
(usually of relatively low frequency) and measuring the reproduced
output with the recorded tone eliminated by a band-suppression fil-
ter of sufficient bandwidth effectively to eliminate the recorded tone.
5.3 The frequency characteristic of the reproducing system used
for measuring noise in recorded sound should be the same as that of the
reproducing systems with which the record is designed to be used,
except that if the reproducing systems include compensation for loud-
speaker characteristics, such compensation should be omitted for the
measurement.
5.4 A statement of the frequency range covered by the repro-
ducing system used in a noise measurement should accompany a
report of the measurement. If the frequency characteristic is ap-
proximately flat between the droops at the ends, the range may be
stated as between the frequencies at which the response has dropped 6
decibels below the average within the effective range. If the re-
producing characteristic used in the noise measurement is one that
has been standardized (as in the case of theater systems) reference to
the standard may serve as description of the reproducing characteristic.
5.5 If a high-pass filter is used for excluding hum and rumble when
measuring record noise, this should be stated, and the cutoff point of
the filter.
5.60 Maximum Signal— Maximum signal is a pure tone of the
maximum level that can be recorded without overload.*
* This is to be distinguished from Recording Reference Level employed in mechani-
cal systems, which has been standardized as 2 inches or 5 centimeters per second
(maximum velocity of cycle) and is about 10 decibels below maximum signal.
460 DISTORTION MEASUREMENT November
5.61 The level of maximum signal is the highest compatible with
the condition that if the input is varied as specified in 5.62, at no
frequency will the distortion exceed a specified amount (for example,
10 per cent intermodulation) .
The distortion permitted in this determination should be specified
in reporting the value of maximum signal.
5.62 In view of the fact that in nearly all program material com-
ponents of high frequency are rarely present in magnitudes as great
as the components of lower frequency, it is permissible for establish-
ing maximum signal level to reduce the input level in the high-fre-
quency range in accordance with the following rule : the input shall
be maintained at constant level from the lowest frequency comprised
n the recording range, up to 1500 cycles, above which the input
level may be reduced at the rate of 4 decibels per octave.
NOTE 1 : The purpose of Specification 5.62 is to permit the employ-
ment of a safe amount of pre-emphasis of high frequencies in record-
ing. By "safe" is meant that only in exceptional cases will material
to be recorded have peak high-frequency components with magnitudes
(relative to the peak magnitudes in the low-frequency range) any
greater than indicated by the characteristic described in 5.62.* It is
anticipated that the figure, 4 decibels per octave, and perhaps, also,
the 1500-cycle transition point, should be reviewed from time to
time in the light of accumulated experience. It should be recognized
that this characteristic is to be determined solely by the nature of the
spectra of music and speech, and that it does not and should not make
any allowance for properties of recording systems which may cause
overload to occur at lower levels in one portion of the frequency range
than in others. It is, however, in order, that this permitted droop
should take account of distortion tolerance, provided experience justi-
fies such an allowance. (For example, harmonics of high-frequency
tones may not be reproduced, and if cross-modulation is not en-
countered, the tolerance may be greater in this range.)
5.7 Signal-to-Noise Ratio — Signal-to-noise ratio is the ratio of
maximum signal to noise.
* Allowable pre-emphasis is discussed in a report by J. K. Hilliard and J. P.
Maxfield in Audio Engineering for April, 1948. Permissible pre-emphasis has
been well expressed by Hilliard as "that which causes equal probability of dis-
tortion at any part of the important frequency range covered by the record."
It is also pointed out that compensation for reduced response of some element of
the recording channel (such as high-frequency droop of a microphone) is not a
part of pre-emphasis.
1948 DISTORTION MEASUREMENT 461
5.71 The limiting distortion (as, for example, 2 per cent total
harmonic distortion or 8 per cent intermodulation distortion) per-
mitted for the purpose of determining maximum signal should be
stated when reporting signal-to-noise ratio, and also the information
with respect to the noise measurements, called for in 5.4 and 5.5.
NOTE 1 : It is evident that a high-quality system which sets a low
limit to permissible distortion, and which reproduces a wide fre-
quency range, cannot realize from a given record material or surface
quality, as large a signal-to-noise ratio as a system in which more dis-
tortion is permitted and a limited frequency range covered.
APPENDIX
(Discussion of some of the proposals on which there have been differ-
ences of opinion.)
SPECIFICATION OF FLUTTER IN TERMS OF ROOT-MEAN-SQUARE
Although the proposal to define flutter as the root-mean-square
deviation in frequency has had the approval of both the SMPE sound
committee and the ASA subcommittee, a number of engineers have
expressed the belief that the peak deviation is likely to be a better
measure of the quality damage than either the root-mean-square or
average deviation. I believe that the thought behind this view is
that we tolerate small fluctuations in speed and may be quite un-
aware of them, but the larger deviations from average speed are
quickly noticed ; therefore, it is reasonable to suppose that maximum
deviations are what count. This reasoning is quite logical, but when
comparing a peak-reading meter with a root-mean-square meter, the
question is not whether extra importance should be attached to the
largest deviations, but whether everything else should be ignored
and no consideration given to the duration of the large deviation.
Fig. 1 shows several types of flutter curve, all having the same peak
value. In curve B, there is little beside the one high peak, while
curve A shows numerous other fluctuations of only slightly lower
amplitude than the peak, and curve C shows deviations which last
many times longer. Should these three be rated alike?
Someone comparing these curves may bring up the point that if the
flutter-measuring system is provided with a weighting network such
as the "Flutter-Index" idea suggests, thus emphasizing flutter of
rather slow rates, the peak-reading meter would give much larger
readings for the flutter shown by curve C than curve J5, since the
462
DISTORTION MEASUREMENT
November
short sharp peak of B could be produced only by high-frequency com-
ponents. The reply to this is that if and when such networks come
into general use, there will be much less difference in the actions of
peak, root-mean-square, and average-reading meters. Such a net-
work narrows the band of nut-
ter rates to which full weight
is given (with attenuation on
both sides of the range 1 to 5
cycles). Were only a single
nutter rate to be considered in
any one measurement, it would
make no difference which type
of meter is used, for their indi-
cations would be in fixed ratio,
and agreement in readings is
only a matter of calibration or
scale. (A pure sine wave with
a peak value of unity has a
root-mean-square value of
0.707 and an average value of
0.64.) But grave errors would
result from assuming that these
V
Fig. 1 — Three frequency-deviation
curves for which a peak-reading meter
would give the same reading.
ratios Avould hold for the ragged waves shown by ordinary flutter
records.
The characteristic of a root-mean-square meter is admirably adapted
to giving special weight to the larger frequency deviations. A 1 per
cent deviation of given duration has four times the effect on the meter
that a 0.5 per cent deviation of the same duration would have. Why
then does it not read four times as high? The answer is that, in
general, the effect which deflects the needle does go up as the aver-
age square of the quantity being measured, but the meter is provided
with a nonuniform scale which in effect extracts the square root.
There is reason for believing that the square lawr is a fairly good
approximation to the noticeability of flutter of any given rate. At
least we can say this, there is a threshold magnitude below which no
one notices the flutter, but with small increases, it seems to get
bad very rapidly, and the number of people who notice it increases
rapidly. For example, in a letter on the subject, Kenneth Lambert
wrote "We have observed here that 24-cycle flutter may become
rapidly objectionable with an increase from perhaps 0.1 to 0.13 per
1948 DISTORTION MEASUREMENT 463
cent (on a root-mean-square basis) and 96-cycle flutter in much the
same degree." If you were to attempt to draw a curve which would
represent the harmful effect of flutter as affected by its magnitude,
you would draw something that looks a good deal like a parabola.
But what about a meter of the rectifier type, which reads the
average of the waves? I think I can give an illustration of a case in
which such a meter would be decidedly in error. Suppose you were
comparing two machines, one of which produced every second a
1 per cent deviation that lasted a tenth of a second, while the other
produced every second two deviations of 0.5 per cent each lasting a
tenth of a second. The root-mean-square meter would rate the for-
i mer as twice as bad. The averaging meter would say they were equal.
You will notice that the flutter specifications as they now stand, al-
though stating the standard as a root-mean-square reading, sanction
| the use of meters of the rectifier type. Meters of this type actually
t show something between the root-mean-square and average values
j of the waves, and therefore it is believed that the difference ordinarily
will be small. In view of that consideration, and the fact that
! equipment is in wide use having various types of indicating instru-
ments, it has not appeared practical to call for strict adherence to the
root-mean-square standard, but it is an urgent requirement that the
type of measurement be clearly stated when a measurement is reported.
Examination of a flutter oscillogram (or "wowgram") does not enable
' one to arrive at a root-mean-square figure, hence the widespread prac-
f tice on the part of those who have recording meters, of reporting flutter
L as the peak-to-peak value. It is hoped that future recording meters
will also be equipped with root-mean-square indicating meters, so
that simultaneous readings may be taken. Meantime it is important,
in order to avoid confusion, that statements of flutter magnitude,
based on inspection of a wowgram, should be accompanied by enough
information to make clear how the wowgram was read, such as
"peak deviation from average," or "range positive to negative peak,"
iand to make certain that the figure will not be taken as a root-mean-
square measure.
FLUTTER INDEX
The subject of "Flutter Index" as proposed by the SMPE sound
committee has met with many questions and indications of doubt.
•Not that many doubt the correctness of the tests for the conditions
under which they were made, but they doubt the applicability of the
formula to the actual conditions under which reproduced sound is heard.
464 DISTORTION MEASUREMENT November
For what bearing it may have on the questions in people's minds, I
would like to call attention to a curve published by Shower and Bid-
dulph in a paper on "Differential pitch sensitivity of the ear," in the
Journal of the Acoustical Society of America, October, 1931. Their
curve showed the threshold of perception of rhythmic frequency
changes of a 1000-cycle tone at rates from 0.7 to 5.5 cycles per second,
the listening being done with headphones. The interesting thing is
that this curve, although having a minimum some 10 times higher
than the minimum for a 1000-cycle tone as found in the live-room
listening tests, was very nearly the same shape.
If nutter rate affects perception threshold in the same manner under
the two extreme conditions of complete absence of reverberation on
the one hand (headphones) and live-room listening on the other, the
general shape is probably not far off for intermediate conditions.
Thus, although the evidence is still insufficient ( the Shower and the
Biddulph data should be supplemented with tests at higher nutter
rates, and with other tones, and tests also made in moderately damped
rooms) , there is a presumption that a curve of the general shape in-
dicated by the Flutter-Index formula would, to a fair degree of ap-
proximation, express the relative perceptibility of nutter under aver-
age listening conditions.
Ultimately, it is hoped, such a formula and weighting factor can be
"proved in" for music. Recent experience has demonstrated to the
writer's own satisfaction, that the tolerance for rapid nutter is much
higher than for slow nutter. Thus, so far as this small item of in-
formation goes, it points to a similar relation for music as for steady
tones, although again probably with a higher threshold throughout
than for steady tones.
RATE OF REDUCTION OF INPUT OF HIGH FREQUENCIES
In many systems of recording, the practice has been followed of
increasing the amplification in the recording channel with increasing
frequency. This has been called "pre-emphasis." In order that the
final sound shall not have exaggerated high frequencies, the reproduc-
ing system is given a drooping characteristic and this has been called
"postequalization." The result is a desired over-all frequency char-
acteristic, but a large reduction in surface noise or in graininess noise
due to the imperfections of the record. Pre-emphasis is possible
without excessive overloading at high frequencies, for the reason that
DISTORTION MEASUREMENT
465
except in rare instances the program material itself contains the high
frequencies, only in very much reduced amplitudes. There are two
possible ways of ascertaining how much pre-emphasis is safe or in
other words how much may be employed without resulting in any
more overloading at high frequency than in the lower range. The
first method is to make extensive measurements such as those re-
ported by Sivian, Dunn, and White (Journal of Acoustical Society,
April, 1931) and by Dunn- and White in the January, 1940, issue of
that Journal.
Fig. 2 — Input droop for determination of maximum signal.
I. Orthacoustic-recording characteristic. II. Assumed safe pre-emphasis.
III. Permissible high-frequency reduction.
The other method is one of cut-and-try, namely, to make record-
ings with various amounts of pre-emphasis or " tip-up," and learn
from general experience how much may be used. Because of the very
extensive use that has been made of the "orthacoustic" recording
characteristic (see Fig. 2), this appears to be the best general guide
to possible tip-up. I have endeavored to obtain expressions from
engineers who have had experience with this system, and while some
express the opinion that the tip-up is excessive there does not seem to
be any overwhelming evidence to that effect. The overloading at
high frequency seems to be rather because of the fact that in disk
recording the actual overload level (in terms of velocity) is lower at
high frequency because of curvature effects, especially near the
466 DISTORTION MEASUREMENT November
inside of the records. If the overload point were always at the same
velocity the orthacoustic rate of tip-up probably would not be exces-
sive, or in other words, it does not more than offset the droop in input
as found in a variety of programs. Following this line of reasoning
the writer suggested to the members of our ASA subcommittee that
we propose, for testing purposes, holding the input constant to 1500
cycles and then dropping at the rate of 5 decibels per octave. This
would be slightly less than the complement of the orthacoustie-
recording characteristic. However, some of the committee thought
we should be more conservative, suggesting 3 instead of 5 decibels
par octave above 1500 cycles. For purposes of discussion I have
called, in the present draft, for 4 decibels per octave. I hope to re-
ceive more expressions in regard to this question. The pre-emphasis
used in film recording is too small to constitute a test of this factor.
If a recording system is designed with a large amount of pre-
emphasis it may in practice be quite capable of handling high levels
of recording, but if tested by means of an oscillator with constant-
voltage input throughout the frequency range, obviously it would be
badly overloaded at the high-frequency end, unless the input through-
out the entire range were dropped to a level far below the power-
handling capacity of the system. We therefore think it justified to
suggest that the determination of maximum signal (especially for
signal-to-noise determinations) be made with an input which droops
at high frequency, as shown, for example, by curve /// of Fig. 2.
If the actual overload level of the system is the same at all fre-
quencies (as, for example, in a variable-area recording) a tip-up of 4
decibels per octave above 1500 cycles could, if curve /// is what it
should be, be employed without any greater likelihood of overload at
high frequency than at low.
If a system overloads at a lower level in a certain frequency range
than in another, then either the over-all level will have to be dropped
in order to meet the specifications for determining maximum signal,
or else the recording characteristic should be modified to reduce the
recording level in this critical range. For example, assuming 4
decibels per octave above 1500 cycles to be the maximum safe pre-
emphasis with constant overload level, disk-recording systems would,
on account of the curvature troubles, either have to establish their
"maximum signal" somewhat below the level which constitutes
overload in the low- and middle-frequency range, or else stop short of
applying the 4 decibels per octave all the way up to 12,000 cycles.
1948 DISTORTION MEASUREMENT 467
DISCUSSION
J. P. MAXFIELD: In connection with the matter of this equalization, I had the
opportunity to obtain simultaneous cuts of records running 15 decibels at 10,000,
around nine and around six. Those were all played with the standard droop for
the higher present equalization. Astonishingly enough, the one with the 6-
decibel rise in the recording reproduced, more highs and cleaner highs than the
one with the 15, indicating that the overload on the latter was so bad that it
was not being tracked. Unfortunately we have only some three sets of such rec-
ords, but I think the situation should be carefully looked into before we pick as
high a rise as 15 decibels to 10,000.
J. K. HILLIARD: I should like to make some comment along this line, in support
of Mr. MaxfiekTs statement. There is other information that also verifies that we
should be cautious in the matter of equalization. I think the experience of the
studios in having parallel disk and film channels would indicate that if equaliza-
tion is provided higher than the six or eight or possibly even ten at a maximum, we
get into this bootstrap lift, the effect of having to lower the level or use excessive
limiting at times and since present equalization is used primarily to increase the
signal-to-noise ratio, if we have to lower the level on the disk or the film in a
general case, then we are defeating the purpose we went out to obtain, and we
have the two problems of having to lower the signal on the record at the time if
we used higher amounts of equalization, and the effect that Mr. Maxfield talked
about, improper tracking, leads to inferior results over that used with lower
amounts of equalization.
CHAIRMAN C. R. DAILY: There is another point, that of excessive modulation
at high frequency. Most types of recording systems lead to increase of cross-
modulation production which may do more harm than the increase in signal level.
E. W. KELLOGG (by letter) : The experiences recounted by Mr. Maxfield and
Mr. Hilliard are just the kind of evidence which we are seeking, and I hope more
experiences bearing on this subject will be reported. The suggested 4 decibels
per octave above 1500 cycles gives a level difference between 1000 and 10,000 cycles
of about 11 decibels which is 4 decibels less than that given by the orthacoustic
curve, and much nearer the conservative figure which I think Maxfield and Hilliard
would approve. But please notice that to say that the input may be drooped at
a certain rate for test purposes, is not by any means equivalent to recommending
a tip-up of equal amount. Our purpose is to try to arrive at a curve which repre-
sents the droop which may be expected to occur in average program material.
If, as in disk recording, overload tends to occur at lower levels in the high-frequency
range (due to curvature) it obviously would be inviting overload, to use a tip-up
which completely offset the normal droop. Hence the figure suggested here may
not be out of line at all with the observations just reported. The question then
arises, how have the people who have used the orthacoustic- recording characteris-
tic gotten by with it as well as they have? There are, of course, differences in
microphones, but, more important, the recordist can exercise a wide control by
such factors as orchestra arrangement, microphone placement, and room acous-
tics. The influence of these factors makes it impossible to say what the average
or normal droop is, but some more or less arbitrary specification seems to be
needed for putting testing systems on a common basis, and unless experience
indicates clearly that it should be revised (and it is open to revision), the figure
proposed in the paper seems to me to be reasonable.
Magnetic Recording
for the Technician*
BY DOROTHY O'DEA
FORMERLY, RCA VICTOR DIVISION, HOLLYWOOD, CALIFORNIA
Summary — The first half of this paper will present to the motion picture
technician a review of magnetic-recording theory; the second half consists
of experimental data taken with the new magnetic-recording equipment of
the Radio Corporation of America. Input-output, frequency-response, and
distortion data, which were taken under test conditions familiar to motion
picture technicians, are presented. There are many excellent articles
available which treat the various aspects of this subject. Those who are
interested in the detailed scientific explanations are referred to these articles,
listed in the bibliography, and to the extensive patent literature. This
paper attempts to consolidate the information in these articles in simplified
form and to provide a useful picture of the phenomena in magnetic record-
ing and reproduction for those whose primary interest is in the application of
the theory.
PART I— THEORY
PERHAPS A GOOD starting place for a discussion of magnetic re
cording would be the two distinct types of magnetic material
which are essential parts of the system. Materials are classed in thi
groups from a standpoint of magnetism. A vacuum has a perm<
ability of unity, which means that the ratio of magnetic induction
to magnetizing force H is 1.0. Diamagnetic materials have a perni(
ability less than unity. Paramagnetic materials have a permeability
somewhat greater than unity. The ferromagnetic materials have
permeability very much higher than unity and this permeability
is variable depending on the particular material and the magnetizii
force applied to it. We are concerned only with these ferromagnetii
materials. They can be subdivided further into two types; hai
and soft, both of which are used in magnetic recording. Soft rm
netic materials, which incidentally are usually soft physically, have
low retentivity; that is, they are easily affected by a magnetic fielc
and easily lose the effect when the exciting field is removed. This
type of material is commonly used in transformers, galvanomet
pole pieces, and magnetic heads.
* Presented May 18, 1948, at the SMPE Convention in Santa Monica.
468 NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
MAGNETIC RECORDING
469
The other type of magnetic material, which is called hard, has high
retentivity; that is, it is not so easily affected by a magnetic field
as the soft materials, but when affected, has the property of retaining
a major portion of this effect after the exciting field is removed. An
example of this is the familiar permanent magnet. Now it is ob-
vious that if a long strip of this hard material were to be subjected to
a magnetic field which was varying according to a voice signal, while
the material was moving at a constant rate, a sound record would be
impressed on the material. In order to confine the signal to a reason-
able length or a reasonable speed for a given length, it is desirable to
PERPENDICULAR
TRANSVERSE
Fig. 1 — Types of recording — longitudinal, perpendicular, transverse.
limit the exposing effect at a given instant to a small portion of the
hard magnetic material. This is, of course, obvious, as in photo-
graphic recording.
The difficulties of confining the signal to a small portion when
using a coil as the exciting field are apparent. Therefore, the coil is
wound around a core which also increases its efficiency considerably.
This core contains an air gap so that the magnetic lines of flux going
around the core leak out at the gap. The hard material is pulled
over the top of the gap, thereby recording and retaining the leakage
flux which varies according to the audio signal. There are various
methods of placing the film in the region of influence of the core or
magnetic head, as shown in Fig. 1. With the perpendicular method,
470 O'DEA November
the film is actually passed between two pole pieces or is pulled through
the gap. Here the length of the magnet from north to south is con-
stant for a given film thickness. One advantage of this method is
that the aspect ratio of the individual wavelengths, which make up
the signal, is less effective in controlling the high-frequency response.
One disadvantage of this method is that it is desirable for reasons of
efficiency and quality to have the gap as narrow as possible, thereby
necessitating a very thin film which then introduces the problem of
strength and durability of the film.
The next method, shown in the right-hand corner, transverse
recording, has the same advantage as the perpendicular method of
constant-aspect ratio, but has the disadvantage of using a narrow
tape. Here the length of the magnets from north to south is constant
for a given film width. Very little information has been found in
the literature on this type. There seems to be no practical use for it
in tape recording and in wire recording it is the same as the perpen-
dicular method.
The method of longitudinal recording is more commonly used at
the present time. It is the method which RCA is using and will be
discussed in more detail. Three types are shown here. The first
type uses only one pole piece; the second uses two offset pole pieces;
and the third method uses the ring-type head. In all three types the
length of the magnets from north to south depends on the wavelength
of the signal. It is apparent that with this method, wide, strong film
can be used, as only the thickness of the coating is important mag-
netically. Sturdy pole pieces can also be used and the gap width
can be very small. The principal disadvantage seems to be that the
aspect ratio decreases with increasing frequency, resulting in high-
frequency losses.
Now that we have covered very simply the method of recording,
let us return to the type of materials used and a brief discussion of
the properties of materials best suited for this purpose. The most
desirable magnetic material for the core would be one which has a
very high initial permeability, very low hysteresis loss, and very
low eddy-current losses. .
For commercial purposes, MU metal meets these requirements
satisfactorily. It is listed as having an initial permeability of 7000,
hysteresis loss is acceptable, and the eddy-current losses are mini-
mized satisfactorily by laminating the cores.
The characteristics of the high-retentivity material used for the
1948 MAGNETIC RECORDING 471
film appear somewhat more involved. Any investigation of films is
very complicated unless certain factors are kept constant such as tape
speed, thickness of emulsion layer, and construction details of the
heads. The obvious starting place for an investigation of films would
be a study of the hysteresis loop of the iron-oxide material. However,
there seems to be considerable controversy about the value of these
curves for magnetic recording. It seems reasonable to doubt that
the theory of magnetic recording could be traced out directly on this
curve, if only because of the secondary effects introduced by the bias
frequency (which will be discussed later) . We believe that it will be
essential to find some modification of this curve with which we can
correlate practical experiments with theory. Several investigations
are being carried out along these lines, but to date, we know of no
theory which successfully explains all the factors involved in record-
ing. The following hypothesis is presented with knowledge of the
many deficiencies it contains. It is hoped it will provide a useful
picture and a stimulus for more comprehensive work. The stand-
ards book1 of the American Society of Testing Materials has been
use as a reference for the definitions of magnetic terms.
Fig. 2 shows in a highly simplified manner the over-all transfer
characteristics of magnetic recording and reproduction. This figure
does not show any of the phenomena peculiar to the type of film, fre-
quency, or speed of recording, but some of these factors will be taken
up as individual steps.
Starting with the audio current, shown in the lower left corner,
which is combined with a high-frequency biasing current, the currents
produce magnetic flux by the action of the coil. The flux which then
flows in the core and across the gap is proportional to the current if
the core material has low losses and has a saturation value above the
range required. For our purpose, we can assume that the head has a
straight-line transfer characteristic, and the 45-degree line in the
lower left corner represents this characteristic. The signal on the
left represents the flux in the core. The magnetic flux across the
gap causes a magnetic force to affect the film which is passing over
the gap and this force results in the magnetic flux flowing into the
film. The line in the upper left corner represents the film characteris-
tic which is nonlinear. It is effectively straightened out by the
biasing current. As the film is a permanent magnet and is being
moved past the gap, this flux leaves the film in a magnetized condi-
tion. Now the film is a magnetized body and is effectively the same
472
O'DEA
November
as the record head in that its force will produce a magnetic flux in t
air surrounding it. It differs from the record head in that its for
is permanent and varies with distance along the film while the force
in the record head varies with time. The curves at the top show the
forces on the film and the dashed line shows the resultant force.
The film is now ready for immediate reproduction if desired. When
it is passed over another head the flux in the air around the film passes
into the head and induces in the coil an electromotive force which is
amplified in the customary manner. The line in the upper right
s
INPUT TO RECORDING HEAD INPUT TO REPRODUCING AMPLIFIERS
Fig. 2 — Over-all transfer characteristics.
corner represents the transfer from flux around film to flux in the
head. The line in the lower right corner represents the transfer
from flux in the head to induced voltage.
Fig. 3 shows a way in which the recording action can be represented
simply although it does not show some important effects. For the
purpose of illustration, the input signal in the lower left corner is
drawn as if it were a 9000-cycle audio wave with 27 kilocycles bias.
The numbers indicate representative points on the signal which can
be found on the magnetization curve. The arrows on the magnet-
ization curve show the travel of the biasing action. The waves on
1948
MAGNETIC RECORDING
473
the right indicate the magnetization on the film and the numbers
represent the same points as on the original input wave. The re-
sultant signal is formed by these two waves and is seen to be an un-
distorted signal. Directly above the biased input signal is an un-
biased input wave and it is apparent that the jog in the magnetization
curve will cause the output wave to be distorted as shown on the
right. The bias current keeps the audio current above this jog or
effectively straightens out the curve. If insufficient bias is used,
distortion results and the output level drops. If too much bias is
used the output level also drops.
MAGNETIZATION CURVE
Fig. 3 — Enlarged view of recording characteristic.
The effect of demagnetization, which was mentioned earlier, oc-
curs in the recording operation. In longitudinal recording, the
wavelength is recorded along the length of the film and at 9000 cycles
the wavelength would be 0.002 inch at normal motion picture film
speed. The width of the recorded track is 0.200 inch so a half wave
consists of a rectangular magnet 0.200 inch wide and varying from
north to south along a length of only 0.001 inch. This ratio is not at
all to the liking of a magnet and it has a strong tendency to demagnet-
ize itself. This ratio has been expressed2 as "aspect ratio"
and the example given above would have an aspect ratio of 0.005.
474
O'DEA
November
Manildi2 has shown that an efficient magnet should have an aspect
ratio of at least 8 for a certain permanent-magnet material. In order
to determine with any accuracy the proper aspect ratio for our re-
cording films, it would be necessary to know the effective hysteresis
curve of the material. However, it is probable that a ratio of 0.005
is low enough to cause high-frequency losses comparable to slit losses
or greater.
Camras3 presents an interesting study of this effect on wire re-
cording and points out that fortunately some of the effect is counter-
EQUALIZED FILM
s
900 1000
FREQUENCY
10,000
Fig. 4 — Frequency-response characteristic with and without equalization.
acted by the action of the reproducing poles. Concerning this coun-
teraction, Wetzel4 shows that the present recovery during playback
increases with increasing frequency, but that the sum of these oppos-
ing tendencies is a net decrease in output with increased frequency.
Equalization is provided in recording to compensate for this loss as
well as slit loss. A third factor which influences the high-frequency
response is the penetration effect. Kornei6 also discusses this.
When the thickness of the recording layer is decreased, the output
level of the low frequencies drops, while the relative high-frequency
response improves. The penetration of the magnetization depends
1948 MAGNETIC RECORDING 475
on wavelength, permeability, and gap width. However interesting
this effect may be, there is little need to go into it further at this
time, because, as Kornei5 points out, the effect is extremely small with
thicknesses in the order of 0.0005 or 0.001 inch. The amount of
equalization required for demagnetization and slit loss is quite small
compared to that required for the next effect.
In reproduction, the output voltage theoretically is directly pro-
portional to frequency for constant-amplitude input. The pickup
can be considered a generator with the head itself acting as a coil
which remains stationary in a changing magnetic field. As the flux
surrounding the film enters the region of the gap, the force causes
flux to flow in the core of the head and as this changing flux links
the electrical circuit of the coil, there is induced in the coil a
voltage which tends to oppose the change. This voltage will be
out of phase with the flux because when the flux passes through zero,
the rate of change is maximum and, therefore, the voltage is maxi-
mum. The voltage is proportional to frequency because in a given
distance the rate of change for a high frequency is greater than for a
low frequency. The equalization required to compensate for this
effect is 6 decibels per octave from whatever frequency is set as the
low-end limit to the frequency at which demagnetization and slit
loss take effect.
Combining these several causes of frequency changes, we have the
familiar frequency-response characteristic without equalization (Fig.
4) . As the equalization required for the low end is the larger amount,
it will affect signal-to-noise ratio. Our experience has been that with
most good films, the equalization brings the output down to about
50 decibels above system noise with the film noise below system noise.
Care must be taken to keep system noise as low as possible. Another
way to increase signal-to-noise ratio, is to tolerate a higher frequency
as the lowest limit. For example, if the lowest frequency desired is
100 cycles, the output and consequently the signal-to-noise ratio is 6
decibels higher than it would be if the lowest frequency were 50
cycles.
PART II— EXPERIMENTAL DATA
All the test data presented here have been taken with RCA's new
magnetic-recording kits. In the interests of users of these kits, they
were designed, not to obtain optimum conditions for this type of
recording, but to obtain the least possible difficulty in the studios
476
O'DEA
November
caused by changing from photographic to magnetic film. Using an
existing recorder created many constructional difficulties such as
space for the heads, plugs, and mounting arrangements. Also, in
the interests of users, the decision was made early to use 35-mm film
with standard dimensions and perforations. In the fall of 1946,
duPont was approached with a request for some experimental film
on 35-mm base. In February, 1947, the first samples of this film
were received and tests were started with a Brush head. These first
tests indicated that the general idea of 35-mm magnetic film was
+ 16 OBM INPUT
BIAS CURRENT-MA
Fig. 5 — Bias input versus audio output.
practical and construction on RCA heads was started by our develop-
ment-engineering group in Camden. These new heads were then
mounted on a PB-36 soundhead with some of the same features it
was intended to incorporate in the final design. During this time,
other film manufacturers were becoming interested in this field and
the original maker was constantly improving his samples. It is
not the intent of this paper to make comparisons between films made
by different manufacturers. However, it is difficult to avoid dis-
tinguishing between films, because of the difference in equalization
required. Complete tests have been made on only duPont and
1948 MAGNETIC RECORDING 477
Minnesota Mining films, both of which could be used for recording.
For the purpose of demonstrating the capabilities of this system, all
the data in this paper were made on duPont SW4 film.
BIAS INPUT VERSUS OUTPUT
A good starting place for film tests appears to be a curve of bias
input versus audio output (Fig. 5) shown here for two input levels.
As mentioned earlier, the bias current has the effect of straightening
the transfer characteristic of the film, thereby decreasing distortion.
+ 8 +16
AUDIO INPUT— DBM
Fig. 6 — Audio input versus audio output.
Output is increased with increasing bias current up to a certain point,
after which output decreases with increasing bias current. Some
investigators think of this decrease as being caused by the bias current
being high enough to erase the audio signal.
AUDIO INPUT VERSUS OUTPUT
Now that the optimum bias current has been chosen, a curve is
plotted of audio input versus output (Fig. 6) using this bias current.
Linearity is very good up to the overload point of the film. This
overload is gradual, being somewhat similar to the characteristic of
478
O'DEA
November
I 5-°
Fig. 7 — Audio output versus total harmonic distortion.
TPUT{+ I60BM INPUT)
STORTION(-H6D8M INPUT)
DISTORTION (+4 OBM INPUT)
23 4 5 6
BIAS CURRENT-MA
9 10
Fig. 8— Bias input versus total harmonic distortion.
1948
MAGNETIC RECORDING
479
variable-density recording and the 100 per cent track level is deter-
mined by the maximum allowable distortion.
DISTORTION
Most of our tests have been made with harmonic-distortion equip-
ment. Fig. 7 shows the effect of varying the audio level on dis-
tortion. In this case, distortion is plotted against audio output.
The optimum bias was used here.
BIAS CURRENT A-
500 1000
FREQUENCY
Fig. 9 — Frequency response.
(0,000
Fig. 8 shows curves of distortion versus bias for two audio inputs.
The two upper curves are output curves and the lower ones are
corresponding distortion curves.
FREQUENCY RESPONSE
Using the optimum bias current and an audio input which is
safely below the film-overload point for the high frequencies, we ob-
tain the frequency-response curve of Fig. 9. Since our equipment was
designed especially for motion picture use, the upper frequency limit
was set at 8000 cycles. However, we know from experiments that it
can be held flat to 10,000. The dashed curves show the frequency
480 O'DEA
response which would be possible if a lower bias current had been
used. Had a lower current been used, however, the distortion would
have gone up.
NOISE
In order to make accurate statements concerning signal-to-noise
ratio for magnetic recording, it is necessary to specify not only the
maximum harmonic distortion, but also the permissible frequency
limits and the characteristics of the measuring channel.
BIBLIOGRAPHY
(1) "Book of American Society of Testing Materials Standards," Philadelphia
Pa., 1936.
(2) J. F. Manildi, "Multiple magnetic circuits," Electronics, pp. 160-163;
November, 1946.
(3) Marvin Camras, "Theoretical response from a magnetic-wire record,"
Proc. I.R.E., vol. 34, pp. 597-603; August, 1946.
(4) W. W. Wetzel, "Review of the present status of magnetic recording
theory," Audio Eng., pp. 14-17; November, 1947.
(5) Otto Kornei, "Frequency response of magnetic recording," Electronics,
pp. 124-128; August, 1947.
(6) L. C. Holmes and D. L. Clark, "Supersonic bias for magnetic recording,"
Electronics, pp. 126-136; July, 1945.
(7) D. E. Wooldridge, "Signal and noise levels in magnetic tape recording,"
Trans. A.I.E.E., vol. 65, pp. 342-352; June, 1946.
(8) Lynn Holmes, "Factors influencing the choice of a medium for magnetic
recording," J. Acous. Soc. Amer., vol. 19, pp. 395-403; May, 1947.
(9) R. A. Power, "The German magnetophon," Wireless World, pp. 195-198;
June, 1946.
(10) C. N. Hickman, "Sound recording on magnetic tape," Bell Sys. Tech.
Jour., vol. 16, pp. 165-177; April, 1937.
(11) James Z. Menard, "High frequency magnetophon magnetic sound re-
corders," Final Report No. 705, Published by Office of Military Government
for Germany (U. S.), Office of Director of Intelligence, Field Information Agency,
Technical, January, 1946.
(12) Heinz Liibeck, "Magnetic sound recording with film and ringheads"
(Translated by W. F. Meeker), Akus. Zeits., vol. 2, pp. 273; November, 1937.
(13) O. W. Eshbach, "Handbook of Engineering Fundamentals," John Wiley
and Sons, Inc., New York, N. Y., 1936.
(14) "Magnetic Recording," Reprint from /. Soc. Mot. Pict. Eny., vol. 48,
pp. 1-62; January, 1947.
35-Mm Magnetic Recording System*
BY EARL MASTERSON
RCA VICTOR DIVISION, CAMDEN, NEW JERSEY
Summary — Ail idea was conceived of designing and building a number of
kits to add magnetic sound-recording facilities to a standard photographic
recorder. It is believed that by starting magnetic recording in this manner
it will enable the studios to obtain some practical experience without the
expense of a complete film-handling mechanism and yet will not interfere
with photographic sound-recording production work. The construction of
the mechanical and electrical components of the kit and the operational
features are discussed as well as the performance characteristics that can
be expected of this system.
i LTHOUGH MAGNETIC RECORDING is one of the oldest recording
J\. methods known, it was only during the last war that this form of
recording came into its own. In more recent years the high fidelity
obtainable by properly designed recording and reproducing equip-
ment, together with improved recording media, made magnetic re-
cording actually a competitor in several sound-recording fields.
Demonstrations of the quality of performance of a laboratory 35-mm
recorder were given to several groups of Hollywood people in Cam-
den. New Jersey. It was felt that the quality of reproduction might
be of great interest to the studios, although it appeared that consider-
able experience would be necessary to determine how successfully
this entirely new medium could be fitted into the operations. It
was also decided that the best method of gaining this experience
would be to design a conversion kit for a standard photographic sound
recorder, so that either photographic or magnetic recordings could be
made on the same machine. In this way operational experience could
be gained without seriously interfering with regular production work.
The design and features of a magnetic conversion kit to adapt a well-
known 35-mm sound recorder for magnetic recording wirl be de-
scribed.
Although almost all of the early work in magnetic recording had
been with solid wire or solid tape, it was found during the war, both
in this country and especially in Germany, that a better recording
f* Presented May 18, 1948, at the SMPE Convention in Santa Monica.
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 481
482
MASTERSON
November
medium could be made in the form of a magnetic coating on a non-
magnetic support or base such as paper or plastic. This form of mag-
netic-recording medium shows several advantages over the solid
type. The two major requirements of a solid recording medium are
that it be both ductile and a permanent magnet. These requirements
are not normally compatible. As a result of considerable work on the
part of wire manufacturers, a satisfactory wire has been developed
for the average quality requirements of sound recording. One
Fig. 1 — Magnetic conversion kit in PR-23 recorder.
company has solved the problem of incompatibility by using a brass
wire which has excellent drawing properties and plating a magnetic
coating upon this wire which has the proper recording properties.
This still does not meet the requirements of the strictly high-quality
sound-recording field.
By suspending a magnetic oxide in a binder and coating this mix-
ture on a film-support base, it is possible to combine both the correct
magnetic qualities and a support having desirable mechanical quali-
ties. Since the magnetic particles of a coated tape can be extremely
fine and also well distributed and suspended in a binder, it is possible
1948
35-Mn MAGNETIC RECORDING
483
to record and reproduce wavelengths of much smaller physical length
than that possible on solid wire or tape, the factor being something
like 3 to 1. This means that a magnetic track at the standard 35-
mm recording speed of 90 feet per minute is capable of producing
excellent frequency response.
Hollywood, unlike other sound-recording industries, is quite for-
tunate in having had years of experience in the handling of long tapes
or films. This makes the conversion from present recording methods
to magnetic recording a relatively simple one, so simple, in fact, that a
conversion kit appears to be a very good answer to the immediate
Fig. 2 — Electrical units of magnetic conversion kit.
! problem of introducing magnetic recording to Hollywood. In order
I* to make the kit available to most studios, it was decided that it
\ should be designed to fit the widely used RCA PR-23 sound recorder.
Fig. 1 shows the installation of the complete mechanical conversion,
kit. In applying this kit to the PR-23, a minimum of work is re-
quired. Several roller studs are pressed out of the frame, and some
new studs, supplied with the kit, are pressed in, to secure the head and
idler mounting plate. This conversion mechanism, includes a tight-
loop sprung idler film filter system and the magnetic-recording and -re-
producing heads complete with all adjustments. A nonmagnetic
484
MASTERSON
November
sound drum and shaft is also supplied. Electrical connections for
the heads are made to small Cannon sockets in the side of the new film-
compartment housing.
It is assumed that most users will erase the film as a separate opera-
tion not connected with recording. However, if it is desired, an
erasing head can be furnished as a separate kit. This head is pivoted
on its own mounting plate and rides on the film on the top of the
sound drum. The sound drum is relieved in the area under the head
to prevent its bounding or hammering.
Fig. 3 — Film path and magnetic heads.
Iii order to make the electrical kit as flexible as possible for various
installation requirements, it was decided to use individual plug-in
chassis which could be mounted either on a single shelf in a rack or in
two boxes as shown in Fig. 2. By means of adjustable brackets fur-
nished with the kit, it is possible to mount these boxes on a wall, on a
table, or underneath a table. A door is provided in the end of the
box so that the chassis may be readily removed for servicing. One
box contains a single chassis which provides an oscillator with its
1948
35-Mn MAGNETIC RECORDING
485
own power supply for high-frequency bias and erasing currents.
This chassis also contains the necessary recording equalization cir-
cuits. A meter on the front panel enables an operator to measure
bias current. The other box contains two preamplifiers and a chas-
sis for the playback equalization network.
Fig. 3 shows the film path with the kit mounted on the PR-23
recorder. A drawing of the new film path and the mounting arrange-
ment for the recording and reproducing heads is shown in Fig. 4.
As the film enters the compartment from the magazine, it passes the
sprocket in the usual manner and then passes around a sprung idler
which acts both as a filter and as a film guide for the photographic-
track placement. After the film passes around the sound drum, it
KI&W FILM
COMPARTME-NT
SPRUNG
IDLE-R.S
ERASING
HE-AD \
OPTICAL
BAR.R.E-L
SOUND
DRUM '
FILM
RE-CORDING HEAD -^ / <- RE-PRODUCING HE-AD
PULL PIN^
Fig. 4 — Film path showing tight-loop system and mag-
netic heads in position.
then goes immediately over the magnetic-recording head and then
over the magnetic-reproducing head. These heads are arranged on
movable arms and are linked together. By means of a pull pin, such
as is sometimes used on pad rollers, it is possible to drop or retract the
two heads for threading. The heads also remain in the retracted
position during photographic recordings. The film then passes
around a fixed idler -and then to a second sprung idler which completes
the tight-loop filtering system, then around the second side of the
driving sprocket into the film magazine. The spacing of the record-
ing and reproducing heads is such that for monitoring purposes the
486
MASTERSON
November
reproduced signal will lag the actual signal by approximately 1/16 of a
.second. The azimuth of each head is independently adjustable by
means of a setscrew. By loosening a screw in the center of the head,
it is possible to rotate head for best contact between gap and film.
The magnetic-track placement can then be adjusted by means of
adjusting nuts on the end of the studs around which the head-mount-
ing arms pivot. Since no standard has been established on the place-
ment of a magnetic track, it was decided to utilize the equivalent print
dimensions of the photographic wide-track push-pull system. There-
fore, the centerline of the magnetic track is identical with that of the
STANDARD PR.E- AMP
POW&R SUPPLY £.
AUDIO LINE-
LINE- FROM
RE-CORDING | 110 V
AMP. — > —
PR23
F-ILM R.E-CORDE-R
WITH MASNE-TlC
CONVE-RSION K.IT
RECORDING
RE-PRODUCING
MHI
Fig. 5 — Block diagram of electrical units.
double-width photographic track. The recording-head width is
184 mils and the reproducing head is 176 mils. In order to maintain
a balanced support of the film over the two heads, a skid the diameter
of the head laminations is provided near the back side of the film so
that the support is symmetrical. Fig. 5 is a block diagram of the
electrical part of the kit as the units are arranged with the recorder, j
The line from the recording amplifier which normally plugs into the
rear of the recorder is now removed and plugged into the rear of the
recording chassis box. A new cable then goes from this box to the
recorder to provide connections for photographic recordings. A
switch is provided on the front of the recording box to select either
1948
35-Mn MAGNETIC RECORDING
487
photographic or magnetic recording. If it is desired to make
magnetic recordings, the audio signal is equalized and then mixed
with the proper amount of high-frequency bias and goes to the re-
recording head. For playback, either as monitoring or as a film
phonograph, the output of the playback head is connected to the
left-hand box by means of another cable and enters the first pre-
amplifier. The output of this amplifier is then equalized and passes
to the second preamplifier, the output of which is at mixer level.
The preamplifiers are powered, and the audio is returned by the stand-
ard preamplifier connecting cable.
x
N.
X
N
X
vP
E
N
^^^^
*' c
.
^
A*
/
/A
/
2
/
0 50 100 500 1000 5000 10.0
Fig. 6 — Response curve of normal system and equalization
required.
The solid curve A in Fig. 6 is the uncompensated response of a
typical magnetic film at 90 feet per minute. If constant current is
maintained in the recording head at all frequencies, substantially con-
stant magnetization of the tape will result at all frequencies, but this
does not mean that a film so recorded will produce a flat response when
reproduced . Two maj or effects prevent the response from being at all
flat. The first is that a magnetic-reproducing head is a "rate-of-
chaiige" device which means that the voltage output is proportional
to the rate of change of the flux rather than to the amplitude of the
flux. This effect accounts for the 6-decibel-per-octave rise starting
at the low-frequency end and, at this film speed, extending up to
488 MASTERSON
about 4000 cycles. The second effect which causes the output above
4000 cycles to fall is due to several factors, one being the in-
ability of any magnetic material to maintain extremely short, closely
spaced magnets. This is called self-demagnetization and would,
perhaps, be equivalent to loss of resolution in a photographic film.
Another cause of this high-frequency loss is again equivalent to a
photographic system and that is slit loss or, in this case, gap loss.
The magnetic gap is and should be smaller than the photographic slit
and consequently, is of less importance in regard to the high-fre-
quency loss. A third effect contributing to high-frequency drop is the
eddy-current loss in the iron in the heads, although this is of minor
importance, especially if thin laminations of good iron are used.
It is obvious that considerable equalization will be necessary to
obtain a flat over-all response, but the uncompensated response is not
greatly unlike that of an uncompensated disk-record reproducer and
can be treated in a like manner. It is well known that the energies of
actual voice and music sounds in the high-frequency range are well
down in output as compared to their low-frequency output. This
fact allows a certain amount of high-frequency pre-emphasis. If con-
stant current in the recording head produces curve A in Fig. 6, and if
the constant current is obtained by the use of a relatively high resist-
ance in series with the recording head, it can be seen that a record-
ing-current curve like curve B in Fig. 6 can be obtained by by-passing
the constant-current resistor with capacitor of proper size. The
amount of tip-up or pre-emphasis and how far it is extended is deter-
mined by the intended frequency response. In this case the specifi-
cations were for flat response to 8000 cycles. Curve C is then the
sum of recording pre-emphasis added to the normal playback response.
Since voice and music energies are at their maximum at low fre-
quency, it is not possible to use pre-emphasis over this range, so this
correction must be made during playback. Somewhere in the play-
back system a characteristic similar to curve D must be inserted.
Since considerable amplifier gain must be used during playback, this
equalizer may be used not only to obtain proper response, but also
to reduce greatly first-tube hiss and noise. This means that this
equalizer should be used after the first tube or tubes of the playback
system. In the conversion kit the equalizer is arranged in the circuit
to follow the first two-stage amplifier. When the response of this
equalizer is added to the normal response, a flat curve will result, asj
shown by curve E.
Optimum High- Frequency Bias
in Magnetic Recording*
BY G. L. DIMMICK AND S. W. JOHNSON
RCA MANUFACTURING COMPANY, CAMDEN, NEW JERSEY
Summary — An experimental study was made of magnetic tapes and
films produced by several manufacturers. The effects of bias current upon
the frequency characteristic, the reproducing level, and the harmonic dis-
tortion are shown. Conclusions are drawn as to the best method of testing
a given tape for the optimum value of high-frequency bias.
A HIGH-FREQUENCY BIAS for magnetic recording was first used by
W. L. Carlson and G. W. Carpenter in 1921. Since that time
there have been differences of opinion regarding the exact cause of the
improved linearity and lower distortion produced by this type of bias.
Many people subscribe to the theory that the action of the high-
frequency magnetic field is to keep the molecules in a constant state
of agitation and thus make them more responsive to the lower fre-
quencies required for the recording of speech and music. Others
believe that the improved results can be accounted for by the action
of the combined high- and low-frequency magnetic fields upon the
normal magnetic characteristics of the material in question. Toomin
and Wildfeuer1 attempted to explain the action of a high-frequency
bias upon a sound-recording system using a recording medium having
permanent-magnet characteristics. Later, Holmes and Clark2 gave a
different explanation of the same phenomena and showed how a
magnetic-recording system is analogous in some respects to a push-
pull amplifier. The writers of the present paper are of the opinion
that the theory advanced by Holmes and Clark adequately explains
the observed performance of a magnetic-recording system when vari-
ous amounts of high-frequency bias are used. The purpose of this
paper is to review briefly the above-mentioned theory and to show the
effects of high-frequency bias upon the total harmonic distortion, the
frequency response, and the output level for four coated magnetic tapes.
Fig. 1 is a simplified diagram showing how the high-fre-
quency bias acts to reduce distortion and noise reproduced from a
* Presented May 18, 1948, at the SMPE Convention in Santa Monica.
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 489
490
DIMMICK AND JOHNSON
November
permanent-magnet recording medium. The dotted line K shows one
half of one of the major hysteresis loops for the magnetic material.
The complete loop is symmetrical about the point 0. The solid
lines OF and OG represent the virgin characteristics of the material
plotted in both the positive and negative directions from the mag-
netically neutral point 0. The curves F and G are the ones with which
we are most concerned, since the material does not pass through a major
loop during the recording process. A high-frequency sine wave L
of amplitude S has superimposed upon it lower-frequency waves E
and D which are identical and
which represent the speech or
music being recorded. It is
assumed that the magnetic ma-
terial on which a record is to
be made is in a magnetically
neutral state before it comes
under the influence of the re-
cording head. As a particular
point on the magnetic tape ap-
proaches the recording air gap,
it is magnetized along a series
of minor loops which occur at
the frequency of the bias. These
loops start at point 0 and progress up curve F. The amplitude
of the minor loops increases until the point on the tape reaches the
entering edge o.f the recording gap. The minor loops remain con-
stant in amplitude during the passage across the recording air gap,
but they may vary in position if the amplitude of the low-frequency
recorded signal varies appreciably during the time the point on the
tape is passing across the gap. When the point leaves the gap, the
amplitide of the minor loops starts decreasing and finally reaches
zero. If the amplitude of the recorded signal was of such value as to
cause the ends of the minor loops to reach position a (Fig. 1) when the
point on the tape reached the exit edge of the gap, the loops would
then decrease in amplitude and recede down curve a-b until point b
is reached. This is the residual induction left in the tape at the par-
ticular point in question after it has passed over the recording head.
From Fig. 1 it can be seen that one of the functions of the high-fre-
quency bias is to eliminate the effect of the "kink" in the normal char-
acteristics of a permanent-magnet material. This can be done if
Fig. 1 — Diagram showing how virgin-
tape characteristic is modulated.
1948
MAGNETIC-RECORDING BIAS
491
Fig. 2 — Distortion-versus-reproducing-level, German Type
C tape.
the amplitude S of the high-frequency bias is about equal to the dis-
tance t between the straight portions of the curves F and G. Another
very important function of the high-frequency bias is to reduce noise
from the reproduced signal. It is well known that the amount of
noise reproduced from a magnetic-recording medium increases with
the residual induction left on the medium after recording. This
RELATIVE OUTPUT LEVEL IN DB
Fig. 3 — Distortion-versus-reproducing-level, duPont SW4
tape.
492
DIMMICK AND JOHNSON
November
RELATIVE OUTPUT LEVEL IN DB
Fig. 4 — Distortion-versus-reproducing-level, Minnesota Min-
ing black-oxide plastic tape. •
noise is of a random nature and sounds to the ear very^nuch like
thermal noise from a resistor, or like "shot effect" from an amplifier
tube or phototube. The ear is very sensitive to this type of noise
when there are no other reproduced signals to mask it. The high-
frequency bias, therefore, serves the important purpose of keeping the
recording medium in a magnetically neutral state when no signal is
recorded.
RELATIVE OUTPUT LEVEL IN 08
Fig. 5 — Distortion-versus-reproducing-level, Minnesota Min-
ing Type RR tape.
1948
MAGNETIC-RECORDING BIAS
493
BIAS IN AMPERE TURNS
Fig. 6 — Output-level-versus-bias, German Type C
tape.
In order to determine experimentally the effect of high-frequency
bias upon distortion and output level, a Y^mch tape recorder and re-
producer was so arranged that many of its characteristics could be
held constant throughout the tests. The tape speed was set at 15
inches per second, and the over-all frequency characteristic was ad-
justed to be flat within 1 decibel from 50 cycles per second to 10,000
cycles per second when using German Type C tape with the bias set
Fig. 7 — Output-level-versus-B?as, duPont SW4 tape
494
DIMMICK AND JOHNSON
November
Fig. 8— Output-level-versus-bias, Minnesota Min-
ing black-oxide plastic tape.
at its optimum value. The frequency of the bias was 100 kilocycles.
The recording characteristic was flat from 50 to 3000 cycles per second
and rose 10 decibels between 3000 and 10,000 cycles per second.
Three-foot loops of each of the tapes were used for the tests, and the
recorded material was continuously erased before new material was
recorded. The erasing frequency was also 100 kilocycles. Ring-type
heads of RCA design were used for recording, reproducing, and eras-
ing. The recording gap was 0.001 inch while the reproducing gap was
§26
I
BIAS IN AMPERE TURNS
Fig. 9 — Output-level-versus-bias,' Minnesota Mining
Type RR tape.
1948
MAGNETIC-RECORDING BIAS
495
0.0005 inch. The total distortion introduced by the recording and
reproducing amplifiers was about 0.25 per cent. A General Radio
Type 732-A total-harmonic-distortion meter was used, and the
distortion and output-level measurements were made at a frequency
of 400 cycles per second. The gain of the reproducing amplifiers was
held constant throughout the tests, in order to have a direct com-
parison between the output levels for the four tapes tested. These
were German Type C, duPont Type SW4, Minnesota Mining black
plastic tape, and Minnesota Mining Type RR tape.
Fig. 2 shows a family of curves in which the total harmonic distor-
tion in per cent is plotted against output level in decibels for various
5 8 8 1
DECIBELS -OUTPUT
-
^**^
— ^=5:
^
i £0
/
x
-
s\
AIM
PERE
/*
N
\ TU
?NS BIAS
x
'
\v
42
7.0
U
*,«
0
X
X
10 5
S i 1 § P 8 § § § I
\ § §
Fig. 10 — Frequency response, German Type C tape.
values of high-frequency bias. This family of curves was made with
German Type C tape under the test conditions described above. A
range of bias values extending above and below the optimum value
was chosen. It is quite apparent that the curve made with a bias of
4.2-ampere turns will result in the greatest output with the least dis-
tortion. The distortion is less than 1 per cent for output levels be-
low 22.5 decibels, after which the curve breaks sharply. For higher
or lower bias values, the distortion is greater for a given output level.
In view of the explanation of the recording process based upon Fig.
1, it seems reasonable to expect the distortion curves to be as they
are shown in Fig. 2. For low values of bias we should expect the dis-
tortion to be relatively low for small recorded-signal amplitudes.
This is because the curve GOF (Fig. 1) is relatively straight where it
496
DIMMICK AND JOHNSON
November
passes through the point 0. When the signal is increased until the
peaks extend beyond the toe portions of curves G and F, the distor-
tion will reach a maximum value. For still greater values of recorded
signal the distortion will decrease because the distortion effect of the
"kink" in the curve GOF will become a smaller percentage of the total
signal. For even greater values of recorded signal, the distortion will
rise again because the peaks of the waves will begin to occupy the
"knee" portions of curves G and F.
It is assumed that the best value of bias is the one which just elimi-
nates the "kink" in the over-all characteristic and results in a con-
stant slope through the origin 0. Any further increase in bias re-
? § §1
AJJJgERE
TJRNS BIAS
FREQUENCY IN CYCLES PER SECOND
Fig. 11 — Frequency response, duPont SW4 tape.
suits in the slope being greater near the origin, and it also results in a
reduction in the total effective length of the characteristic curve, due
to partial erasing. These two effects account for the fact that for
high values of bias the distortion is increased and the overload level is
decreased.
It should be pointed out here that the absolute values of the output
levels and bias values shown in Figs. 2 to 9 have no significance.
A bias of 1.0-ampere turn used with a recording head of one design
would not necessarily produce the same effect as the same bias value
used with a recording head of another design. Since in this case the
same recording head was used throughout the tests, the numerical
values of bias are necessary in order to compare one tape with an-
other. The gain of the reproducing amplifiers was held constant
1948
MAGNETIC-RECORDING BIAS
497
throughout the tests in order to make it possible to compare output
levels for different tapes.
Fig. 3 shows a family of output-versus-distortion curves for duPont
SW4 tape. It may be seen that the best bias value is 3.5-ampere
turns. Comparing Fig. 3 with Fig. 2, it will be observed that the
overload levels (for optimum bias) are about the same for German
Type C and duPont SW4 tape. The distortion values below overload
are, however, appreciably higher for the duPont tape.
Fig. 4 shows a family of output-versus-distortion curves for Minne-
sota Mining black plastic-base tape. It will be observed that the
f
s
.
I Ml §
rREOUENCY IN CYCLES PER SECOND'
Fig. 12 — Frequency response, Minnesota Mining black-oxide plastic
tape.
best bias value is 4.9-ampere turns, which is somewhat higher than
for the other two tapes. If we arbitrarily assume the overload point
to occur at 3 per cent distortion, the overload level (for optimum bias)
of Minnesota Mining black tape is 4 decibels higher than for German
Type C tape. It will be noticed that the curve for 5.6-ampere turns
is much poorer than for 4.9-ampere turns. At 12.2-ampere turns, the
curve gets better again but not so good as for the optimum bias.
The overload point drops 3V2 decibels when the higher bias value is
used, but this level is still higher than for the German tape.
Fig. 5 shows a family of output-versus-distortion curves for Minne-
sota Mining Type RR tape. The best bias value occurs at 3.5-
ampere turns, which is the same as for duPont SW4 tape. The over-
load level (for 3 per cent distortion) is 3 decibels lower than for
498
DIMMICK AND JOHNSON
November
German Type C tape and 7 decibels lower than for the Minnesota
Mining black plastic-base tape. The Type RR tape has compensat-
ing advantages over the black type in that it is much easier to erase,
requires a lower bias value, and has a lower noise level.
The effect of bias current upon output level for four different tapes
is shown in Figs. 6 to 9, inclusive. These curves were made with a
constant recording level, and the output level is plotted as a function
of bias in ampere turns. The recording level was set to give an out-
put level of 23 decibels for German Type C tape, 23 decibels for du-
Pont SW4, 28 decibels for Minnesota Mining black tape, and 21
FREQUENCY IN CYCLES PER SECOND
Fig. 13 — Frequency response, Minnesota Mining Type RR tape.
decibels for Minnesota Mining Type RR tape. By comparing
values on Figs. 2 and 6, it can be seen that for German Type C tape
the bias value required for lowest distortion is the same value re-
quired for maximum output level. This is a very desirable condi-
tion because it means that slight variations in bias about its optimum
value will not cause corresponding variations in the output level.
For duPont SW4 and Minnesota Mining Type RR tapes (Figs. 7
and 9) the bias required for least distortion occurs slightly below the
point of maximum output level. For Minnesota Mining black plas-
tic-base tape, the bias which gives the maximum output is nearly
twice the value which is optimum from the standpoint of distortion.
The effect of bias upon frequency characteristic for four tapes is
shown in Figs. 10 to 13, inclusive. The purpose of these curves is
1948 MAGNETIC-RECORDING BIAS 499
to show that the bias current produces some erasing of the signal and
this effect is greatest at the highest signal frequencies. It will be
observed that Minnesota Mining black tape is less affected by bias
. than the other red-oxide tapes. This is probably because the co-
ercive force for this tape is higher and the effect of demagnetization at
high frequencies is less. All three of the American-made tapes have
better high-frequency response than the German Type C tape. The
curves shown in Figs. 10 to 13 were made without compensation and
represent the variation in output voltage of the reproducing head
with recorded frequency. The current in the recording head was
held constant at all frequencies.
REFERENCES
(1) Hershel Toomin and David Wildfeuer, "The mechanism of supersonic fre-
quencies as applied to magnetic recording," Proc. I.R.E., vol. 32, pp. 664-668;
November, 1944.
(2) Lynn C. Holmes and Donald L. Clark, Electronics, July, 1945.
DISCUSSION
CHAIRMAN JOHN G. FRAYNE: Mr. Pettus, is the performance of the film drive
from the standpoint of velocity or speed variation as good for the position occu-
pied by the magnetic-recording head which is not on the recording drum as com-
pared to the speed variation or nutter which exists in optical recording in which
the point of translation is on the drum?
. MR. J. L. PETTUS: Are you referring to the recorded reproduction?
CHAIRMAN FRAYNE: I am referring to the position of the record head. How
does the speed-recording variation compare to the speed for normal optical film?
MR. PETTUS: Tests of flutter made by optical recording have indicated this
drive to be considerably improved over the former machines. With the recording
reproducing head in place, which would be adding to the film, our measurements
show no serious difficulty in that respect. In brief, we believe that the drag of the
recording reproduction head offers nothing objectionable.
CHAIRMAN FRAYNE: Do you have any factual data?
MR. PETTUS: Not at this time.
MR. CRONIN: Mr. Dimmick, in connection with the curves in which you
described the percentage of distortion of total distortion versus ampere turns —
is it not a problem to know the volume of material that is being magnetized; for
example, was there any difference in the thickness of the magnetic film or the
pigment? I presume the width of the regions of influence were the same. Could
not that be normalized to some quantity stated to unit volume, for example?
MR. G. L. DIMMICK: We have no data on the exact thickness of the four films
presented. I believe that the German film is thicker than the films used in the
American-manufactured films, but I am not certain. To get that data, we should
have to go to the manufacturers of the film itself.
MR. C. R. KEITH: Do you have any particular reason for choosing the position
500 DIMMICK AND JOHNSON
of the sound track that was mentioned in the paper? Tt appeared that the sound
track is placed in exactly the same position as the 200-mil photographic sound
track. That is quite possible, but it seemed to me there would be some advan-
tage in having the magnetic track farther away from the sprocket holes.
MR. DIMMICK: That is a very good question, and one for which we probably
do not have a complete answer, but one on which the various committees are
working. All other things being equal, it would seem best to have the track loca-
tion in the same place as formerly used for wide-track photographic recording.
One might think that it would be best £o put the track down the center. I
think there is no doubt that you would be freer from the effect of the sprocket
holes if you did this; however, you do sacrifice the ability to turn the film around
and put two tracks on. I think that there will have to be much work done on
this before the final standardization is given for the location of magnetic track.
CHAIRMAN FRAYNE: Have you done any work on ITVs-mm film?
MR. DIMMICK: No.
MR. L. D. GRIGNON: Do you have any information comparing the impreg-
nated tapes versus the coated tapes that you have shown here today?
MR. DIMMICK: Are you thinking of tape like the German tape? No,' we have
not, and I think the reason for that is that the Germans themselves became dis-
couraged by the results they obtained from the Type L. The reason for that
was that the print-through was too great. You would have one layer against
the other, and you get printing too high to tolerate.
MR. GEORGE LEWIN: Is there any optimum value for the width of the track?
It would seem to me the wider you make it, the better the signal-to-noise ratio.
MR. DIMMICK: Yes, the same consideration holds for magnetic recording and
variable-density recording.
MR. LEWIN: Why did you use only 180 or so? Why not the full width?
MR. DIMMICK: Once again, we did take exactly the same dimensions as had
previously been standardized for photographic recording. There is a record-
ing head over the same area formerly covered by the recording light beam, and
the reproducing head is the same width as the reproducing light beam.
QUESTION: Do you feel then that if you went to full width that the amount of
gain is not sufficient to justify it?
MR. DIMMICK: There are two factors. We believe that the signal-to-noise
ratio obtainable from the present width of track is so much greater than for-
merly available in photographic recording that it is not necessary to go to wider
tracks, and tolerance on angle of both recording and reproducing heads gets
much worse as the width of the track goes up in width.
MR. LAW: Will the bonding agent in the tape, when used, say in 16-mm work,
stand developing without any signal loss?
MR. DIMMICK: I cannot answer that specifically, but I am of the opinion that
developing will not harm the tapes. I believe there are in the audience represen-
tatives from the tape manufacturers. Possibly they could answer that better.
MR. R. R. HERR: I cannot speak from experience, but certainly nothing
that I know of has much effect. There is a plastic oxide binder in which the mag-
netic oxide is placed, and I am quite sure it will have no effect on the oxide itself
which is impregnated in that binder.
Variable- Area Recording
with the Light Valve*
BY JOHN G. FRAYNE
WESTERN ELECTRIC COMPANY, HOLLYWOOD, CALIFORNIA
Summary — Various types of variable-area track, including standard
and push-pull, may be obtained by various arrangements of the light-valve
ribbons. A mathematical analysis is made of the effect of various light-valve
constants on the magnitude of the resonance-peak and frequency-response
measurements of an improved light valve with high magnetic damping are
given. A theoretical study of the effect of azimuth deviation on unilateral,
dulateral, and bilateral tracks is included in the paper and is illustrated with
graphical charts of the distortion produced by various amounts of azimuth
deviation for these types of tracks.
INTRODUCTION
THE NEW WESTERN ELECTRIC variable-area light valve recently
introduced to the motion picture industry is based on a design
initiated by Wente and Biddulph1 of the Bell Telephone Labora-
tories in the application of the light valve to variable-area recording
in connection with the development of a stereophonic sound-film
system. While the light valve has been ordinarily associated with
the variable-density type of sound track it is only necessary to rotate
the valve through an angle of 90 degrees to obtain the variable-width
type of modulation. While many more or less unsuccessful attempts
have been made over a period of years to adapt the light valve to
variable-area recording its successful application in this type of
recording has only been made possible by improvements in light-
valve design and in the development of a suitable anamorphote opti-
cal system to magnify the relatively small movements of the ribbons
into full sound-track-width modulations.
A cross section of the Wente-Biddulph valve is shown in Fig. 1.
The outer shell forms the permanent-magnet structure of the valve
while the permandur pole pieces form a sealed structure when the as-
sembly is magnetized. A condenser lens is mounted in one pole
piece while an objective lens is mounted in the other pole piece.
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 501
502
FRAYNE
November
The ribbon-support structure is carried by one of the pole pieces.
An optical schematic of the associated variable-area light-valve
modulator is shown in Fig. 2. It will be noted that in addition to the
spherical objective lens LI mounted in the valve, a cylinder lens L2
is mounted near the film plane. The former serves to magnify the
ribbon movement so that the desired track width is obtained for 100
per cent modulation while the latter serves to reduce the height of the
slit S to a suitable value at the film plane.
CONDENSER LENS
v-POLE PIECES
L.V. RIBBONS
OBJECTIVE LENS
Fig. 1 — Cross-section Wente variable-area valve.
The new variable-area light valve which is described in another
paper2 follows the general basic design of the Wente valve and all
details of slit height, lens magnification, and ribbon arrangements are
given in that paper.
TYPES OF TRACKS
(a) Two-Ribbon Valve .
The various types of tracks that may be recorded by an area light
valve depend on the number and disposition of the ribbons employed.
The simplest valve employs only two ribbons; and, by reference to
Fig. 3, three types, namely, (a) bilateral, (b) unilateral, and (c) push-
pull, may be obtained by connecting the signal input circuits as shown
in the figure. In these three applications only one bias line appears
1948
VARIABLE-AREA RECORDING
503
on the film with the application of noise-reduction bias current as
indicated. In the bilateral method the noise-reduction currents are
superimposed on the signal currents in both ribbons in a manner
analogous to that used in the well-known double-ribbon variable-
density light valve.3 For a ten times magnification of the light-
valve aperture a 0.2-mil biased spacing will produce a 2-mil bias line.
Since a 3.8-mil spacing is indicated to produce a 38-mil "half" or
unbiased track, the noise reduction for 2-mil bias line obtained is
theoretically given by4
NR = 10 log— = 12.8db.
FILM PLANE
CONDENSER
LENS
Fig. 2 — Schematic for Wente variable-area light-valve modulator.
In the unilateral method the signal is applied to one ribbon only
while noise reduction is applied to the second ribbon. This requires
double the movement of the signal ribbon obtaining in the bilateral
connection, the effect of which will be discussed below.
Class A push-pull operation is obtained by the connections shown
in Fig. 3(c). The signal currents flow in the same direction in each
ribbon causing them to move in the same direction due to the reac-
tion of these currents to the magnetic field. On the other hand,
the noise-reduction currents, flowing in series through the ribbons,
produce the normal biasing action.
Reproduction of the three types of tracks resulting from the above
three methods of connecting the ribbons to the signal and bias cir-
cuits is shown in Fig. 4 (a) . These are all based on a 76-mil maximum
504
FRAYNE
November
modulation the half-track width being 38 mils for the duplex and
unilateral tracks. In order to permit a 6-mil septum on the push-
pull track, a "half" track of 41 mils, with modulation limited to 94
per cent of the ribbon aperture, is employed. In all cases the maxi-
mum track width is limited only by the modulation of the light- valve
ribbons.
In addition to the above types of track that may be laid down by
the two-ribbon light valve it is possible to produce a double-bilateral
or a double-unilateral track by means of an image doubler. This is
readily accomplished by inserting an optical biplate in the objective
lens of the light-valve modulator in a manner described in another
O SIGNAL O
(o) BILATERAL
NO?SE
REDOgTION
; SIGNAL
(b) UNILATERAL
6 SIGNAL 6
(c) PUSH-PULL
Fig. 3 — Connections for two-ribbon variable-area valve.
paper.2 The use of either of these types of track obviates the dif-
ficulty encountered from reproducing either the bilateral or unilateral
tracks in reproducers equipped with push-pull reproducing optics,
since the septum in these optics may result in severe distortion in low-
level signals. The double-unilateral track, known as dulateral
track, and shown in Fig. 4(b), is produced by the same type of
valve as that used for obtaining the unilateral track, the only physi-
cal difference being in the spacing of the ribbons which is reduced by
one half. This valve has the advantage over the double-bilateral
type of separating noise reduction and signal circuits and offering
greater damping by reducing the over-all resistance of the light-valve
signal circuit.
1948
VARIABLE-AREA RECORDING
505
(b) Three-RMon Valve
While a push-pull track may be recorded by a two-ribbon valve a
preferred method employs a three-ribbon valve, the circuit connec-
tions of which are shown in Fig. 5. In this case the speech currents
are confined to the center ribbon while the noise-reduction currents
are confined to the outer two ribbons. This provides complete
separation of the two circuits which previously5 has been shown to be
desirable for stable operating conditions. A sample negative track
laid down by this valve is shown in Fig. 6. In this case the clear
central area of modulation is caused by 'the movement of the image
of the central speech ribbon while the dark portions are produced by
the varying spacing between the noise-reduction ribbons and the
(a) BILATERAL (b) UNILATERAL (c) PUSH -PULL
Fig. 4(a)
Fig. 4(b)—
Dulateral
track.
central ribbon in accordance with the impressed signal modulation
and noise-reduction bias currents. If a 76-mil push-pull track with a
o-mil septum is required a 4.0-mil wide ribbon with ten times magni-
fication is provided. On the other hand if a double-width track is
required the width of the ribbon must be increased proportionately.
In practice, a 10-mil ribbon is required for a double-width track in
order to preserve the centerline separation of each component track
previously established for 200-mil push-pull density6 sound tracks.
The three-ribbon structure may also be employed to obtain Class B
type push-pull operation. This is shown schematically in Fig. 7.
In this case the signal currents flow in the same direction in the outer
ribbons while the return current flows in the opposite direction in the
center ribbon. In this valve ribbons 1 and 3 must not be coplanar
506
FRAYNE
November
with ribbon 2 in order to permit passage behind the latter. If the
three ribbons are adjusted so that no light is transmitted through
the aperture between them for the condition of zero input, then rib-
bons 1 and 3 on moving to the right under an impressed signal will
permit a half wave of exposure to be transmitted between 2 and 3.
Vice versa, the other half wave will be transmitted between 1 and 2
when the ribbons 1 and 3 move to the left of the normal position.
Thus a Class B track may be laid down with this valve. A Class
A-B track may be recorded by permitting some exposure between
the ribbons for the unmodulated condition. Since noise reduction is
automatically obtained with either Class B or A-B push-pull no provi-
sion for bias currents is required with this valve.
0 NOISE °
Fig. 5 — Connections for
three-ribbon Class A push-
pull variable-area valve.
Fig. 6—
D o uble-
width class A
push-pull.
Fig. 7— Connection for
three-ribbon Class B variable-
area valve.
FREQUENCY CHARACTERISTIC OF LIGHT VALVE
(a) Theoretical
The maintenance of precise sound-track and bias-line dimensions
depend on a high order of stability of ribbon position and sensitivity.
This called for the elimination of the high resonance peak which had
been found to cause considerable instability in the earlier type of
light valves.7
While this peak may be eliminated by the use of feedback,8 in the
design of a modern variable-area valve a high magnetic-flux density
in the air gap is relied on mainly to provide efficiency electromagnetic
1948 VARIABLE-AREA RECORDING 507
damping of the ribbons at resonance. In order to evaluate the
factors which control the frequency response of a light valve the fol-
lowing mathematical analysis is employed.
The velocity v of the ribbon at any instant when subject to a force
F is given by
„ _ g .(J^_(1/aC)) - f. CD
where
Rm = mechanical resistance
M = mass
Cm = compliance
w = 2ir X frequency
Zm = mechanical impedance.
The force on the ribbon is given by
F = Eli (2)
where
E = magnetic flux in the air gap
I = length of the ribbons
i — the current in the ribbons.
Now the counter electromotive force generated by the ribbon
motion in the magnetic field is Blv. Hence
E = Blv + Roi (3)
where
E = the voltage applied to the ribbon
RO = electrical resistance.
Substituting (1) and (2) in (3), we obtain
B
(4)
fl'Q ~T~ (/> I /Zrn)
From (2)
' F . j,-™^. • (5)
Therefore, from (1)
v = ^ (BlE/Rj
(BlE/Ro)
(ft)
Rm + (BH2/Ro) +j(<aM— (1/wCw))
The two real terms in the denominator constitute the effective
508
FRAYNE
November
mechanical resistance. The first term which is mainly due to windage
resistance may be neglected in valves where the magnetic flux B is
large as in the case of the modern permanent-magnet light valve.
The displacement is given by integrating (6)
The displacement at zero frequency is given by
j BlECm
do = — n - •
KQ,
Equation (7) may therefore be rewritten in terms of dQ as
_
d0
+ j(a>*M - (I/Cm))'
Expressed in terms of Q which is defined as
fined as l/\/MCm
-JQ
where cor is de-
(10)
The valve frequency-response characteristic for a constant-current
input can therefore be computed from (10) simply by inserting as-
signed values of co/cor and Q. The latter may be determined experi-
mentally by comparing the sensitivity (displacement) at the me-
chanical resonance frequency to that at zero frequency or direct cur-
rent as indicated by (10) which reduces to
* = Q for 2. = 1.
d0
(ID
Now from (7)
Hence •
(13)
Since it is customary to express frequency response in terms of deci-
bels, (10) may be written
db = 20 log 5- = 20 log X
(to
or
(14)
A series of characteristic curves based on (14) are shown in Fig. 8
1948
VARIABLE-AREA RECORDING
509
Fig. 8 — Light-valve theoretical response for various values
of Q.
for values of Q ranging from 10 to 0.5. The latter curve which shows
a loss of 6 decibels at resonance corresponds to the critical damping
condition.
(6) Experimental Characteristic
It will be noted that the Q of the light valve varies inversely as the
square of the flux density in the air gap and directly as the resistance
100
10000
1000
FREQUENCY IN CYCLES PER SECOND
Fig. 9 — Actual response of variable-area light valve.
510
FRAYNE
November
of the valve circuit. In the design of the new variable-area light
valve, the flux density has been built up to a value in excess of 30,000
gauss. By applying the signal to a single ribbon and avoiding the
additional resistance of the simple circuit required in two-ribbon
valves the resistance of the light-valve circuit is kept to a minimum.
The resistance of the light-valve circuit may be reduced further by
placing a low-resistance shunt across the signal ribbon. As a result
of increased flux and reduced circuit resistance it has been possible to
design the new variable-area light valve with a circuit Q less than 2.0.
Actual frequency-response curves
of a modern light valve are shown
in Fig. 9. The upper curve rep-
resents the response for con-
stant-voltage input at A in the
circuit and corresponds to the
condition existing in a zero-im-
pedance generator. The actual
resonance peak is somewhat less
than 2 decibels, corresponding to
a Q of 1.26. When measured for
a constant input voltage in the
600-ohm circuit at B the peak
is increased to slightly over 4
decibels, since in this case the
voltage across the ribbon is not constant but rises with the impedance
of the valve near resonance frequency. This peak is eliminated in
practice by the insertion of an equalizer with a transmission char-
acteristic which is roughly complementary to that of the valve. The
resulting frequency-response characteristic measured at C in Fig. 9
shows the actual recording characteristic for constant input voltage
into the recording channel. The square-wave response of a typical
valve with 0.5-ohm shunt across the signal ribbon and with the com-
plementary equalizer in the circuit is shown in Fig. 10. It will be
noted that the response approaches that to be expected from a criti-
cally damped device.
PEAK CHOPPER
As pointed out in another paper by Browder, 2 it is necessary to have
the ribbons mounted accurately in the same plane in order to secure
sharp definition of the images of the ribbon edges at the film plane.
Fig. 10 — Square-wave response of
area light valve plus equalizer.
1948
VARIABLE-AREA RECORDING
511
Since physical clashing of the ribbons at overload should be avoided,
it has been found desirable to limit the modulation by a chopping
circuit. The use of a conventional limiter in addition to the noise-
reduction circuit has not proved to be completely satisfactory as the
combination gives a pumping quality which is not noticeable when
either is used alone. Instead the use of a driving amplifier with an
abrupt overload characteristic has been found quite satisfactory.
The load characteristic of this amplifier is shown in Fig. 11. It will
be noted that at the overload output of +28.5 decibels, peak chop-
ping begins and no further increase in output results. The harmonic
content of this amplifier is 0.5 per cent at 2 decibels below the peak-
25
20
;,5
>
>
10
5
PEAK CHOPPING +26 5 OBM
/
/"
S*s
^
LI
^EAR CHX
kRACTERISTIC
/
^
>*
/
fa
' NC
(
RMAL CO
;HARACTE
MPRESSIO
RISTIC
N
/
/
/
/
7
45 40 35 30 25 20 15 10
INPUT DBM
Fig. 1 1 — Load characteristic of main amplifier.
chopping level which should be set just below the overload point of
the light valve. Listening tests have confirmed that the distortion
resulting from overload of the amplifier is quite similar to that as-
sociated with clipping of the signal peaks in an overmodulated track
by the reproducer scanning slit. It is preferable to clashing the rib-
bons at overload, since in this case a transient is induced because the
valve is not completely critically damped.
EFFECT OF AZIMUTH DEVIATION ON TRACKS
In evaluating the performance of the various types of variable-
area tracks laid down by the light valve, the rigorous mathematical
512
FRAYNE
November
Fig. 12 — Off-azimuth scanning of unilateral track.
analysis given in the Appendix has been made to determine the effect
of azimuthal deviation of the reproducer scanning slit on distortion
in the various types. This problem has been studied previously
by Cooke9 and Foster.10
In Fig. 12, there is shown a unilateral-type track with amplitude aQ
scanned by a slit S at angle a with the true azimuth of the scanning
Fig. 13 — Off-azimuth scanning for bilateral track.
1948
VARIABLE-AREA RECORDING
513
beam. The instantaneous amplitude of the unilateral track so
scanned is
F = a0 + «o sin - (x + w)
A
(15)
where
«o = the amplitude of the wave
x = the distance from the origin
X = the wavelength of the sound, and
u = the reflection of the scanning line S along the x axis.
Fig. 14 — Off-azimuth scanning of dulateral track.
In the Appendix it is shown that as a result of scanning this wave
form with the off-azimuth slit the resultant wave form is given by
(.6)
where
r =.the deviation of the scanning beam.
The expression within the brackets on the right-hand side repre-
sents an infinite series of harmonic components resulting from scan-
ning a sinusoidal wave form by a slit with a deviation r over the nor-
mal track width.
514
FRAYNE
November
The coefficients of the first three components are as below :
2
fundamental
second harmonic
third harmonic =
^2(^—1; and
A /
(37rr/X)
(17)
(18)
(19)
where the J's are the well-known Bessel coefficients.
It will be noted that the term r/X appears in the expression for the
amplitude of the fundamental and various harmonic components
derived from (16). This indicates that the distortion produced by
0 O.I 0.2 0.3 0.4 05 0.6 0.7
r/x
Fig. 15 — Effect of azimuth on fundamental and harmonics in
various tracks.
aperture misalignments is a function only of this ratio. This means
that as frequencies on the sound track increase, that is, as X decreases,
the deviation r decreases proportionately for constant harmonic dis-
tortion. In other words, aperture misalignment is a more serious
source of distortion in the high-frequency end of the sound spectrum.
A similar analysis of the duplex- or bilateral-type track based on the
construction shown in Fig. 13 indicates that the effect of aperture
misalignment in this case also produces an infinite series of harmonics,
the values of the first three terms of the series being
fundamental =
cos
(20)
1948 VARIABLE-AREA RECORDING 515
second harmonic = ^^ J2 (?) sin (?) ; and (21)
third harmonic = * (£) COS (£> <22>
The analysis of the dulateral track based on the sketch of Fig. 14
similarly shows that the values of the first three terms of the series are
given as follows :
fundamental = ^ /, (?) cos (?) ; (23)
second harmonic = j^y^x) Ji (?) cos (?) ; and (24)
third harmonic = J, (f?) cos . (25)
The relative amplitudes of the fundamental and second and third
harmonics have been computed for the three cases analyzed above and
are shown in Fig. 15 for values of r/\ up to 0.7. This latter value
corresponds to a slit deviation of 1.4 mils for 9000 cycles or 2-mil
wavelength. In practice a deviation of the order of 0.2 mil is about
the maximum to be tolerated. This gives a value of r/X of 0.1 at
9000 cycles. For this value of r/X the second-harmonic content is
about 2.5, 6, and 15 per cent, respectively, for the bilateral, dulateral,
and unilateral tracks. For a very severe misalignment resulting in
values of r/X > 0.25, the second harmonic of the bilateral exceeds
that of the dulateral. It will be noted that the amplitudes of the
fundamental and third harmonic are identical for both the bilateral
and dulateral tracks.
From the equations derived for the amplitude of the fundamental
and second and third harmonics, these values may be determined as a
function of frequency for any assigned value of azimuth deviation.
Thus, assuming an angular deviation of 0.25 degree, the curves of
Fig. 16 show the reproduced characteristic for a unilateral track
while Fig. 17 shows the same characteristics for the bilateral and
dulateral tracks. All of these curves are based on 100 per cent modu-
lation.
CONCLUSION
The light valve which has long been used to produce the variable-
density type of track lends itself readily to the variable-width type of
modulation. Increase in magnetic flux in the air gap adds to the
516
FRAYNE
November
200 500 1000 10000
FREQUENCY IN CYCLES PER SECOND
Fig. 16 — Frequency and harmonic response for bilateral and dulateral
tracks.
stability of the valve, making possible the maintenance of accurately
located bias and track centerlines. A variety of single tracks as well
as a Class A push-pull track can be laid down by a two-ribbon-type
valve while both Class A and Class B push-pull readily can be re-
corded by a three-ribbon structure. Both double-bilateral and double-
unilateral (dulateral) can be obtained from the single tracks by image
doubling. An analysis of the effect of azimuth deviation on distor-
tion indicates the dulateral track to be superior to the unilateral for
all azimuthal deviations and superior to the bilateral for large values of
the deviation.
FUNDAMENTAL
500 1000
.FREQUENCY IN CYCLES PER SECONDS
Fig. 17 — Unilateral track, frequency and distortion characteristics.
1948 VARIABLE-AREA RECORDING 517
ACKNOWLEDGMENTS
The writer wishes to express his appreciation to Dr. V. Pagliarulo
for his assistance in obtaining the analytic solution for distortion in-
duced by azimuth deviation of the scanning beam, and to Mr. L. B.
Browder for his invaluable help in obtaining basic information used in
many of the figures.
REFERENCES
(1) E. C. Wente and R. Biddulph, "A light valve for the stereophonic sound-
film system," J. Soc. Mot. Pict. Eng., vol. 34, pp. 397-406; October, 1941.
(2) L. B. Browder, "Variable-area light-valve modulator," J. Sdc. Mot.
Pict. Eng., this issue, pp. 521-534.
(3) H. C. Silent and J. G. Frayne, "Western Electric noiseless recording,"
/. Soc. Mot. Pict. Eng.,. vol. 18, pp. 551-571; May, 1932.
(4) G. R. Crane, "Variable matte control for variable-density recording,"
/. Soc. Mot. Pict. Eng., vol. 31, pp. 531-539; November, 1938.
(5) J. G. Frayne, T. B. Cunningham, and V. Pagliarulo, "An improved 200-
• mil push-pull density modulator," /. Soc. Mot. Pict. Eng., vol. 47, pp. 494-519;
December, 1946.
(6) E. M. Honan and C. R. Keith, "Recent developments in sound-tracks,"
/. Soc. Mot. Pict. Eng., vol. 41, pp. 127-136; August, 1943.
(7) T. E. Shea, W. Herriot, and W. R. Goehner, "The principles of the light
valve," /. Soc. Mot. Pict. Eng., vol. 18, pp. 697-732; June, 1932.
(8) W. J. Albersheim, "Stabilized feedback light valve," /. Soc. Mot. Pict.
Eng., vol. 38, pp. 240-256; March, 1942.
(9) E. D. Cook, "The aperture alignment effect," J. Soc. Mot. Pict. Eng.,
vol. 21, pp. 390-403; November, 1933.
(10) D. Foster, "Effect of orientation of the scanning image on the quality of
sound reproduced from variable-width records," /. Soc. Mot. Pict. Eng., vol.
33, pp. 502-517; November, 1939.
APPENDIX
Referring to Fig. 12 for the off-azimuth scanning of the unilateral
sound track, the instantaneous amplitude of the light transmitted
by the unilateral modulation5 as scanned in the figure is given by
Y = a0 + a0 sin (x + «) (26)
where
a0 = the amplitude of the wave
x = the distance from the origin
X = the wavelength of the sound wave
u = the reflection of the scanning line S along the x axis, and
r = the deviation of the scanning beam.
518 FKAYNE November
The voltage developed across the track in the phototube is propor-
tional to the length S uncovered in the clear area of the track. The
shaded area is considered sufficiently opaque to contribute no volt-
age.
Now
S = Y sec « (27)
where
S = the length of the scanning line
a = the azimuthal deviation of S, and
u = S sin a (28)
so that
S = a0 sec a (l + sin ^ (x + S sin u)\ (29)
Multiplying each side by sin a, and adding x to each side
x + S sin a = x + a0 tan a + a0 tan a sin — (x + S sin a). (30)
A
Let
y = -£ (x + S sin a)
? (x + a0 tan «) (31)
A
ft
Then (30) becomes
or
y = z + ft sin ?/.
This simplified form is identical with that previously developed7 for
the exposure of a variable-density sound track with a two-ribbon light
valve and the solution takes the same form. Thus, y — z can be ex-
panded into a Fourier series of z having sine terms only.
Thus,
y — z = ^ an sin nz. (33)
l
1948 VARIABLE^AREA RECORDING 519
Hence,
An = \ (y — z) sin nz dz. (34)
Integrating by parts,
An = - \ — (y — z) cos nz * + / cos nz d(y - z) . (35)
nir\_ o JO _|
The integrated term disappears since y = z f or the limiting values of
0 and TT as is evident from (31). Also J^ cos ny dy = 0.
Hence, on substituting
y — ft sin y = z
An = —'- I cos n (y — ft sin y)dy (36)
= \Jn (na) (37)
by the Bessel integral. The solution thus becomes
y — z = - J^ -Jn(np) sin nz. (38)
i
Substituting from (31)
\ 2 \-~» 1 T / n&ir&o \
+ ao tan a) -\ — > - Jn I - — tan a ]
n "Y n \ X /
sin^ (x + aotan a). (39)
2?r . 2?r
— (x + £ sin a) = — -
A A
T
Since tan a = -
,...
(40)
From this equation the coefficients of fundamental, second, third,
etc., harmonics may be computed for assigned values of r/\ and known
values of Jn.
Solutions for the bilateral and dulateral cases may be obtained in an
analogous manner.
DISCUSSION
MR. C. R. SKINNER: -The Skinner Company has been manufacturing a unit
like that since 1928. There is one slight difference. To get the slit effect, a
520 FRAYNE
double slit was used instead of a separate slit for the purpose. Other than that, it
is identical to the machine that has been on the market since 1928.
ME. P. E. BRIGANDI: Many times I have heard Dr. Frayne repeat at meet-
ings that variable area has a certain characteristic noise that identifies it, and
that it is difficult to project because of its susceptibility to uneven slit elimination.
Is the Western Electric type- of variable area less susceptible to those noises and
problems, and does the hush-hush he mentioned indicate that density is a less
desirable medium for recording than area?
DR. J. G. FRAYNE: The answer to the first part of Mr. Brigandi's question
should be referred to Eastman and duPont for an answer.
With regard to hush-hush, twenty years ago that was a big problem, when we
had very coarse-grained films. As you know, the noise on a variable-density
track is primarily grain noise. The noise on variable-area track, on the other
hand, if it is processed correctly, is not grain noise, but noise contributed by
scratches and dirt which increase with running. The higher grain noise on film
shows up in hush-hush or breathing in variable density much more than it does
on variable area.
To offset that, the wider track was brought in several years ago, so that on
originals that has not been particularly bothersome, because of the increase on
the wider track. On density release, the problem still remains to a certain extent,
although with the finer-grain films made by the film industry the problem is no
longer serious for variable density. That, I think, is borne out by the large
number of Academy Sound Awards won by variable-density films.
Variable-Area Light- Valve
Modulator*
BY LEWIS B. BROWDER
WESTERN ELECTRIC COMPANY, HOLLYWOOD, CALIFORNIA
Summary — A variable-area modulator is described which employs a
ribbon light valve as the basic modulating element. Double-width push-pull,
' variable-area sound track or standard width dulateral sound track may be
recorded at will by inserting the appropriate light valve into the modulator.
The light valve is registered in place in the modulator by indexing dowels
and securely locked by means of lever-controlled clamping springs.
The light-valve ribbons are oriented so as to be parallel to the direction of
motion of the film. The ribbon edges are projected at ten times magnifi-
cation onto the film to define the amplitude co-ordinate of the recording
image while the image height is determined by a narrow rectangular stop
which is imaged onto the film at a 70:1 reduction in height by a cylindrical
lens system.
The modulator is a completely self-contained unit embodying the basic
components for the recording optical system, an optical system for rear pro-
jecting an enlarged image of the ribbon aperture onto a viewing screen, a
photoelectric monitoring system, and an exposure meter.
INTRODUCTION
THE WESTERN ELECTRIC variable-area sound-on-film recorder to
be described represents a reduction to motion picture studio
practice of the recording apparatus employed in the stereophonic-
sound-film system demonstrated by the Bell Telephone Laboratories
before the Society in 194 1.1 In particular, the ribbon or light-valve
type of modulator and the basic recording optical arrangement as
used in that system have been carried over into the new recorder.
As used in the variable-area modulator, the light-valve ribbons are
supported so that their edges are parallel to the direction of motion of
the film. The ribbons thus serve to define the vertical edges of an
illuminated aperture, an enlarged image of which is projected onto
the sensitive surface of the film. These aperture edges move toward
and away from each other in response to speech and noise-reduc-
tion currents to vary the lateral extent of the illuminated image
on the film. The height of the recording image in the direction of
film travel is fixed, being defined by a horizontal slit located on the
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 521
522 BROWDER November
film side of the above lens system. The aperture enclosed by the
horizontal slit edges is imaged on the film by a cylindrical lens system
whose axis is parallel to that of the horizontal slit. This lens system
is located adjacent to the film so that the height of the recording im-
age is greatly reduced from the actual spacing of the horizontal slit.
The two optical systems involved, that for defining the amplitude
co-ordinate of the recording image and that for defining the recording
image height are completely independent, since the cylinder lens
does not participate in the magnification of the vertical ribbon edges
and the slit defining the image height is located on the image side of
the lens system employed in projecting the ribbon edges. This op-
tical system permits the recording of a considerable variety of vari-
able-area sound tracks, the electrical connections and the arrange-
ment of ribbons to secure the various types of sound tracks being de-
scribed in another paper.2
Two of the variations have been developed for use with this film
recorder. The first of these employs three ribbons with speech
current applied to the central one of the three and noise-reduction
current applied to the outer two ribbons. Such a valve when used
with this recorder will lay down a double-width push-pull, variable-
area sound track. The other light valve employs but two ribbons
with speech current impressed upon one and noise-reduction current
on the other. An image-doubling device incorporated within the
magnifying objective lens forms two images of the aperture defined by
the light-valve ribbons. These images are formed side by side giving a
sound track consisting of two, inphase unilateral, variable-area sound
tracks with a total width at full modulation of 76 mils. This track
will be referred to as a "dulateral" type of variable-area track. -Such a
sound track has been shown to be adequately free of the type of dis-
tortion introduced by scanning with a reproducing optical system
improperly adjusted for azimuth.2
GENERAL DESIGN OF THE VARIABLE-AREA LIGHT VALVE
The ribbons used in the variable-area light valves are sheared from
0.0005-inch thick duraluminum sheets. They are suspended between
pairs of clamp carriages which are mounted so as to insulate each
ribbon electrically from the case of the valve and from the other rib-
bons. These carriages provide a means for achieving an accurately
coplanar setting of the several light-valve ribbons so that they will
all be in sharp focus at the film, and also allow of adjustment during
1948 VARIABLE-AREA LIGHT VALVE 523
the assembly of the light valve of the spacing between the ribbons
and of the tension to which they are stretched. The carriages are
shown assembled on the light-valve pole piece in Fig. 1. The mag-
netomotive force is supplied by an Alnico V magnet which forms the
case of the valve. The magnetic flux is concentrated at the ribbon
gap by means of permandur pole pieces which also constitute the ends
of the light valve as shown in the cross-sectional view of Fig. 2. It is
possible with such a structure to achieve flux densities in the ribbon
gap of the order of 30,000 gauss. The mechanical force developed
between the two pole pieces after the valve is magnetized is adequate
Fig. 1 — Area light-valve pole piece showing
arrangement of ribbon carriages. Fig. 3— Assembled area light valve.
to hold the poles and the magnet together without resort to conven-
tional fastening devices. The ribbon carriages and the magnetic cir-
cuit employed in the variable-area light valves are identical to those
developed for use with the new Western Electric variable-density
light valves and have been fully described elsewhere.3
The aperture between the light- valve ribbons is projected on the
film at a magnification of ten times by the light-valve objective lens.
The necessity always present in variable-area recording for obtaining
sharp photographic boundaries between the exposed and unexposed
portions of the sound track imposes a rather severe requirement on
the degree of resolution expected of this lens system. Fortunately it
has been possible to arrange the optical working distances so that a
high-quality standard microscope objective lens can be employed
524
BROWDER
November
in this application. A ten-power apochromatic lens is used because
of the completeness of the correction for spherical aberration with
this type of lens in the neighborhood of 4200 angstroms at which the
film is most sensitive. The lens system is mounted within the struc-
ture of the light valve which enables the light valve to be removed
from the modulator and reinstalled or replaced without the neces-
sity for refocusing or readjustment of sound-track position. A plane-
glass window is employed at the light-source end of the light valve
RIBBON CARRIAGE •/ *- MAGNET
Fig. 2 — Cross-section view of area light valve.
so that, it, the objective lens, the pole pieces, and the magnet struc-
ture provide complete sealing of the ribbon gap against the entrance
of moisture or magnetic particles. Fig. 3 shows the completely as-
sembled variable-area light valve.
The RA-1247, or push-pull, light valve employs three ribbons which
are directly imaged onto the film by means of the objective lens.
The RA-1405 dulateral light valve employs an image doubler, the
functioning of which is depicted schematically in Fig. 4. The light
flux emerging from the ribbon aperture is divided by a biplate image
doubler which consists of two plane-parallel glass slabs inclined in
1948
VARIABLE-AREA LIGHT VALVE
525
opposite senses on the two sides of the optical axis and so arranged
that each section intercepts half of the light flux and laterally dis-
places the image of the ribbon aperture formed by the first two lens
elements. The third lens element thus sees two images of the ribbon
aperture which it projects onto the film. The amount of deviation is
such that at full modulation the two images completely fill the 76-mil
RIBBON APERTURE
^- APOCHROMAT LENS ELEMENTS
Fig. 4 — Action of biplate in forming dulateral recording image.
360 BET SPEECH fs
760 MAX TRACK WIOT
GUID£0 EOGE-
THAVEL
PULATERAL PUSH-PULL
A B
Fig. 5 — Dulateral and double-width push-pull sound tracks.
wide sound track. At low levels of modulation the two unilateral
sections of the track are symmetrically related to the sound-track
centerline so that the sound track may be played on reproducers
with push-pull optics without danger of distortion occurring when
low-level modulation crosses the centerline due to misalignment of
the reproducer or weaving of the film. Any interference with a
septum in the reproducer will occur only at full modulation where the
526
BROWDER
November
effect is not so serious. In this respect the dulateral sound track is
similar to the duplex bilateral type of variable-area sound track ex-
cept that the two halves of the sound track are identical rather than
being mirror images as in the case of the duplex bilateral. The
dulateral-type sound track is shown in Fig. 5 A while Fig. 5B shows
the push-pull sound track obtained with the R A- 1247 light valve.
GENERAL DESIGN OF THE MODULATOR
The various components associated with the RA-1243 variable-area
modulator are all mounted on a cast-aluminum plate which is in-
Fig. 6 — Assembled variable-area modulator.
stalled in the recorder proper by means of four mounting bolts. The
general arrangement as seen from the operating side of the modulator
is shown in Fig. 6 while Fig. 7 shows the complete recorder. Provi-
sion has been made for. easy removal of the light valve which slides
into an opening in the center of the modulator and indexes by means
of dowel pins. The condenser lens as seen in Fig. 6 is located in the
housing immediately to the left of the light valve, while the record-
ing lamp is mounted on a bracket at the extreme left of the modulator.
An exposure meter is incorporated into the modulator structure and is
located below the light-valve opening. The visual monitor optical
system is located immediately above the light-valve opening. The
1948
VARIABLE-AREA LIGHT VALVE
527
monitoring phototube is located in an assembly on the rear face of the
modulator mounting plate. The modulator is thus completely self-
contained and may be removed easily from the film recorder and
interchanged with the Western Electric variable-density modulator.3
A double-width push-pull sound track or a standard 76-mil wide
sound track is recorded with this modulator depending on whether
an RA-1247 or an RA-1405 light valve, respectively, is installed
into the light-valve compartment. The change in light valves is
accomplished simply and quickly without the use of tools. The
difference in sound-track position between the push-pull sound track
Fig. 7 — Variable-area modulator as installed in film recorder.
and the standard sound track is taken care of in the optical adjust-
ment of the light valve. Thus the push-pull light valve is arranged to
record a track symmetrical about the optical centerline of the modula-
tor, while the centerline of the standard sound track is displaced op-
tically by adjustment of the light-valve objective lens to the standard
position for release sound track. This adjustment is made at the
time the light valves are assembled. The circuit connections to the
light valves are taken care of in the contact arrangement at the edge
of the valve, while the optical offset automatically takes care of the
change from push-pull to standard operation in the photoelectric-
monitoring system.
528 BROWDER November
OPTICAL SYSTEM
The recording optical system is shown schematically in Fig. 8.
The recording light source is a 7.8-ampere, 10.5-volt curved-filament
lamp equipped with a prefocused-type base. The lamp works into
an aspheric surfaced anamorphic condenser-lens system, which im-
ages the height of the filament at the plane of the light-valve ribbons
and the length of the filament between the ribbon plane and the light-
valve objective lens. Such an arrangement has proved quite effec-
tive in obtaining uniform exposure across the width of the sound track
Fig. 8 — Optical schematic of area modulator.
despite the inherently discontinuous structure of the filament image.
A mask is located between the ribbons and the valve objective lens
which is so proportioned that light leakage does not occur around the
outer, edge of the light-valve ribbons at the extreme of their inward
excursion. The objective lens, as mentioned previously, is a 10-
power apochromatic microscope objective especially mounted and
coded for this application. The focusing and adjustment of this
lens is carried out in a fixture designed to simulate a nominal in-
stallation in the modulator. The horizontal slit which defines the
1948
VARIABLE-AREA LIGHT VALVE
529
height of the recording image is located immediately to the right of
the light- valve opening and is integrally mounted with the modulator
structure. The slit is a clear ruled section in an aluminized mirror
surface. The mirror is inclined so as to direct upward to the visual
monitoring and the photoelectric-monitoring facilities the excess of
the light over that passing through the recording slit. The cylin-
drical lens, which images the ruled slit on the film, is a two-element
system designed to minimize spherical aberration when working at an
aperture of approximately f/1.7. The azimuth of the recording im-
age is adjusted by rotation of the cell in which the cylindrical lenses
are mounted, while focus of the recording slit onto the film is achieved
by axial movement of this cell by means of a threaded focusing ring.
a M RELATIVE RESPONSE DB
±5
0
• — — — «i.
— .—
•— -W
*"*«*,
-*>
-10
x
0 50 100 200 500 1000 2000 50OO 10000 200
FREQUENCY IN CYCLES PER SECONDS
g. 9 — Experimental film-frequency characteristi
The light projected upwards by the slit mirror is directed into the
visual-monitor compartment directly above the light valve where the
beam is folded for the sake of mechanical compactness. The spatial
image of the ribbons formed by the light-valve objective lens is
picked up by a 6-power achromatic relay lens, further folded and pro-
jected upon a ground-glass screen located so as to be visible through a
corresponding opening in the door of the modulator compartment of
the recorder. The image formed at the screen is erect, i.e., with the
light- valve ribbon edges vertical. Its height is made greater than
the opening for the visual-monitor screen so that the excess is inter-
cepted by a reflecting-lens system and projected rearward through the
modulator-mounting plate to the plates of the monitoring phototube.
The separation of the two halves of the push-pull monitor image is
effected by dividing the above-mentioned reflecting lens into two
530
BROWDER
November
PRINT DENSITY
Fig. 10 — Experimental cross-modulation curves obtained
with push-pull light valve.
sections so arranged that the two parts of the image are directed to
the appropriate plates of the RCA-920 monitoring phototube. The
reflecting separator lenses project upon the phototube plates an image
of the exit pupil of the light-valve objective lens which produces a
variable-intensity spot at the phototube.
An exposure-meter photocell is incorporated in the modulator for
measuring the light flux delivered to the film. This meter is brought
into operation by means of a solenoid-operated mirror which swings
or
^
?o
*>
^
/<
y
V
>
x
Q/
'
Y
25
^
.
\
7
\
\
7
j
«
/
^n
\
/
\
/
\
7
\
/
'
/
\
/
/v
-7
•jg
—
S
\
\
Y
2
V
J
\
/\
/
40
\^
^_^
4<S
PRINT DENSITY
Fig. 11 — Experimental cross-modulation curves obtained
with dulateral light valve.
1948
VARIABLE-AREA LIGHT VALVE
531
into place to deflect the recording light beam down and to the ex-
posure-meter cell. A blue filter is used in this optical train in order to
complement the spectral response of the RCA-929 phototube to
match that of the film so that exact correlation between exposure-
meter reading and density of the sound track is obtained. This
exposure-meter facility is also available for setting up noise reduction
on an exposure basis although the graduated scale of the visual moni-
tor screen is ordinarily used for setting up noise reduction on the
basis of width of the bias line. The graduations on this scale are
0.030 inch apart and represent 0.005-inch intervals at the film so that
i.o
22 2.4 2.6 2.8 3.0 32
NEGATIVE DENSITY
Fig. 12 — Experimental processing tolerances for dulateral
sound track based on minimum of 30-decibel cancellation.
2-mil bias lines may be estimated easily. A glow-discharge type of
bloop lamp is located so that the light therefrom is directed off the
surface of the solenoid-operated mirror onto the film.
OPERATING CHARACTERISTICS
The height of the recording image is such as to record a sound
track on a variable-area negative emulsion such as EK-1372 or
duPont 226, which will require a print density of 1.4 to 1.5 on release
stock such as EK-1302 or duPont 225 to give maximum cross modu-
lation cancellation.4 The frequency characteristic of such a film re-
corded at constant-amplitude modulation of the light valve is depicted
in Fig. 9. Typical cross-modulation curves obtained with high-gamma
negative developer and normal release-print developer are shown in
Fig. 10 for the push-pull sound track and in Fig. 11 for the dulateral
532
BROWDER
November
release sound track. Corresponding processing tolerances for a
minimum of 30 decibels cancellation with the dulateral sound track
are shown in Fig. 12.
The RA-1247 push-pull light valve requires a level of +21 dbm*
for 100 per cent modulation, while the noise-reduction ribbons re-
quire 300 milliamperes for complete closure. The level into the
speech ribbon includes that absorbed by a shunt connected across
the ribbon. The use of this shunt adds to the effectiveness of the
damping besides simplifying the design of the light-valve equalizer.
With the shunt, the resonant rise of the light valve as seen
50 100 200 500 1000 2000 5000 10000 20000
FREQUENCY IN CYCLES PER SECOND
Fig. 13— Frequency response of area light valve with and
without light-valve equalizer.
by a 600-ohm generator looking into the primary of the light-
valve matching coil is as shown in Fig. 13. An equalizer, coded
RA-1275, has been developed for use with this and the RA-1405 re-
lease light valve, which gives an over-all response as shown in the
second curve of Fig. 13. A photograph of the cathode-ray trace of
the monitoring-phototube output as the light valve is driven with a
square-wave generator is shown in Fig. 14. The relationship between
the magnetic-flux density, the input voltage, and the ribbon dynamics
has been fully developed in another paper.2
The RA-1405 standard light valve requires for operation of its
speech circuit a level of approximately +14.0 dbm,* while noise-
reduction ribbon closure is obtained with a current of 200 milli-
amperes. The frequency characteristics and the square- wave
* Decibels with respect to 0.001 watt.
1948
VARIABLE-AREA RECORDING
533
response are identical to those shown for the RA-1247 on Fig. 13.
In the case of the RA-1405 a shunt is used across both the speech and
noise-reduction ribbons, that across the noise-reduction ribbon is
employed in order to limit and damp out the transient excursions of
this ribbon due to any accidental mechanical shock excitation.
The original model of this modulator has been used in making
original production and release recordings at Twentieth Century-Fox
Studios since August, 1947. The recordings made with this machine
exhibit a degree of cleanness and naturalness which has seldom, if
ever, been attained in sound-on-film recordings.
CONCLUSION
The modulator described in
this paper adapts the ribbon light
valve to the recording of variable-
area sound track. By incorporat-
ing the light- valve ribbons, the
magnetic structure, and the light-
valve objective lens into an in-
tegral, easily replaced unit, an
extremely flexible design of modu-
lator has been obtained. The
possibilities of the light valve in
producing distortion-free record-
ings having been demonstrated
in the stereophonic system, the basic philosophy governing the
design of the new variable-area modulator has been convenience of
operation. Toward this end, the modulator is arranged so that it
can be set up for either push-pull or single operation in a matter of
seconds by merely interchanging light valves.
REFERENCES
(1) H. Fletcher, "The stereophonic sound-film system," /. Soc. Mot. Pict.
Eng., vol. 37, pp. 331-353; October, 1941.
(2) J. G. Frayne, "Variable-area recording with the light valve," J. Soc. Mot.
Pict. Eng., this issue, pp. 501-521.
(3) J. G. Frayne, T. B. Cunningham, and V. Pagliarulo, "An improved 200-
mil push-pull density modulator," J. Soc. Mot. Pict. Eng., vol. 47, pp. 494-519;
December, 1946.
(4) J. O. Baker and D. H. Robinson, "Modulated high frequency recording
as a means of determining conditions for optimal processing," /. Soc. Mot. Pict.
Eng., vol. 30, pp. 3-18; January, 1938.
Fig. 14 — Cathode-ray oscillogram of
area light-valve square-wave response
with equalizer.
Nine Recent American Standards
NINE ADDITIONAL American Standards on Motion Pictures appear
in the following pages, bringing to 49 the number of new and re-
vised standards made available to the motion picture industry since
January, 1946. With the help of many SMPE and Motion Picture
Research Council committees, the ASA at that time embarked on an
expanded standards program calling for a review of all motion picture
standards approved prior to the recent war and reappraisal of all
temporary war standards developed for the use of the military services
during the intervening years.
All 49 of the standards have appeared in issues of the SMPE Jour-
nal for April and September, 1946, August and December, 1947,
March, 1948, and the current issue. A
complete subject index to these stand-
ards has been printed in S1/^- X 11-inch
size and copies were mailed to all who
purchased the SMPE Standards Binder
shown here. If you have the Binder
and your loose-leaf index has not been
received, you either are not listed or
are incorrectly listed on our records. If
that is the case, please send your correct
address marked "for the Standards
Binder mailing list" to BOYCE NEMEC, Executive Secretary. A few
•complete sets of all standards approved to date, with binders, are
still available from the Society office for $8.50, postpaid, when mailed
to an address within the United States or $9.00 in U. S. funds when
mailed to a foreign country. The nine new standards which appear
on the following pages may be purchased, as a group, from the SMPE
for $1.00. Individual copies of the standards, however, must be
bought directly from the American Standards Association, 70 East
45th Street, New York 17, N. Y. The ASA will also furnish a catalog
of American Standards in all industrial fields upon request and without
charge.
534
(continued on page 540}
NOVEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
NINE RECENT AMERICAN STANDARDS 535
American Standard Dimensions for
Theatre Projection Screens
KI-Z. V. S. I'al. Off.
Z22.29-1948
Revision of
Z22.29-1946
>UDC 778.55
1. Scope and Purpose
1.1 This standard specifies dimensions for projection screens used for view-
ing motion pictures.
2. Screen Size
2.1 Sizes of screens shall be in accordance with the table below.
2.2 The over-all size shall be measured from the outer edge of border to the
outer edge of opposite border. The ratio of the over-all width to over-all
height shall be 4 to 3.
3. Border
3.1 A fabric reinforcing border shall surround the screen. The width of this
border shall be from 2.5 to 3 inches.
4. Grommets
4.1 Metal mounting grommets, size No. 3 or No. 4, shall be securely
fastened through the fabric border.
4.2 Grommets shall be spaced on 6-inch centers, starting from grommets
located at the centers of the four sides of the screen, except that there shall
also be a grommet in each corner of the screen. Grommets shall be set in a
line parallel to the edge of the screen, with their centers from 1.0 to 1.31
inches inside the outer edge of the border.
5. Selection of Screen Size
5.1 The width of the screen should be not less than 1/6 of the distance
from the center of the screen to the most remote seat.
5.2 The distance between the screen and the front row of seats should be
not less than 0.87 foot for each foot of screen width.
Page 1 of 2 pages
536
NINE RECENT AMERICAN STANDARDS November
American Standard Dimensions for
Theatre Projection Screens
K--Z. V . S. I'at. Off.
Z2 2. 29- 1948
Revision of
Z22.29-1946
Size No.
of Screen
Over-all
Width
(feet)
Over-all
Height
% (feet)
Minimum Effective
Picture Size
(feet)
8
8.00
6.00
7.50 5.50
9
9.00
6.75
8.50 6.25
10
10.00
7.50
9.50 7.00
11
11.00
8.25
10.50 7.75
12
12.00
9.00
11.50 8.50
13
13.00
9.75
12.50 9.25
14
14.00
10.50
13.50 10.00
15
15.00 11.25
14.50 10.75
16
16.00
12.00
15.50 11.50
17
17.00 12.75
16.50 12.25
18
18.00
13.50
17.50 13.00
19
19.00
14.25
18.50 13.75
20
20.00 15.00
19.50 14.50
21
21.00 15.75
20.50 15.25
22
22.00 16.50
21.50 16.00
23
23.00 17.25
22.50 16.75
24
24.00
18.00
23.50 17.50
25
25.00
18.75
24.50 18.25
• 26
26.00
19.50
25.50 19.00
27
27.00
20.25
26.50 19.75
28
28.00
21.00
27.50 20.50
29
29.00
21.75
28.50 21.25
30
30.00
22.50
29.50 22.00
NOTES:
1. Masking on each of the four sides of the screen is recommended as follows:
1 inch of Vnasking within the projected picture area on each of the four sides
of the picture for every 12 feet of picture width, with a minimum of 1 inch
for pictures less than 12 feet in width.
2. Screens larger than Size No. 30 are not specified as such screens are usually custom
built or not in 4 by 3 ratio due to projection angle.
Page 2 of 2 pag
1948
NINE RECENT AMERICAN STANDARDS
537
American Standard Specification for
Buzz-Track Test Film
for 16-Millimeter Motion Picture Sound Reproducers
Krg. V. S. Pat. Off.
Z22.57-1947
1. Scope and Purpose
1.1 This specification describes a buzz-track test film used for checking the
position of the sound scanning beam in 16-mm motion picture sound repro-
ducers.
2. Test Film
2.1 The test film shall have originally recorded 300-cycle and 1000-cycle
signal tracks on either side of the central exposed strip as shown in Fig. 1 .
The position of the tracks, weave in running film on the recorder included,
shall be in accordance with Fig. 1.
UNIFORMLY OPAQUE STRIP
1000 CYCLES
+ 0.0005
0.020-o.oooo IN.
0.096-o.ooo5 IN.
2-44 1°-°°° MM
0.012
2.2. The central exposed strip and the exposed portions of the two signal
tracks shall have a density of 1.6 + 0.4.
2.3 Film Stock. The film stock used for the test film shall be cut and per-
forated in accordance with the American Standard Cutting and Perforating
Dimensions for 16-Millimeter Sound Motion Picture Negative and Positive
Raw Stock, Z22.12-1947 (Revision of Z22.12-1941), or the latest edition
thereof approved by the American Standards Association.
Page'J of 2 pages
538 NINE RECENT AMERICAN STANDARDS November
American Standard Specification for
Buzz-Track Test Film
for 16-Millimeter Motion Picture Sound Reproducers
Rfg. V. S. Pat. Off.
Z22.57-1947
2.3.1 Resistance to Shrinkage. The film stock used for the test film shall
have a maximum lengthwise shrinkage of 0.50 percent when tested as follows:
At least 20 strips of film approximately 31 inches in length shall be cut for
measurement of shrinkage. After normal development and drying [not over
-f-80 F ( — 26.7 C)], the strips shall be placed at least Vi inch apart in racks
and kept for 7 days in an oven maintained at + 1 20 F ( + 49 C) and a relative
humidity of 20 percent. The strips shall then be removed, reconditioned
thoroughly to 50 percent relative humidity at +70 F ( + 21.1 C), and the
shrinkage measured by an adaptation of the pin-gage method outlined in
Research Paper RP-1051 of the National Bureau of Standards. The percent
shrinkage shall then be calculated on the basis of deviation from the nominal
dimension for the length of 100 consecutive perforation intervals given in
American Standard Z22.12-1947.
2.4 Each test film shall be provided with suitable leader and trailer, and a
title or other markings to identify the film.
2.5 The standard length of test film shall be 100 feet.
NOTE:
A test film in accordance with this standard is available from the Society
of Motion Picture Engineers.
Page 2 of 2 pa
1948
NINE RECENT AMERICAN STANDARDS
539
American Standard
Theatre Sound Test Film for
35-Millimeter Motion Picture Sound
Reproducing Systems
K'-a. U. S. I'ul. Off.
Z22.60-1948
>UDC 778.5
1. Scope and Purpose
1.1 This standard describes a film for quali-
tatively checking and adjusting 35-millimeter
motion picture sound reproducers and for
judging the acoustical properties of the audi-
torium in which the sound is reproduced.
2. Test Film
2.1 The film shall have a sound track and
accompanying picture. The sound track shall
comply with American Standard Sound Rec-
ord and Scanned Area, Z22.40-1946, and
the film stock used shall be cut and per-
forated in accordance with American Stand-
ard Cutting and Perforating Dimensions for
35-Millimeter Motion Picture Positive Raw
Stock, Z22.36-1947, or any subsequent revi-
sions thereof.
2.2 The test film shall contain samples se-
lected from studio feature pictures by an
appropriate engineering committee of the
Motion Picture Research Council and the
Society of Motion Picture Engineers. The fol-
lowing sound samples are typical of those
which may be included:
(a) Main title music
(b) Dialogue
(c) Piano
(d) Orchestral music
(e) Vocal music
2.3 The assembled film shall include appro-
priate samples of typical release-print mate-
rial intended to provide a qualitative check
of such reproducing system characteristics
as:
(a) Frequency response
(b) Volume range
(c) System noise
(d) Power-handling capacity
(e) Flutter
2.4 Each film shall be provided with head
and tail leaders as specified in American
Standard Specifications for 35-Millimeter
Sound Motion Picture Release Prints, Z22.55-
1947, or any subsequent revision thereof.
The main title shall include the issue number
of the film so that revised versions, which
may be issued periodically to conform to
changing studio practices, may be easily
identified. Subtitles superimposed over each
sample shall indicate the particular sound
characteristic demonstrated by that portion
of the film.
2.5 The length of the film shall be approxi-
mately 500 feet.
3. Method of Use
3.1 From a typical location in the audito-
rium the observer should determine whether
or not the frequency-response characteristics
of the complete reproducing system are nor-
mal by listening to the sound reproduced
from the test film at normal sound level.
NOTE: A test film in accordance with this stand-
ard is available from the Motion Picture Research
Council or the Society of Motion Picture Engineers.
540 NINE RECENT AMERICAN STANDARDS November
(continued from page 534}
THEATER PROJECTION SCREENS, Z22. 29-1948
This 1948 revision of the theater-screen standard establishes over-
all screen dimensions in addition to minimum effective picture size
and includes border dimensions with more specific location of grom-
mets. Screens wider than thirty feet have not been included since
they are made to order, and because of keystoning do not usually
have the standard height-to-width ratio of 3 to 4.
16-Mn BUZZ-TRACK TEST FILM, Z22.57-1947
Critical dimensions of this new standard remain exactly the same
as the previous American War Standard Z52. 10-1944. Performance
of the film has been improved by making both the 1000-cycle and
300-cycle buzz signals an integral part of the center opaque strip.
They had each been separated from the center strip by a clear area
0.002 inch wide, but films made that way give a somewhat less critical
indication of lateral placement of the reproducer scanning slit.
35-MM THEATER SOUND TEST FILM, Z22.60-1948
The familiar Academy Test Reel is described in a general way
without specifying detailed requirements, either qualitative or quan-
titative. This is an apparent departure from accepted standardiza-
tion practice but is very practical because it includes a statement
of this film's purpose and a description of the several types of typi-
cal release-print sound samples that are included.
35-MM SOUND FOCUSING TEST FILM (9000-CycLE), Z22.62-1948
This 9000-cycle film is an original recording intended for use by
35-mm reproducer manufacturers or theater equipment maintenance
shops for focusing sound optical systems. It provides a critical test in
these applications but is not recommended for use in theaters because
theater amplifiers normally have low-pass filters that cut off somewhat
below 9000 cycles.
35-MM SCANNING-BEAM TEST FILM (SERVICE TYPE), Z22.65-1948
The familiar continuous "snake track" used in the theater for
checking the uniformity of illumination across the width of the 35-
mm reproducer scanning beam is described. In service it is used as a
loop, joined preferably with a butt splice held together with scotch
(continued on page 546}
1948 NINE RECENT AMERICAN STANDARDS 541
American Standard
Sound Focusing Test Film for
5-Millimeter Motion Picture Sound Reproducers
(Laboratory Type)
H:;. U. S. I'm. Off.
Z22.62-1948
>UDC 778.5
1. Scope and Purpose
1.1 This standard describes a film which may be used for precise focusing
of the optical systems in 35-millimeter motion picture sound reproducers. The
recorded frequency shall be suitable for use in laboratories and factories.
2. Test Film
2.1 The film shall be a print from an original negative and shall contain a
9000-cycle, sinusoidal, variable-area track recorded at 1 decibel below 100-
percent modulation. The variation in power output level from the film shall
be not more than ± 0.25 decibel.
2.2 The sound track shall comply with American Standard Sound Record
and Scanned Area, Z22.40-1946, and the film stock used shall be cut and
perforated in accordance with American Standard Cutting and Perforating
Dimensions for 35-Millimeter Motion Picture Positive Raw Stock, Z22.36-1947,
or any subsequent revisions thereof.
NOTE 1: This test film is not recommended for theater use because the reproducing
amplifiers ordinarily installed in theaters normally have low-pass filters which cut off below
9000 cycles.
NOTE 2: A test film in accordance with this standard is available from the Motion
Picture Research Council or the Society of Motion Picture Engineers.
542
NINE RECENT AMERICAN STANDARDS
November
American Standard
Scanning-Beam Uniformity Test Film for
35-Millimeter Motion Picture Sound Reproducers
(Service Type)
ASA
«••;. U. .S. I'M. OIL
Z22.65-194E
1. Scope and Purpose
1.1 This standard describes a film which may be used for determining the
uniformity of scanning-beam illumination in 35-millimeter motion picture
sound reproducers. The recorded sound track shall be suitable for use in the
routine maintenance and servicing of the equipment.
2. Test Film
2.1 The film shall be a print from an original negative. It shall consist of a
1000-cycle, variable-area recording at full modulation of the 0.007-inch
width and shall be approximately sinusoidal. The track shall move uniformly
0.077 inch from one edge of the scanned area to the other as shown in Fig. 1 .
2.2 The scanned area shall comply with American Standard Sound Record
and Scanned Area, Z22.40-1946, and the film stock used shall be cut and
perforated in accordance wtih American Standard Cutting and Perforating
Dimensions for 35-Millimeter Motion Picture Positive Raw Stock, Z22.36-1947,
or any subsequent revisions thereof.
2.3 The length of this film shall be approximately 8 feet.
NOTE: A test film in accordance with this standard is available from the Motion
Picture Research Council or the Society of Motion Picture Engineers.
1948
NINE RECENT AMERICAN STANDARDS .
543
American Standard
Scanning-Beam Uniformity Test Film for
•-Millimeter Motion Picture Sound Reproducers
(Laboratory Type)
K.'i-. V. S. /'at. Off.
Z22.66-1948
'UDC 778.5
Page 1 of 2 Pages
1. Scope and Purpose
1.1 This standard describes a test film which may be used for determining
the uniformity of scanning-beam illumination in 35-millimeter motion picture
sound reproducers. The recorded sound tracks shall be suitable for use in
laboratories and factories.
2. Test Film
2.1 The test film shall contain a number of -1000-cycle, variable-area tracks
of narrow width, recorded at 100-percent modulation.
2.2 The test film shall contain 17 individual sound tracks, each with the same
amplitude of approximately 0.007 inch. These tracks shall appear on the film
in succession, the first so placed that its center line shall be not more than
0.197 inch from the guided edge of the film, and the seventeenth so placed
that its center line shall be not less than 0.292 inch from the guided edge of
the film. The intermediate tracks shall be spaced at equal intervals between
the first and seventeenth tracks, similar to that shown in Fig. 1.
FIG. 1
2.3 The film stock used shall be cut and perforated in accordance with
American Standard Cutting and Perforating Dimensions for 35-Millineter
Motion Picture Positive Raw Stock, Z22.36-1947, or any subsequent revisions
thereof.
2.4 Each test film shall be provided with a suitable head leader identifying
the film.
544
NINE RECENT AMERICAN STANDARDS
November
American Standard
Scanning-Beam Uniformity Test Film for
35-Millimeter Motion Picture Sound Reproducers
(Laboratory Type)
Kif. u. a. t'ui. og.
Z22.66-1948
UDC 778.5
Page 2 of 2 Pages
2.5 The length of this test film shall be approximately 230 feet.
2.6 Each of the 17 tracks shall be identified by an appropriate spoken
announcement. The track modulated by the voice shall be limited to the same
track width as a single 1000-cycle test signal.
2.7 Each film shall be accompanied by a calibration sheet showing the
center-line position of each of the 17 tracks measured from the guided edge.
The accuracy of these dimensions shall be within 0.002 inch.
NOTE: A test film in accordance with this standard is available from the Motion
Picture Research Council or the Society of Motion Picture Engineers.
1948 NINE RECENT AMERICAN STANDARDS 545
American Standard
1000-Cycle Balancing Test Film for
Millimeter Motion Picture Sound Reproducers
Rug. V. S. Fat. Off.
Z22.67-1948
*UDC 778.5
1. Scope and Purpose
1.1 This standard describes a film which may be used for balancing the
respective power-level output from two or more 35-millimeter motion picture
sound reproducers.
2. Test Film
2.1 The film shall be a print from an original negative containing a 1000-
cycle, variable-area track recorded at 50-percent modulation. It shall be
accompanied by a statement of the percent modulation of the incident light
in the reproducer. The accuracy of calibration shall be within — 1 decibel.
2.2 The harmonic distortion of the recorded 1000-cycle note shall not ex-
ceed 2 percent.
2.3 The sound track shall comply with American Standard Sound Record
and Scanned Area, Z22.40-1946, and the film stock used shall be cut and
perforated in accordance with American Standard Cutting and Perforating
Dimensions for 35-Millimeter Motion Picture Positive Raw Stock, Z22.36-1947,
or any subsequent revisions thereof.
3. Instructions
3.1 An instruction sheet, describing the manner in which this film is to be
used in various types of reproducing equipment, shall be provided with
each film.
NOTE: A test film in accordance with this standard is available from the Motion
Picture Research Council or the Society of Motion Picture Engineers.
546 NINE RECENT AMERICAN STANDARDS November
(continued from page 540}
tape. An 0.007-inch wide, 1000-cycle track moves across the 0.084-
inch standard scanned area in 8 feet of film or approximately 5x/2 sec-
onds. Any nonuniformity of illumination will appear as cyclic varia-
tions of output power level.
35-MM SCANNING-BEAM TEST FILM (LABORATORY TYPE) Z22.66-1948
This laboratory type of scanning-beam film is a more precise meas-
uring tool than the "snake track" and is used by manufacturers of
theater and studio sound reproducers for adjusting new equipment or
as a final check of new sound installations. It has seventeen succes-
sive individual 1000-cycle sound tracks each 0.007 inch wide and about
12 feet long, equally spaced across an area slightly wider than the
standard sound-track scanned area covered in American Standard
Z22.40-1946, SMPE JOURNAL for April, 1946, page 292.
35-MM BALANCING FILM (1000-CYCLE), Z22.67-1948
The output power level from all soundheads in any theater must be
balanced so that normal release prints will reproduce at proper volume
with a single fader setting. This film provides a standard reference
signal required for balancing power-level output from standard sound
reproducers but is not intended for use in balancing the outputs from
the two halves of push-pull reproducer systems.
DOUBLE-WIDTH PUSH-PULL SOUND TRACK, Z22.69-1948
AND Z22.70-1948
Double- width push-pull sound tracks, sometimes called "200-mil
push-pull," that are now in commercial use follow two "standards."
One is the Normal Centerline Type, Z22.69, wherein the half of the
track nearest the perforations will play back on a reproducer intended
for track in the normal 35-mm release-print position (American
Standard Z22.40-1946, SMPE JOURNAL for April, 1946, page 292).
The other is the Offset Centerline Type, Z22.70, which has its center-
line 0.021 inch nearer the center of the film and will not play back on
conventional theater equipment. These double- width tracks are now
used only in studio production prior to re-recording the final release
negative.
•
Motion Picture engineers who have been following the standards programs of
the Society, the Motion Picture Research Council, and the American Standards
Association will be interested in the reference to the Universal Decimal Classi-
fication System recently adopted by the ASA, which appears on page 552 of this
issue. — Editor
NINE RECENT AMERICAN STANDARDS
547
American Standard
Sound Records and Scanning Area of
Double Width Push-Pull Sound Prints
Normal Centerline Type
ASA
K>-f. V. .<. I'al. Off
Z22.69-1948
'UDC 778.534.4
AREA PRINTED
IN SOUND PRINTER
OUTER EDGE OF
PRINTED AREA '
INNER EDGE OF
PRINTED AREA
ID
SHADED AREA TO BE EFFEC-_
TIVELY OPAQUE ON PRINT
-GUIDED EDGE
0.191 ±0.002 IN.
4.85±0.05 MM
0.420 IN. MIN
10.66 MM MIN
VARIABLE
DENSITY RECORD
WIDTH OF SEPTUM-
WIDTH OF EACH HALF -
OF PUSHPULL SOUND
RECORD
VARIABLE
AREA RECORD
WIDTH OF SEPTUM
WIDTH OF EACH HALF
OF PUSHPULL SOUND
RECORD
RD
2
0.010 ±0.0015 IN.
0.25 ±0.04 MM
0.095t0-002 IN.
ND
D REC
D
ID
2.41 ±0.05 MM
0.293± 0002 IN
ORD
»
7.44+0.05 MM
0024 ±0.001 IN.
\
I .
5
\
i
0.61 ±0.02 MM
0076 ±0.001 IN.
ND
L
u
1.93 to. 02 MM
0.293 ±0.001 IN.
/ 7.44 ±0.05 MM
<t SCANNED AREA
SEPTUM
WIDTH OF EACH HALF OF
PUSHPULL SCANNED AREA
0.016 ±0.001 IN.
0.41 ±0.02 MM
0.084 tO.002 |N.
2.13 ±0.05 MM
548
NINE RECENT AMERICAN STANDARDS
American Standard
Sound Records and Scanning Area of
Double Width Push-Pull Sound Prints
Offset Centerline Type
AREA PRINTED
IN SOUND PRINTER
OUTER EDGE OF
w
//////
ID
D
-^GUIDED EDGE
0.|9|±0.002 IN.
PRINTED AREA
INNER EDGE OF
PRINTED AREA ~~
SHADED AREA TO BE EFF
TIVELY OPAQUE ON PRIN
VARIABLE
DENSITY RECORD
WIDTH OF SEPTUM
— >
1
»,
w
4.85*0-05 MM
^ 0.420 IN. MIN
10.66 MM MIN
0.010 ±0.0015 IN.
EC-
\
T
r -^
'1
3
WIDTH OF EACH HALF
0.25*0.04 MM
0.095±0-°°2 IN.
OF PUSHPULL SOUND
RECORD
<L SOUND RECC
VARIABLE
AREA RECORD '
WIDTH OF SEPTUM
Z)_
2.41 ±0.05 MM
0314+ 0.002 IN
)RO-
— >
7.98i0.05 MM
0.024*0.001 IN.
1
3
,
0.61 ±0.02 MM
0.076±°-OOI IN.
OF PUSHPULL SOUND
RECORD
«L SCANNED AP
SEPTUM
!
ZD
1.93 ±0.02 MM
0.314*0-001 IN.
EA-
— >-
— H
/"7.98 ±0.02 MM
0.016 ±0.001 IN.
0.406*0.02 MM
WIDTH OF EACH HALF OF
PUSHPULL SCANNED AREA"
0.084*0.002 IN.
2.13 ±0.05 MM
Section Meetings
Atlantic Coast
On September 22, 1948, the Atlantic Coast Section of the SMPE held its first
of the regular fall series of meetings. Two papers were presented before a group
estimated at 300 people in Studio 3-A of the National Broadcasting Company.
The first paper by Robert M. Fraser of the Development Engineering Group of
the National Broadcasting Company was titled "Recording Television Programs
of Motion Picture Films." The paper dealt with the history and the growth of the
methods used to record television images onto 16-mm film for subsequent reuse.
Both the technical and the practical engineering aspects were covered by Mr.
Fraser. A demonstration reel showing a comparison of early experimental work
with the present stage of the art was well received by the audience.
The second paper, "Flicker in Motion Pictures," by Lorin D. Grignon,
Sound Department of Twentieth Century-Fox Corporation, was presented for the
third time to an SMPE audience. This was a timely review of the causes, meth-
ods of analysis, and suggested remedial steps concerned with various types of
flicker in motion pictures.
Midwest
The September 16, 1948, meeting of the Midwest Section was held jointly with
a group forming the Optical Society of Chicago in the auditorium of the Metallurgy
and Chemical Engineering Building, Illinois Institute of Technology in Chicago.
Dr. Robert A. Woodson addressed the 125 members, guests, and those interested
in forming the optical group who were in attendance.
First on the program was a 16-mm English Technicolor film titled "Colour"
produced by the Imperial Chemical Industries. It dealt with color fundamentals
and research in the development in coal-tar dyes. "Adopting Motion Picture
Equipment to the Needs of Medical Teaching" by Mervin W. LaRue, Sr. covered
the reasons for the great need of medical films to keep practicing physicians
abreast of the latest techniques and practices as well as student teaching of medi-
cine. Special apparatus for microcinematography and macroscopic work was
described and the actual equipment was on display. This equipment is covered
in a paper published in the JOUBNAL for September, 1947. A reel of Koda-
chrome motion pictures of medical subjects made on the described equipment was
projected, and appropriate comments were made by the author.
"Seeing Light and Color" by Ralph M. Evans in charge of Color Quality, Color
Control Department, Eastman Kodak Company, Rochester, was accompanied
by over two hundred color slides which prove beyond a doubt that you "only see
what you think you see. . ."
EMPLOYMENT SERVICE
POSITION WANTED
CAMERAMAN: Eight years* experience in 35- and 16-mm cinematography,
color, black-and-white. Experienced in educational, documentary, training, ad-
vertising, and entertainment films — studio and location. References and com-
plete record of experience available. Willing to relocate anywhere. Write air
mail, Peter H. Swart, 28 Webber Road, Germiston, South Africa.
549
Book Reviews
The Diary and Sundry Observations of Thomas Alva Edison,
Edited by Dagobert D. Runes
Published (1948) by the Philosophical Library, 15 E. 40 St., New York 16,
N. Y. 244 pages + XII pages + 3-page index. 5 illustrations. 6X9 inches.
Price, $4.75.
This volume is in all respects the most extraordinary of all the presentations in
print pertaining to Mr. Edison. It is quite as remarkable and quite as interesting
for what it is not as for what it is. It tells something about him, but with the
quality of an image projected through a diffusing screen and picked up by a
ground-glass mirror. There is enough of him there for the reader to be conscious
of his presence but not convinced of his actuality and substance. To the re-
searcher of tomorrow who would know about Edison, it is a document to be con-
sidered and read only after he has seen everything else that has been printed. In
that respect, this book would have the same interest which it now has to those
intimate with both the real and the traditional Edison and with a measure of his
poignant realism and dynamic place in the industrial scene.
From the particularized and technological point of view of the readers of this
journal, the specific attentions attributed to Mr. Edison pertaining to motion
pictures are positively somewhat less than negligible. The document contains
nothing informative pertaining to Mr. Edison's invention of and contribution to
the motion picture which is not either in casual error, in casual misunderstanding,
or, at best, susceptible of misinterpretation. For motion picture engineers, this
book can be an object lesson in the perils of pseudo-literary adventure in the art
of expression for persons who do not carry over into that art the skills and criteria
that they bring to bear upon their science. For this peculiar state of affairs, there
are two discernible reasons. The first of these pertains to the frequently non-
chalant manner in which Mr. Edison discussed his works and his charming willing-
ness to talk about anything which might be a passing topic of interest. Second
is the fact that the volume appears to have been assembled and edited by Dr.
Dagobert D. Runes, a writer of distinction about philosophical subjects, for his
Philosophical Library, Inc. The approach is scholarly-mannered and with a
brave effort at categoried analytical presentation of the omniferous miscellany of
Mr. Edison's interests, all of them expressed in his declining years.
There is, unhappily because of that play upon the words "the diary," an early
disappointment. The "diary" element hazily covers one week in 1885. Very
little of the real Ediso a comes through, anywhere. We get no picture of that salty
fellow, dynamic, belligerent, collarless, with tobacco stains on the bosom of his
hard-boiled shirt, mildly profane and belligerently positive, talking behind that
big roll-top desk at West Orange about what he really thought. We have here an
Edison sandpapered, shellacked, and waxed. We do not have, in any part of the
book, the "Old Man," ebullient, ironic, and vital.
For purposes of specification to this engineering audience, it is appropriate to
cite a single but painfully indicative specimen of error occurring on page 77 of the
volume, in which Mr. Edison is caused to say apparently that his Kinetoscope,
550
Book Reviews
the name of which, incidentally, is misspelled, "attracted quite a lot of attention at
the World's Fair in Chicago in 1893." The date alleged and the facts pertaining
to the Kinetoscope involved are all important to motion picture history. The
Kinetoscope, as Mr. Edison had very decided reason to know, was under contract
for exhibition at the World's Fair in Chicago in Edisonia Hall, and was in fact
not delivered there, being unavailable until after the Fair closed. That was be-
cause the mechanic he had assigned to the building of the battery of mechanisms
•decided to spend most of his time at an adjacent West Orange tavern, playing
dominoes or something. Edison cared very little. So, as has been competently
recorded and is historically documented, his Kinetoscope made its first appear-
ance to the public on the night of April 14, 1894, at 1155 Broadway.
The great Edison was great enough not to be deceived about himself and he
took neither the motion picture nor Edison too seriously. This reviewer once
took Mr. Edison to task because of a piece of Sunday -supplement journalism
which had gone to extravagant lengths in a sort of interview indicating that the
Wizard of West Orange was building a machine with which to communicate with
the dead.
The "Old Man" looked puzzled a moment and then flung out with a defense.
He said, "Don't be too hard about it. That reporter was a space writer. He
came over here without any raincoat and there were holes in his shoes. He
needed a story in the worst way and I gave him the best one I could think of."
You need a touch of that to understand this book.
TERRY RAMSAYE
Motion Picture Herald
New York 20, N. Y.
L'Annuaire du Cinema 1948 (Motion Picture Yearbook for 1948)
Published by Editions Belief aye, 29 Rue Marsoulan, Paris (12eme), France.
1230 pages. 5l/z X Sl/2 inches. Price, $6.00. United States Representative,
Andr£ Harley, 15 E. 40 St., New York 16, N. Y.
This French yearbook on the film industry is the first revised and re-edited
book of its kind to be published since the Liberation of France.
The book is divided into seventeen sections, subdivided as follows: Paris
addresses, out-of-town addresses, general information on French film industry,
list of motion picture theaters in Paris and environs with number of seats, man-
ager's name and address, same list for out-of-town theaters, 16-mm section with
names of all people interested (laboratories, distributors, synchronizers, and
theater owners), films (information on 440 films presented in France between
January, 1946, and June, 1947), producers, distributors for Paris, distributors for
other regions, export, foreign countries, newspapers and magazines, technicians,
artists, suppliers, studios, and laboratories.
551
Current Literature
nnnE EDITORS present for convenient reference a list of articles dealing with
J- subjects cognate to motion picture engineering published in a number of se-
lected journals. Photostatic or microfilm copies of articles in magazines that are
available may be obtained from The Library of Congress, Washington, D. C., or
from the New York Public Library, New York, N. Y., at prevailing rates.
American Cinematographer
29, 9, September, 1948
Television Camera Operation
(p. 302) H. I. SMITH
British Kinematography
13, 1, July, 1948
Colour Vision and the Film Industry
(p. 1) W. D. WRIGHT
13, 2, August, 1948
Reversal Processing (p. 37) IR. H.
VERKINDEREN
Set Construction Methods (p. 46)
J. Gow
Light Efficiency of 16- Mm Projectors
(p. 50) W. BUCKSTONE
Electronics
21, 10, October, 1948
Television Transcriptions (p. 68)
T. T. GOLDSMITH, JR., AND H.
MlLHOLLAND
International Photographer
20, 9, September, 1948
Motion Picture Industry in Sweden
(p. 10) W. J. BARNDALE
International Projectionist
23, 9, September, 1948
Magnetic Recording Advances Prom-
ise Extensive Use for Film Work
(p. 6)
Basis of the Schmidt Optical System
(p. 8)
Television: How It Works (p. 17)
W. BOUIE
The Transistor: Amplifier-Oscil-
lator May Supplant Vacuum Tube
(p. 19)
Radio and Television News
40, 3, September, 1948
Something New in Color Television
(p. 40) R. CROSMAN
The Recording and Reproduction of
Sound. Pt. 19 (p. 48) O. READ
40, 4, October, 1948
The Recording and Reproduction of
Sound. Pt. 20 (p. 56) O. READ
Tele-Tech
7, 9, September, 1948
New Design for Medium Definition
TV Camera System (p. 52) J. B.
SHERMAN
ASA Adopts Universal Decimal Classification System
The American Standards Association
has decided to adopt the practice fol-
lowed by other national standardizing
bodies and to classify American
Standards in accordance with the
Universal Decimal Classification sys-
tem. By means of this classification,
American Standards can be easily in-
corporated into libraries and identified
as part of the technical literature in all
552
parts of the world. The UDC num-
bers, which are in Arabic and thus can
be read without difficulty regardless of
the language of the country, will appear
on the front cover of all standards
approved by the American Standards
Association and distributed through
the ASA office.
Reprinted by permission of the ASA from
Industrial Standardization, for August, 1948.
New Products
TJurther information concerning the material described below can
JL be obtained by writing direct to the manufacturers. As in the case
of technical papers, publication of these news items does not consti-
tute endorsement of the manufacturer's statements nor of his products.
Heavy-Duty Splicer
A combination 8-mm and 16-mm
semiprofessional splicer for heavy-duty
use in schools, film libraries, and labora-
tories has been announced by Bell and
Howell Company, 7100 McCormick
Road, Chicago 45, Illinois.
In one operation, the new splicer
shears both ends of the film diagonally
and applies pressure to the film ends
while they are being cemented. An
electrical element in the base, operat-
ing on alternating current only, heats
the shear blades, thus shortening ce-
ment-setting time.
In addition to the usual provision for
scraping emulsion from the left-hand
film end, the right-hand shear blade
and arms of the new splicer are designed
to permit scraping the emulsion from
the right-hand film end, a process neces-
sary for splicing certain types of prints
and titles.
A gauge block on the splicer base
simplifies setting the scraper blades
at the proper working depth. Extra
scraper blades may be stored in a
covered receptacle on the right side of
the base.
This splicer is 63/4 X 51/* X 3Va
inches and weighs but three pounds.
The base and three operating arms are
made of cast aluminum; the four shear
blades are hardened, ground, stainless
steel.
The splicer base has been designed
so that it may be screwed to a work-
table; an accessory subbase has been
designed to accommodate the -splicer
combined with a Filmotion Viewer and
heavy-duty rewinds, to provide a com-
plete heavy-duty editing outfit.
Synchronous Tape Recorder
The Hallen Development Company
of Burbank, California, is reported to
be manufacturing a magnetic tape, slit
and perforated to 16-mm dimensions,
and a gear-driven recorder which will
stay in synchronization with any
camera equipped with a synchronous
drive.
Additional features include top-
quality amplifiers, a recording capacity
greater than sound film, and an im-
mediate playback. It is adaptable to
16-mm and 35-mm commercial film
production, as well as in television
broadcasting.
553
SECTION OFFICERS
Atlantic Coast
Chairman Secretary-Treasurer
William H. Rivers Edward Schmidt
Eastman Kodak Co. E. I. du Pont de Nemours & Co.
342 Madison Ave. 350 Fifth Ave.
New York 17, N. Y. New York 1, N. Y.
Midwest
Chairman Secretary-Treasurer
R. T. Van Niman George W. Colburn
Motiograph George W. Colburn Laboratory
4431 W. Lake St. 164 N. Wacker Dr.
Chicago 24, 111. Chicago 6, III
Pacific Coast
Chairman Secretary-Treasurer
S. P. Solow G. R. Crane
Consolidated Film Industries 212—24 St.
959 Seward St. Santa Monica, Calif.
Hollywood, Calif.
Student Chapter
University of Southern California
Chairman Secretary-Treasurer
Thomas Gavey John Barn well
1046 N. Ridgewood PI. University of Southern California
Hollywood 38, Calif. Los Angeles, Calif.
Office Staff— New York
EXECUTIVE SECRETARY OFFICE MANAGER
Boyce Nemec Sigmund M. Muskat
STAFF ENGINEER JOURNAL EDITOR
William H. Deacy," Jr. ' Helen M. Stote
Helen Goodwyn Thelma Klinow
Dorothy Johnson Ethel Lewis
Beatrice Melican
554
Journal of the
Society of Motion Picture Engineers
VOLUME 51 DECEMBER 1948 NUMBER 6
PAGE
Flicker in Motion Pictures: Further Studies
LORIN D. GRIGNON 555
Video Distribution Facilities for Television Transmission
ERNST H. SCHREIBER 574
Improved Optical Reduction Sound Printer J. L. PETTUS 586
Films for Television JERRY FAIRBANKS 590
Sensitometric Aspect of Television Monitor-Tube Photography
FRED G. ALBIN 595
Colorimetry in Television WILLIAM H. CHERRY 613
Origins of the Magic Lantern. . . J. VOSKUIL 643
Report of the Studio Lighting Committee 656
Samuel Edward Sheppard 667
ARTHUR C. DOWNES HELEN M. STOTE GORDON A. CHAMBERS
Chairman Editor Chairman
Board of Editors Papers Committee
Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in
their annual membership dues; single copies, $1.25. Order from the Society's general office.
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers,
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office,
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930,
at the Post Office at Easton, Pa., under the Act of March 3, 1879.
Copyright, 1948, by the Society of Motion Picture Engineers, Inc. Permission to republish
material from the JOURNAL must be obtained in writing from the General Office of the Society.
Copyright under International Copyright Convention and Pan-American Convention. The
Society is not responsible for statements of authors or contributors.
Society of
Motion Picture Engineers
342 MADISON AVENUE— NEW YORK 17, N. Y.— TEL. Mu 2-2185
BOYCE NEMEC . . . EXECUTIVE SECRETARY
OFFICERS
1947_1948
PRESIDENT EDITORIAL VICE-PRESIDENT
Loren L. Ryder Clyde R. Keith
5451 Marathon St. 233 Broadway
Hollywood 38, Calif. New York 7, N. Y.
PAST-PRESIDENT CONVENTION VICE-PRESIDENT
Donald E. Hyndman William C. Kunzmann
342 Madison Ave. Box 6087
New York 17, N. Y. Cleveland, Ohio
EXECUTIVE VICE-PRESIDENT SECRETARY
Earl I. Sponable G. T. Lorance
460 West 54 St. 55 La France Ave.
New York 19, N. Y. Bloomfield, N. J.
1948-1949
ENGINEERING VICE-PRESIDENT FINANCIAL VICE-PRESIDENT
John A. Maurer David B. Joy
37-01—31 St. 30 E. 42 St.
Long Island City 1, N. Y. New York 17, N. Y.
TREASURER
Ralph B. Austrian
25 W. 54 St.
New York, N. Y.
John W. Boyle
1207 N. Mansfield Ave.
Hollywood 38, Calif.
David B. Joy
30 E. 42 St.
New York 17, N.
William H. Rivers
342 Madison Ave.
New York 17, N. Y.
Governors
1947-1948
Robert M. Corbin Charles R. Daily
343 State St. 5451 Marathon St.
Rochester 4, N. Y. Hollywood 38, Calif.
Hollis W. Moyse
6656 Santa Monica Blvd.
Y. Hollywood, Calif.
1948
S. P. Solow R. T. Van Niman
959 Seward St. 4431 W. Lake St.
Hollywood, Calif. Chicago, 111.
1948-1949
Alan W. Cook Gordon E. Sawyer
4 Druid PL Lloyd T. Goldsmith 857 N. Martel St.
Binghamton, N. Y. Burbank, Calif. Hollywood, Calif.
Paul J. Larsen
Los Alamos Laboratory
University of California
Albuquerque, N. M.
Flicker in Motion Pictures:
Further Studies*
BY LORIN D. GRIGNON
TWENTIETH CENTURY-FOX FILMS, BEVERLY HILLS, CALIFORNIA
Summary — Flicker is defined for the general case and additional infor-
mation on subjective effects and analysis presented. The subject is then
restricted to those types of flicker which are the result of equipment de-
ficiencies and quantitative methods for measuring such effects are described.
The application of methods to specific equipments, the results obtained, and
certain remedial measures are discussed. Finally, recommendations for
future work in this field are submitted.
TJILICKER IN MOTION PICTURES is a visual, random, or periodic change
JL in the brightness of a projected picture which is not deliberately
introduced for its suggestive or dramatic effect. The periodic frame
or shutter rates which are fundamental to the methods of motion pic-
tures are included in the definition when these rates produce visual
sensation. Such effects are, in general, sufficiently high in frequency
to be detrimental only to the second order at presently used illumina-
tion levels. In the following material the subject matter will be re-
stricted to flicker which is caused by periodic rates in addition to the
basic frame or shutter frequency except in instances where remarks on
the general subject are pertinent.
Several years ago a paper1 was presented before the Society on this
subject. Most of the data given were not quantitative due, princi-
pally, to lack of suitable apparatus and testing methods for such a
complex problem. The purpose of the present paper is to review the
basic problem of flicker, to describe methods of analysis applicable to
the problem, present the results of the application of indicated meth-
ods to specific situations, to discuss the data, indicate certain remedial
measures, and submit recommendations for future work.
DISCUSSION
Even though some basic facts concerning flicker have been pre-
viously stated it seems expedient to restate the information and, per-
haps, add thereto to further an understanding of the underlying problem.
* Presented May 17, 1948, at the SMPE Convention in Santa Monica.
DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 555
556
GRIGNON
December
Luckiesh and Moss2 state Porter's law to the effect that the maxi-
mum observable frequency of an intermittent visual stimulus in-
creases proportionately as the logarithm of the brightness but Hecht
and Smith3 show that this is only true below certain light levels and
that above these levels no further increase in frequency is noted, as
shown in Fig. 1. The discontinuity in the curves at low levels is
caused by the transition from rod to cone vision. In considering
60
50-
I
4O-
30 1
tr
o
LOG RETINAL ILLUMINATION-PHOTONS
1 2345 6
Fig. 1 — Relations between perceptible frequency,
illumination, and visible area.*
* Figures 1 and 2 reprinted from the Journal of General Physiology, July, 1937,
with the permission of the copyright owners.
periodic flicker we are principally interested in perceptible brigh
ness differences. We can get some idea of this matter from other da
Luckiesh and Moss2 also show a curve by Hecht of the minimum
perceptible brightness difference as related to the logarithm of the
incident brightness. This curve, Fig. 2, shows the minimum to be be-
tween 1 and 100 millilamberts which is within the range of present
screen-brightness practice. The same authors present data concern-
ing the relation between minimum perceptible brightness difference
ht-
ta.
im
1948
FURTHER FLICKER STUDIES
557
and the brightness ratio of a central area to its surroundings. As
shown in Fig. 3, the minimum perceptible difference is less than 0.5
per cent for ratios greater than unity and to something above 10.
Although we do not have data on periodic brightness differences it
should be noted that projection practice has established conditions
which are desirable for motion picture presentation but have made it
manyfold easier to observe flicker.
3.5
RODS
3.3
3.2-
LOG B
MILLILAMBERT5
.4
Fig. 2 — Relationship between field brightness and
minimum perceptible brightness difference.
The relations between modulating flicker frequency brightness
and surrounding conditions are apparently complex. Broca and Sul-
zer4 examined this subject indirectly by studying the growth and de-
cay of visual sensation. They established that the apparent bright-
ness of a periodic pulse of light grew from zero, went through a maxi-
mum, and then approached a constant value somewhat less than this
maximum as the duration of the pulse was allowed to increase from
zero. Their data were collected for various pulse intensities but at
only one repetition rate.
558
GRIGNON
December
None of the earlier studies mentioned considered the basic problem
of a periodic amplitude modulation of an effectively steady light.
Data corroborating that given in the original paper on this subject
regarding perceptible nicker will be given later in this report.
It is interesting and instructive to consider mathematically nicker
due to frame-by-frame differences caused by shutter action. Consider
0.5
BRIGHTNESS RAT.O ^^J^
NlDINGSj
.0.1
Q.I
1.0
10.
Fig. 3 — Relationship between minimum per-
ceptible brightness difference and the ratio of the
brightness of the central visual field to the bright-
ness of the surrounding field. *
* Reprinted from "Science of Seeing" by Luckiesh and Moss, © 1937, by D.
Van Nostrand Company, Inc., with the permission of the copyright holders.
the case of a camera and assume the exposure to be the product of a
constant light intensity and the open time of the shutter, the latter
being a function of a constant angular velocity plus a superimposed
sinusoidal disturbance in angular velocity. Also assume the ampli-
tude of the modulating velocity small compared to the constant veloc-
ity and neglect the finite included angle of the picture frame
and aU second-order and frequency-modulation terms. With these
1948 FURTHER FLICKER STUDIES 559
assumptions, the shutter open time may be expressed as follows:
where
X = open time of shutter
/ = modulating frequency
a = constant angular shaft velocity
<£ = shutter open angle
b/a = velocity modulation index
co = 2irf.
Note that the maximum flicker amplitude is represented by
1 1 1 1
1Z CPS
SUCCESSIVE FRAME EXPOSURE VARIATION
|l
6 CPS
Fig. 4 — Illustrating possible nicker due to frame-by-frame
transmission differences for 12 cycles per second at 3 initial
phase angles and 6 cycles per second at 2 phase angles.
which reveals the relations between flicker amplitude, the shutter
opening, relative velocities, and corresponding initial amplitudes. If
^ = mr (n is any integer)
then flicker does not exist even though nonunif orm shutter velocity is
in evidence.
The above equation represents a continuous periodic wave, whereas
shutter operation is a sampling process. The use of this equation re-
quires obtaining values when t is successively incremented by the
frame period. Hence, it is obvious that the initial value of t deter-
mines the actual wave shape of the disturbance. This effect is shown
in Fig. 4.
560 GRIGNON December
The analysis further states that shutter motion should not be in-
vestigated by other sampling schemes such as stroboscopes if accurate
data are required, unless such tests include a number of different ini-
tial phase angles.
MEASURING METHODS
In the following discussion on measurements, it is sometimes con-
venient to use communication terms and in this sense the frame or
shutter rate can be considered as the carrier frequency, periodic
nicker as an amplitude-modulating frequency, and random variations
as noise.
Required measuring techniques fall into three general classes, as
follows: methods for determining angular velocities, photographic
densities, and light intensities. The difficulties associated with any of
these appear when applied to the measurement of nicker rates super-
imposed on frame or shutter rates. In other words, the problem is to
measure signals 35 to 45 decibels below a carrier of 24 cycles per sec-
ond, or some multiple of same, with the frequency ratio as small as two,
to one. Incidentally, with projector shutter rates of 48 or 72 cycles
per second, the problem is somewhat ameliorated but the cross-modu-
lation products of 24 cycles remain because of the initial taking frame
rate.
Consider the measurement of light nicker in a projector. Basically,
this must be a light-responsive device whose output must be inte-
grated and/or recorded and must be capable of accurately indicating
signals under the above-stated conditions. Such equipment is feasible
provided the ratio of signal to noise is favorable which is hardly the
case under the circumstances.
Densitometric measurements must be made frame by frame and be
accurate and reproducible to density differences of 0.002. The varia-
tions due to exposure, dirt, and abrasions, analogously, constitute a
serious noise problem and, further, the same area in each frame must
be used. Man'ually, this is a very laborious and questionable process
unless unusually large deviations exist. Suitable recording or auto-
matic equipments to meet these requirements of analysis apparently
do not exist.
Apparatus for angular-velocity measurement is more readily avail-
able than either of the aforementioned classes of equipments. The
carrier frequency for the signal may be selected, means provided for
generation, stable amplification applied, and the resultant rectified,
1948
FURTHER FLICKER STUDIES
561
filtered, and measured with common apparatus. A similar class of
apparatus is represented by sound-equipment flutter-measurement
devices. The carrier frequency may be generated by tone wheels on
the shaft in question and suitable pickups or by stable oscillators ex-
citing the fields of various types of velocity generators. The diffi-
culties of velocity measurement reside in the necessary equipment
precision and variations equivalent to noise.
Given apparatus to measure modulating signals, some method must
be devised to introduce such signals under controlled and reproducible
conditions to enable engineering analysis.
For the work at han4, efforts were concentrated on angular-velocity-
measurement methods. Velocity generators of the drag-cup induction
type were chosen as the translating device. Such units are available
with very low moment of inertia, utilize a carrier frequency of 400
cycles per second, and produce an output which is linearly proportional
MODULATED
400 CPS
li
AMPL
RECT
LOW PASS
FILTER
BUCKING
CCT
RECORD-
ER
Fig. 5 — Block schematic of velocity generator receiving cir-
cuits and recording device.
to instantaneous angular velocity over a considerable range. They
have the disadvantage that mechanical coupling to the part to be
measured must be devised and the signal-to-noise ratio is dependent
upon velocity, the ratio being poorest at low velocities. The signal-
to-noise ratio at 1440 revolutions per minute of commercially obtain-
able units is about 30 decibels, making it possible to measure modu-
lating signals of 5 to 6 per cent only, unless the noise signal can be re-
duced. It is possible to reduce noise by 15 or 20 decibels by careful
and judicious trimming of the drag cup and the electrical balance;
hence, it is possible to measure modulating signals of 1 to I1/ 2 per cent
of the carrier. Care must be exercised when coupling the device to
any rotating member to consider relative moments of inertia, avoid
f rictional loading, and nonunif orm velocity or play in the mechanical
coupling. For hand-held contact coupling the most suitable method
is to provide a collet-type member on the generator shaft fitted with a
562
GRIGNON
December
hemispherical rubber point of 60 to 70 Shore hardness. This contact
should preferably, mate with a 90-degree shaft center but the common
shaft centers have been generally suitable.
The electronic apparatus (Fig. 5) consists of a stable 400-cycle oscil-
lator and associated power amplifier for energizing the generator
fields, an amplifier for the generator output, linear rectifier, carrier
filter, and indicating or recording means.
To provide reproducible modulating signals, a magnetic brake was
devised and arranged for attachment to various rotating parts. This
brake is energized by a stable, very low-frequency oscillator, wave-
shaping circuits, and power amplifier, as in Fig. 6.
LF OSC
I^TO 25
CPS
WAVE
SHAPING
AM PL
GROUND HARD CHOME
SURFACE
BRAKE
Fig. 6 — Magnetic brake for test loadings and
associated driving circuits.
For thorough analysis it was considered advisable to have three
complete receiving systems available to provide simultaneous meas-
urement at three points; and, further, permanent records were de-
sirable. A multielement recording oscillograph provided the latter
facility.
MEASUREMENTS AND RESULTS
Before proceeding with a program of tests, a flicker-free projector
was needed to evaluate" the flicker samples visually. Flicker in pro-
jectors is caused by (a) shutter rate, (b) nonuniform shutter velocity,-
(c) arc-supply ripple, and (d) arc-burning characteristics.
The effects of (a) are generally known. Most projectors are sup-
plied with two-bladed shutters producing a 48-cycle shutter rate which
is sufficiently high, at present illumination levels, to be of second-
ary importance. In studio review rooms a three-bladed shutter
1948 FURTHER FLICKER STUDIES 563
frequently is used. No studies were made of two- versus three-bladed
shutters. One point concerning shutters should, however, be made.
Any multibladed shutter must be symmetrical, otherwise the 24-cycle
frame rate is reintroduced and frame flicker becomes apparent.
Analytically, shutters can be expressed by a Fourier series and the
effects on flicker rate of various designs studied very simply.
The shutter used for all visual work consisted of three 93-degree
blades and three 27-degree openings. It is currently used in all studio
review rooms at Twentieth Century-Fox.
Nonuniform shutter velocity can result from poor driving motor
operation or excessive mechanical backlash between driving point and
shutter shaft. Analogously, backlash can be considered as a complex
nonlinear compliance which in concert with the masses involved can
be resonant. Many cases of long gear trams with large backlash on
currently used equipment have been noted with attendant flicker
observable in the projected picture. In order to evaluate the serious-
ness of the shutter nonuniformity, a temporary filtered shutter was
devised and applied. The design was not wholly satisfactory because
of an insufficiently low cutoff frequency but it served to demonstrate
that the higher-frequency flicker components could be noticeably at-
tenuated. This part of the work has not progressed beyond this point.
It is, however, clear that projector improvements either should in-
clude shortened gear trains with a minimum of backlash or some type
of damped or filtered shutter.
Arc-supply ripple is the worst source of flicker. Any 60-cycle com-
ponent greater than 0.15 per cent will cause a 12-cycle flicker resulting
from beats between the 60 and the 48 cycles of a two-bladed or the 72
cycles of a three-bladed shutter. Supplies operating from 50-cycle
sources do not produce the same result since the beat frequencies are
2 and 22 cycles, respectively. Unfortunately, it has generally been
considered that three-phase rectifiers and motor-generator sets require
filtering only for the theoretical higher frequencies prevalent and that
any 60-cycle components are low enough to be neglected. This is
not true. Consider the three-phase rectifier. Either unbalanced line
voltages or differences in rectifier element voltage drops will introduce
line-frequency components and these must be safeguarded against by
some filtering which is effective at such frequencies. Motor-generator
sets can also contain line-frequency components resulting from arma-
ture slots or rotational effects. Six-phase rectifiers are superior in this
regard because they tend to contain less 60-cycle components. In
564 GRIGNON December
view of the above remarks the cure is obvious but it is surprising to
learn that single-phase rectifiers having insufficient filtering are used
for arc supplies.
To eliminate the difficulties from 60-cycle ripple, all studio projec-
tors employing three-phase rectifiers for arc supply are being equipped
with additional series inductance. The increased series impedance
offers a further advantage in that the arc stability is improved.
Independently, engineers involved in the frequency conversion to
60 cycles in the southern California area discovered the same trouble
when flicker appeared in theaters after the 50- to 60-cycle change.
The inductance design has been supplied them and, to date, five thea-
ters have been equipped. Reported results state that the improve-
ment amounts to 75 to 90 per cent elimination of visible flicker due to
this source. Specifications for the inductance used for 50-ampere sup-
plies are stated in the Appendix.
To forestall serious arc flicker, maintenance men have been sup-
plied with a ripple meter which is arranged principally to measure
60-cycle components.
A good visual tool having been obtained, attention was next di-
rected to flicker sources in motion picture production. These include
(a) illumination, (b) camera, (c) film, (d) negative processing, (e) print-
ing, and (f ) positive processing. It was felt that processing was not a
major factor at present, so no work has been done in this category.
It has been demonstrated1 that printers can introduce flicker but,
again, the present investigation has not progressed to that point.
Hence, there will be no further discussion of this element. It is also
known that set-lighting generator outputs contain 60-cycle compo-
nents but to date no evidence exists that illumination periodic flicker
has been important. The remaining iteniSj camera and film, have had
further study, particularly the former, and now will be discussed.
Flicker has been isolated to film difficulties in a few cases in the past.
The drying-rack effects are easy to determine but the rate here is so
low that it is not significant as a flicker effect. Drying-rack rate is
well known to the manufacturers and it is presumed that eventually
this trouble will be corrected. The changes in film which cause peri-
odic troublesome flicker are hard to separate from the other elements in
the system. Assuming that processing is free from variation, flicker
should be visually apparent if a length of film is uniformly exposed by
some means whereby it is not subjected to the strains common to cam-
era mechanisms, processed, and then projected. With careful handling
1948 FURTHER FLICKER STUDIES 565
it would be possible, though tedious, to measure and plot the den-
sity variations over a given section. To expose film in 50- or 100-foot
lengths in this manner requires elaborate equipment but such lengths
are necessary to permit good visual study or catch intermittent cases.
Such equipment does not exist in Hollywood but several attempts
were made to expose stock in this manner. The tests were made in a
relatively crude manner and showed random variations but, to date,
no serious periodic rates have been noted in the stocks tested. This
statement does not imply that film is always free from the subject de-
fects but is, at least momentarily, free from suspicion. For future
investigation, or to isolate trouble, suitable apparatus to provide the
uniform exposure is highly desirable.
Flicker introduced by or within the motion picture camera results
from (a) nonuniform shutter motion, (b) changes in film sensitivity
by certain film stresses, and probably (c) nonuniform register or film
lay in the picture aperture. The last two have been given only a su-
perficial examination in this work. The most promising approach to
(c) appears to be the use of high-speed photography and has been con-
templated but not yet performed.
Periodic nonuniform shutter velocity is certainly a source of flicker,
being the basic element in determining exposure when frame speed
and illumination are fixed.
The flicker equation given earlier in this paper was set up particu-
larly for camera-shutter operation and the explanatory statements
made then specifically apply here.
For further discussion of camera mechanisms, it is convenient to
consider all elements in terms of electrical analogies. If damping is
ignored the driving-motor system constitutes one oscillatory mesh and
further, since cameras generally include compliances either in the form
of flexible couplings or mechanical backlash, additional oscillatory
modes exist. In the specific case under discussion two flexible cou-
plings, other than motor system compliance, existed. These were in-
troduced for mechanical reasons and to minimize noise, and for such
purposes served very well. The only damping present in the system
was provided by the frictional losses, loading, and a small inherent
damping in the coupling material. The designers of sound-recording
apparatus know the necessity of damping but these techniques have
not been applied as yet to cameras. As a result, oscillatory conditions
can exist when excited by a suitable signal.
To analyze the camera conditions, a velocity generator was properly
566
GRIGNON
December
connected directly to the shutter, a second unit was used to indicate
motor-shaft velocity, and when necessary a third unit was used on the
distributor of the driving interlock system. In this way simultaneous,
instantaneous measurements of velocities were made for steady-state
and deliberately introduced signals. A typical record, simplified for
purposes of illustration, is shown in Fig. 7. The transient response
due to sudden loading or unloading also is shown. Note that the dis-
tributor is not affected although it displays a small amount of periodic
variation. By varying signal frequency and recording the amplitudes
at all points the frequency characteristic may be obtained and the
resonant frequencies determined. A typical record is shown hi Fig. 8.
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
DISTRIBUTOR
I I I I I I I I
Fig. 7 — Velocity changes at three points in
camera system. Average velocit}^ 1440 revo-
lutions per minute for each curve.
From data taken in the above manner, two significant results ob-
tained. First, camera motors and associated systems have a resonant
frequency between 3 and 6 cycles per second. This results from semi-
standardized dimensions of frames and armatures which determines
the moment of inertia and the pull-out power requirements of camera
motors which determines the stiffness. Note that all types of con-
ventional speed-controlled camera motors will fit into this category
because of prevalent motor sizes and power and, further, that the
frequency of resonance is within the critical flicker region. To move
the resonant frequency of motor systems to less critical regions would
require at least a tenfold decrease in moment of inertia, a correspond-
ing increase in stiffness, or a threefold change in both. This is a very
1948
FURTHER FLICKER STUDIES
567
impractical solution. The best answer to this situation would employ
some form of damping or complete isolation of the shutter from prob-
able disturbances from this source. In general, it may be said that the
greater the power capability of a camera motor with respect to a given
load, the less the possibility of flicker from various excitation signals.
Second, one of the auxiliary couplings in the mechanism tested
showed resonance at frequencies between 4 and 10 cycles, depending
upon the material used, the amount of usage, temperature, and the
preciseness of the various fits. As is obvious, these frequencies are
also in the critical region of flicker. In such cases the moments of
Fig. 8— Frequency characteristic for two motors having different
power capability on same supply system.
inertia and stiffnesses must be suitably chosen to avoid trouble or,
alternatively, damping applied.
In determining the data discussed so far sine-wave signals were used
and applied to the camera motor shaft by means of the brake pre-
viously described.
Attention was next directed to the determination of the effect of
periodic loadings when applied to various points in the camera mecha-
nism. Only two such places were readily available; namely, the mo-
tor shaft and the film take-up. The results of motor-shaft loading
have already been discussed except in the quantitative sense. With
sine-wave signals, it was found possible to introduce easily shutter
568
GRIGNON
December
rates from 1/z to 12 cycles in magnitudes varying from 1 to 7 per cent.
This magnitude is sufficient to cause noticeable flicker. Sine-wave
signals applied to film take-up points did not appreciably disturb the
shutter until the applied load was great enough to cause take-up
clutch slippage. This latter result was not in keeping with simple
exploratory tests made at this point by loads applied 'with the fingers.
By trying signals of other shapes, it was found that impulsive waves
of approximately the shape of the peak portions of sine waves would
disturb the shutter sufficiently to cause flicker and still not be appar-
ent visibly in the camera operation. A similar signal applied to the
motor shaft also disturbed the shutter with a magnitude greater than
ever achieved with sine-wave loading. The resulting instantaneous
Fig. 9 — Typical velocity differences in a camera when
artificially loaded at 3 cycles per second. Average
velocity is 1440 revolutions per minute.
velocities of one such test are shown in Fig. 9. The applied load
was small enough so as not to be noticeable to operators of present
equipments.
At this point it was decided to correlate the data obtained analyti-
cally by an attempt to introduce flicker into a photographic test.
The exposure was made while applying periodic impulsive loads, of
the type previously described, to the motor shaft and a record of the
shaft velocity was taken simultaneously. The record is shown in
Fig. 10, indicating peak-to-peak velocity differences as great as 7y2
per cent. The resulting picture contained flicker at the applied fre-
quency and in greater magnitude than any previously obtained in
1948
FURTHER FLICKER STUDIES
569
production work. This result generally corroborates the data of the
original paper which stated that 3 per cent variation was sufficient to
cause perceptible flicker. If any revision is needed it would seem to be
a downward one, perhaps to 2 per cent, especially when considering
Fig. 10 — Velocity differences of camera motor
shaft which introduced excessive nicker in a
photographic test. The 8-cycle-per-second
result was visually much worse than the 3-cycle-
per-second test.
negatives, because of the common usage of over-all gamma around
1.6. Definite correlation is thus obtained. Additional tests with
spiked impulsive wave forms gave visual flicker effects generally cor-
responding to the applied wave shape but sine-wave signals did not
570 GRIGNON December
produce as great a visual effect for a given load application as for the
impulsive cases. This would indicate that periodic binding or tight
spots in the camera mechanism can readily produce flicker if occurring
at the proper rates.
The next question concerns remedial measures. Only four immedi-
ate remedies are available: (1) use motors with the best possible in-
herent damping and the greatest practical pull-out power, (2) reduce
backlash to the absolute minimum, (3) keep all film propelling or
handling mechanisms free of binding or other irregularities, and (4),
where flexible couplings are used, redesign to include damping and
constancy of characteristics or make the resonant frequency outside of
the critical frequency region. Since the above methods are mostly
precautionary, the best and ultimate solution resides in the use of a
properly filtered or damped shutter. Again, this is analogous to the
course followed successfully in sound-equipment design. It is prob-
able that reduction of velocity variations in the 4- to 10-cycle region
need be only to 0.5 per cent for the present, although future technical
improvements may require a revision of this figure. To date no work
has been done toward providing a filtered shutter for existing cameras
but some thought has been given to the feasibility of such a design.
It is hoped that this preferable course can be pursued in due time.
CONCLUSIONS AND RECOMMENDATIONS
(1) Steady technical improvement in illumination level and theater
presentation has reached the point where future attention must be di-
rected toward improving mechanical motions and providing better
auxiliary apparatus and materials to minimize flicker and/or allow
greater latitude in usage before flicker becomes apparent. Nearly all
elements in the basic technical motion picture production and exhibi-
tion system have insufficient margin for flicker-free operation.
(2) The sum total of all variations in the system which produce
flicker should not exceed 2 per cent. However, 3 per cent variations
may be temporarily acceptable.
(3) Analytic examination of film processing and incident or re-
flected light (with projector running) of a theater screen is involved
and difficult. Mechanical motions can be best studied by measure-
ment of instantaneous velocity.
(4) Symmetrical two-bladed projection shutters producing a 48-
cycle rate are presently acceptable but may require revision if illumi-
nation is increased appreciably.
1948 FURTHER FLICKER STUDIES 571
(5) Arc-supply apparatus should not contain more than 0.15 per
cent of 60-cycle components for usual line-supply conditions and prac-
tical limits of equipment aging. Series inductance is desirable for
filtering to meet these requirements and further stabilizes arc burning.
Arc-supply ripple should be measured periodically to indicate proper
operation and forestall serious flicker from this source.
(6) Film stock has introduced flicker but such cases are apparently
random in nature.
(7) Nonuniform shutter velocities, either camera or projector,
cause flicker. Variations up to 7 per cent, peak to peak, have been
measured and reproduced for analysis. Shutters should be damped or
filtered. Consequently, future work should be directed along this line.
(8) Conventionally controlled camera motors should be supplied
for the greatest practical pull-out power. This includes synchronous
motors controlled by line frequency.
(9) In so far as possible, flexible couplings having torsional compli-
ance should be avoided but if this is impractical or impossible, suitable
damping must be provided. Flexible couplings used for angled drives
introduce nonuniform motion in the driven member. Therefore, this
type of mechanism must be avoided.
(10) All film propelling or handling mechanisms must be kept free
of small periodic bindings, tight spots, or other irregularities. This
rigid requirement can be lessened if filtered shutters are provided.
(11) Work should be initiated to investigate the effects of periodic
supply variation on photographic illuminants and the flicker resulting
therefrom.
(12) A study should be made to provide accurate data on periodic
perceptible brightness differences as a function of brightness, fre-
quency, and surroundings. This could be done by a university or
medical school, but since the information is peculiarly applicable to
motion pictures, it may be that the Society should undertake to spon-
sor such a program.
Undoubtedly, in the foregoing material it has been noticed that
many branches of this subject have not been explored and others only
superficially examined. This is an indication of the amount of work
still to be done and emphasizes the need for broadened and acceler-
ated activity in this problem of motion picture production and
presentation.
572 GRIGNON December
APPENDIX
Derivation of equation for exposure, or open shutter time, when the
included angle of the picture aperture is small compared to the
shutter opening.
Let X = exposure, or open shutter time
a = constant component of angular velocity of shutter shaft
b = peak velocity value of disturbing frequency
/ = frequency of disturbance
b/a = modulation index of velocity
<t> = angle of shutter opening
e = angular displacement of shutter shaft.
Then, for sinusoidal disturbing frequencies, the shutter velocity is
-T = a -\- b sin ut = a ( 1 -\ — sin ut J. (1)
Assume b <C a, then
dt
it If. b . \
vr = -{ 1 — - Sin ut I.
18 a\ a /
dd "v * " "*" """ ;'
Now, since 6<C a and if second-order terms are neglected, B ~ a£.
Therefore,
dt I/, 6 . co0\
50-aV1 ~aanT>
The exposure time, or open shutter time, X is
= 1 p+^fj _ 6 { «?1 rf5
a J I a oj
where
00 = angle at which shutter opens.
aX = <f> + I — cos — (6n -}- d>) — - cos
a
26
Now, (4) is a statement pertaining to single frames but we desire
an expression for consecutive frames. This we obtain by increment-
ing 0o by 2w and writing the corresponding values of Xi,Xz,Xs,...Xn.
Further, having defined b -4C a and B ~ at we may write
1948 FURTHER FLICKER STUDIES 573
It should be observed that (5) expresses a continuous periodic
wave which is most suitable for analysis but to obtain data for each
frame, t is successfully incremented by the frame time.
Filter Inductance
63 turns No. 7 twin square copper (83/4 pounds)
Window area — 37/32 X P/4 inches
Tongue — 2 inches
Build — 33/8 inches
Optimum gap (center leg only) — u/32 inch
Grade A transformer iron
Inductance at zero direct current — 3.5 millihenries.
BIBLIOGRAPHY
(1) L. D. Grignon, "Flicker in motion pictures," J. Soc. Mot Pict. Eng., vol.
33, p. 235; September, 1939.
(2) M. Luckiesh and F. K. Moss, "Science of Seeing," D. Van Nostrand Co.,
New York, N. Y., 1937.
(3) Selig Hecht and E. L. Smith, "Intermittent stimulation by light," J. Gen.
PhysioL, vol. 19, p. 979; 1936.
(4) A. Broca and D. Sulzer, /. de Physiologic et de Pathologic Generate, no. 4,
p. 632; July, 1902.
(5) T. C. Porter, "Contributions to the study of flicker," Proc. Roy. Soc., vol.
A63, pp. 347-356; 1898; vol. A70, pp. 313-329; 1902; vol. A86, pp. 495-513,
911-912.
(6) H. E. Ives, "A theory of intermittent vision," J". Opt. Soc. Amer., vol. 6,
pp. 343-361; 1922.
(7) Percy W. Cobb, "The dependence of flicker on the dark-light ratio of the
stimulus cycle," /. Opt. Soc. Amer., vol. 24, p. 107; 1934.
Video Distribution Facilities
for Television Transmission*
BY ERNST H. SCHREIBER
THE PACIFIC TELEPHONE AND TELEGRAPH COMPANY, Los ANGELES,
CALIFORNIA
Summary — This paper describes the Bell System's plans for furnishing
network and local video facilities. The Telephone Company is now using
broad-band coaxial cable and microwave radio systems to provide regular
message telephone service on a number of principal intercity routes through-
out the nation. These facilities can be used to provide television trans-
mission channels when properly equipped. Video service between Wash-
ington, D. C., New York, and Boston over these two types of facilities has
been demonstrated. New facilities are rapidly being extended. Local
video channels for pickup and metropolitan-area networks are provided by
ordinary paper-insulated cable pairs, special shielded polyethylene-insulated
pairs, by microwave radio systems, or by combinations of these systems.
Amplifier and equalizing arrangements for providing wide-band transmission
over these facilities are described. Present Bell System views of the avail-
ability of microwave and coaxial cable facilities on the principal routes, types
of circuits, bandwidths, bridging and terminating arrangements, and general
information concerning the provision of television circuits are covered.
THE PHENOMENAL ADVANCE which has taken place in the field of
television during the past few years has come in no small measure
from many developments and contributions which have been fur-
nished by the motion picture industry. As the industry develops
further it appears that television will establish itself on a firm eco-
nomic basis when the same programs are released to many audiences
throughout the nation in a manner similar to that used so effectively
in the motion picture field through extensive film-distribution organ-
izations. With this focus we may examine the general types of
facilities which are being provided by the Bell System for the trans-
mission and distribution of television video signals on an area or
nation-wide network basis.
Since the inception of network broadcasting 25 years ago, the BeL'
System has been furnishing program-transmission service to an ever-
increasing number of stations. Today more than 1000 stations
* Presented May 17, 1948, at the SMPE Convention in Santa Monica; May 20,
1948, at the Second Annual Broadcast Engineering Conference of the National
Association of Broadcasters in Los Angeles.
574 DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
VIDEO DISTRIBUTION FACILITIES 575
receive service over their lines and more than 150,000 miles of pro-
gram circuits are in use.
Within the last few years audio program network channels have
been provided in many instances by means of carrier systems oper-
ating over broad-band circuit facilities. These program-carrier sys-
tems are designed with phase and attenuation equalizers and provide
high-quality channels. With about 600,000 miles of these broad-band
systems in our plant, many of the future 8000- and 15,000-cycle pro-
gram channels will be provided by this means.
Recently a 15,000-cycle network circuit using carrier terminal
equipment was placed in service between Washington, D. C., and a
suburb of New York City. By an extension of these same methods
video circuits can also be provided.
In contrast to ordinary telephone-message circuits which require a
bandwidth of about 3000 cycles, the video circuits require a band of
frequencies several million cycles wide, depending upon the picture
detail and definition which are desired. Such factors as echoes or
ghosts, attenuation, phase distortion, noise, cross talk, and modula-
tion products have an important bearing on the quality of picture
images which are received. As in the case of message facilities, the
necessary controls and protective measures are employed to maintain
the quality of the picture transmission.
CHANNELS FOR TRANSMISSION OF LOCAL VIDEO SIGNALS
Local channels for television video pickup, distribution, and studio-
transmitter connections are usually arranged for a bandwidth of
about 4 megacycles. Because of the wide variety of locations at
which these channels may terminate, advantage is often taken of the
availability of ordinary paper-insulated cable pairs which can be used
for video transmission. The transmission losses of these facilities in
the video range are shown in Fig. 1.
Because of the relatively high losses which are involved, it is neces-
sary to provide amplification on these circuits at intervals of from
0.6 to 1.5 miles depending upon the gauge of the conductors and
physical layout of the cable.
These paper-insulated conductors are used most frequently where
the loops are short or special low-loss pairs are not available. Their use
entails the removal of bridge taps, or multiple conductors, to avoid
echoes, and the pairs must be carefully selected to avoid interference.
576
SCHREIBER
December
The second type of facility for local service is a special 16-gauge,
shielded, polyethylene-insulated, balanced pair. This structure has a
low loss, is relatively free from interference and noise, and is very
stable.
These polyethylene pairs are being provided over relatively short
distances to many locations where it is expected that recurring or per-
manent television channels will be required. With a loss of about 18
too
too
10 KC
50 KC 100 KC 500 KC IMC
FREQUENCY
SMC 10 MC
Fig. 1 — Typical attenuation losses of facilities used for video
transmission.
Regular telephone-cable pairs.
Polyethylene-insulated shielded pair.
decibels per mile at 4 megacycles, amplifiers can be spaced at intervals
up to 3.5 miles in length. This wider spacing greatly reduces the re-
quirement for locating video repeaters at locations outside of estab-
lished telephone central offices.
Where long-range planning is possible, polyethylene pairs can be
located to advantage in the same sheath with full-sized cables pro-
vided for telephone requirements with consequent economy of duct
usage.
1948
VIDEO DISTRIBUTION FACILITIES
577
In addition to the channels which may be established by means of
wire facilities, the telephone companies are utilizing microwave radio
in many instances. The field of use for this latter facility is generally
the longer loops which would otherwise require several wire sections
in tandem.
As television programming develops further, it may be that micro-
wave facilities will become economical to an even greater extent in
furnishing the shorter temporary loops to points which may not
normally be reached over the basic local wire networks.
Amplifier and equalizer arrangements for video channels are shown
in Fig. 2. Transmitting and receiving amplifiers are available for
UNBALANCED
INPUT
AMPL
TRANSMITTING
f- LINE EQUALIZERS
INTERMEDIATE
/- LINE EQUALIZERS-^
<EtJ
UNBALANCED
OUTPUT
OUTPUT
AMPL
RECEIVING
Fig. 2 — Video amplifiers — block schematic.
use at the circuit terminals when required. Intermediate amplifiers
are provided along the circuit route to make up for the attenuation
losses. Predistorting and restoring networks are available for use to
minimize the effects of high-frequency noise when required. The am-
plifiers are arranged for connection to 75-ohm unbalanced circuits
such as are in common use at studio or pickup locations or for con-
nection to 110-ohm balanced-cable pairs. Attenuation and phase
equalization for the video cable circuits are provided at intermediate
or receiving locations.
At transmitting terminals where amplification is not required, a
repeating coil capable of passing video frequencies is connected be-
tween the 75-ohm output and 110-ohm balanced-cable circuit to
effect the transition. A similar coil is used at the receiving end to
make the transition from balanced-cable pair to unbalanced output.
57S
SCHREIBER
December
Because of low-frequency transmission characteristics of these coils
it is necessary to employ a clamper to reinsert the low-frequency in-
formation which has been removed in transmission. The transmis-
sion characteristics of the repeating coil, which has recently been made
available, are shown in Fig. 3. An amplifier is associated with the
clamper to assist in its operation and to increase the output signal to
the desired level. In this case amplification and equalization for the
balanced-loop facilities are provided at the intermediate central
offices through which the circuit passes.
The application of the repeating coils and clamper circuit at the
circuit terminals are shown in Figs. 4 and 5, respectively. Typical
frequency characteristics of equalized video circuits provided by
local cable facilities are shown hi Fig. 6.
«»
**
TD
52
</>
CO
3'
0
IOC
\
\
V
\
V
^
;PS too CPS IKC to KC 100 KC IMC ION
FREQUENCY
Fig. 3 — Loss versus frequency characteristics of 197A
video repeating coil.
Local distribution networks for transmitting video programs from
one location to a number of other locations, such as chains of depart-
ment stores or theaters, can be provided by using the same general
circuit and equipment arrangements.
VIDEO INTERCITY NETWORK FACILITIES
As a consequence of the elaborate and costly arrangements which
are required for producing television studio programs, it appears that
there will be an even greater economic need for video network facilities
than in sound broadcasting. As in the case of sound broadcasting, it
is believed that the demand for network facilities will develop with
the expansion of the television industry and that nation-wide net-
works will soon become a reality.
1948
VIDEO DISTRIBUTION FACILITIES
579
Present techniques for providing long-haul telephone circuits make
use of broad-band facilities such as are provided by coaxial-cable or
microwave-relay systems. At the present time there are about 7000
miles of these two types of facilities completed or under construction
§t-
31=3 C=3
HWW-I
->/WW '
_J
and it is expected that this amount will be nearly doubled in the next
few years. The same facilities that are used for deriving ordinary
message circuits can also be used for providing audio program chan-
nels or for television video channels.
The requirements of the circuits, namely, ordinary message, audio
program, or video program, determine the selection of the terminal
580
SCHREIBER
December
h!
BT , »
> so >
-WMA-I
z$c
d°
ISJ"
—
IB
Clomper
Amplifier
Rectifier
i
ti
s
1948
VIDEO DISTRIBUTION FACILITIES
581
and intermediate apparatus to be applied to the broad-band system.
While the basic techniques are relatively simple, considerable develop-
ment work is required to provide the greater refinement and improve-
ment which wide bandwidth and extreme length of circuit entail.
COAXIAL-CABLE SYSTEM
The layout and circuit details of the coaxial-cable system have been
covered quite fully elsewhere so they will only be touched on briefly at
this time. The <?oaxial conductor consists of a copper tube 3/8 inch in
diameter with a central conductor insulated from the outer tube by
insulating disks. The bandwidth which can be transmitted over a
conductor of this type depends upon the amplifier and equalizer ar-
rangements which are used and the spacing at which they are placed.
-LENGTH 1.6 MILES
ATTENUATION AT 4 MC - 102 db
1 INTERMEDIATE AMPLIFIER
____
FRCQUENCY KILOCYCLES
- LENGTH 9.0 MILES
ATTENUATION AT 4 MC- 356 db
5 INTERMEDIATE AMPLIFIERS
Fig. 6 — Video circuits — frequency characteristics.
The amplifiers now in use have a gain of about 50 decibels at a fre-
quency of about 3 megacycles and are used at 8-mile intervals. For
telephone service, 600 message circuits are operated in the band be-
tween 68 and 2788 kilocycles. For television transmission, the signal
is translated through a system of double modulation and transmitted
as an upper sideband and a vestigial lower sideband on a carrier fre-
quency of about 311 kilocycles.
The associated high-quality television audio channel is transmitted
in the range below 200 kilocycles. This lower range is not used for
video transmission because of the inadequacy of the shielding at low
frequencies and the difficulty of equalizing a 3-megacycle band which
includes these lower frequencies.
The present coaxial system provides a video band about 2.8 mega-
cycles wide which can be used to transmit television pictures or as
many as 600 message telephone circuits. Development work is being
582
SCHREIBER
December
carried on for a system of the same general type which will have a
useful bandwidth of about 7 megacycles and which will permit trans-
mission of a 4-megacycle television channel as well as about 600 tele-
phone circuits over two one-way coaxial tubes.
In order to insure high circuit stability and continuity of service
hi the coaxial system, many features have been provided to insure
that incipient troubles are quickly corrected and that interruptions
and variations are kept to a minimum. The underground type of
construction employed reduces the likelihood of service interruption
and greatly reduces the effect of temperature changes. Spare line
facilities arranged for automatic switching in the event of cable trouble
add to the reliability of the coaxial circuits.
Fig. 7 — Bell System radio-relay route between New York
and Boston.
MICROWAVE-RELAY SYSTEMS
In order to determine the practicability of using microwave radio
as a means of providing wide-band circuit facilities, a system of this
type was constructed between New York and Boston. It was com-
pleted in November, 1947, and used to demonstrate the feasibility of
this medium. The equipment operates in the 4000-megacycle range
and requires line-of -sight transmission between relay points.
Following the customary wire-line technique of using repeaters at
intervals along a circuit, seven intermediate radio repeaters or relays
are employed as shown in Fig. 7. At each repeater location the in-
coming energy from each direction is converted to a frequency of 65
1948
VIDEO DISTRIBUTION FACILITIES
583
megacycles for amplification and then raised to the transmitting fre-
quency. A typical microwave repeater station is shown in Fig. 8.
On the roof of each radio-relay station are four 10- by 10-foot horns
each incorporating a shielded lens for focusing the radiation into a
narrow beam, so sharp that it is 10,000 times more powerful than an
unfocused signal. This increases the effective power very substan-
Fig. 8 — Birch Hill microwave repeater station.
tially and permits long-range transmission with very small radiated
power. Two of the four horns face New York and two face Boston.
This allows two-way operation, one antenna of each pair being used
for transmitting and one for receiving.
These microwave radio-relay systems have been designed to afford
a transmission band about 4.5 megacycles in width and provide an
excellent means of producing high-quality television video circuits or
584
SCHREIBER
December
1948 VIDEO DISTRIBUTION FACILITIES 585
hundreds of message-telephone circuits. The use of the New York-to-
Boston system for television was demonstrated in November, 1947,
and the results were very gratifying. By looping the two two-way
circuits back and forth between New York and Boston, a circuit
about 880 miles in length and having 32 repetitions was realized.
Picture signals received back at New York were difficult to distin-
guish from the original material even on a direct comparison basis.
FUTURE PLANS
To provide additional capacity for television and telephone cir-
cuits, a wide-band New York-Chicago radio-relay system is planned
for service late hi 1949. The initial installation will care for one work-
ing channel and one stand-by channel in each direction and will be
capable of extension to a substantially larger number of channels
when required.
Early in 1949, the gap between the southern transcontinental
coaxial cable and eastern coaxial network is scheduled to be closed
by the completion of the St. Louis- Jackson (Mississippi) section.
Although novel equalizing problems will be involved in equipping
this length of cable for television, it is believed that transcontinental
television can be provided about one year after completion of the
cable if the demand justifies proceeding at once with this work. Fig.
9 indicates present and proposed television network routes.
Whether cable or radio-relay circuits will emerge as the better
means for providing the various services in a particular area is still
unknown. Present indications are that both systems will play their
part and that the use of either or both will be determined by the
particular needs and particular geographical conditions.
BIBLIOGRAPHY
(1) L. Espenchied and M. E. Strieby, "System for wide-band transmission over
coaxial lines," Bell Sys. Tech. Jour., vol. 13, p. 654; October, 1934.
(2) M. E. Strieby, "Coaxial cable system for television transmission," Bell
Sys. Tech. Jour., vol. 17, p. 438; July, 1938.
(3) Lawrence G. Woodford, Keith S. McHugh, and Oliver E. Buckley, "The
Bell System's progress in television networks," Bell Sys. Mag., vol. 25, p. 147;
Autumn, 1946.
(4) W. E. Bloecker, "Interconnecting facilities for television broadcasting,"
Electronics, vol. 20, p. 102; November, 1947.
* (5) J. F. Wentz and K. D. Smith, "A new microwave television system,"
Trans. A.I.E.E., vol. 66, p. 465; 1947.
(6) H. T. Friis, "Microwave repeater research," Bell Sys. Tech. Jour., vol. 27,
p. 183; April, 1948.
Improved Optical Reduction
Sound Printer*
BY J. L. PETTUS
RCA VICTOR DIVISION, HOLLYWOOD, CALIFORNIA
Summary — An improved 35-mm to 16-mm optical reduction sound printer
embodying improvements in image quality and film motion is described.
I. INTRODUCTION
THE OPTICAL REDUCTION sound printer was developed for the in-
dustry because it afforded one of the most practical methods of
producing 16-mm sound-track prints from original 35-mm sound nega-
tives.1 A number of these printers were produced by the Radio Cor-
poration of America and have been successfully used by processing
laboratories for the past twelve years. During this time a number of
improvements in image quality and speed regulation have been re-
ported before this Society by Drew and Sachtleben.2 These and other
features have now been incorporated into a new printer known as the
RCA Type PB-177.
II. DESCRIPTION
As illustrated in Fig. 1, the mechanism consists essentially of a
driving motor, mounted on the rear side, with associated enclosed
gearing, 35-mm and 16-mm film paths appearing left to right, respec-
tively, the optical system, and a control panel. The driving motor is
specially designed for smooth starting, with torque reduced for the
first half second of operation. An additional soft-starting circuit
which employs a time delay and resistive network provides an adjust-
ment which insures optimum starting voltage for the first five seconds
of operation. Film feed and take-up sprockets are geared to the mo-
tor. The two sound drums are film-driven. Each sound drum is indi-
vidually stabilized by its own damping wheel.3 The take-up spindles
are belt-driven from pulleys geared to the driving motor. Spring-
loaded idlers, with continuously adjustable tension, are employed to
* Presented May 20, 1948, at The SMPE Convention in Santa Monica.
586 DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
OPTICAL REDUCTION SOUND PRINTER
587
maintain optimum belt tension. The 35-mm feed and take-up brack-
ets are designed to accommodate 1000 feet of film. The 16-mm raw-
stock feed bracket and take-up bracket accommodate 1200 feet and
400 feet of film, respectively.
Fig. 1
The printer optical system includes an exposure lamp, illuminating
system, and an objective lens barrel. The latter contains two objec-
tives and a cylindrical lens assembly, which provide for a 2V2-to-l
reduction along the line of travel and approximately 85 per cent re-
duction laterally. Facilities are provided for vertical and horizontal
positioning and for rotation of the exposure lamp.
The control panel contains a rheostat for adjusting exposure-lamp
voltage, a 0- to 150-volt direct-current meter for indicating exposure
588
PETTUS
December
lamp voltage, a lamp off-on switch, and a motor off-on switch. Addi-
tional facilities mounted on the printer are a Ruby pilot lamp for use
while threading and a footage counter which reads in feet of 35-mm
film.
The operating mechanism of the printer is mounted with supporting
columns on a four-footed pedestal, which insures stability.
III. TECHNICAL DATA
The operating speed is 180 or 150 feet per minute on 60- or 50-cycle
power supplies for the 35-mm negative. Likewise it is 72 or 60 feet
per minute for the 16-mm print. Driving power is obtained from a
Fig. 2
60/50-cycle, 230-volt synchronous-type motor operating at 1200/1000
revolutions per minute and producing approximately 100 mechanical
watts. Stabilizing time from start is 10 to 12 feet of 35-mm film or
4 to 5 feet of 16-mm film. Stopping time is 6 to 8 feet of 35-mm film
or 21/2 to 3 feet of 16-mm film. Speed regulations as shown by motion
studies indicate a total flutter content of 0.1 per Cent in the 16-mm
print. Fig. 2 illustrates a photomicrograph of a 6000-cycle, 16-mm
print exposed at 80 volts lamp voltage on Eastman Type 5302 fine-
grain positive stock having a density of 1.4 using a fine-grain negative
master with a density of 2.6 to 2.9. The exposure lamp is a 120-volt,
100-watt, CC-13 filament, bayonet base, projection type. Track-
placement alignment of both 35-mm and 16-mm films are controllable
1948 OPTICAL REDUCTION SOUND PRINTER 589
by lateral positioning of the film at the respective sound drums by
means of adjustable guide rollers. All optical adjustments are fac-
tory-set and sealed for normal operation whereby the emulsions of each
film face each other and the 16-mm stock is type-B wound. Varia-
tions in this procedure are possible by refocusing and realignment.
IV. CONCLUSION
The Type PB-177 printer permits the transfer of sound recording
from 35-mm to 16-mm films with superior quality. It is capable of
high daily output due to its high operating speed. Film motion com-
pares very favorably with that of high-quality 35-mm studio-type
recorders.4 Color-corrected optics and coated lenses provide excellent
image definition and in conjunction with a 120-volt, 100-watt expo-
sure lamp make possible the exposure of Kodachrome, the slowest film
likely to be used in the printer.
ACKNOWLEDGMENT
Grateful acknowledgment is tendered to Messrs. C. E. Hittle, L. T.
Sachtleben, A. W. Freeman, and George Worrall whose efforts con-
tributed to the design and construction of this printer.
REFERENCES
(1) M. E. Collins, "Optical reduction sound printer," /. Soc. Mot. Pict. Eng.,
vol. 27, pp. 105T107; July, 1936.
(2) R. O. Drew, and L. T. Sachtleben, ' 'Recent laboratory studies of optical
reduction printing," J. Soc. Mot. Pict. Eng., vol. 41, pp. 505-514; December,
1943.
(3) E. D. Cook, "The technical aspects of the high-fidelity reproducer,"
J. Soc. Mot. Pict. Eng., vol. 25, pp. 289-314; October, 1935.
(4) M. E. Collins, "A de luxe film recording machine," /. Soc. Mot. Pict.
Eng., vol. 48, pp. 148-157; February, 1947.
Films for Television*
BY JERRY FAIRBANKS
JERRY FAIRBANKS, INC., Los ANGELES, CALIFORNIA
Summary — Film will be the backbone of television programming, re-
quiring the motion picture industry to prepare pictures for this new medium.
When television becomes as advanced as radio is today, video stations will
devote more than 50 per cent of their program time to film because film
is the only proved method by which a show can be planned, rehearsed,
staged, edited, previewed, and telecast with professional perfection. Film
photographed for theatrical release is not satisfactory for television. New
lighting techniques must be used, long shots avoided, and television's small gray
screenkeptin mind during production if quality transmission is to be obtained.
WITH TWENTY-SEVEN stations'^ on the air in May, 1948; with con-
struction permits for another 66 issued; with more than 229
applications (146 of which are in hearing) now pending before the
Federal Communications Commission ; and with these figures climb-
ing rapidly weekly — it is apparent that television is no longer just
around the corner.
With the National Broadcasting Company planning to have 32
stations and affiliates on the air by December; with other networks
now lining up stations across the nation; with the sale of sets ex-
pected to reach the 700,000 mark this year ; and with the number of
stations predicted to number 65 in 42 cities by the end of the year,
1000 by 1953 — it is obvious that television is here to stay — and des-
tined to become one of the nation's major industries.
It is equally obvious that film will be one of the main sources of
video programming and that a tremendous new motion picture in-
dustry to supply this entertainment is certain to develop.
When television is as .advanced as radio today, video stations will
devote more than 50 per cent of their program time to motion pictures
because film is the only proved method by which a show can be
planned, rehearsed, staged, edited, and previewed and telecast with
professional perfection.
* Presented May 21, 1948, at the SMPE Convention in Santa Monica.
fNoTE: As of October 14, 1948, there were 40 stations on the air, 83 additional
construction permits, and 311 applications (186 in hearing).
590 DECEMBER, 1948 JOURNAL OP THE SMPE VOLUME 51
FILMS FOR TELEVISION 591
As you are well aware, film eliminates the "human errors'- which
constantly threaten "live" productions. A mistake is easily edited on
film; scenes done badly can be photographed again. It is the only
safe method of presenting a sponsor's message. The advertiser knows
his blades will not cut the shaver ; his aspirins will dissolve immediately ;
no gadgets from his refrigerator will come loose and clatter to the floor.
All of our tricks of the trade — process shots, miniatures, slow and
accelerated motion, animation, optical tricks — are either difficult or
impossible to do in "live" telecasts. But they can be accomplished
easily on film and add tremendous production value to any program.
Exterior scenes, always of vital importance whether for westerns,
dramas, or comedies, are extremely difficult to do "live," especially if
interspersed with interiors. Film, however, makespossible the use of as
many exterior scenes as are desired, addinglif e and realismto allprograms.
Film is the only method by which unlimited action can be obtained.
A television program using "live" talent has the same limitations as a
stage production. All action at any given time is on a single set and
because the action is continuous, the players are held to a single
costume. With film there are no intermissions, locale is not re-
stricted, and as many sets and costumes as needed can be used.
Films free the writer, the director, and the producer from the shackles
of the "live" stage.
Film overcomes the present-day broadcast problems of time.
Sponsors using film will be able to book shows at whatever hour they
wish and at whatever stations they desire. It is the most practical
way for a sponsor to achieve immediately a national network; it
will be especially important in tying together small and remotely
situated stations during the early stages of television.
Film photographed for theatrical release, however, is not, and
never will be, satisfactory for television. Many stations now are
telecasting old pictures because of the scarcity of program material.
Once films shot especially for television are shown, however, the
telecasting of these old pictures will stop because of their poor repro-
duction quality. In televising theatrical pictures, deep-shadow
effects lose their effectiveness and sometimes turn white on video re-
ceivers. Long shots blur and it is difficult to recognize players.
Television film should and must be shot expressly for telecasting.
New lighting techniques must be used, long shots avoided and tele-
vision's small screen and limited gray scale kept in mind during pro-
duction if quality transmission is to be obtained.
592 FAIRBANKS December
In preparing our "Public Prosecutor" and other television series
for NBC we are using new techniques developed during nearly three
years of research. This series, the first to be filmed especially for
video, is being photographed in a very high key with back lighting
emphasized. We are using much more camera movement than we
would use in theatrical filming because of the close grouping of
players. This camera movement, of course, gives the viewer the
feeling he is seeing more action than actually was photographed.
Close grouping of actors is a must because of the smallness of
video screens. If large, sharp images and facial reactions are to be
clearly received on video sets, players must remain closely grouped.
Half figures are the rule, not the exception. Pan and dolly shots
should be emphasized and there are many more, and larger, close-ups
than would be used in preparing film for theatrical distribution.
In filming for television, the camera must be carefully centered on
the players and action because of the curvature of the television re-
ceiving tube. Any action on the edge of the picture is likely to be
distorted because of this curve. Extreme blacks and whites should be
avoided because they will not televise satisfactorily. Rapid pan
shots will blur and large sharp lettering should be used in all titles.
Even the technique for sound recording is different. In theatrical
pictures the voice is recorded "big" to go with the "large" picture
being projected on the theater screen. For television, the sound
should be "small" for the small receiving screen. Sound for regular
pictures is designed for large auditoriums. Sound for television, like
radio, should be recorded for hearing in an average living room.
Sets for the series we are filming for NBC are constructed smaller
than the sets we build for theatrical pictures. This is being done so
that a larger section of the background is visible to the viewer, thus
creating more atmosphere and more quickly establishing a locale.
If larger sets were used, the video audience would see a smaller section
of the background because filming for television necessitates the
camera's being closer to players and sets.
In an effort to provide television films with the same intimacy of
radio, we have borrowed the technique of making the camera a person.
The camera is "y°u>" the television audience. In brief, "you" ex-
amine the clues; "you" accompany the Prosecutor as he makes his
investigation; "you" are a part of the drama. Players frequently
talk to "you" and "you" are given the same opportunity to solve the
crime as the players in the film.
1948 FILMS FOR TELEVISION 593
All timing is faster for video film. The theatrical film is designed for
an audience of hundreds. The television picture should be made for an
audience of five, five people at home, with all the distractions of home.
Scripts should be prepared in such a way so that the viewer can follow
the plot by listening, and not be required to remain glued to the set at
all times to follow the action. On the other hand, the dialog should not
explain every happening. In short, television-film writing should be a
careful blending of radio and motion picture scripting.
The acting technique for television motion pictures is a combina-
tion of stage and screen. Long shots for television require the cast
to play scenes somewhat "broader" than would be necessary for
theatrical film. The reason is obvious. Facial expressions are lost in
longer shots because of the small video screen. In close-ups, however,
the technique is the same as for regular motion pictures.
It is our belief that television film also will differ from theatrical
pictures in format and running time. Video executives now believe
that the basic time periods of television will be ten and twenty
minutes rather than the fifteen- and thirty-minute programs of radio.
This is understandable because on television so much more can be
told and shown in the same period of time. As a result, our programs
are scheduled for ten- and twenty-minute times. The " Public Pros-
ecutor" series runs twenty minutes including the sponsor's message.
An "open-end" technique is being used, allowing space for the ad-
vertisement at the beginning and at the end of programs. Shows
of the series are designed to play individually or serially and each
program is a complete show.
Television will create thousands of new positions and opportunities
in the motion picture field. It will be responsible for the development
of a tremendous new film industry, an industry devoted to the making
of quality entertainment especially for television.
How large will this industry be? Only time will tell. But if tele-
vision requires films for fifty per cent of its programming, as we be-
lieve, it will eventually total a need for more than 300 hours of film a
week. This figure, when compared to the present Hollywood output
totals a tremendous new prosperity for all motion picture employees.
It should be remembered, however, that there will be no "box
office" for television films. Motion picture standards of this new film
industry must be, and will be, scaled down to junior size. The bill for
these films will be footed by the sponsor and, obviously, no sponsor
ever will be able to afford two million dollars for a program. The
594 FAIRBANKS
extravaganza movie, the two- or four-million-dollar showcase picture,
never will have a place in video.
Television film will be a new and completely separate industry from
the present motion picture industry. Lower wage scales throughout
all crafts will be necessary to get the industry rolling. Although these
scales will be lower, the total annual earned income of employees will be
as large or larger because the men will be working twelve months a year ,
not as they now get calls from producers, on a picture-to-picture basis.
Footage cost of video films must be lowered from current "theatrical"
picture footage costs if this new industry, this new prosperity for
movie workers, is to get under way properly. Heavy labor demands
at the present time will greatly endanger the development of video films.
Will television, and this new video film industry, be harmful to the
present motion picture industry? Obviously it is far too early to say
for certain. Most television executives and many film producers,
however, do not think that it will hurt box-office returns. There are
many of us, on the contrary, who believe that television will provide
exhibitors with a greater drawing card, with larger returns.
Television, for one thing, provides a means of publicizing motion
picture coming attractions in the home. Most exhibitors will agree
that the trailer is one of the most important means of advertising new
products. Thanks to video, the exhibitor now can show a sample of
his wares to thousands of potential theater goers, thousands who do
not now go to shows regularly.
It has been estimated that a top-quality picture only attracts an
aggregate audience of 25 million admissions. Television, if it follows
the development of radio, as is now believed, will eventually be seen
by more than 90 per cent of the population of the nation. In short,
exhibitors will enter 37 million homes with their trailers and will have
the opportunity of attracting roughly 50 million more ticket buyers
to their box offices.
There also are many of us who see the day when the theater goei
will be able to see a television newsreel of the day's events. And wt
think that it will not be too long before the regular feature is halted
and a special telecast is shown audiences of the winning run of the
Rose Bowl game or the World Series.
Television will enhance theater programs, provide a greater draw-
ing card, and prove a greater incentive to go to a show.
Film, for theater and for television, has a very bright future.
Sensitometric Aspect of Television
Monitor-Tube Photography*
BY FRED G. ALBIN
RCA VICTOR DIVISION, HOLLYWOOD, CALIFORNIA
Summary — The performances of the iconoscope and orthicon pickup tubes
and kinescope monitor tubes constituting a television system are considered
in regard to the response versus level characteristics. A nonlinear elec-
trical network is advocated for combination with the iconoscope to equalize
the gamma variations to a constant gamma approximately complementary
with the monitor-tube gamma. Another nonlinear electrical network is
advocated for combination with the orthicon to reduce the gamma of this
camera to the same gamma as the corrected iconoscope camera.
A direct positive photographic technique is described using a negative
monitor picture obtained by electrical phase reversal, and the toe region of
the positive film characteristic. A general mathematical expression for the
shape of the film toe as a function of the gammas of the television camera
and monitor as required for linear over-all performance is derived.
The merits of such a photographic technique are economy, simplicity,
rapidity of processing, and greater average screen brightness.
INTRODUCTION
PHOTOGRAPHY FROM A cathode-ray tube onto motion picture film
of television pictures which were primarily intended for direct
viewing on a home receiver is an increasingly popular procedure. The
television broadcast studios and stations require photographs of their
programs as a record of past shows and for use by their own or other
stations for repeated or delayed broadcasts. Also, theaters require
photographs of televised show material on standard motion picture
film which may be subsequently projected before their patrons with
standard film projection equipment.
I. TELEVISION SYSTEM CHARACTERISTICS
In order to incorporate a television system into an over-all repro-
ducing system involving photography, a knowledge of the perform-
ance characteristics of the television system as well as the photog-
raphy is required. The sensitometric method of measuring perform-
ance lends itself well to television as it does to photographic film.
In photography, the mensuration of the response characteristics
*Presented May 21, 1948, at the SMPE Convention in Santa Monica.
DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 595
596 ALBIN December
makes use of the H-axid-D curve, which is a plot of log-exposure in-
crements as abscissas against the negative log transmission incre-
ments or density increments as ordinates. Similarly, a measure of
the over-all photographic system involving a negative and positive
photograph is a "print-through" H-and-D curve, which is a plot of
log negative exposure increments against the resulting log positive
transmission increments. The slope of the straight portion of either
of these curves is a measure of contrast and is the popularly known
term "gamma." This concept may be expanded to represent the
slope at any point on the curve and identified by the term "instan-
taneous gamma."
It can be proved by a rather simple mathematical treatment and
illustrated graphically by means of the "tone-reproduction quadrant
diagram" of Jones1 and Mees2 that the gamma of the over-all photo-
graphic system is equal to the product of the gammas of the negative
and positive which constitute the over-all system. Jones and Mees
go further to show that faithful reproduction of the original is achieved
only when the over-all gamma is unity, except that in practice an
adjustment is made in this value to an over-all gamma of approxi-
mately 1.5, to allow for unaccounted-for factors or to satisfy sub-
jective requirements.
In the photographic art the exposures are confined largely to the
straight-line portions of the .H-and-D curve where the gamma is a
constant single value. Where the mean exposure of the negative is
determined by the mean illumination of the subject and this together
with the printer light determines the mean exposure of the positive, a
certain latitude of variation of these mean values is allowed without
affecting the gamma or linearity of the over-all performance. This lati-
tude is a virtue of the constant gamma of the negative and positive films.
A corollary to the photographic theorem, applicable to a television
system might be stated as follows : "A system containing two or more
nonlinear elements is linear over-all if the over-all product of gammas
at corresponding points on the response curves of all nonlinear ele-
ments is unity." This -general statement covers a photographic
system which is usually linear over-all and is comprised of a negative
and positive, both nonlinear. It covers an ideal television system,
and a system combining television with film which might involve four
or more nonlinear elements.
The present-day television system with which we are dealing in-
volves (1) a pickup tube, (2) an amplifier and transmission system,
1948
TELEVISION-TUBE PHOTOGRAPHY
597
and (3) a reproducer tube. For the purpose of this paper, the second
item may be considered as linear. The response characteristics of the
pickup tubes and the reproducer tubes taken from published data3
are illustrated by Fig. 1. Here, the abscissa is a log scale of relative
input levels and the ordinate is a log scale of relative output levels.
For the pickup tubes (a) and (6), the input is instantaneous light
RESPONSE VS INPUT LEVEL
TYPICAL CHARACTERISTICS
80 100
flux and the output is electrical volts in response thereto. For the re-
producer tubes (c), (d), and (e), the input is instantaneous electrical
volts and the output is light flux in response thereto. Plots (6), (c),
(d), and (e) being straight lines represent constant gammas. Plot
(b), having unity slope or gamma with a positive sign, represents a
linear function. But (a) is a curve having variable gamma ranging
from 1 at low levels to l/& at high levels. This variable garnma must
be reckoned with because the over-all system gamma is consequently
variable as a function of the reference level of illumination of the
scene being reproduced by the camera containing this tube.
In Fig. 1, (d) is a popular direct-viewing type of kinescope used in
home television receivers. It exhibits a gamma in excess of 3. Fig.
1 (c) represents a projection-type kinescope currently used in larger
598
ALBIN
December
home receivers, with a gamma of approximately 2.75. Fig. 1 (e) rep-
resents a kinescope similar to (c), but provided with a flat face and a
blue-light phosphor. Fig. 2 (a) illustrates the spectral characteristics
of this kinescope and (b) the spectral sensitivity of nonsensitized
monochromatic photographic film of the release-print type. The
effect of this narrow spectrum of blue light in photography is to lower4
the gamma approximately 15 per cent below the white-light charac-
F I CURE 2
SPECTRAL CHARACTERISTICS
RELATIVE RESPONSE OR SENSITIVITY
VS ANGSTROMS WAVE LENGTH
t eristic. Hence the gamma of 1 (e) used for photographic exposures is
illustrated as being 2.35, lower than the otherwise similar white-light
tube 1 (c).
If a television system were to reproduce faithfully the scenes in
front of the camera, the over-all gamma should be constant at unity,
or to be in accordance with professional photographic practice, about
1.5. Accepting a predominating gamma for reproducers at a con-
stant of approximately 3, the camera gamma should be a constant at
0.5. The over-all gammas resulting would be approximately 1 .7 for the
direct-viewing home receiver and 1.18 for the photographic exposures.
Thus 0.5 gamma for the camera appears to be a good compromise.
1948 TELEVISION-TUBE PHOTOGRAPHY 599
The iconoscope pickup tube tends toward a low gamma character-
istic and the kinescope reproducing tube tends toward a high gamma
characteristic.5 Thus, when used in one system, these two tubes are
complementary, but only connotatively speaking because of the
manner and extent to which the gamma of the iconoscope varies each
way from an optimum value of 0.5. The present effect of the com-
bination of iconoscope and kinescope is excessive gamma in the
shadows or lowly lighted scenes, and deficient gamma in the high
lights.
To complicate television reproduction further, it is present practice
to use iconoscope cameras and orthicon cameras interchangeably in
the same televised show. The picture quality from such dissimilar
cameras cannot possibly match, and if one is considered proper with a
certain receiver characteristic or adjustment, the other cannot be
with the same receiver characteristic or adjustment. Thus, to correct
the over-all system gamma by lowering the receiver gamma is not
only uneconomical because of the large number of receivers involved,
but would result in inconsistent performance when different types of
cameras are used.
This leads to the following required system modifications in order
to obtain consistently good performance :
Correct the characteristic of each camera so that the gamma func-
tion has a single value over the range used. This value should be com-
plemental to the reproducer gamma. There are several alternatives:
(a) Adjust all cameras (transmitting system) to some arbitrary
value, say 7C=1. Adjust all receivers to the complemental value,
ym=l. This becomes uneconomical when a large number of re-
ceivers is involved.
(b) Accept a value of receiver gamma at ym =3, thus avoiding
changes in a large number of receivers. Adjust transmitter gamma to
a compromise value yc= l/i.
Alternative (b) is advocated because besides being more economi-
cal, it improves the signal-to-noise ratio of transmission, since for a
given high-level limit of video signal, the low levels will not drop as
low with a low gamma characteristic of transmitted signal. Keeping
the low levels up thereby keeps them higher relative to the noise and
maintains good signal-to-noise ratios during transmission.
II. GAMMA-CONTROL NETWORKS
Electrical networks can be built to exhibit a nonlinear level re-
sponse characteristic as required, within limits.
600 ALBIN December
Nonlinear networks must contain one or more nonlinear elements.
In all cases, if a response curve obtained by plotting output levels as
ordinates and input levels as abscissas is curved, the network is non-
linear. A concave upward curve denotes high gamma and a concave
downward curve denotes low gamma. These data for output and
input levels may be plotted on logarithmic scales in which case the
slope of the curve at any point is the gamma of the network at that
point. For a circuit element, a plot of applied potentials as ab-
scissas and the resulting currents as ordinates will be straight for a
linear element, indicating constant conductance, and curved for a
nonlinear element indicating conductance variable as a function of
voltage or current. A concave upward curve designates a high gamma
element and a concave downward curve a low gamma element.
A high gamma element in a series arm of an electrical network im-
parts a high gamma characteristic to the network. Similarly a low
gamma element imparts a low gamma. Conversely, a high gamma
element in a shunt arm imparts a low gamma characteristic to the
network and, similarly used, a low gamma element imparts a high
gamma characteristic.
In practice, a single nonlinear circuit element exhibits a single value
of gamma other than unity for only a relatively narrow range of level.
Outside of the range of constant gamma, the gamma may gradually
change toward unity gamma, or it may suddenly become discontinu-
ous and become zero or infinite, depending upon the nature of the
nonlinear element employed. To achieve a wide range of constant
gamma other than unity, the elements may be arranged in stages, the
stages operated in a cascade order of levels so that over the level range
for which the network is designed, one or more stages operate, each
stage within its narrow level range of gamma control.
The instantaneous gamma of a network is the product of the in-
stantaneous gammas of the individual stages. To achieve a network
gamma control greater than that of a single stage, two or more stages
may be arranged to control the gamma simultaneously by overlapping
the level ranges over which the individual stages are operative.
Linear circuit elements, when associated with nonlinear elements,
tend to veer the gamma toward unity, except when a bridge circuit is
formed. In a bridge circuit, the network gamma may exceed the ele-
ment gamma, especially near the level when a null in output level
occurs.
1948 TELEVISION-TUBE PHOTOGRAPHY 601
Such a network has a loss, the maximum value of which is a func-
tion of the amount of gamma correction and the latitude L2/Li of
signal range at the input or where the gamma is unity. This decibel
loss is expressed by the product (7 — 1) times the decibel signal range
and must be compensated by equivalent amplifier gain.
Some nonlinear circuit elements are :
(1) High Gamma Devices
(a) Vacuum-tube diode 3/2-power-law characteristic
(b) Amplifier-tube grid-toe characteristic
(c) Copper-oxide, selenium, germanium, or other contact
devices
(d) Silicon carbide, aluminum silicate, etc.
(e) Gas-discharge tubes or ionization devices
(f) Carbon-arc or thermal-resistance device
(g) Discontinuously functioning device such as a rectifier
(h) Any saturable insulation
(2) Low Gamma Devices
(a) Vacuum-tube shoulder characteristic
(b) Metallic thermal-resistance device
(c) Any saturable conductance
For the use of a nonlinear network referred to in this paper, the
action must be practically instantaneous to act well within the period
of the highest video-frequency cycle. For this reason the susceptance
must be low relative to the conductance. This requirement rules out
many of the items in the above list such as the thermal and gas de-
vices having inertia.
On the other hand, reactive components can be employed in con-
junction with nonlinear components to produce a desired new param-
eter of gamma control which is a function of frequency. By this
means the gamma for high frequencies may be made greater than the
gamma for low frequencies, thus improving detail contrast which
tends to be impaired by halation and aberrations in the optical sys-
tems, and by other high-frequency losses.
Fig. 3 (a) illustrates in symbolic form a popular form of nonlinear
network which provides flexibility in the shape of the characteristic
response curve. It employs vacuum-tube diode or contact rectifier
discontinuously functioning devices Xi, X2, X3, primarily as switches
or valves to open or close the respective circuits in the ladder-type
602
ALBIN
December
attenuator in response to changes of electrical level in the circuit . The
threshold controls EI, EZ) E3, for the respective circuits are adjusted
to cause these circuits to close in succession as the input voltage of the
network is increased. The input and output voltages, EA and EB,
respectively, are adjusted simultaneously to establish the threshold
of the network.
Fig. 3 (6) is the idealized resulting response curve attainable only
with an infinite number of stages. The sharp discontinuities, which
otherwise would be evident as each successive stage becomes opera-
A OUTPUT VS.INPUT LEVEL CHARACTER
CONSTANT GAMMA REDUCTION.
tive, are smoothed out by the toe in the characteristic curves of the
nonlinear elements X\, XZ) X$, etc. Curve 3 (6) is the characteristic
of an electrical network which is advocated to be combined with the
orthicon or other linear pickup tube to become an integral part of the
camera. The over-all gamma of the camera thereby is constant with
a value of l/z or so as desired to fit into the system. Curve 3 (b) thus is
characteristic of the corrected camera also.
Curve 3 (c) is the characteristic of an electrical network which is
advocated to be combined with the iconoscope pickup tube to become
an integral part of the camera. The configuration of Fig. 3 (a) might
1948 TELEVISION-TUBE PHOTOGRAPHY 603
be used with different constants and threshold adjustments. The
over-all gamma of the camera is thereby made constant with a value
the same as with all other cameras of the same system, as illustrated
by 3 (6). In this case, since the curvature of the iconoscope gamma
is compensated by the curvature of the network gamma, it is impera-
tive that the level references of the iconoscope correspond with the
level reference of the network such that the two characteristic curves
are complementary throughout. Hence, these level references must
be maintained exactly in the system setup and operation. It is
largely for this reason that the network should be closely and in-
separably associated with the iconoscope in the camera. The camera
having constant gamma then enjoys independence of reference level
relative to other system components.
III. PHOTOGRAPHY
The television system, corrected as described in the foregoing
section and operating in the normal manner, will produce a positive
picture which is a faithful reproduction of the original scene within
the limits of capability of the system. This picture may then be
photographed by a photographic technique which in itself is linear,
thereby reproducing the television picture faithfully.
The conventional negative-positive film technique, for example, is
capable of excellent results and preferred if several positives are re-
quired. It requires a separate negative film and processing, printing
to one or more positive films, and processing of the positive filnis.
If only one film is required, the commonly used chemical reversal
technique is more economical. The reversal and redeveloping tech-
nique is more involved than the simple processing of a separate negative.
Electrical Reversal
For producing films of televised scenes for theater use where only
one film is required, it is feasible and more economical to reverse the
television picture from a positive to a negative by merely reversing
the phase of the electrical circuit preceding the picture-reproducing
cathode-ray tube and employing a direct positive type of photography.
Only one film is required, and the processing facilities and time are
reduced to a minimum, resolved to the minimum basic steps, develop-
ing, fixing, washing, drying.
But the character of a negative television picture obtained by
electrical phase reversal is not the same as the character of a negative
604 ALBIN December
photograph of a positive subject in respect to exposing the photo-
graphic positive. If we assume that the television-system character-
istics are predetermined by the direct- viewing conditions already out-
lined above and the negative picture obtained by phase reversal, we
can then determine the response characteristics of a positive photo-
graphic film necessary to reproduce the televised picture faithfully.
As will be seen from the description which follows, the over-all
television and photographic system under consideration involves the
three nonlinear elements or steps, the television camera, television
monitor, and the photographic film. The empirical characteristics
of the uncorrected camera and monitor can be expressed mathe-
matically only by higher-order terms which cause the analysis on this
basis to become cumbersome. But with cameras which are corrected
to a constant gamma approximately complemental to the inherent
constant gamma of the reproducers or monitors, the mathematical
analysis is greatly simplified.
For the purpose of this paper, each of the three steps is represented
by a hypothetical case and can be expressed in simple mathematical
terms. The three hypothetical conditions are chosen to be com-
plemental to an over-all linear system condition.
In addition, each hypothetical condition is allowed to vary inde-
pendently both ways from the initially chosen value, whence the end
result in the over-all system and its departure from the linear condition
indicates the effect of varying the characteristic of each step from the ini-
tially chosen value. Thus, the effects of prevailing empirical conditions
on the over-all system performance can be determined by inspection.
A graphical representation, Fig. 4, is used here after the manner
of the tone-reproduction quadrant diagram of Jones, except that six
quadrants are required to represent this over-all system. Fig. 4 (/)
illustrates the relation of the output to the input, and thus the over-all
response characteristic of the system. The six sections of Fig. 4 rep-
resenting the five major component steps of the over-all system plus
an over-all characterization are arranged in the manner of the tone-
reproduction quadrant diagram wherein the output of the first step
is applied as the input to the second step and so on through all steps in
sequence, finally completing a cycle by returning to the input of the
first step. This procedure will be made clear by following Fig. 4 with
the descriptions of the sequential steps as given below :
Fig. 4 (a) represents a television camera which produces an output
voltage Ve in accordance with the input illumination I. This input
1948
TELEVISION-TUBE PHOTOGRAPHY
605
illumination is assumed to be in direct proportion to the scene bright-
ness. The characteristic response curve of this camera is shown as
being concave downward, typifying low gamma. The general ex-
pression for this characteristic is given with the gamma factor yc as
an exponent of I. A corrected camera is represented in this illustra-
tion with a constant value of 0.5 useid for ye, which represents the
characteristic of the pickup tube after correction with a nonlinear
network as described in Section II. The output voltage becomes the
input voltage for the next step.
Fig. 4 (6) represents the television equipment and electrical trans-
mission system involving variable values of contrast C and initial
brightness B. It also involves phase reversal wherein the output
voltage varies inversely with the input voltage. The initial brightness
B is a constant term added to the signal voltage (negative) and is the
voltage producing an initial monitor-tube brightness when the camera
output voltage is absent. The contrast factor C is a coefficient mul-
tiplying the input voltage, or, in other words, is a control of the elec-
trical signal gain of the system. The negative sign indicates phase
reversal. The output voltage of this system involving the initial
brightness voltage, the contrast factor, and the signal voltage from
606 ALBIN December
the camera is expressed by B — CIyc. The nominal values for B and C
are both unity which condition is illustrated by the central plot.
Plots are also given for B = 0.8, illustrated by the lower plot, and
B = 1 .25, illustrated by the upper plot. In these cases C is maintained
at its initial value of unity. Plots are similarly given for two addi-
tional values of C in which B is maintained constant at unity. Thus
the effects of independently varying B and C are illustrated. In all
cases the characteristic curves are straight lines. Therefore, this step
is linear except for reversing the phase.
Item 4 (c) represents the cathode-ray picture-reproducing tube or
monitor. Here the input voltage Vm is the same as the output voltage
of the preceding step and is plotted as the abscissa of this curve.
The output light values are plotted as ordinates. The response
characteristic of a cathode-ray tube of this sort when plotted is a con-
cave upward curve as illustrated and therefore is nonlinear. The ex-
ponent ym, which is the gamma factor, in this case is greater than
unity. A value of 2 for ym is the nominal value used in these plots. In
the plot of the over-all, a value of 3 is also used and the results plotted.
Step 4 (d) represents the film camera having the function of ex-
posing the photographic film with an exposure E in accordance with
the light output L from the preceding step and in accordance with
other factors A embracing several factors such as the optical speed,
exposure time, magnitude, actinic value of the exposure light, or any
factor which affects exposure as a multiplier. The exposure is con-
sequently expressed as a product of A XL. All plots being straight
lines, this step is linear. This exposure is applied as the input to the
subsequent step.
Step 4 (e) represents the exposure of the photographic film and this
figure illustrates the relationship of the exposure of this film and the
consequent transmission after processing. The hypothetical relation-
ship chosen here is the required characteristic for producing over-all
linear performance as will be illustrated in the next step. The rela-
tionship of exposure and transmission of this hypothetical film for the
ideal case of linear over-all performance is given by the general ex-
pression as follows :
- en-
When each constant A, B, and C equals unity,
Tye = 1 - E1/
ym
1948
TELEVISION-TUBE PHOTOGRAPHY
607
In the illustrated case, 7C=0.5 and 7,
illustrated by the parabolic curve
T1/* = 7 - E1/*.
2.0. This relation is
The plot of 4 (e) is made with these values. This step is not only
nonlinear, the gamma is not constant.
Item 4 (/) illustrates the over-all response characteristic of the
system under all of the various conditions provided in the preceding
OVERALL-TELEVISION PHOTOGRAPHY
USING SINGLE FILM TECHNIQUE
2 EXCEPT AS NOTE
A-e-c*i 7c
TRANSMISSION OF FILM
INPUT LIGHT TO TEL. CAMERA
RELATIVE APERTURE- FILM CAMERA
INITIAL BRIGHTNESS, MONITOR
TELEVISION SYSTEM CONTRAST
GAMMA OF CAMERA
7TO-GAMMA OF MONITOR .^
7P= GAMMA OF FILM
steps, and the relationship of the transmission of the film and the input
illumination of the camera. It is assumed that the reproduced picture
brightness is directly proportional to the transmission of the film.
It is also assumed that the input illumination in the television camera
is proportional to the original scene brightness. The ideal curve
therefore is the centermost plot of Fig. 4 (/) which is a straight line.
The other curves, all identified, illustrate the effects on the over-all
characteristic caused by independently varying A, B, or C either way
from their nominal values of unity.
In Fig. 4 (/) linear relationship of input and output of the over-all
system is illustrated by a straight line of any slope passing through
608 ALBIN December
the origin. Fig. 5 is a plot of the same data used by Fig. 4 (/) except
that logarithmic scales are used for both the abscissas and ordinates.
Plotted in this manner, linearity of output versus input of the over-all
system is represented by any straight line with unity slope. Here as
in Fig. 4 (/) the effects on linearity caused by independently varying
A, B, and C each way from their nominal unity values are illustrated
by plots appropriately marked. Several additional plots are also
given in Fig. 5 and will be discussed later.
The characteristic of photographic film is commonly illustrated in
the form of an H-axid-D curve which is a plot of density increments
as a function of log-exposure increments. The slope of the straight-
line portion of the .H-and-D curve is represented by the commonly
used term photographic gamma. Furthermore, the straight portion
has the highest slope of all parts of the H-and-D curve. The region
represented by exposures less than those producing the straight-line
region is known as the toe. In Fig. 6 are four H-and-D curves of the
hypothetical film characteristics included in the over-all system pre-
viously described and expressed by the general relation
Ei/ym + T^ = 1
are plotted. For all cases it is apparent that for a value of E = 1,
T = 0. The slope in this region therefore approaches infinity, and
equals infinity at E = 1. Since the photographic gamma of film is
the slope of the steepest portion of the .H-and-D curve, in each hypo-
thetical case the gamma of the film according to common nomen-
clature would be infinity. The plots of hypothetical film shown in
Fig. 6 represent largely the toe regions of H-and-D curves, and since
in each case the H-and-D gamma is infinity, the differences between
these four H-smd-D curves are differences in the shapes of the toes.
In the ideal case which produces linear over-all performance, the
gamma of the camera was arbitrarily chosen as 0.5, complementary
to the gamma of 2 for the picture-reproducing tube. In present prac-
tice, orthicon cameras are frequently used without corrective net-
works as described in Section II for which then the gamma is prac-
tically unity. The over-all gamma of such a television system used
in a normal manner for direct viewing would thus exhibit an over-all
gamma of 2, which is excessive. In Fig. 5 a plot identified as yc = 1 is
made for this condition in the subject system employing photography
of a negative television picture which results in a straight line but
with excessive slope or over-all gamma of 2, the same as though the
system were for positive viewing.
1948 TELEVISION-TUBE PHOTOGRAPHY 609
In the hypothetical ideal case the gamma of the monitor was taken
as 2. In Fig. 1 it is shown that this value may be 3 or higher. A plot
for this value is given in Fig. 5 and marked ym = 3. It appears from
a comparison of this plot with the ideal case that the departure in
slope is not great except in the high-light region exceeding 50 per cent
transmission. In this region the contrast would be reduced below
normal. This is the opposite effect from what might have been an-
ticipated as a result of increasing the gamma of the monitor. The
paradox is attributed to the reversal of the television-monitor
picture, and to the great nonlinearity of the film over the region used.
If it is desired to obtain linear over-all performance with a monitor
gamma of 3, this value may be applied to the general equation
expressing the film characteristic to obtain the required shape of the
.H-and-D curve.
From an inspection of the curves of Fig. 5 it is apparent that the
values of A and B have a very pronounced effect on the over-all per-
formance of the system. It is also apparent that variations of A and
B have like effects and that one can be varied to counteract the effect
of the other. It is also apparent by inspection that variations of C
affect the brightness but not the contrast or gamma of the over-all.
The significance of the nominal values of unity for each of these
might be described as follows :
Black Level — At black level, or a condition of zero input illumina-
tion to the television camera, the following values result in the suc-
cessive stages of the system (see Fig. 4) :
/ = o, Ve = 0, CI = 0, L = Bym, E = ABym.
Under these conditions the reproduced picture should be just black.
Thus
T = 0; E = 1; ABym = 1.
Thus A and BJm are reciprocal factors. In practice, the combination
of A and B should be so chosen that with no light entering the camera
the exposure of the film should be adjusted to result in the film's being
just black.
White — A white subject, when illuminated to a level lying within
the operating range of the system, will produce an input illumination
to the television camera identified as white level with an arbitrary
magnitude of 1 for I and T, and the following values result in the
successive stages of the system :
/ = 1, Vc = 1.
610 ALBIN December
Under these conditions the reproduced picture brightness should be
maximum. Therefore
T = 1, E = 0, L = 0, Vm = 0, B - CPC = 0, B = CIye, B = C.
Thus, since B was established in its relationship with A by black-level
considerations, C is subsequently set so as to make the reproduced
picture brightness of white level conform with white-subject levels.
In practice then, a white subject which has a brightness value which
is just white, and has a brightness ratio to the picture, black subjects
not exceeding the brightness range capability of the system, say 50
to 1, should be adjusted by variable C to be reproduced just white.
Results of Tests
Photographic tests of the inverted system were made, using the
facilities and factors as listed:
Cathode-ray monitor tube C73103D
Spectral emission peak 4600A
Potential — 2nd anode 28,000 volts direct current
Current — 2nd anode — average 15 microamperes
Raster area 15 square inches
Film camera lens aperture f/2.5
Shutter open period per frame Vso second
Phosphor persistence — effective (Pll) 1A millisecond maximum
Film EK 5302
Developer formula Modified D 16
In modifying the developer, the metol and bromide concentrations
were increased somewhat above D 16 formula.
Fig. 6 illustrates empirical data on the toe region and a portion
of the straight-line region of the H-and-D curve of the above film
developed to a time-scale gamma of 3. Also, in Fig. 5 identified as
7P = 3 is illustrated the effect on the over-all characteristic of a film
with a maximum gamma of 3 in lieu of infinity as theoretically re-
quired. It is apparent that for transmission values from 3 per cent
to 90 per cent corresponding to a picture brightness range of 30:1,
the departure from the hypothetical is within 10 per cent of the ideal
transmission at any point. This equals conventional practice in
motion picture professional photography.
A further advantage in using this toe region of the positive film
is that the mean transmission is approximately 50 per cent as com-
pared with 25 per cent in professional practice, thus providing in-
creased average screen brightness with the same light source.
1948
TELEVISION-TUBE PHOTOGRAPHY
611
Practical Application of General Film Equation
The several factors involved in the general expression for the film
characteristics are recapitulated as follows:
1. Contrasted with conventional 2-film photography, the camera
gamma yc and the monitor gamma ym need not be complemental for
over-all unity gamma.
2. Theoretically, any combination of ye and ym can be accommo-
dated, resulting in over-all unity gamma. However, an H-a,nd-D
FIGURE 6
H
IN
0
E
T
7c
a
i
c
rf
e
a :
DIP
r T
E
= E)
= TR
= G*
-Qf
CURVES OF POSITIVE
?ECT POSITIVE PHOTOGF
ELEVISED PICTURES.
T£+T7c = 1 WHERE
(POSURE OF POSIT VE-RELX
ANSMISSION OF POSITIVE-R
MMA OF TELEVISION CAMEF
MMA OF TELEVISED MONIT
.VERSED a BLACK LEVEL ES
c= '/3 7m =3
c = '/3 7m =2
e = '/2 7m =2
c =l 7m =l
--o--- TP =3 (EMPIRICS
^HOTOGR
JAPHY
TIVE VAL
ELATIVE
(A
OR
TABLISHE
APHS
UES /
TO 100% /
1
1 /
1 /
/
1 L
.02
.04
.06
.08
.1
.2
.4
.6
.8
1.0"
RE
= T
= 7
= 7
" 7
z: _
.0
/
//
1 1 T~
/
/
i L
T
1
1
~"
U)
/
1 '
•*""
/
~T~\
-H
H -j
—j
f-
-
en j
O
3
I
^
/
/
/
/
/
f--
_
in
en
x
/
/*>
2
/
-
</>
^^
?
&
^
<^
&
/
f
»-
<
g ^=5*<
^
^-^
;?:
,/
z
H
-—^
'
^
2
^^
" -----
j^^— «=
> "
3=^=
^i=
— t-r
= -
^!
^
ai
t^"
^
£;
.004 .006.00601 .02 .04 .06 .08
EXPOSURE (LOG SCALE)
.4 .6 .8 1.0
curve approximating that of Fig. 6(c) can be realized practically
more readily than a curve such as Fig. 6(d). Furthermore, to satisfy
subjective requirements, the over-all gamma preferred is approxi-
mately 1.5 rather than unity.
3. jm value is not optional. For economic reasons where a large
quantity of apparatus is involved, the actual value cannot be changed.
Furthermore, the assumption of a value different from actual af-
fects constancy of over-all gamma.
4. yc has a direct effect on the over-all gamma. If the actual
value is used in the general film equation, the over-all gamma will
be unity. An actual value higher than the value assumed for the
612 ALBIN
equation will produce a correspondingly higher over-all gamma, and
higher by the same percentage.
5. The initial brightness of the monitor and the corresponding
film exposure under conditions of zero scene illumination must be
at the exact value which results in an over-all reproduced picture
being just black.
In consequence of the above factors, the recommended procedure
for obtaining an over-all characterization with constant gamma at
a value of 1.5 to satisfy subjective requirements is as follows:
6 . Measure by empirical means the effective photographic 6 gamma
of the monitor. If a C73103D tube is used, a value of 2.35 will be
sufficiently accurate. This value should be used for jm in the general
equation.
7. Assume a camera gamma of 0.5 arbitrarily, and use this value
for 7C in the general film equation.
8. Adjust the camera gamma to an actual empirically measured
gamma 50 per cent higher than the assumed value, or 0.75 to obtain
an over-all of 1.5.
9. Solve the general equation with the values assigned and plot
the curve with logarithmic co-ordinates. The resulting curve is
the H-smd-D curve describing the required film characteristic.
APPENDIX
(1) L. A. Jones, "On the theory of tone reproduction, with a graphic method
for the solution of problems," /. Frank. InsL, vol. 190, p. 39, comm. 88; 1920.
(2) C. E. Kenneth Mees, "The Theory of the Photographic Process," chapter 10.
(3) RCA Tube Handbook HB-3, vol. 1-2 Sec. 3, issue of October, 1947.
(4) J. G. Frayne and V. Pagliarulo, "The effects of ultraviolet light on variable-
density recording and printing," J. Soc. Mot. Pict. Eng., vol. 34, pp. 614-632;
June, 1940.
(5) D. G. Fink, "Principles of Television Engineering," chapter VIII, sections
49, 53.
(6) In determining film characteristics by sensitometry, a time-scale exposure of
relatively low color temperature is generally used while the actual exposures of the
film by a cathode-ray tube have variable intensity of high color temperature and
very short time. Thus, for film, this sensitometry should be used only as a con-
trol and not as the means of determining the exact optimum values. Also, the
data given here for tube characteristics are typical and are not necessarily accurate
for specific tubes, especially with newly designed tubes subject to changes and ap-
preciable tolerances. Fortunately, the tolerance of the observers to over-all per-
formance apparently is large, based upon the present acceptance of conditions so
far from ideal.
Golorimetry in Television*
BY WILLIAM H. CHERRY
RCA LABORATORIES, PRINCETON, NEW JERSEY
Summary— The colorimetrically exact reproduction of color in simultane-
ous television is now possible, through the congruence of the camera spectral
sensitivities to definite characteristics specified by colorimetry and through
the combination of the camera signals in both positive and negative amounts
by suitable circuits and amplifiers. The negative sensitivities of the photo-
pickups formerly required for certain spectral wavelength intervals are obvi-
ated by these signal-mixing circuits, and simultaneous color television, both
as to color range and accuracy of reproduction, is capable of the finest color
reproduction available anywhere. The mixing circuits perform a function
resembling that known in color photography as masking, but whereas the
latter is always approximate and often a hit-or-miss procedure, the television
system can approach perfection without undue complication. The basic
concepts and relations of trichromatic colorimetry are here developed. Many
of these relationships are of immediate importance in color reproduction and
are stated explicitly, with the aid of a concise notation. In addition, certain
rather philosophic aspects of television as a means for the communication
of sense perception are discussed and a plea is made for the extension and
compilation for television purposes of knowledge about the properties of
the eye.
COLOR TELEVISION, when developed, will bring to the public a
new service and a new field of entertainment and artistic ex-
pression. For the communications engineer it will multiply the
problems of the kind with which he is already familiar. It will fur-
ther make necessary his acquaintance with an entirely different sub-
ject hitherto never regarded as belonging in communications. This
subject is colorimetry. It is outlined here with particular regard for
the problems and point of view of the color-television engineer.
THE GENERAL PROBLEM OF COLOR TELEVISION
The object of a television communications system is to provide to
the receiving person the same sense perceptions he would experience
were he present and observing the scene transmitted. The attain-
ment of this ultimate ideal is bounded by considerations of economy
and by the limitations of technology. Consequently a great variety
* The original and longer version of this paper appeared in the 'RCA Review,
volume 8, number 3, pages 427-459; September, 1947.
DECEMBEB, 1948. JOURNAL OP THE SMPE VOLUME 51 613
614 CHERRY December
of sense perceptions regarded of secondary importance have been
omitted entirely from modern television systems. These omissions
have to do largely with the olfactory and tactile senses and with the
perceptions of mechanical motion and vibration. However, a great
deal of information is also deleted pertaining to the two major senses,
sight and hearing, with which television systems are ordinarily con-
cerned, namely, the absolute levels of the sound volume and of the
scene brightness, the absolute scale of size (with the viewing angle
reduced to that of a small screen), and practically all perception of
perspective. In the last category are included not only binaural
and binocular perspective, but the sound and sight perspectives ob-
tained through head motion and the visual perceptions obtained
through the natural process of focus accommodation of the eye and
through its limited depth of focus. Present television systems at-
tempt merely to transmit a single sound signal and the video signal
associated with a simple flat image of the scene.
While the visual perceptions produced by such a flat image are
recreated by the television receiver through the generation of ap-
propriate physical stimuli, it is of course unnecessary for these stimuli
to be physically identical with those of the original subject image.
The eye is notoriously unable to distinguish fine differences of one sort
or another in light images of this type, and in the case of color vision
we shall find that lights of certain very different spectral distributions
have identical appearances. One might suppose that the quality of
the reproduction would depend on how close its physical properties
approached those of the original. However, effort expended in a
precision of reproduction finer than the eye's discriminating ability
certainly is wasted, and the exact or closely approximate duplication
of the physical properties of the original light pattern is quite beyond
the technical capacity of any television system. The information
inherent in the dependence of intensity of light on wavelength, posi-
tion in the image, and time, which constitutes an adequate physical
characterization of the stimuli giving rise to the visual perceptions,
fundamentally exceeds by far the information-transmitting capacity
of a communications channel, even disregarding the usual limitations
of bandwidth and reproducing equipment.
Every television system necessarily must effect a tremendous
reduction in the information to be communicated. This is done by
judicious selection, from a comparatively much smaller number of
sets of physical stimuli which can be generated in the reproduction,
1948 COLORIMETRY IN TELEVISION 615
of that one which produces the same set of visual perceptions as is
produced by the original image of the television scene. A system,
successful in such endeavor for all subjects it is called upon to tele-
vise, certainly may be regarded as perfectly performing its function.
Thus it is possible, through the limited resolving powers of the visual
process, for conventional black-and-white television to create the
appearance of a continuously luminous and continuously moving
image with a single spot of light. This is moved to form a raster
simulating a two-dimensional still image and then moved again to
present successive images portraying motion. It will be seen that
in a similar manner, by the selection of combinations of three stimuli
alone, it is possible for a color-television system to reproduce ac-
curately the visual perceptions produced by original subject images
through very many different spectral distributions.
The television system must select from a relatively limited number
of possible sets of physical stimuli corresponding, in their consequent
visual perceptions, to a great number of possible sets of stimuli pre-
sented in the original subject. The correctness of such selection de-
pends on whether the original and the reproduction are not notice-
ably different. Hence all engineering judgment in the design and
development of new apparatus, and in the comparison and preference
for one system of transmission and reproduction over another, hinges
critically upon the characteristics of the visual process. It is neces-
sary to know just what properties of the reproduction will be per-
ceived by the eye, in what ways the consequent visual perceptions
will be the same for the reproduction as for the original image, in
what ways differences will be perceived and whether these differences
can be ignored or tolerated, and in what ways marked discrepancies
will appear and whether objectionable and disagreeable effects will
take place. For example, the occurrences of flicker and color breakup
in sequential color television were important reasons for its being
deferred in favor of simultaneous systems.
With the correlationships of physical stimuli which give rise to the
same sensation playing so vital a part in the television problem, and
governing choices between technically different modes of transmission
and presentation, there is clearly a very great need for a thorough ex-
ploration and tabulation of the visual properties of the eye. The
various properties, resolving powers, and acuities, and the manner in
which they change with time and differ from one person to another,
should be stated in a systematic way such that they can be connected
616 CHERRY December
immediately with the technical television problems of spot size,
number of lines, field frequency, bandwidth, and so on. Of course
there have been many reliable investigations of the phenomena of
vision, and a large amount of literature is available on the subject, but
the objectives in most of these investigations were sufficiently diverse
and removed from the field of communications that the data are not
suitable for direct application in. that field. Certainly there has
never been an adequate compilation and codification of those data for
television purposes.
The principal categories of visual properties, stated with reference
to a two-dimensional chromatic image, are accounted for according
to the situation. Where there are no time variations and the space
variations are independent of wavelength the properties cover mono-
chrome visual acuity and resolving power. Where there are neither
temporal variations nor significant spatial .variations the properties
include most of colorimetry. Where there are spatial variations but
no temporal variations they include color resolution and acuity.
Where there are time variations but no space variations they include
flicker, sequential color mixing, and after-image effects. In addi-
tion, there are important cross influences as of one type of variation
on the perception of another and as of the interrelation of different
parts of the same class of perceptions. Thus, for example, what the
eye sees in one part of the field of view depends to some extent on
what is present in other parts. Furthermore, the characteristics of
different people's eyes are different, and beyond this the characteris-
tics of an individual eye change from time to time, under different
nonvisual stimuli, and after different previous experience. A knowl-
edge of the magnitude and frequency of such deviations in properties
is also valuable for proper engineering judgments in television.
COLORIMETRY
It has been found quite generally that a substantial portion of the
properties of most people's eyes which have to do with the percep-
tion of color can be tabulated and codified under the comparatively
simple system of data known as trichromatic colorimetry. These
data apply within the broad range of light intensities to which the
eye ordinarily is subjected and except the two extremes of nearly com-
plete darkness and of viewing highly incandescent bodies. Also
they refer to foveal vision, which is of chief interest in television (ex-
cept perhaps for the case of nicker perception), since presumably the
1948 COLORIMETRY IN TELEVISION 617
television screen is the center of interest in the field of view and its
image falls on that central part of the eye's retina called the fovea. .
With the exception of color blindness, which usually constitutes a
marked departure from the norm, the dispersion of visual properties
in different persons and at different times is sufficiently small that the
norm, expressed as the characteristics of a standard observer, is satis-
factorily representative of the majority group. Thus in the discus-
sion which follows, reference to the characteristics of the standard ob-
server may just as well be interpreted as reference to the characteristics
of any ordinary individual, excepting for small quantitative differences.
Largely as a consequence of the tristimulus system of codification
found possible in colorimetry, there has been developed the theory of
three-color vision in which the eye is thought to contain receptors
possessing three different spectral sensitivities. These three receptor
types are all associated with the cones in the retina, for it is the cones
which are believed responsible for color vision. The rods, also light-
sensitive elements in the retina, are responsible primarily for scotopic
or darkness vision, and in any case are believed not present in the
foveal part of the retina. Colorimetry itself, however, can be under-
stood entirely as a phenomenological description of the color proper-
ties of the eye, and is in no way dependent on any explicit theory of vi-
sion . The present discussion is from this phenomenological standpoint .
Colorimetry is concerned with those of the visual properties directly
associated with the wavelength dependence of the intensity of the
light being seen. That is, it deals purely with color perception,
and, strictly speaking, is confined further to the correlation and codi-
fication of color matches and color mixtures. It is concerned with
perceptions of solely spectral origin. However, all of the properties
of the eye are affected in more or less degree by the intensity depend-
ence on wavelength, and conversely the spatial and temporal depend-
ence of the intensity also have their effects on color perception. For
example, visual acuity depends upon the colors and color relation-
ships in the field of view, while in the opposite sense, the color per-
ception of one region of the field of view may be altered by the proper-
ties of the immediately adjacent areas. Colorimetry seeks to avoid
complications of this nature. In the collection of data, as in its
principal instrument which is the colorimeter, time variations and
spatial inhomogeneities are carefully suppressed, with perhaps the
exception of one bifurcation of the field of view to enable color matches
to be made.
618 CHERRY December
The colorimeter consists essentially of optical means for presenting
to the eye two immediately adjacent, uniformly luminous fields,
usually two adjacent semicircular disks, and means for quantitatively
altering the light intensities and combinations in each of the two
fields separately. One of these fields, the sample field, may take a
variety of forms, depending on the kind of sample that is to be ex-
amined. For instance, it may be a pigmented surface illuminated
with a definite light source, the combination of pigment plus il-
luminant constituting the sample as a whole, since the kind of light
emitted by the field is dependent on both. The basic colorimetric
data, however, are concerned only with the visual effects of the light
itself, and not with the origin of its spectral characteristics. As will
appear presently, it is better that the sample distribution come
directly from a source and illuminate a white field in the colorimeter
so that other lights can be mixed in with it. Thus the sample field
and the standard or matching field of the colorimeter may both be
white matte surfaces, capable of being illuminated simultaneously
by several light sources whose intensities are continuously variable
while their respective relative spectral distributions, that is, their
spectral character, are carefully held unchanged.
Measurements with the colorimeter are performed by adjusting the
intensities and mixtures on the standard side until the line of de-
marcation between the sample field and the standard field disappears.
There is thus no visual distinction between one side and the other (ex-
cept location, of course) and the two colors are said to match. Some-
times it is advisable to rotate the demarcation line by some optical
means, so as to allow for possible anisotropies in the eye structures.
There is always a small interval in the intensity settings over which
the match is equally effective. This is an indication of the limited
color-discriminating capabilities of the eye. Furthermore, with the
same adjustment of the instrument, a mismatch may be observed at
some later time or by a different observer. This is indicative of that
variability of the visual characteristics which was mentioned before.
Nevertheless, on a statistical basis a match is a definite observable
relationship.
Following the above outline, two different forms of conventional
colorimetric data are obtained in the colorimeter, depending upon
Whether the light sources for the standard or matching field consist of
three primary sources, as defined below, each with a constant relative
spectral distribution, or whether they consist of one source of white
1948 COLOKIMETRY IN TELEVISION 619
light and a monochromatic source of variable wavelength. A white
light commonly used is the so-called "C" illuminant which has a
certain standardized spectral characteristic and is rather easily re-
produced. As will be seen later, there is in principle no restriction
on the spectral nature of the white light, nor on the three primary
sources, for data taken with them can be, from their spectral distri-
butions, converted over to data corresponding to any other set of
sources. Of course, whenever the color gamut of the samples to be
examined will allow, it is more convenient and more precise if the
matching sources are capable of effecting direct matches. In general,
directly additive matches, in which the matching sources illuminate
the matching field alone, and the sample source illuminates the sample
field alone, are not always possible. In order to effect a match it is
sometimes necessary to divert one or more of the matching sources
from the matching field to the sample field, and, so to speak, dilute
the light from the sample source with the lights from some of the
matching sources. In these cases the matching light intensities
added to the sample are said to be added to the matching field with
negative intensities. With allowance for the use of such a device,
which henceforth always will be assumed, it is possible to match any
sample with light from any set of three primary sources or with light
from the white and monochromatic sources.
Two color sources are said to be primary with respect to each other
if, when they are regarded in the two fields of the colorimeter, no
nonzero adjustment of their intensities can produce a match; three
color sources are said to be primary with respect to each other, and
form a set of primaries, if no one can be matched by any combina-
tion of the other two.
The results of colorimetric investigations with different spectral
distributions and also with monochromatic light of various wave-
lengths have led to the formulation of two fundamental laws which
express the trichromatic basis of color representation and the linearly
additive colorimetric nature of light:
(1) The light from any sample light source, at any given intensity,
which is in its effect a color stimulus or simply a color, can be matched
by one and only one combination of a given set of three primary
sources.
(2) The combination of two colors matches with the combination
of their matches.
In addition to these laws, from which follow the relations that
620 CHERRY December
will be used in the subsequent discussion, the colorimetric data yield
three functions or combinations thereof, which are conventionally
tabulated as the colorimetric distribution functions and often symbol-
ized by x, y, and z, numerical quantities dependent on wavelength.
These functions are to be interpreted as the intensities of the three
ideal primary sources needed to match unit intensity monochromatic
light of the corresponding wavelength. It is important to realize that
although functions of wavelength and at the same time representing
intensities, these quantities by no means represent the spectral inten-
sity distributions of the primary sources. Their connection with the
latter is quite indirect. Furthermore, when the data are converted to
a given set of primaries in actual use, the resulting functions likewise
do not describe the spectral distributions of the primaries. They state
what intensities of the primaries, with all their spectral distributions,
are required to match various monochromatic colors. These colorimet-
ric functions state, therefore, what total intensities of the primaries
are required in a given case, not what spectral distributions in the
primaries are required. Thus it turns out that while a chosen set
of primaries may have zero intensities over considerable portions of
their spectra, the appropriate distribution functions of wavelength
are never zero (within the visible spectrum) except perhaps at a few
singular wavelengths.
The various colorimetric relationships consequential to the two
basic laws and necessary for the proper utilization of the numerical
data are usually expressed in a rather cumbersome notation which
requires the frequent writing of triplets of symbols and formulas of
similar type. Since these relationships are given here in some detail,
a more convenient notation has been adopted. Fortunately this nota-
tion is not new. It has been taken from the absolute calculus and
should be quite familiar to those acquainted with tensor analysis. It
will be described here from the point of view of its special application.
The first colorimetric relationship is an obvious deduction from
law 2, and may be stated as follows:
Relation 1 — If two colors match, and their intensities are changed
by the same factor, the resulting colors also will match.
Accordingly, it is convenient to write an appropriate factor ex-
plicitly. Let S be a sample source which is being matched in the
colorimeter, as by a set A. of three primary sources, A\t A^ and A*.
If E is the total energy intensity of the sample as it illuminates the
field of the colorimeter, let Elf1 be the corresponding intensity of
1948 COLORIMETRY IN TELEVISION 621
the AI, primary source as it effects the match with S. Then I?1 is the
relative intensity of the AI primary, or the actual intensity if the
sample had unit intensity. The superscript identifies the primary
source effecting the match and the subscript identifies the sample
source matched. If the latter is monochromatic, of wavelength X, it
is more convenient to write the relative intensity as I\l. For the
given set of primaries, the three quantities /x1, I\*, I\', depending
on the wavelength of the monochromatic sample, are thus functions
of X. To have the notation consistent, since the spectral intensity
distributions of the various sources are also functions of wavelength,
they will also be represented by letters with X as an index. Again
it is convenient to separate the total intensity and the relative spec-
tral distribution, so the spectral intensity distribution of the sample S
is given by ED$, while the relative spectral distribution of the pri-
mary source AI, or its actual intensity distribution when adjusted to
give unit total intensity, is given by DA,. X is written here as a
superscript and when in connection with /, as a subscript, in part to
show the different dimensionality of the two quantities.
Consider now sets of three primary sources AI, A 2, A3, and ft,
BZ, B3, etc. By a, or by a', is denoted any one of the sources AI, A2,
A3, and by b, or &', any one of ft, Bz, B3. Thus any relation involv-
ing a is to be read first with a = AI, then with a = Az, then with
a = As, and so on. Thus /? stands indifferently for any of the three
relative intensities if1, if*, if', and any equation written in terms
of a holds whether a is replaced by AI (the same throughout the
equation), by A 2, or by A3. Similarly, when X appears as an index,
the equation holds for all values of wavelength. In this manner
there are nine quantities 7? denoting the three sets of each three rela-
tive intensities of the A primary sources matching the B sources and
there are nine different quantities /£ for the relative intensities of the B
primary sources matching the three A sources.
The usual summation convention is also adopted for this present
notation : whenever, in a product, the same index appears twice, the
expression is read as the sum of the several like expressions in which
this index is replaced successively by each of the values it can assume.
For example, EL%R is read a$tE(D\jfl + D\tf* + D\jf>).
Thus there are 81 quantities IJ/J' (with regard for the summation
convention the primed and unprimed indexes are not considered the
same, even though they refer to the same primary set), 9 quantities
7J/a', and 1 quantity /£/*. In the event that X is a repeated index,
622 CHERRY December
since X can assume a continuum of values, the notion of summation is
replaced by that of integration with respect to wavelength. For
colorimetric purposes it is sufficient that the limits of integration in-
clude all visible wavelengths, but the original notion of total energy
intensity must likewise be confined to the same limits, about 380 to
750 m/A*. In dealing with photoelectric devices in connection with
colorimetry it is therefore most essential to ensure that the physical
apparatus is also thus confined. From the definition of D$ as the
relative spectral distribution, it follows accordingly that
/"750mu x
/ D*d\ = 1
y380mM
for any source S. From the present convention a similar integral is
formed by the repetition of the index X, for
Relation 2 — Since any sample color of relative spectral distribution
D* may be regarded as the combination of all the monochromatic
colors with appropriate relative intensities in that distribution, while
the matches to the latter are given by the three functions I\ the
combination of these matches is the match to the original color, by
application of law 2. In terms of the adopted notation, this color-
imetric relationship may be expressed as 7? = I\D* . Should mem-
bers of a primary set B be employed as samples, this expression would
be written as 1% = IJD£ .
Relation 3 — If a sample source S at intensity E is matched by a
set of three A primary sources with intensities EIaS} and if each of
these primariesjwhen adjusted to unit intensity, is separately matched
by a second primary set B with relative intensities Iba, then the sample
S can be matched directly by the B set with intensities EIbs where
Ibs = Ibalas. Replacing S by X for monochromatic samples, one finds
also that /J = /£/£.
Relation 4 — A consequence of the above relations is that if each of
primary set A with unit intensity is matched by a primary set B
with relative intensities Iba and if each of the set B is matched by the
primary set A with relative intensities /?, the two sets of intensities
are related by the equations Illba' = d"' where the symbol baa>t
the Kronecker delta, is 1 or 0, depending on whether a *= a' or a 9^ a',
respectively. The three systems of three simultaneous linear
* mn = millimicron = 10 ~ft meter.
1948 COLORIMETRY IN TELEVISION 623
equations can be solved either for the II or the Iba. For example,
_ IB** IB*** -IB* IB*'
'* ~w\-
where, in the determinant, the superscript denotes the row, the sub-
script the column.
Relation 5 — It follows either from relation 4, or more directly from
the uniqueness rule expressed in law 1, that if two identical primary
sets A are matched with each other, Iaa> — tfa>
Relation 6 — The quantities Iaa' in relation 5 may be computed with
the aid of relation 2, and take the form /£/ = /JD^ = 5J/.
Relation 7 — The condition that a set B of three sources be primary,
that no one can be matched by a combination of the other two, can
of course be 'applied to the matches by set A of each of the set B.
Thus IBI must not be a linear combination of /#, and /#„ that is,
that there be no nonzero solutions xb to the three simultaneous equa-
tions Ilxb = 0. This requires that the determinant of the coeffi-
cient of xb be other than zero : \I I \ 9* 0.
Relation 8 — The luminosity of a color is not defined solely in terms
of colorimetric matches, but some of the most important properties of
the well-known luminosity function or curve of eye sensitivity versus
wavelength are contained inherently in the standard colorimetric
data. This is by virtue of the properties believed of the luminosity of
a color, that it is a single numerical quantity, the same for all color-
imetric matches, and that the luminosity of a combination of colors
is the sum of the luminosities of the components. The luminosity L\
of unit-intensity monochromatic colors is in consequence a linear
combination of the functions /\ where the A sources are any primary
set. This can be seen easily if the luminosity of, a monochromatic
color is expressed in terms of the luminosities of the matching primary
set which in turn are the sums or integrals of the luminosities of
their monochromatic components: Z/x — I\D^L^>. The expression
on the right is obviously a linear combination of the three functions
/x- L\ is also known as the standard visibility function.
Luminosity is one example of the several visual perceptions con-
nected with the spectral distributions of the light received yet recog-
nized as common to, or possessed in different degree by different
distributions which are not colorimetric matches. Another such per-
ception, is that of hue, another that of chroma or saturation. They
are properties of color possessing a certain degree of uniqueness with
624 CHERRY December
respect to other properties and yet not always capable of objective
operational definition. For instance, the ability of hue identification
and discrimination differs very greatly among individuals and at the
same time is very much a matter of experience and training. Con-
ceivably it may be an entirely acquired visual ability, learned from
ordinary experience with the dilution of dyes and pigments by white.
Of these visual perceptions, luminosity appears to be the best defined
and established. Large luminosity differences between colors are
easily recognized, even though the colors are otherwise very different.
When the relative spectral distribution is unchanged, luminosity
varies directly with intensity and this as a simple proportionality.
There is of course substantiating evidence for this assumption. For
example, the luminosity, measured by comparison with a steady
source of the same relative spectral distribution, of light subject to
rapid time variations, is the same as the time average of the in-
stantaneous luminosities. The homochromatic comparison of the
luminosities of the light from two fields may be done very easily in a
colorimeter or similar instrument, which is then regarded simply as a
photometer. Heterochromatic photometry is substantially more dif-
ficult, and since exact comparison when the colorimetric properties
of the Sources are widely different is almost impossible, the measure-
ments are usually done in a step-by-step process in which a series of
slightly different sources are compared. Thus in the determination
of the luminosity-versus-wavelength function, two monochromatic
lights of almost the same wavelength are equated in luminosity by
adjusting their relative intensities until their visual differences in
the colorimeter appear to be minimized. Then this procedure is
repeated, using one of the monochromatic lights already used, and
so on down the wavelength scale.
Although the procedure of step-by-step comparison outlined above
constitutes the necessary operational definition of luminosity, from
which it is subsequently found that the luminosity function is a
colorimetric function in accordance with relation 8, a very much less
tedious method of heterochromatic photometry is available in the use
of the flicker photometer. Here the two light sources to be compared
are alternatively presented in the same field of view and their intensi-
ties are adjusted so as to minimize the critical flicker frequency. This
is the frequency of alternation of the field at which, as the frequency
increases, the flicker arising from the alternation just disappears.
While this method is subject to the criticism that flicker perception
1948 COLORIMETRY IN TELEVISION 625
is a separate phenomenon, connected primarily with time variations
and not necessarily at all with luminosity perception, and further-
more, flicker perception is due, at least in part, to the rods in the
retina, which are perhaps inactive in color vision, nevertheless, the
flicker photometer has been considered by many as giving good re-
sults in heterochromatic photometry.
With the aid of the colorimetric laws and relations which have been
given, all that is necessary for the codification of colors for reproduc-
ing and matching is the numerical values of the functions I\ for
some definitely specified set of primaries A which are accepted as
standard. In accordance with relation 2, therefore, all spectral-
intensity distributions D* would be uniquely characterized in their
colorimetric properties by the appropriate sets of three quantities
I*. Since both the functions I\ and D* are empirical in origin they
are not conveniently represented by simple functions and the process
of computing the integrals I\D$ is usually one of numerical inte-
gration. For this procedure it is very inconvenient to have negative
values of I\ and since the choice of the standard primary set is
arbitrary in the first place, it would be best if a choice could be made
which would, if possible, avoid these negative values. Furthermore,
the luminosity function is a linear combination of the general expres-
sions 1° and it is also always positive. Hence the total luminosity
can be a very convenient extra dividend obtained when computing
the 1° values of a color, if the luminosity function is itself one of
the I\. Now supposing the actual color data to be taken with a
definite set of primaries B, by relation 3, I\ = 7?/x. The quantities
II are determined uniquely from the quantities Iba by the equations
of relation 4, 1 '£/£/ = §£'. The quantities Iba, on the other hand, are
fixed by relation 2, Iba = I\Dl. If the functions Z>£ are chosen arbi-
trarily it is possible to assign any set of values to the quantities II.
Thus it is at once possible to choose 7? so that I\ 2 is the luminosity
function. Since the latter is always greater than zero for visible
light, as has already been pointed out, two more independent com-
binations of the /x can be found which are nonnegative. Of course
the values at either end of the visible spectrum are all always ap-
proaching zero, and thus in particular it is possible to find at least
one such independent combination* which approaches zero somewhere
* In the neighborhood of 503 m/* the X function comes near but does not quite
touch zero. This facilitates computation and saves co-ordinate space in the
chromaticity diagram by bringing the curve of spectral colors close to the y axis.
626 CHERRY December
toward the middle of the visible range. In addition, the function /£
may be multiplied by appropriate numerical factors so that all three
'380mM
have the same value. Hence, a light source whose in-
tensity is independent of wavelength, called the standard "E"
illuminant, will have tristimulus values all of the same numerical
value. It so happens in the colorimetric data that for wavelengths
greater than about 550 mju, one of the functions, say if ', can be ex-
pressed as a linear combination of the other two, to a very good ap-
proximation. It is therefore convenient to choose the /\ so that the
residual third function is approximately zero in this wavelength region.
The above considerations have been taken into account and the
choice of the standard distribution coefficients I\ has then been fixed*
by the specification of the trichromatic coefficients of a certain three
monochromatic wavelengths and the standard "B" illuminant, which
has a specified spectral distribution that, incidentally, gives a white
color. The conventional designation for the distribution coefficients,
which have been standardized and tabulated is, x, y, and z, the y func-
tion (of wavelength) being also the luminosity function (Fig. 1).
It is not possible to state exactly what the spectral-intensity distribu-
tions of the three corresponding ideal primaries are because the distri-
butions are not uniquely specified nor required by the functions /£ or
x, y, z. Relation 6, that I$D^> — 6"', however, does give evidence of
one important condition on the D%, namely, that for all the I\ to be
positive, for some wavelengths the Z)J must be negative. Now since
no light source can have negative intensity, over any part of its
spectral distribution, it is clear that the ideal primary sources are
imaginary. To be sure, they can be obtained directly in the colorime-
ter by the artifice of negative addition but each primary would then
consist of two light sources. Fortunately there is no real need for the
direct use of these imaginary, ideal primaries. Their utility arises in
colorimetric tables, charts, and computations.
In order to secure a graphical representation of the colorimetric
relations and of various color differences and similarities, a color chart
in the form of a chromaticity diagram (Fig. 2) generally is used. While
* The transformation from existing experimentally derived distribution coef-
ficients to the standard distribution coefficients was thus specified by the
Colorimetry Committee of the Eighth Session of the International Commission
on Illumination, Cambridge, England, September, 1931.
1948
COLORIMETRY IN TELEVISION
627
many types of color charts have been invented with this object in
view, the standard chromaticity diagram is capable of a fairly com-
plete, objective representation although it cannot be used for direct
visual comparison with sample colors unless it be printed with an
appropriately shaded colored background. It is constructed from
the tristimulus values of the sample colors to be represented on it, and
since the diagram is two-dimensional it is appropriate to discount
from the color specification that factor having to do primarily with
WAVELENGTH
Fig. 1 — Standard tristimulus values of equal ener-
gies of the spectral colors.
intensity and preserve in the representation those quantities repre-
senting properties seemingly more intrinsic to the notion of color.
In this way the tristimulus values of a sample, which are the quanti-
ties EIas = EI^D^s when the I\ are x, y, z, and are denoted conven-
tionally by X, Y, Z, are formed into two ratios
X 7.Ai Y I,AI
X + Y + Z
X + Y + Z
These two ratios, called the trichromatic co-ordinates of the sample,
x and y, have values between 0 and 1, and together with the value of
628
CHERRY
December
y, the luminosity, completely specify the colorimetric properties of
the sample in the same way as do the quantities 7? . There is a third
trichromatic co-ordinate,
X + Y + Z
but since, always, z = 1 — (x + y\ it can usually be ignored.
Fig. 2 — International Commission on Illumina-
tion standard chromaticity diagram, showing the line of
spectral colors and of pure purples, the location of the
standard "C" illuminant, and Nos. 25, 47, and 58
Wratten filters in combination with that illuminant to
illustrate the reproducible triangle of a receiver with
these as primaries.
The chromaticity diagram is based on Cartesian co-ordinates with
respect to which are plotted the trichromatic co-ordinates of the
samples to be represented, the value of x along the x axis and the
value of y along the y axis. In this manner, all colors which differ in
visual characteristics only in luminosity correspond to the same
1948 COLORIMETRY IN TELEVISION 629
point on the chromaticity diagram. Every monochromatic line of the
visible spectrum has such a point and the points corresponding to suc-
cessive wavelengths turn out to be on a horseshoe-shaped curve.
The red side of this horseshoe is very nearly a straight line. Such a
property is to be expected when, as has been mentioned before, in this
wavelength region, one linear connection exists between the quan-
tities /t*
Relation 9 — One of the most important characteristics of the
chromaticity diagram, which is a simple consequence of its method of
construction, is that the point standing for the additive combination
of two colors lies at the center of gravity of the two points standing
for each of them, they being assigned weights proportional to their
respective quantities £/? or X + Y + Z. This notion similarly
a
applies to the combination of several colors.
It therefore follows that all colors which are real, meaning that
the spectral distributions of the light of the colors are made up of
positive intensity contributions of the various monochromatic wave-
lengths, must lie inside of the area bounded by the monochrome horse-
shoe and the line joining its two ends. Daylight and sunlight, which
each have roughly uniform intensities of light throughout the visible
spectrum, are in this manner represented by points lying near the
center of the horseshoe.
Relation 10 — In consequence of the center of gravity relation 9,
it is clear that the points representing a set of primary colors on a
chromaticity diagram are necessarily noncollinear. Any color
matched by a positive combination of such a primary set, whatever
the relative intensities, must lie on or within the triangle formed by
the three points representing the primaries.
*This linear relationship must be homogeneous, that is, contain no additive con-
stant, since all tristimulus values pass to zero in the infrared. Thus the linear
interdependence of the x and y for yellow and red monchromatic lights may be
shown through algebraic manipulation of their formulas. In the case of the
standard tristimulus values, that z = 0 for these wavelengths makes this linear
interdependence particularly easy to see. Since the homogeneous linear rela-
tionship may be written as Z = aX + bY, the formulas for x and y become
(a + 1) x =
(a + D x + (&
(g + i)++ 1) Y
the last equation showing the linear interdependence of x and y.
630 CHEEKY December
Some additional properties of the chromaticity diagram may be
noted as follows. The purity of color, corresponding very closely to
the visual perception of saturation or chroma, is intended to indicate
how nearly monochromatic a color is, or conversely, how degraded
with white. If a straight line is drawn from the point on the chro-
maticity diagram representing white, for which the standard "C"
illuminant is usually taken, through the point associated with a given
color, on to meet the monochrome line or line of spectral colors, the
purity of the given color is denned as the ratio of the line segment
between white and the color, to the total length. If the given color
is a purple, it is necessary to close the diagram by joining the two ends
of the monochrome line with a straight line which may be thought of
as representing the pure purples. The meeting point with the mono-
chrome line mentioned above indicates a particular wavelength
which is known as the dominant wavelength of the given color sample.
In the case of purples this wavelength is not so denned, but rather is
taken at the intersection with the monochrome line of the line from
the given color extended through white. This corresponds, as will be
seen later, to the dominant wavelength of the color complementary to
the given purple, and therefore is distinguished by the suffix c.
From the properties of the diagram already given, it can be seen
that the combination of white and monochromatic light of the domi-
nant wavelength, or its equivalent negative combination of white and
the complementary dominant wavelength in the case of purples, when
in amounts indicated by the purity and the total luminosity, will
provide an accurate match of the given color. Obviously therefore,
the location of a color on the chromaticity diagram is equally well
specified either by its trichromatic co-ordinates or its dominant wave-
length and purity.
In connection with judgment as to the approximate reproduction
of colors, a very useful representation of color tolerances may be
made on the chromaticity diagram by plotting local contours of least
chromaticity differences perceptible to the normal eye.
The combination of colors as it is understood in colorimetric re-
lationships and data and as it is expressed on chromaticity diagrams,
involves the combination of light fluxes impinging on the eye and has
nothing whatever to say about the manner in which the spectral-
intensity distributions were secured originally. The artifice of nega-
tive addition of light has referred merely to the addition of light flux
to the sample side of the colorimetric balance, rather than to the
1948 COLORIMETRY IN TELEVISION 631
matching side. Unfortunately, from the point of view of terminology,
the notions of primary colors, color combinations, and charts have
been applied to a large group of processes associated with color
production which themselves have received a collective misnomer,
subtractive processes. They are concerned with the ways in which
dyes and pigments modify the light impinging on them and the conse-
quences of mixing and combining different pigments and dyes. For the
present brief discussion of subtractive processes, the actual atomic
properties which give rise to the color characteristics of these sub-
stances will not be considered but the spectral reflectances and trans-
mittances will be assumed.
The spectral quality of light passing through a dyed medium or
incident on a dyed surface is altered by the selective absorption of
some wavelengths of the light. The characteristic transmittance of
the dye, like the reflectance of a pigment, is a proportional function.
While the light which is absorbed by the dye may be thought of as
subtractive (hence the term subtractive process) the subtraction is
really logarithmic, since the amount subtracted is in proportion to
the amount initially present. When two dyes are mixed, the same
ray of light is acted on by both of them and the resultant transmit-
tance is the product of the separate transmittances of the two. Thus,
even if the illuminant is specified and the light passing through a defi-
nite quantity of a given dye may therefore be assigned a unique set
of tristimulus values, the light which passes through a combination of
two dyes is in general totally different from an additive combination
of the light passing through each of the dyes separately.
There is no unique relation between the colorimetric specification
of the original dyes and of their combination. Their mixture char-
acteristics depend on the entirety of their original spectral transmit-
tances, are colorimetrically nonlinear, and do not obey the colori-
metric relations for the combining of colors. For example, there is no
single point on the chromaticity diagram which corresponds to a
given dye, even with a specified light source unless the quantity of
dye passed through is also specified. The light transmitted by zero
density of dye has, of course, the trichromatic co-ordinates of the
source, that is usually in the white, and as the density of the dye
increases, the point moves outward along some generally curved path.
With the density increasing without limit, the point approaches the
position of the monochromatic light of wavelength corresponding to
the peak transmission of the dye, although the total amount of light
632 CHERRY December
transmitted becomes vanishingly small. Another example of the
irregularities of the so-called subtractive systems is encountered when
mixing a dye with each of two dyes which at the given density have
identical appearances. Notwithstanding the colorimetric equiva-
lence of the latter two dyes in combination with the chosen white
source, the final mixtures may yield entirely different colors.
While for the most part pigments or pigmented surfaces share in the
subtractive properties of dyes, their properties in mixtures are more
complex because the modification of the light by absorption is ac-
complished not only during direct transmission but also during mul-
tiple reflection and diffusion among the pigment particles. Conse-
quently the mixture characteristics of a pigment cannot be com-
pletely specified by even an entire spectral-reflectance curve. If the
pigment particles of a mixture are of such a nature that an element
of incident light must, on the average, pass through or be reflected
by many particles before it is 'returned to the surface and emitted
in its modified or attenuated form, then the mixture characteristics
of the components will be substantially the same as those of dyes.
If, however, the particles are so opaque or so sharply back-reflecting
that a single particle can re-emit from the pigmented surface as
modified, most of the light incident on it, each part of the mixture
contributes additively to the emitted light.
Although of perhaps more widespread use, the notion of primary
colors in subtractive processes is much less precise and has more
limited applicability than has been used here for the additive com-
binations. Merely relative differences in the logarithmic trans-
mittances of three dyes are not sufficient to ensure that all colors can
be matched, while in the subtractive processes there is no useful
analog to the negative combinations in the colorimeter of the additive
systems. Through this artifice of negative addition an unlimited
number of different primary sets were possible in the additive com-
binations, even though a comprehensive gamut of colors could be
matched with only positive values by relatively few sets, such as a
high-purity red, a high-purity green, and a high-purity blue. If the
same gamut of colors is to be covered in a subtractive process, these
high-purity primaries themselves must be attainable. Ideal dyes
may be conceived of, which would achieve this, and accordingly are
often identified with the manner in which they control the effective
amounts of appearance of the additive primaries. Thus with white
light at the start, the dye which controls the amount of blue is the
1948 COLORIMETRY IN TELEVISION 633
one which removes blue from the white light, hence "minus blue."
Applying the rules of additive mixtures to white light and a negative
amount of blue, we see from the chromaticity diagram the resulting
color is yellow. Hence a yellow dye which has that color by virtue of
fairly uniform high transmission in the red and green parts of the
spectrum and deep absorption in the blue is capable of being the
"minus blue" subtractive primary. Similarly the "minus red" is a
blue-green or cyan dye, and the "minus green" is a magenta or purple,
which transmits red and blue light and absorbs green. Because the
subtractive characteristics rarely approach the ideal of sharp spectral
selectivity and because pigments sometimes show a little of the addi-
tive characteristics besides, the common primaries in pigments are
generally blue, yellow, and red.
The color produced by a "minus" dye in combination with white
light and the color controlled by this dye in subtractive processes are
said to be the complements of each other. An additive combination
of the two produces white. They appear conjointly in the well-
known eye-fatigue phenomenon in which a bright design of one color
is viewed fixedly against a white background, and then a blank white
surface is viewed, wherein the same design in the complementary
color appears for ft time as an illusion. As has been indicated before,
what is called white requires an arbitrary standardization, but all the
properties of complementary colors are consistent if the same white is
referred to. From this point of view, the dominant wavelengths of
two colors determine their complementary properties, although the
artistic values of this relationship usually refer to complementary
properties of hues, there being an almost one-to-one correspondence
of hue and dominant wavelength.
With regard to subtractive processes it must be remembered, how-
ever, that all true colorimetric specifications are referred to additive
combinations of light sources, and that the conditions on additive
combinations must always be satisfied to achieve color matches
whether the intermediary steps by which the colors are produced are
additive or subtractive. The simultaneous television receiver is an
excellent example of the additive method of forming color with super-
posed light from three different light sources.
COLOR REPRODUCTION
From the foregoing section on colorimetric matching it should be
clear that the visual identity of colors may be stated uniquely in terms
634 CHERRY December
of the identity of their tristimulus values. Therefore the function of
a color-reproduction system is to provide colors with the same tri-
stimulus values as occur in the original. While reproducing the
original spectral distributions would indeed satisfy this condition,
such a procedure is clearly not necessary, even for exact color repro-
duction, and it is not possible in a practical sense. Admittedly a
three-color system which is exactly right for one person may not be
quite perfect for someone else, and in addition engineering tolerances
may prevent exactly correct reproduction in any case. In general,
however, for any three-parameter system, the precision of color re-
production possible on the basis of colorimetric matching is vastly
better than can be achieved by any method of approximately re-
producing spectral distributions. Therefore, in all systems of color
reproduction, in photography, printing, and color television, and
wherever a three-color system is used, no attempt is made, nor would
it be desirable, to approximate the spectral distribution of the original
colors. This is an important example in communications of the selec-
tion of physical stimuli of markedly different physical specification from
that of the original in order to give rise to the same sense perception.
In terms of the facts and relationships of colorimetry, to achieve
accurate color reproduction in a color-television system, the correct
signals must be expressed in the light intensities of the three color
sources in the receiver and this requires, in effect, that the television
system operate as an electronic colorimeter, individually for each
picture element. Assuming linearity of the three receiver primary
light sources, or appropriate correction for nonlinearity, the electrical
signals controlling these sources must, with two exceptions, corres-
pond, when the transmitting camera is viewing a color sample S,
precisely to the quantities EIbs, which have been discussed in the pre-
ceding section, and represent the intensities when matching S in the
colorimeter of the three primary light sources B which also are the
receiver light sources, i.e., kinescope phosphors plus light filters.
The first of these exceptions is merely that in conventional tele-
vision practice no attempt is made to duplicate in the receiver pre-
cisely the same brightness level as occurs in the scene being televised,
but the adjustment of scene brightness is left to the selection of the
individual viewer. Thus the quantities EIbs may be in error by some
proportionality factor common throughout the picture.
The second exception arises from the difference in purpose of color-
imetric matching and color reproduction. In the former the securing
1948 COLORIMETRY IN TELEVISION 635
of quantitative relationships was the prftnary purpose, and the modi-
fication of the visual effect of the sample by the negative addition of
the primaries was quite admissible. In reproduction the sample is
not available to be so modified nor is such desired, for the sample
should be reproduced like the original. Therefore complete accuracy
of reproduction will be confined to those colors which require only
positive combinations of the three primaries. As was shown before,
these colors lie on or within the triangle on the chromaticity diagram
determined by the three primaries and at the same time their luminosi-
ties are limited to the luminosity range of the receiver primaries.
Therefore in the design of - television receivers it is of considerable
importance to select receiver primaries which will cover the desired
gamut of accurately reproducible colors. In the present three-
kinescope-type simultaneous receivers this is very largely a problem in
phosphor composition. Fortunately phosphors are available with
very satisfactory spectral characteristics although there is ever a need
for more luminosity. Some of these phosphors, in combination with
optical filters, are capable of saturation so nearly approaching that of
monochromatic light, at well-separated points of the chromaticity
diagram, that simultaneous receivers can accurately reproduce the
colors from virtually all natural and artificial dyes and pigments.
They are not able to reach pure monochromatic colors such as a
spectroscope will produce but they can easily reproduce rainbow
colors which are, as usual, diluted with a small amount of white light.
The available color gamut is greater than in most other color-repro-
duction systems and appears adequate.
Within the range of reproducible colors, the function of the camera
and transmitting equipment is to derive from the light of the sample
S the quantities EIbs and impress them on the primary sources at the
receiver. The unique physical specification of the sample S is its
spectral distribution ED* which is all the signal information available
to the camera. Conventional ideas as to how the camera should ob-
tain the proper signals from this spectral distribution have, until re-
cently, followed the pattern that each of the three signals EIbs should
be obtained by a separate operation on the light ED*. Thus in the
sequential color camera the plan was for the scene first to be scanned
and the "red signal" Elf1 to be derived for all elements of the picture,
then for the scene to be scanned with the "green signal" Elf 2 derived,
and so on. In the simultaneous color camera the plan was for all
three of these operations to be done at the same time but each by a
636 CHERRY December
separate camera tube. That, as they are, these methods are funda-
mentally unworkable has not always been appreciated, and when it
has, the attitude has been largely to do as well as possible with the
system and accept the compromise. While not wholly responsible
for this unfortunate viewpoint, two erroneous ideas have contributed
to it and it is well to repudiate them explicitly before describing
systems which will fully accomplish the desired result.
The first of these misconceptions is a consequence of attempts to
approximate the spectral distribution of the original color. This dis-
tribution was to be divided into three adjacent blocks, roughly red,
green, and blue, by being scanned through three filters with rectangu-
lar spectral cutoff characteristics. With the three receiver primaries
having similar rectangular spectral distributions and being controlled
each by one of the signals derived from scanning, the reproduced
spectral distribution would approach the original in the manner of a
block diagram approaching a continuous curve. If it were otherwise
feasible to increase the number of blocks by having perhaps ten or
more receiver colors and signal channels, color reproduction by this
method would have good quality, but with only three parameters
available, the results are necessarily far inferior to colorimetric
reproduction.
The second common misconception involves some concession to
colorimetry in that it is recognized that spectral-sensitivity curves
peculiar to the nature of the receiver primaries must be obtained in
the color camera. The mistake lies in supposing that the weight by
which light of a certain wavelength is to be counted in the camera
must be the same as it appears in the receiver primary; in other
words, that the contours of the spectral sensitivities of the camera
should be the same as the spectral-intensity distributions of the re-
ceiver colors which they each control. Actually the correct sensitivi-
ties are the functions /x, which scarcely resemble the primary distribu-
tions Dl. If the reasoning of colorimetry is to be abandoned momen-
tarily for the sake of a crude intuitive explanation, the reason that the
above idea is wrong can be seen from the fact that a combination of
the light from the receiver primaries can give the same visual im-
pression as light whose wavelengths are totally absent from all these
primaries.
The fundamental idea as to how the color camera is to derive the
three signals EIbs is based directly on the colorimetric relations by
which these same quantities may be computed, namely, by a process
1948
COLORIMETRY IN TELEVISION
637
of multiplication and then integration with respect to wavelength.
Thus EIbs = ED*I^, with the integration with respect to wavelength
indicated by the repeated index A. This process might be performed
very easily by a photosensitive camera tube, for its total response
is the integral of its response at each wavelength, and this spectral
response is characteristic of the nature of the photosurface. The
incoming light distribution may be further modified by a multiplica-
tive function which is the transmission characteristic of an optical
filter so that the effective sensitivity T{, where c may stand for cam-
Fig. 3 — Transmitter camera sensitivities required for
the primaries of Fig. 1, assuming that each receiver
primary is separately derived. (The curves show
negative values for some spectral wavelengths.)
era tube and filter combinations Ci, €2, or €3, is the product of the filter
transmission and camera photosurface sensitivity. Where, as in
usual ideas for either the simultaneous or sequential color camera and
transmitting equipment, each camera tube and filter combination has
sole control over one receiver primary, so that each C may be identi-
fied with one B, the effective sensitivities may be written as TX.
To satisfy the unique colorimetric requirement on the camera,
therefore, it must be that T\ — /*, except for constant propor-
tionality factors which apply in effect to the amplification gains in
the three channels. Herein lies the difficulty with the idea of the
separate derivation of the three signals. The functions I\ necessarily
638 CHERRY December
must be negative over some part of the wavelength spectrum (Fig. 3) .
This is a logical consequence of the colorimetric relationships and
is most easily seen in relation 6, where it must be remembered that
the receiver primaries here considered are perfectly real light sources
and have only positive spectral intensity distributions D$. Thus, for
example, the requirement that
/"750 R ^
/ I?1D o d\ = 0
7380 >
means, since D\ is always positive and of appreciable amplitude over
a considerable range of wavelengths and since if 1 is, as shown before,
a linear combination of the linearly independent functions x, y, and z,
and hence can be zero only at a few discrete points, that I?1 must have
some negative values. To provide an intuitive background for this
fact it should be remembered that on the chromaticity diagram, colors
outside the triangle of a primary set require negative amounts of at
least one of the primaries for a match, and monochromatic colors come
under this category. The television camera operates along this pro-
cedure, that it finds out the amount of the receiver primaries re-
quired to match each of the monochromatic wavelengths in the sample
and then adds up the result, and thus along the way these negative
amounts have to be taken into account even though the final sum
represents a color which is well within the triangle of the primaries.
While by the choice of suitable optical filters it is possible to ap-
proach the positive portions of the curves /£ with the camera sen-
sitivities T*, as yet no camera tubes have been developed which
yield, in effect, negative photoresponse in some wavelength intervals.
Hence it has not been possible to realize the negative portions of the
/£ functions which are essential to accurate color reproduction when
the three receiver signals are derived separately. The frequently
suggested remedy, that a constant or bias be added to the I\ functions
so as to make them positive, the camera sensitivities be adjusted to
the result, and then subsequently a constant signal be subtracted, is of
course useless because the "constant" signal is a function of the in-
coming light characteristic. On the other hand, if there were pro-
vided another camera tube for each of the ones in the conventional
arrangements, such that the effective sensitivity of one corresponds
to the positive part of the /x, and the sensitivity of the other to the
negative part, the differences of the outputs would give precisely the
desired signals.
1948 COLORIMETRY IN TELEVISION 639
In contrast to the cumbersome, objectionable ways available to
attain the camera spectral sensitivities I\ necessary to accurate color
reproduction when the receiver signals are each derived by a separate
photopickup arrangement, an elegant method of deriving the signals
in simultaneous systems becomes evident upon a more thoroughgoing
application of the colorimetric relationships. From relation 3 it is
apparent that the I\ may be expressed as independent linear com-
binations of analogous functions applying to other primary sets, as
for instance the C set of primaries later to be identified with the
separate camera tubes: I\ = Ibcl\. Similarly the I\ may be ex-
pressed as linear combinations of I\ where the A set of primaries are
the standard ideal set and hence the I\ are also the conventional
x, y, and z. Now if three camera tubes with effective sensitivities
T{ = /x are used to scan the sample color, their separate outputs
will consist of the integrals EI£DXS just as in the conventional schemes
in which the receiver contained the primary set C. These signals
then are combined by algebraic addition in three different ways by
linear networks in accordance with the three transformations denned
by the coefficients Ibc which are fixed into the network by potentiom-
eters or the like. The output signals thus have the magnitude
J%EI\Ds which are exactly the same as the desired signals EI\D*.
Clearly the C set of primaries have played an entirely intermediate
role and they do not have to be actually realized. There is no re-
quirement therefore that their spectral-intensity distributions D*
be entirely positive quantities and consequently there is no reason
why the functions 7\ have to have some negative values. The only
requirement on the I{ is that they be independent linear combina-
tions of the positive functions z, y, and z. Hence there is com-
parative freedom to choose these functions I\, first so that they are
positive and make possible the design of optical filters which will
give the proper camera sensitivities T{ = /x and yet so that the
subtraction of signals which occurs in the combining networks repre-
sents, at least for white light, a comparatively small correction on
the main signal strength going through each channel. This is asked
because in subtraction, signal strengths may be diminished but noise
always increases. Finally, in the interest of convenience the com-
binations of x, y, and z should be chosen such that, when the spectral
sensitivities of the phot osurf aces are divided out, the resulting filter
characteristics which are required can be secured with obtainable dyes.
In the design of the light filters to provide the camera tubes with
640
CHEERY
December
73
2
a
2
o
05
1
O
I
I
C
O
1948 COLORIMETRY IN TELEVISION 641
the effective sensitivities T\ which have been selected, additional
characteristics are frequently incorporated to compensate for the
light source illuminating the subject being televised. It very often
happens that the light source is limited in its spectral characteristics
by other considerations, as, for example, in the flying-spot slide and
motion picture scanner where the phosphor must be of high inten-
sity and very short decay time, and where the slides and motion pic-
tures are ordinarily intended for projection with light from a tungsten
lamp. Rather than applying an optical filter directly to the light
from the phosphor, a very substantial saving in intensity can be ob-
tained by modifying the effective sensitivities of the camera tubes
so as to give the equivalent over-all characteristics as if the proper
illuminant were used.
The network used for the algebraic addition of the camera signals
usually requires some means of phase inversion to secure at the same
time signals of both polarities. (See Fig. 4.) The combination of
the signals is then easily accomplished by purely resistive elements
which may have already fixed in them the appropriate constants Ibc
as computed from colorimetric relations 2 and 4 and the camera sensi-
tivity and receiver-intensity distributions J\ and £)&• For reasons of
avoiding these calculations and of avoiding the need for reliance on
precision in the circuit components, it is perhaps better to provide
adjustable circuit elements which can be set when the system is in
operation. In spite of the large number of variables, usually nine or
more, a rapidly converging procedure of adjustment can be secured
through the obvious facts that the transmission of white must yield
white and that the camera, when viewing a light source* of the same
tristimulus values as any one of the three primaries in the receiver,
together with the combining networks, must yield a signal in the
corresponding channel controlling that primary, and in no other
channel. The adjustment of the white requires a balancing of the
three signals output to the receiver and for this purpose it is more
convenient to provide separate gain controls in the output channels
even though there is then a duplication of variables.
As a result of the rigorous application of colorimetric information
to simultaneous color television which is admirably well suited to the
purpose, a new medium for the reproduction of color is becoming
available. It is capable of by far the finest performance yet known
in commercial processes, having at the same time a wide gamut of
* This source may conveniently be a dummy receiver.
642 CHERRY
colors, colors of very high saturation, and an 'intrinsically accurate
means of adjusting these colors automatically.
BIBLIOGRAPHY
(1) W. D. Wright, Researches on Normal and Defective Colour Vision, Kimp-
ton, London, 1946.
(2) Massachusetts Institute of Technology, "Handbook of Colorimetry," The
Technology Press, Cambridge, Mass., 1936.
(3) A. C. Hardy and F. L. Wurzburg, "Theory of three color reproduction/"
/. Opt. Soc. Amer., vol. 27, pp. 227-240; July, 1937.
(4) Simultaneous all-electronic color television," RCA Rev., vol. 7, pp. 459-
468; December, 1946.
(5) R. D. Kell, "An experimental simultaneous color-television system — Part
I, Introduction"; G. C. Sziklai, R. C. Ballard, and A. C. Schroeder, "Part II,
Pick-up equipment" ; K. R. Wendt, G. L. Fredendall, and A. C. Schroeder, "Part
III, Radio-frequency and reproducing equipment," Proc.I.R.E., vol. 35, pp. 861-
875; September, 1947.
Origins of the Magic Lantern
BY J. VOSKUIL
RESEARCH CHEMIST, GELDERMALSEN, HOLLAND
Summary — A critical survey of old and new literature reveals the develop-
ment of the modern slide projector out of the old "art of mirror writing"
which in its turn can be derived from the silhouette. A wrong interpreta-
tion of a passage in an old book which describes a camera obscura caused
the wrong opinion that the slide projector must have developed from the
camera obscura. This contrivance, however, has been the forerunner of the
modern photographic camera.
IT is USUALLY HELD that the slide projector, formerly called the
"magic lantern" has its origin in the "camera obscura" and in
this .connection the names of Porta (1538-1615) and of Athanasius
Kircher (1602-1615) are mentioned. The latter was alleged to have
constructed the magic lantern in the middle of the seventeenth
century. He described it in the second edition of the voluminous and
abundantly illustrated "Ars Magna Lucis et Umbrae" ("The Great
Art of Light and Shadow," 1671) which publication is accompanied
with two pictures, one of which is reproduced in Fig. 1.
A closer study of the literature of this subject, however, reveals an-
other origin of the magic lantern, which may be traced back to the
very old "silhouette show" and in this development the importance of
Kircher and Porta is not so great as is generally accepted.
Before we continue with the subject an explanatory remark should
be made on the principles of the camera obscura and the modern pro-
jector. The latter forms by means of a lens, the objective, a real in-
verted image of an object, which therefore can be projected on a screen.
The nearer the object (slide, film) to the focus of the objective, the
larger the image on the screen and the larger the distance between
screen and objective. Thus in slide and film projectors the slide
or the film is placed practically in the focus of the projecting lens.
By moving the object from the objective the image will become
smaller and smaller until it stands practically in focus when the object
is at a great distance from the lens. In this way we have changed the
projector into the camera obscura and therefore the essential differ-
ence between the camera obscura and the magic lantern lies in the
position of the object before the lens.
The Italian Porta, who lived long after the invention of the camera
obscura and thus is not the inventor as is often supposed, deserves,
DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51 643
644
VOSKUIL
December
however, the merit of having popularized it in his famous book
"Magia Naturalis" (first edition 1558, second edition 1589 in Naples),
a curious mixture of science and charlatanry. The result was a wide
application of the camera obscura, which in those days indeed had the
dimensions of a "camera" (room, see Fig. 2), as a contrivance for per-
formances of various character. In one of the walls a simple spectacle
lens was placed and a hollow mirror was used to reflect the images of
Fig. 1— Kircher's "magic lantern" (1671).
the objects outside the room in this lens; the pictures were thus pro-
jected on the screen, the opposite wall, right side up. As in the
modern theater, the spectators sat facing this screen with their backs
to the lens which was more or less hidden, making the performance a
mysterious affair.
On an open space in full sunlight outside the "camera" and before
the lens the different scenes were played. For instance, hunting
parties were very popular in which the game was represented by dis-
guised boys or wooden effigies. War scenes and passion plays were
also presented. At night statues and large pictures painted on canvas
lighted by torches were shown. In this way the public saw pictures
1948
ORIGINS OF THE MAGIC LANTERN
645
of the emperor, scenes of the rising and setting of the moon and stars,
and last but not least images of the devil to frighten the spectators
who still looked mpon the whole performance as an uncanny and
supernatural affair. It must be noted that in its application the
camera obscura came quite near the performances which were given
with the "laterna magica" some seventy years later.
When in addition the well-known English chemist Priestley (1733-
1804) in his work on the history of optics (1776) wrote that Porta also
used transparent drawings as "slides," the close connection between
the projecting lantern and the camera obscura seemed to be certain
Fig. 2 — The camera obscura in the sixteenth and seventeenth centuries.
and for a long time it was held that the former developed out of the
latter. As has already been mentioned a careful study of the avail-
able old literature shows that another development is more probable.
In Priestley's description Porta is said to have traced drawings on
transparent paper attached to one of the sides of a hollow cube the
opposite side of which was open and turned to the lens. From this
transparent drawing, placed outside of the room, an enlarged picture
was formed on the screen. The necessary light came from the sun.
By making the slide movable Porta is said to have been able to attain
effects which seemed positively uncanny to his contemporaries.
Priestley further supposed that the German Jesuit, Athanasius
Kircher, following up on Porta's device, later invented the magic
lantern (Fig. 1) which did the job of the camera obscura at night.
646 VOSKUIL December
Thus Priestley refers to the "Magia Universalis" (1657) a work of
Kaspar Schott, an assistant of Kircher. But in turn Schott refers to
the first edition of Kircher's "Ars Magna Lucis et Umbrae" of 1646 in
which the camera obscura was described, and after an explanation of
the apparatus Schott wrote ("Magia Universalis," volume 1, page
198, Wiirzburg 1657) : "In order that the spectator does not notice
the small hole with the lens so that the effect is more mysterious, one
attaches inside the room in front of the lens a hollow cardboard cube
' with blackened sides except the side turned to the lens which is open
and the opposite side which is made of transparent paper. On this
transparent paper one projects the image of cardboard objects placed
outside the room which are turned upside down in order to get the
pictures right side up on the screen . . . . "
Priestley thus made the mistake of supposing the cube to be at-
tached outside the room and supposing the screen to be a "slide" for
the drawings. Second, Porta was accredited with the technique,
which, however, was developed long after Porta by Kircher to whom
Schott refers. Owing to the authority of Priestley his mistake was
repeated in the historical works of Joh. Carl Fischer and Poggendorff
and from these in the more modern books.
Of course it may be imagined that the development of the magic
lantern from the camera obscura had taken place after systematic
research work on geometrical optics as is the practice today. By
moving the object from a point far off to the focus of the lens and by
constructing a device for optical lighting (condenser) which is a char-
acteristic feature in a projector, one had changed the camera obscura
into a magic lantern. But in Porta's days there was no question of
systematic research. All investigation was more or less guesswork, for
the greater part done by adventurers and charlatans who looked for
ways in which to deceive the credulous public and to make quick
money. As for the very small scientific world of those days, this was
only interested in the magnifying power of lenses and their use in
microscopes (Hooke van Leeuwenhoek) and telescopes (Galilei).
So the principles of optic projection were invented by way of trial and
error and in the just mentioned first edition of Kircher's "Ars Magna"
of 1646 we can find a good starting point for the historical develop-
ment of the slide projector, for on pages 907-917 we read about ex-
periments which, traced backwards, point to the "silhouette show"
and on the other side directly lead to the first magic lanterns of
Christiaan Huygens and Thomas Walgensten.
1948 ORIGINS OF THE MAGIC LANTERN 647
Kircher, who was at that time in Rome, carried out these experi-
ments because he was fascinated by the old "art of mirror writing."
From passages in Agrippa von Nettesheim's works on occult philoso-
phy (sixteenth century) we learn that this art is very old, even the
name of Pythagoras (500 B.C.) being connected with it. The ancients
seem to have experimented with a system of long-distance communi-
cation by writing on a plane or concave mirror which was reflected on
a screen placed at some distance. Thus a kind of optical telegraphy
was constructed to be used for messages to army leaders in battle or
for other emergencies. Von Nettesheim tells us the fantastic story
that Pythagoras, while in Italy, in this way communicated with his
friends in Byzantium. He wrote the letters with his blood and re-
flected the mirror to ... the moon!
Figs. 3b and 3c show schematically how the Ancients planned —
and perhaps put into practice — their "art." On a mirror Sp the re-
versed letters (represented by the figure F) were traced which did not
reflect the rays of the sun, thus forming shadowy figures on a screen S
to which the mirror directed these rays. In fact mirror writing really
is nothing but using a mirror to direct a certain shadow (Fig. 3a) to a
certain spot (Fig. 3b) . By using a concave mirror the Ancients tried to
get larger images (Fig. 3c) .
We learn in the primers on optics that the shadow of an object
which is lighted by a light of some dimensions becomes less sharp as
the distance between the object and the screen grows. The inner
shadow grows narrower and the penumbra broader. Consequently
the reflected writings are somewhat blurred at relatively small dis-
tance and badly blurred at greater distance. Now Kircher tried to
improve on this method by means of a lens~ That he chose a lens was
very probably not the result of scientific reflections (see the end of
this article), but of the fact that the lens, as an optical implement, was
becoming more and more popular. The seventeenth century in which
Kircher lived was the century of the rise of optical science and practice.
Snellius (1580-1626) had worked out his well-known law of refrac-
tion sin i/sm r = n by which a rational construction of optical instru-
ments had become possible. Christiaan Huygens (1629-1695) and
later Isaac Newton (1642-1727) published their famous treatises on
the nature of light and moreover constructed different optical ap-
paratuses. The study of microscopical objects and celestial bodies
went through a boom period and so it can be easily understood that
minor scientists like Kircher tried the lens as an improvement in their
648
VOSKUIL
December
Fig. 3 — Scheme of the development of the modern projection systems from
the silhouette.
y = object; s = screen; I = source of light; la = slide; fi = film; sp = mirror;
and o, le, co = lens (o as objective, co as condenser)
1948
ORIGINS OF THE MAGIC LANTERN
649
optical devices, boasting on priority when they had made an "inven-
tion." Moreover Kircher had read about experiments of another
Jesuit, the Italian Bettini (Marius Bettinus, 1582-1657) which, ac-
cording to Kircher might be very useful for his research. Bettini's ex-
periments can be found on pages 26 and 27 of the "Apiaria Universae
Philosophiae Mathematicae" (1642), which, translated freely, means
"a miscellany of mathematical philosophy." Under the heading
"Shadow Projection with the Lens" (Fig. 4), Bettini dealt with a
"secret method with which, during the night, one can communicate
Fig. 4 — Bettini's shadow projection with the lens (1642).
with a friend in another place with the aid of a hyperbolic lens,
painted figures, and a source of light." The figures had to be made of
materials which did not affect the polished surface of the lens, for in-
stance, wax or clay.
It must be noted that the lens did not project an image of the
figures. These appeared — as was the case with the figures on the
mirror — as shadows. But the lens did achieve a concentration of the
light and we may consider it as the first primitive condenser. The so-
called "hyperbolic" lens only existed in the fantasy of the inventor,
because the grinding of this kind of lens is, even with modern tools,
almost an impossibility. So much for the experiments of Bettini who
was able to perform "the art of mirror writing" at night.
650 VOSKUIL December
The first thing Kircher did was to extend the distance between the
mirror and the screen because, as he wrote, "it was hardly 20 steps ..."
He did this by placing a lens in the reflected rays, which produced a
sharp, enlarged, and inverted image on the screen (Fig. 3d and Fig. 5) .
The plane mirror had a diameter of 4 centimeters and the lens a di-
ameter of 3 centimeters and we may conclude from Fig. 5, which is a
copy of the picture in the "Ars Magna" of 1646, that the lens must
have had rather a great focal length. For it must be noted that the
distance between the mirror and the lens is rather long and that the
enlargement on the screen is rather small. Considering the technical
possibilities of the seventeenth century this was the only means of
avoiding the spherical and chromatic aberration. Kircher first pro-
jected texts which now "were clearly visible on a distance of 500 feet."
With two assistants, namely, Kasper Schott (well known for his
book "Magia Universalis Naturae et Artis") and Georgio de Sepi (who
acted as an instrument maker), Kircher industriously went on with
his experiments. The mirrors were made of a special alloy because
normal steel mirrors were affected by the ink. Neither were mirrors of
glass of any use as the double reflection of the light rays in the glass
produced a blurred image.
It was found that concave mirrors worked better than plane ones
and this is understandable as the concave mirror reflects the rays in a
convergent bundle on the center of the lens and so produces a sharper
image than the plane mirror where the rays reflected in the margin of
the lens are more refracted than those which are transmitted nearer the
center (Fig. 3d). Further it was pointed out that it was very im-
portant to have a well-ground lens which had to be spherical or better
still "hyperbolical."
The many performances which were given by Kircher and his assist-
ants excited a lively interest and made a profound impression. First
texts were projected, then the dial of a clock which was painted on the
mirror, a pointer, made of paper indicating the correct time. Later
geometrical line drawings, filled in with transparent paint were pro-
jected and Kircher was surprised at the fact that the colors appeared
unchanged on the screen. He was pleased with this kind of projection
and relieved his feelings in circumstantial treatises.
The experiments went on. Right in front of the mirror a cardboard
puppet was placed, the limbs of which could be moved by invisible
threads and. . .tfce spectators saw the first moving pictures! Then a
fly was fixed on the mirror with honey and a terrifying monster
1948
ORIGINS OF THE MAGIC LANTERN
651
appeared on the screen! By sticking a needle into the fly and moving
a magnet behind the mirror, which in this case could not be made of
iron, it looked as if the fly moved and were alive. This apparatus may
be considered as one of the first primitive solar microscopes.
The spectators were profundly impressed by this kind of per-
formance and the "moving pictures" even frightened them. However
WA B C D £ P G H I K L M N Q P Q R S T V X Y Z
./, Y g c D H L Q H j j, j, yj. ^ o ^ (-y"jr .g j^ Ay ; X
n v i £ v MS V D * 3 D D ^ i j » o n .n n •? .^ 2
u &> j b i> 7. & 3 A P f a D £ j> J ? * Q th J u J ? s
»h$r*e*.^*5 A B' r A>£ z |j O I K A M N F.CXn P I T Y 0> X
.•r/u**-infifsto* ¥ H ;L v> F r ij 0 ^ K: V "W M' H O II B 2 XX'* X
Fig. 5 — Kircher's projection methods as an improvement of the old art of
mirror writing (1646),
652 VOSKUIL December
Kircher was not yet satisfied with these results. The public had to be
impressed much more and the idea occurred to him that this might be
achieved if the performances were given at night. Now he had to
work with an artificial source of light and a burning candle was
chosen which was placed in front of a concave mirror because a plane
mirror would only reflect part of the divergent rays into the lens (Fig. 3e) .
As he had not always the correct concave mirrors at his disposal,
Kircher invented another device. Thinking of Bettini's method with
which he had experimented before, Schott wrote about this "third
art," "I have tried it with Kircher and it came out well" — the con-
cave mirror was substituted by one of those rather large spherical
flasks filled with water, which in the seventeenth century, were on
hand in all sorts and sizes and were used by the physicians as "urine
receptacles." The drawing or text was fixed or painted inverted
and reversed on the water-filled flask at the side turned to the projec-
tion lens but, as the surface of the flask was spherical, it was not
possible to focus all points of the figures on the screen and the image
as a whole remained blurred (Fig. 3j). It is remarkable that Kircher,
instead of using a second lens, chose a flask. Therefore we may draw
the conclusion that lenses still were rare in the seventeenth century.
Thus we may consider the device of Fig. 3e as the first primitive pro-
jector with the "reflector lamp" (Fig. 3g) and the device of Fig. 3j as
the first with a condenser (Fig. 31) .
As the light of a candle is very feeble compared to that of the sun,
only figures and short words as "Pax" and "Salve" could be shown, but
the influence of the darkness onvthe spectators was so undeniable that
the simple words made a more profound impression than the moving
puppet in the sunlight. Kircher considered this kind of projection a
very useful means to convert godless people. Therefore he took great
pains to project on the windows of houses in Rome, the "panes" in
those days being of paper. We may imagine the feelings of the sinful
Roman citizens when suddenly they saw the bright figures^ in the
darkness and supposed an ominous resemblance withsthe "Mene
tekel. . . " of king Belshazzar.
Schott wrote in his "Magia Optica" (one of the volumes of the
"Magia Universalis") that "these performances of images in darkened
places were more alarming than those .in the daylight." By this art
godless people might easily be kept from committing sins, especially
if one should fix a picture of the devil on the mirror and project
this in a dark place. It is a pity that such views have hampered the
1948 ORIGINS OF THE MAGIC LANTERN 653
development of the projection lantern. When, namely, in the course
of time, the profound seriousness of the priests, concerned about the
souls of the godless, had vanished, the "laterna magica" remained in
the hands of charlatans and necromancers who used it to impress
simple and superstitious minds. During almost two centuries the lan-
tern had only been used to project terrifying figures and for other
mysterious purposes. The development toward an instrument for
the use of scientific instruction and education was for this reason
hampered until about 1850.
However, in the seventeenth century there was somebody who, in
Kircher's improved art of mirror writing, saw not only an instrument
of wonder but something more. It was the Belgian Jesuit and mathe-
matician Andreas Tacquet (1612-1660) of Lou vain who was the
first to give a lantern lecture. He had met friar Martin Martini, a
Jesuit missionary, who had undertaken a far and dangerous journey
to China and after his return to Louvain (1653) visited Tacquet and
had formed the plan to hold a lecture upon his adventures. Tacquet
who was acquainted with Kircher's method of projection, probably by
having read the "Ars Magna," realized how Martini's lectures could
gain in importance if they were accompanied by projected illustra-
tions. And so it happened that during Martini's narrative, before
the eyes of the interested and astonished spectators, pictures of foreign
countries and peoples appeared which seemed to come out of nothing
and also vanished into nothing ....
It is again Schott who tells us something about this lecture in his
"Magia Optica" (page 426) and in spite of the fact that he does not
mention the use of slides, Tacquet must have used them, as it is im-
possible to accept the fact that the pictures were painted on the mirror
and wiped out again during the lecture (Fig. 3f ) . Thus Tacquet must
have introduced the first lantern slides.
All elements for a complete slide projector now were present and
only the mind was lacking to unite these elements into a whole. This
mind proved to be the well-known Dutch scientist Christiaan Huygens
(1629-1695). He constructed a complete projector with a condenser
and a calculated projection objective. Moreover, he had made
separate slides (Fig. 3k). It is outside the scope of this article to de-
scribe Huygens' interesting work in this field because it is a chapter in
itself. A new period in the development of the projection lantern now
begins in which the names of Walgensten, Dechales, Zahn, Robert
Jlooke, and William Molyneux must be mentioned. That the name of
654 VOSKUIL December
Huygens has receded into the background is the result of the fact that
afterwards he felt somewhat ashamed of his activities connected with
the projection lantern. Gradually the charlatans began to frighten
the public with this instrument of wonder and picking its purse.
That such a person as Huygens was above any form of charlatanry
is understandable and he tried to forget the "incident" (as he called
it) as soon as possible.
A final word about Kircher's "laterna magica" in the second edition
of the "Ars Magna" (Fig. 1) : When this picture met the eyes of the
readers the projection lantern had already existed for ten years and
therefore it is certainly incorrect to consider the device in Fig. 1 as the
first magic lantern and Kircher as its inventor.
At the utmost he discovered the principle of optic projection when
he, more or less by chance, used a lens to improve the art of mirror
writing. In fact he had not the slightest idea of the importance of his
"invention" and he was not able to develop it in a logical and sys-
tematic way. When, for instance, he writes about the clearness with
which the projected writing can be seen, he claims that with the
means at his disposal, a sharp image is obtained at a distance of 500
feet. "Thus" an instrument 24 times larger would give a sharp image
at a distance 24 times greater, namely, 12,000 feet. Of course there
would be "some difficulties," he mentions drily, and the images would
be too -large and too faint but many improvements could be intro-
duced, for instance, by "using more concave mirrors." In which way
this had to be performed Kircher did not mention "as time to make
further experiments failed him." However, he "recommended his
idea to other scientists for further reflections." And when Huygens
had finished his lantern in 1659, one of his acquaintances, a certain
Guisony, wrote him a letter from Rome (1660) remarking that
Kircher was not yet very familiar with "the invention of the lantern."
"The good old Kirkher (Kircher)," Guisony wrote, "is performing a
great number of tricks with his magnet in the Collegium Romanum
but if^he had the invention of the lantern he should frighten the
Cardinals with ghosts all the time."
According to this it appears that Kircher's knowledge about "the
art of light and shadow" had not made much progress in the years
after he had experimented with Schott and de Sepi and that the
application of the magic lantern, namely, "frightening the Car-
dinals" was nearer to his heart than its construction. Indeed, if we
have a critical look at Fig. 1, we notice that, for instance, the objective
1948 ORIGINS OF THE MAGIC LANTERN 655
is in the wrong place and apparently Kircher had Bettini's art in his
mind, so we are justified in concluding that Huygens in 1659, Walgen-
sten in 1660, and Dechales in 1665 with their "lanterns" were nearer
the goal than Kircher in 1671. Kircher's complaint in the second
edition of the "Ars Magna," that "Walgensten had sold copies of his
lantern at high prices in France and Italy to many prominent people"
is therefore unfounded.
BIBLIOGRAPHY
(1) M. Bettini, "Apiaria Universae Philosophiae Mathematicae," Bobgna,
1642.
(2) C. A. Crommelin, "The Grinding of Lenses in the 17th Century," Am-
sterdam, 1929.
(3) J. M. Eder, "History of Photography," 4th ed., Halle, 1932, p. 52, chapter
V, History of the Camera Obscura.
(4) Christiaan Huygens, "Collected Papers" (Oeuvres Completes), vol. 3, p.
45.
(5) J. S. Kestler, "Physiologia Kircheriana Experimentalis," Amsterdam,
1680, p. 125.
(6) A. Kircher, "Ars Magna Lucis et Umbrae," 1st ed., Rome, 1646; 2d ed.,
Amsterdam, 1671, p. 788.
(7) F. P. Liesegang, "The relation of the old silhouette show to the invention
of the laterna magica," Prometheus, vol. 30, p. 345; 1919.
(8) Idem, "The camera obscura of Porta," Mitteilungen zur Geschichte der
Medizin and der Naturwissenschaften, vol. 18, p. 1; 1919.
(9) Idem, "Performances with the camera obscura in ancient times," Op-
tische Rundschau, nos. 31-33; 1919.
(10) Idem, "The oldest projectors," Centralzeitung fur Optik und Mechanick,
vol. 39, pp. 345 and 355; 1918.
(11) Idem, "Origins of the projection lantern, "Die t/wsc/iaw,vol.23,p. 107; 1919.
(12) Idem, "The camera obscura and the origins of the laterna magica,"
Photographische Industrie, p. 197; 1920.
(13) Idem, "The oldest lantern lecture," Photographische Industrie, p. 39; 1919.
(14) Idem, "Historical development of the laterna magica with the condenser
from the 'art of mirror writing' with a lens and without a mirror," Central Zeitung
fur Optik und Mechanik, vol. 43, p. 79 ; 1922.
(15) Idem, "The laterna magica of Athanasius Kircher," Deutsche Optische
Wochenschrift, vol. 7, p. 180; 1921.
(16) R. H. Mayor, "Athanasius Kircher," Annals of Medical History, 3d series,
p. 105; 1939.
(17) J. G. Poggendorff, "History of Physics," Leipzig, 1879.
(18) J. B. Porta, "Magia Naturalis," Libri XX, 1st ed., Naples, 1558; 2d ed.,
Naples, 1589.
(19) P. Reinhardt. "The inventor of the projection apparatus," Prometheus,
vol. 15, p. 304; 1904.
(20) M. von Rohr, "On the historical development of the magic lantern," Zeit-
schrift der Deutschen Gesellschaft fur Mechanik und Optik, pp. 49 and 61; 1919.
(21) K. Schott, "Magia Optica," Wiirzburg, 1657, pp. 426 and 440.
(22) E. Wiedemann, "On the invention of the camera obscura," Verhandlungen
der Deutschen Physikalischen Gesellschaft, vol. 12, no. 4; 1910.
Report of the
Studio Lighting Committee
Summary — This report contains information on three phases of studio
lighting as follows: I. Various types of new equipment which have been
recently developed and used are described and illustrated. II. New uses
for standard commercial lamps are discussed. III. The results of tests
conducted to determine the effect of variation of power-supply voltage and
lamp maintenance upon color rendition of 16- mm commercial Kodachrome
are given.
I. NEW EQUIPMENT
Special Effects with Remote-Controlled Shutters
SHUTTERS OF THE Venetian-blind type have been used for some
time on set lighting lamps for dimming, color-changing, simulat-
ing the turning off and on of lights in a room, and various other types
of special effects. Wherever they have been used in groups a need
for exact synchronization has been indicated. With manually
operated shutters a man must be located at each lamp to operate his
respective shutters upon receipt of a cue and with this mode of
operation it is impossible to have all of the shutters operate in unison.
A remote-control system was indicated where all the shutters would
operate in synchronism and from one control point.
In response to this need the electrical department at the Metro-
Goldwyn-Mayer Studios developed a direct-current self-synchronous
system for remote control of the shutters in groups. They have ob-
tained interesting results in special-lighting effects such as the simu-
lation of the sunrise following the rainy season in "The Yearling," and
the Easter sunrise service in the "Three Darling Daughters." The
lighting setup for the sunrise effect in "The Yearling" involved re-
mote control at an extremely slow speed with as many as 40 individual
shutters operating in exact synchronism.
Direct-current self -synchronous remote-controlled shutters are now
available (see Fig. 1) and are being successfully used in the studios.
The shutter is an improved design of the Venetian-blind type which
has been used in the studios for several years. The shutter motor is
* Original manuscript received by the Society August 9, 1948.
656 DECEMBER, 1948 JOURNAL OF THE SMPE VOLUME 51
STUDIO LIGHTING REPORT
657
essentially a direct-current self-synchronous motor constructed inte-
gral with a gear-reduction unit. The transmitter is in the form of a
rheostat whose brushes can be manually rotated by a handle external
to the rheostat housing. Approximately 180-degree motion of the
handle is required to rotate the remote-controlled shutter blades
through their full travel.
Fig. 1 — Remote-controlled shutter shown
mounted on a set-lighting arc lamp, connected
through cables and plugging box to a transmitter
capable of positioning up to 15 shutters, as manu-
factured by the Mole-Richardson Company.
Friction slip clutches are provided in the shutter-motor gear-
reduction boxes for synchronization. After a group of shutters
have been set up and connected to the direct-current supply, one
complete movement of the transmitter operating handle from its open
to closed position, or vice versa, will result in the synchronization
658 STUDIO LIGHTING REPORT December
of all the shutter blades of all connected shutters. A reversing switch
is included on each shutter motor. With this switch thrown to the
"normal" position the shutter blades will follow the movement of the
transmitter operating handle in the normal fashion; that is, moving
the transmitter handle to the open position will result in the blades'
opening, and movement of the transmitter handle to the closed
position will cause the shutter blades to close. If the switch on a
particular shutter motor is thrown to the "reverse" position, the
operation of the blades on that shutter will be out of phase with the
remainder of the units. Thus it is possible to cause some shutters to
close as others open and vice versa in any desired combination.
Each of the transmitters is capable of operating one to fifteen
shutters. The standard equipment includes a plugging box con-
nected to the transmitter with 50 feet of 3-conductor rubber-covered
cable. Various shutters can be plugged into this plugging box with
their respective 25-foot cables, or a group of shutters can be inter-
connected in a series-parallel arrangement to the plugging box.
If it is desired to operate more than 15 shutters in a unit, provisions
can be made to operate more than one transmitter from a single
handle.
A direct-current system has advantages over an alternating-current
system in the following respects :
(1) Direct current is always available as a power supply on the
sets whereas in some locations there might be no alternating current.
(2) The direct-current control motors produce no noise such as
might be present in the form of a 60-cycle hum in the alternating-
current motors.
(3) The direct-current self-synchronous motors cannot be dam-
aged by overload if stalled out of correspondence.
(4) The direct-current system requires three conductors to each
motor whereas an alternating-current system would require five.
The position of the shutter blades follows the position of the trans-
mitter and the speed of the shutter-blade movement follows the
speed of the movement of the transmitter. For dimming purposes
the blades can be made to move as slowly as it is possible to turn the
transmitter, there is a limiting maximum speed at which the system
can be operated without having the shutter blades fall out of step or
lose synchronism with the transmitter. However, in actual service
with the shutters simulating the turning off and on of lights in a room
1948
STUDIO LIGHTING REPORT
659
from a snap switch, it has been determined that the shutter blades
will travel from their full open position to their full closed position
or vice versa in four frames on the film. Hence, the time required for
the blades to rotate through their full travel is one sixth of a second.
This speed is adequate for any situation which is anticipated in studio
lighting effects.
"Snap-switch" operation of the remote-controlled shutters was
well executed in a recent production at the Warner Brothers studios.
Fig. 2 — A group of remote-controlled shutters mounted on arc lamp
rigged for studio set lighting. Courtesy of Electrical Department,
Warner Brothers Pictures, Inc.
In a scene in this picture an actress walks out of a living room,
turns off the living-room lights, and enters a bedroom. She walks
to the bed and turns on the bed lamp, then back to the wall switch
and snaps off the main bedroom lights. The actress goes to bed and
shortly thereafter turns off the bed lamp, at which time moonlight
appears through the bedroom window. The set lighting for this
660 STUDIO LIGHTING REPORT December
rather complicated sequence of light changes was accomplished by
means of the remote-controlled shutters on the set-lighting lamps
operated by special motor-driven transmitters developed by the
Warner Brothers electrical department. The control circuits for
the shutters were connected to and operated by the various light
switches which the actress operated in the scene. The synchroniza-
Fig. 3 — Mole-Richardson Type 400 arc lamp.
tion of the variations in lighting with the operation of the various
light switches was perfect, arid there was no possibility of missing a cue.
The remote-controlled shutters were again used in a recent Warner
Brothers production to simulate the illumination in a room coming
through a window from a flashing neon sign. The neon sign, com-
plete with its flashing mechanism, was installed on the set outside
of the window. Its operation was such that it flashed from red to
white, to green, to white, to red, to white, and so on. Lamps were
1948 STUDIO LIGHTING REPORT 661
rigged on the set with red filters to simulate the red portion of the
neon-sign lighting, green filters to simulate the green portion, and
no filters to simulate the white portion. Remote-controlled shutters
operated by the specially constructed motor-driven transmitters
were mounted on the lamps. The circuits between the transmitter
and the shutters passed through relays which were electrically oper-
ated by the neon flashing mechanism. Thus the operation of the
various shutters was automatically synchronized with the flashing
mechanism of the neon sign, and a perfect lighting illusion was
created.
Small Arc Lamp for Close-to-Camera Work
A small lightweight arc lamp is now available which can be located
close by the camera lens or concealed behind relatively small objects
(see Fig. 3) . It is essentially one half of a Type 40 Duarc and when
supplied with a reflector and diffusing glass will produce an intensity
of 125 foot-candles at 10 feet with a spread of about 140 degrees.
It can also be equipped with a spherical mirror and Fresnel lens to
produce an intensity of 250 foot-candles at 10 feet with a narrower
spread of about 80 degrees. The arc current is 40 amperes. The
lamp weighs about 35 pounds and is capable of being operated in
either the vertical or horizontal position. A separate grid unit is
located about 25 feet from the lamp.
This small arc lamp has been used in several productions close by
the camera for close-ups to produce a soft front fill light to wash out
undesirable shadows, or in locations slightly on one side of the
camera to give a close-in key light. Being small in size it lends itself
to concealment behind relatively small objects, columns, or beams.
The "Brute" Lamp
The Mole-Richardson Type-450 "Brute"1 (see Fig. 4), having
twice and in some cases more than three times the illumination of
any single source previously used, has proved itself to be an ex-
tremely valuable tool. One of its chief uses is to create an illusion
of "one-source" lighting, casting single well-defined shadows through
the complete scene of action. The Brute has in numerous cases
made it possible to illuminate sets adequately with fewer lighting units
than otherwise would have been required. For large, deep sets
this lamp can provide the required level of illumination through-
out the full depth of the scene. It has also been found to be
useful for providing "booster light" on outdoor sets.
662
STUDIO LIGHTING REPOKT
December
Small Incandescent Bulbs for Special Effects
One of the lamp-manufacturing companies has developed and
introduced a small line of incandescent bulbs for special-effect light-
ing. One of these lamps is known as catalog No. 25S6 (see Fig. 5)
and has a 25-watt, 115- to 125-volt filament placed in a bulb 3/4 inch
in diameter and about l!/2 inches long. This motion picture effect
lamp has an average life of 50 hours. Another has a 25-watt, 50-hour
filament placed in a small bulb iy2 inches in diameter having a built-
Fig. 4— Mole-Richardson Type 450
"Brute" Molarc.
Fig. 5 — Small special-effect
incandescent lamp. Catalog No.
25S6.
in reflector and is known as catalog No. 25R12DC (see Fig. 6). It is
in effect a miniature reflector photoflood. This lamp will produce a
light intensity of approximately 75 foot-candles at 20 inches.
Both of the above midget-size lamps lend themselves to being
concealed behind small objects and have a number of special uses
for small lighting effects.
II. NEW APPLICATIONS OF STANDARD COMMERCIAL LAMPS
Airplane Landing Lights Used for Automobile Headlamps
Out-of-door scenes simulating night conditions are frequently made
in broad daylight with a filter over the camera lens in order to ob-
tain a night effect but still have the entire scene illuminated so as to
1948
STUDIO LIGHTING REPORT
663
Fig. 6 — Small incandescent reflector
lamp, Catalog No. 25R12DC.
produce good definition on the
screen. Ordinary automobile
headlamps under these condi-
tions do not appear lighted unless
perhaps the headlights are aimed
directly at the camera. Many
attempts have been made to
illuminate more brilliantly the
headlight lens by paralleling fila-
ments in a bulb, using 50-candle-
power bulbs, overvoltaging fila-
ments, etc., with mediocre results.
An outstanding successful
method is to use sealed beam-type airplane landing lamps developed
for military craft which are rated at 450 watts and fit the regular
sealed beam-lamp assemblies on automobiles (see Fig. 7). Such a
lamp is so powerful that in a re-
cent color picture the headlight
beams on an actress's dress
showed clearly on the screen as
she walked in front of the auto-
mobile even though the picture
was actually taken in sunlight
with the camera filtered to simu-
late a nighttime setting. This
lamp is available as a No. 4540
which is rated at 450 watts, 13
volts and No. 4541 which is
rated at 450 watts and 28 volts.
Both lamps have a 25-hour
average life.
Photo Reflector Lamps Applied for
Fill-Light Illumination
A reflector photoflood desig-
nated as the RFL2 (see Fig. 8), a
500-watt, 115- to 120-volt flood-
lamp, was used in 1947 much
more extensively than in the past
Fig. 7— Airplane landing lamp Catalog . . .. ,
No 4540 for delivering a flood of light
664 STUDIO LIGHTING REPORT December
used for fill-in purposes on locations. Occasionally where greater
distances were involved or where small key-lighting effects were
desired, the RSP2 photospot was used. The photospot lamp is
identical in size, shape, wattage, and color temperature with the
photoflood but has a much narrower and several times more power-
ful beam. Generator capacity is often at a premium on locations
and although these reflector lamps have a short life, they can be em-
ployed advantageously under such conditions to obtain a relatively
large amount of light with the limited power supplied. Being light
in weight these lamps simplify the transportation problems but,
Fig. 8— The RSP2 photospot, left, and the RFL2 photoflood, right.
however, do not allow the flexibility of control of illumination which
is characteristic of the focusable Fresnel-lens units.
III. RESULTS OP TESTS PERTAINING TO COLOR RENDITION OF
16-MM COMMERCIAL KODACHROME
Effect of Arc-Lamp Supply . Voltage Upon Color
Tests were recently conducted at the Mole-Richardson Company
in conjunction with the Eastman Kodak Company to determine the
effect of variations in arc-lamp conditions upon the color rendition of
16-mm Eastman commercial Kodachrome (3200-degree Kelvin)
film. The tests were made using an M-R Type 170 Molarc lamp
with a new Y-l filter for illumination with a Wratten No. 83 filter
1948 STUDIO LIGHTING REPORT 665
and the proper emulsion color-correction filter over the camera lens.
The normal current drawn by a Type 170 arc is 150 amperes with a
line voltage of 115 volts. Photographic tests were made under the
following three sets of conditions :
(1) The line voltage was varied from 108 to 118 volts with the
carbons adjusted so that the arc current was maintained at the nor-
mal value of 150 amperes in each take.
(2) The line voltage was maintained at 118 volts and the arc
current varied from 134 to 158 amperes by adjustment of the position
of the carbons.
(3) The arc lamp was adjusted for normal operation of 150 am-
peres with a line voltage of 115 volts, and then the line voltage was
varied from 108 to 118 volts with arc current varying in correspond-
ence with the variations in line voltage.
In each take the lamp was spotted or flooded as necessary to
maintain the same light intensity of approximately 1200 foot-candles
on the subject. No noticeable visual effect in color was observed
under the above variations of arc-lamp illumination.
The Committee plans to make similar tests, the results of which
can be published in a subsequent report, to determine the effect of
variations of incandescent lighting on the color rendition of commer-
cial Kodachrome film. Information published in the Photo-Lab
Index2 indicates that the color temperature of incandescent illumina-
tion will not visually distort the color on the film unless it departs as
much as approximately 100 degrees Kelvin from the correct value.
The color temperature of photographic incandescent lamps changes
from the rated value about 10 degrees Kelvin for each volt difference
between the actual supply voltage and the rated voltage of the lamp.
Hence a 115-volt lamp operated at 125 volts will have a color tempera-
ture which is 100 degrees Kelvin higher, or if operated at 105 volts its
color temperature will be 100 degrees Kelvin lower than the rated
color temperature.
The above tests would indicate that color is not appreciably af-
fected by the usual expected operating variations encountered with
illumination. However, the importance of maintaining correct line
voltage should not be minimized. Even though line voltage can
vary to a certain extent without appreciably affecting the color,
such variations definitely affect the intensity of illumination and
the efficiency of arc operation. Line voltage therefore should be
STUDIO LIGHTING REPORT
maintained as closely as possible to the normal value in order that
variations in light intensity and abnormal arc operation are kept
to a minimum.3
Effect of Maintenance of Arc Lamps Upon Color
The importance of maintaining clean arc-lighting equipment was
demonstrated in split-screen tests recently conducted at the Mole-
Richardson Company in conjunction with the Eastman Kodak Com-
pany. A subject was illuminated with a clean M-R Type 40 Duarc
lamp and photographed on Eastman commercial Kodachrome (3200
degrees Kelvin) film with a Wratten No. 83 filter and the proper
emulsion color-correction filter on the camera. This exposure was
made on one side of the film. The other half of the film was later
exposed with all conditions remaining the same except that the clean
Duarc was replaced by one whose front-door glass and reflector were
considerably contaminated with the arc-flame residue material which
accumulates with time if the lamps are not properly maintained.
This split-screen test clearly indicated that the color in the picture
tends strongly toward the yellow if dirty lamps are used for illumina-
tion. All arc-lighting equipment should be kept clean to avoid such
off -color effects.
STUDIO LIGHTING COMMITTEE
1948
M. A. HANKINS, Chairman
W. E. BLACKBURN C. W. HANDLEY
RICHARD BLOUNT C. R. LONG
J. W. BOYLE W. W. LOZIER
KARL FREUND D. W. PRIDEAUX
REFERENCES
(1) M. A. Hankins, "Recent developments of super-high-intensity carbon-arc
lamps," /. Soc. Mot. Pict. Eng., vol. 49, pp. 37-47; July, 1947.
(2) Photo-Lab Index No. 10-ILL-20, Quarterly Supplement No. 28 (replacement
page), pp. 10-13, published by Morgan and Lester, New York, N. Y.
(3) "Report of Studio Lighting Committee," J. Soc. Mot. Pict. Eng., vol. 45,
pp. 249-260; October, 1945.
SAMUEL EDWARD SHEPPARD
1882-1948
SAMUEL EDWARD SHEPPARD was born in Catford, England,
and was educated at St. Dunstan's College and University
College, London. At University College he obtained the degree
of B.Sc. by research in 1903, for a thesis dealing with the theory of
the photographic process, and involving a repetition and extension
of the earlier work of Hurter and Driffield. This work was greatly
extended in his research for the D.Sc. degree, which was granted in
1906 for a thesis which was published in 1907 jointly with that of
C. E. K. Mees under the title of "Investigations on the Theory of the
Photographic Process/'
In 1913 Dr. Sheppard accepted an invitation to take charge of the
sections of physical and colloid chemistry in the Kodak Research
Laboratory, which had just been organized under the direction of
Dr. Mees at Rochester, N. Y.
His work on photography covered the whole of photographic
chemistry including the study of the process of development, the
structure of the light-sensitive emulsion, the nature of the latent
image, and the causes of sensitivity in photographic emulsion. In
addition, he made some important advances in colloid chemistry,
including the use of mixtures of powdered coal and oil as a fuel, and
methods of electroplating rubber in thin coatings. His early work
in the United States dealt principally with the physicochemical
properties of gelatin, but about 1920 he turned his attention to the
action of light on the silver halides and the nature of sensitivity, and
published a series of papers on the theory of photographic sensitivity
and the formation of the latent image.
The well-known sensitizing property of gelatin in the photographic
process led to a systematic study of the difference between photo-
graphic gelatins in their sensitizing power and the nature of the
667
668 SAMUEL EDWARD SHEPPARD December
substance in gelatin which conferred sensitivity. By a painstaking
series of analyses, it was found that the sensitizer inherent in natural
gelatin was concentrated in the liquors obtained by the acid treat-
ment of the raw material after liming, and eventually it was found
that the chemical properties of the sensitizer corresponded to those
of allyl thiourea and that therefore the gelatin sensitizer was essen-
tially one which could produce silver-sulfide specks in the silver-
bromide crystals.
This discovery is perhaps the major advance made in Sheppard's
scientific career. All further study of the photographic properties of
gelatin, of the nature of the sensitivity of silver halides, and of the
latent image have been conditioned by it. Its publication won for
Sheppard instant recognition. He was awarded the honorary
fellowship of the Royal Photographic Society in 1926, the progress
medal of the Royal Photographic Society in 1928, and the Adelskb'ld
medal of the Swedish Photographic Society in 1929. In 1928 he
delivered the Hurter and Driffield Memorial Lecture to the Royal
Photographic Society, and in 1930 received the Nichols medal of the
American Chemical Society. He was made a Fellow of the Society
of Motion Picture Engineers in 1944 and an honorary Fellow of the
Photographic Society of America in 1946.
Since 1930, Dr. Sheppard's scientific work covered a prodigious
range of knowledge. Besides the work on the latent image, he
studied such matters as the photovoltaic effects, that is, the electri-
cal response of silver halide to light, the colloidal structure of film-
base materials and their physicochemical and elastic properties,
the nature of development, and particularly the nature of dye sen-
sitizing, the adsorption of sensitizing dyes to silver halides, the
structure of the layers which they formed, and their sensitizing
effects.
More than any other single worker, Dr. Sheppard has been re-
sponsible for our present knowledge of the theory of the photo-
graphic process. He explored every section of the chemistry of
that process, and everywhere his studies brought light.
C. E. K. MEES
'ournttl of the
"ociety of Motion Picture Engimei
INDEX
Volume 51
July— December., 1948
TlllS ISSUE IN TWO PARTS
I'art I -December Journal
Part II - Index to Volume 51
\\«!
•w\v\vw\\ "\\i\VvfV \ tl
INDEX TO AUTHORS
Volume 51
July, 1948, through December, 1948
ALBIN, F. G.
Sensitometric Aspect of Television
Monitor-Tube Photography
December, p. 595
AUDIGIER, L., with ROBERTSON, R.
Gaumont-Kalee Model 21 Projector
September, p. 269
BACK, F. G.
Zoomar Lens for 35-Mm Film
September, p. 294
BENHAM, H. J.
Projection Equipment for Screen-
ing Rooms September, p. 261
BOON, J. L., with FELDMAN, W., and
STOIBER, J.
Television Recording Camera
August, p. 117
BOWDITCH, F. T. (Chairman)
Report of Standards Committee
September, p. 230
BROWDER, L. B.
Variable- Area Light- Valve
Modulator November, p. 521
BUTTOLPH, L. J.
Ultraviolet Air Disinfection in the
Theater July, p. 79
COILE, R. C.
Parabolic Sound Concentrators
September, p. 298
CONTENT, E. J.
Quieting and Noise Isolation
August, p. 184
COOK, R. K.
Behavior of Acoustic Materials
August, p. 192
COTT, W. B.
Service and Maintenance of Air-
Conditioning Systems July, p. 92
CRANE, G. R., with MILLER, W. C.
Modern Film Re-Recording Equip-
ment October, p. 399
CHANDLER, J. S.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion) October, p. 437
CHERRY, W. H.
Colorimetry in Television
December, p. 613
DIMMICK, G. L., with JOHNSON, S. W.
Optimum High-Frequency Bias in
Magnetic Recording
November, p. 489
ERDE, BERNARD
Color-Television Film Scanner
October, p. 351
FAIRBANKS, JERRY
Films for Television
December, p. 590
FELDMAN, W., with BOON, J. L., and
STOIBER, J.
Television Recording Camera
August, p. 117
FORDYCE, C. R.
Improved Safety Motion Picture
Film Support October, p. 331
FRAYNE, J. G.
Variable-Area Recording with the
Light Valve November, p. 501
GOLDSMITH, T. T., JR., with MIL-
HOLLAND, HARRY
Television Transcription by Motion
Picture Film August, p. 107
GOTSCHALL, G. D.
Light Modulation by P-Type Crys-
tals July, p. 13
GRIGNON, L. D.
Flicker in Motion Pictures: Further
Studies December, p. 555
HANDLEY, C. W. (Chairman)
Report of Studio-Lighting Commit-
tee (1947) October, p. 431
HANKINS, M. A. (Chairman)
Report of Studio Lighting Com-
mittee (1948) December, p. 656
HOPKINS, H. F., with KEITH, C. R.
New Theater Loudspeaker System
OCTOBER, p. 385
ISOM, W. R.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion)
October, p. 437
JOHNSON, S. W., with DIMMICK, G. L.
Optimum High-Frequency Bias in
Magnetic Recording
November, p. 489
KEITH, C. R., with HOPKINS, H. F.
New Theater Loudspeaker System
October, p. 385
KELLEY, W. F.
Motion Picture Research Council
October, p. 418
KELLOGG, E. W. (Chairman)
Proposed Standards for the
INDEX
Measurement of Distortion in Sound
Recording . November, p. 449
KELLOGG, E. W.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion)
October, p. 437
KIMBALL, D. W.
Motion Picture Theater Air Con-
ditioning July, p. 52
LANKES, L. R.
Historical Sketch of Television's
Progress September, p. 223
LITTLE, R. V., JR.
Developments in Large-Screen Tele-
vision July? P- 37
LOGAN, H. L.
Brightness and Illumination Re-
quirements July, p. 1
LYMAN,D. F.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion)
October, p. 437
MALOFF, I. G.
Optical Problems in Large-Screen
Television July, p. 30
MANDEBFELD, E. C., . .with, MILLER,
HAROLD
35-Mm Process Projector fg;
October, p. 373
MARTIN, L. R.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion)
October, p. 437
MASTERSON, EARL
35-Mm Magnetic-Recording System
November, p. 481
MAXFIELD, J. P. .
Auditorium Acoustics
August, p. 169
MlLHOLLAND, HARRT,Wlth GOLDSMITH,
T.T.,jR. ,
Television Transcription by Motion
Picture Film
August, p. 107
MILLER, HAROLD, with MANDERFELD,
E. C.
35-Mm Process Projector
October, p. 373
MILLER, W. C., with CRANE, G. R.
Modern Film Re-Recording Equip-
ment October, p. 399
, DOROTHY
Magnetic Recording for the Techni-
cian November, p. 468
PETTUS, J. L.
Improved Optical Reduction Sound
Printer December, p. 586
JliNG, LESTER
Display Frames in the Motion Pic-
ture Theater July, p. 101
ROBERTSON, R., with AUDIGIER, L.
Gaumont-Kalee Model 21 Projector
September, t>. 269
RYDER, L. L.
Report of the President
September, p. 221
SACHTLEBEN, L. T.
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion)
October, p. 437
SCHREIBER, E. H.
Video Distribution Facilities for
Television Transmission
December, p. 574
SCHULTZ, E. W.-
Use of- 16-Mm Motion Pictures for
Educational Reconditioning
October, p. 424
SINGER, KURT
Continuously Variable Band-Elim-
ination Filter August, p. 203
STOIBER, J., with FELDMAN,' W., jand
BOON, J. L.
Television Recording Camera
August, p. '117
SPISELMAN, J. W.
Air Purification by Glycol Vapor
July, p. 70
VOSKUIL, J.
Origins of the Magic Lantern
December, p. 643
WASHER, F. E.
Errors in Calibration of the/ Number
September, p. 242
WEST, A- G. D.
. Development pf Theater Television
in England August, p. 127
WILSON, H,.H. .
Portable 16-Mm Sound Projector
July, p. 21
INDEX TO SUBJECTS
Volume 51
July, 1948, through December, 1948
ACOUSTICS
Behavior of Acoustic Materials,
R. K. Cook August, p. 192
Quieting and Noise Isolation, E. J.
Content August, p. 184
Auditorium Acoustics, J. P. Max-
field August, p. 169
AIR CONDITIONING
Service and Maintenance of. Air-
Conditioning Systems, W. B. Cott
July, p. 92
Ultraviolet Air Disinfection in the
Theater, L. J. Buttolph July, p. 79
Air Purification by Gly col Vapor,
J. W. Spiselman July, p. 70
Motion Picture Theater Air Condi-
tioning, p. D. Kimball July, p. 52
FILMS
Educational and Documentary
Use of 16-Mm Motion Pictures for
Educational Reconditioning, E. W.
Schultz October, p. 424
General
Films for Television, Jerry Fairbanks
December, p. 590
Improved Safety Motion Picture
Film Support, C. R. Fordyce
October, p. 331
GENERAL
Display Frames in the Motion Pic-
ture Theater, Lester Ring
July, p. 101
Light Modulation by P-Type Crys-
tals, G. D. Gotschall July, p. 13
HISTORICAL
Origins of the Magic Lantern, J.
Voskuil December, p. (543
Historical Sketch of Television's
Progress, L. R. Lankes
September, p. 223
LOUDSPEAKERS
New Theater Loudspeaker System,
H. F. Hopkins and C. R. Keith
October, p. 385
OPTICS
Improved Optical Reduction Sound
Printer, J4 Lv Pettus
December, p. 580
Zoomar Lens for 35-Mm Film, F. G.
Back September, p. 29 1
Errors in Calibration of the /-Num-
ber, F. E. Washer
September, p. 242
Optical Problems in Large-Screen
Television, I. G. Maloff July, p. 30
Light Modulation by P-Type Crvs-
tals, G. D. Gottschall July, p. 13
PRINTING
Improved Optical Reduction Sound
Printer, J. L. Pettus
December, p. 586
PROJECTION
Background
35-Mm Process Projector, Harold
Miller and E. C. Manderfeld
October, p. 373
35-Mm
Flicker in Motion Pictures: Further
Studies, L. D. Grignon
December, p. 555
PROJECTORS
35-Mm Process Projector, Harold
Miller and E. C. Manderfeld
October, p. 373
Gaumont-Kalee Model 21 Projector,
L. Audigier and R. Robertson
September, p. 269
Projection Equipment for Screen-
ing Rooms, H. J. Benham
September, p. 261
16-Mm
Portable 16-Mm Sound Projector,
H. H. Wilson July, p. 21
5
6
INDEX
December
RESEARCH COUNCIL
Motion Picture Research Council,
W. F. Kelley October, p. 418
SCREEN BRIGHTNESS
Flicker in Motion Pictures: Further
Studies, L. D. Grignon
December, p. 555
Brightness and Illumination Require-
ments, H. L. Logan July, p. 1
SMPE ACTIVITIES
Committees
Report of Studio Lighting Commit-
tee (1948), M. A. Hankins, Chairman
December, p. 656
Proposed 16-Mm and 8-Mm Sprocket
Standards (Discussion), E. W. Kel-
logg, W. R. Isom, L. T. Sachtleben,
J. S. Chandler, D. F. Lyman, and
L. R. Martin October, p. 437
Report of Studio-Lighting Commit-
tee (1947), C. W. Handley, Chairman
October, p. 431
Report of SMPE Standards Com-
mittee, F. T. Bowditch, Chairman
September, p. 230
President
Report of the President, L. L. Ryder
September, p. 221
SOUND RECORDING
Variable-Area Light- Valve Modula-
tor, L. B. Browder
November, p. 521
Variable- Area Recording with the
Light Valve, J. G. Frayne
November, p. 501
Optimum High-Frequency Bias in
Magnetic Recording, G. L. Dimmick
and S. W. Johnson
November, p. 489
35-Mm Magnetic-Recording System,
Earl Masterson November, p. 481
Magnetic Recording for the Techni-
cian, Dorothy O'Dea
November, p. 468
Proposed Standards for the Measure-
ment of Distortion in Sound Record-
ing, E. W. Kellogg, Chairman
November, p. 449
Modern Film Re-Recording Equip-
ment, W. C. Miller and G. R. Crane
October, p. 399
SOUND RECORDING (Continued)
Parabolic Sound Concentrators,
R. C. Coile September, p. 298
ij Continuously Variable Band-Elimi-
nation Filter, Kurt Singer
August, p. 203
Light Modulation by P-Type Crys-
tals, G. D. Gotschall July, p. 13
SPROCKETS
Proposed 16-Mm and 8-Mm
Sprocket Standards (Discussion),
E. W. Kellogg, W.'R. Isom, L. T.
Sachtleben, J. S. Chandler, D. F.
Lyman, and L. R. Martin
October, p. 437
STANDARDS
Nine Recent American Standards
November, p. 534
Proposed Standards for the Meas-
urement of Distortion in Sound Re-
cording, E. W. Kellogg, Chairman
November, p. 449
Proposed 16-Mm and 8-Mm
Sprocket Standards (Discussion),
E. W. Kellogg, W. R. Isom, L. T.
Sachtleben, J. S. Chandler, D. F.
Lyman, and L. R. Martin
October, p. 437
Report of SMPE Standards Com-
mittee, F. T. Bowditch, Chairman
September, p. 230
TELEVISION
Colorimetry in Television, W. H.
Cherry December, p. 613
Sensitometric Aspect of Television
Monitor-Tube Photography, F. G.
Albin December, p. 595
Films for Television, Jerry Fairbanks
December, p. 590
Video Distribution Facilities for
Television Transmission, E. H.
Schreiber December, p. 574
Color-Television Film • Scanner,
Bernard Erde October, p. 351
Historical Sketch of Television's
Progress, L. R. Larikes
September, p 223
1948
INDEX
TELEVISION (Continued)
Television Recording Camera, J. L.
Boon, W. Feldman, and J. Stoiber
August, p. 117
Television Transcription by Motion
Picture Film, T. T. Goldsmith, Jr.,
and Harry Milholland
August, p. 107
THEATER
General
Display Frames in the Motion Pic-
ture Theater, Lester Ring
July, p. 101
Lighting
Brightness and Illumination Re-
quirements, H. L. Logan July, p. 1
THEATER (Continued)
Maintenance and Operation
Service and Maintenance of Air-
Conditioning Systems, W. B. Cott
July, p. 92
THEATER TELEVISION
Development of Theater Television
in England, A. G. D. West
August, p. 127
Developments in Large-Screen Tele-
vision, R. V. Little, Jr. July, p. 37
Optical Problems in Large-Screen
Television, I. G. Maloff July, p. 30
INDEX TO NONTECHNICAL SUBJECTS
Volume 51
July, 1948, through December, 1948
Book Reviews
Camera and Lens, by Ansel Adams
(Reviewed by L. E. Varden)
October, p. 443
Developing — Technique of the Nega-
tive, by C. I. Jacobson (Reviewed
by J. S. Friedman) July, p. 105
Diary and Sundry Observations of
Thomas Alva Edison, Edited by
Dagobert D. Runes (Reviewed by
Terry Ramsaye) November, p. 550
Enlarging — Technique of the Positive,
by C. I. Jacobson (Reviewed by
J. S. Friedman) October, p. 443
Informational Film Year Book 1947,
Published by Albyn Press, Edin-
burgh, Scotland, (Reviewed by-G.
E. Matthews) October, p. 444
Magic Shadows, by Martin Quigley,
Jr. (Reviewed by J. E. Abbott)
August, p. 214
L'Annuaire du Cinema 1948 (Motion
Picture Yearbook for 1948), Pub-
lished by Editions Belief aye, Paris,
France November, p. 551
Photographic Facts and Formulas, by
E. J. Wall and F. I. Jordan (Re-
viewed by H. A. Miller)
August, p. 214
Preparation and Use of Visual Aids,
by K. B. Haas and H. G. Packer
(Reviewed by W. A. Wittich)
September, p. 330
Conventions
Convention Papers July, p. 104
64th Semiannual Convention
August, p. 212
64th Semiannual Convention
September, p. 323
July, p. 106
August, p. 217
Current Literature
October, p. 445
November, p. 552
Miscellaneous
ASA Adopts Universal Decimal Classi-
fication System November, p. 552
Czechoslovak Film Standards
August, p. 211
Incorporation of American Standards
Association October, p. 440
International Scientific Film Congress
August, p. 211
Technical Societies Council Elects
Officers July, p. 104
Obituaries
Armat, Thomas October, p. 441
Barrows, Thad C. October, p. 442
Lumiere, Louis October, p. 442
Sheppard, Samuel Edward
December, pp. 667-668
Section Meetings
Atlantic Coast November, p. 549
Midwest August, p. 216
September, p. 327
November, p. 549
Society of Motion Picture
Engineers
Committees of the Society
September, p. 312
Correspondence August, p. 216
Journal Exchange July, p. 104
October, p. 446
New Products August, pp. 218-219
November, p. 553
35-Mm and 16-Mm Test Films
October, p. 446
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11