gave way to continuous machine proc-
essing.161-163
How generally the distinctions be-
tween specular and diffuse density,164 and
between exposure modulation by varying
time (light valve) and varying intensity
(as by glow lamp)‘65 were understood at
first is a question, but these points were
well covered in the literature. The
Eastman Capstaff Densitometer,166 which
was developed primarily for measuring
picture negatives for contact printing,
reads diffuse densities. This would be
appropriate for measuring the densities
cf sound negatives for use in contact
printers, but not for densities of sound-
track prints, for it is the specularly trans-
mitted light which reaches the photocell
in
a
reproducer.
The widely used EastmanIIb Sensitom-
eter, brought out about 1932,1‘j7 which
gives an accurately standardized series of
test exposures in the form of a step tablet
with exposure time increasing in the
ratio
d?
per step, and ranging from
about
0.004
sec to
4
sec, has been of
utmost vahe in maintaining controls.
However, it does not simulate sound-
track recording conditions, where the
intensity is extremely high and the time
for average exposure was approximately
1/18000 sec (1/36000 sec with a later
light-valve system and in present prac-
tice about 1/90000 sec) and still shorter
for low exposures. The 1934 paper by
Jones and Webb165 gives an indication
of the magnitude of the error. The East-
man Sensitometer on the other hand gives
exposures which approximate sufficiently
well those which
a
print receives, and
are thus suitable for determining gamma
of contact prints. For many purposes it
has been satisfactory to draw conclusions
by applying correction factors, if needed,
to the readings of these instruments.
In the course of a few years densitom-
eters employing photocells were de-
veloped which had the advantages of
greater accuracy and much faster op-
eration than the Capstaff visual-balance
type.168-1i2 For exposing sound negatives
for sensitometry purposes, the light valve
itself, with suitable calibration, can be
used. The subject is again discussed
under “Intermodulation Test.”
While the conditions for low distortion
were to keep both negative and positive
exposures on the straight parts of the
H
&
D characteristics, studies reported
in 1931 by D. MacKenzie17*~ showed that
low distortion was still possible while
using the “toe” range of both films
(“toe recording”) or that of the positive
only (“composite”). Toe recording using
positive stock for the sound negative
might, if the recording-system light was
limited, be preferable to resorting to
faster and coarser-grained recording
stock. In the case of single-film systems
(sound recorded on the picture negative)
where the development of both the nega-
tive and positive soundtracks is fixed
by picture requirements, MacKenzie
found that the composite system offered
best promise of low distortion. Both the
toe and composite systems give higher
output than
a
classical
or
straight-line
system, but poorer signal-to-noise ratios.
It took a number of years to bring
about the full transformation from the
methods (depending much on visual
judgments) which had been employed
for making silent pictures, to the close
controls and scientific precision needed
for satisfactory and consistent sound.
The constant and close checking of
every element exerted
a
pressure for
improvement along the whole front,
including the manufacture of the film,
in which departures from uniformity
were quickly detected. The story is inter-
estingly told by
J.
I.
Crabtree.146 An
early account is given by
J.
W. Coff-
man.153
Processing, Variable-Area.
Since the
ideal variable-area track
is
part clear
and part black with a sharp boundary
between, there is no question of preserv-
ing correct shades of gray, but in general
the higher the contrast (or gamma
product) the better.
As
in the case of
variable-density tracks it must be as-
sumed that the print development will
be that which is wanted for the picture,
and that has been taken in general to
give a gamma of about 2.0. Variable-
area negatives as well as the prints are
processed in high-contrast developers.
The variable-area system is noncritical
with respect to gamma product but, for
a given positive emulsion and processing,
there is for any given negative
a
best
setting of printer light.
A comprehensive study of available
sound-recording films and their process-
ing was published by Jones and Sand-
~ik.’?~ Another study was made by
J.
A.
Ma~rer.’~~ From his curves it ap-
peared that negative densities of 1.3 or
higher were desirable, and the prints
which gave maximum outputs were the
ones having densities (in the dark areas)
about equal to those of the negatives
from which they were made. This held
true for negative densities ranging from
0.6 to
1.3
and higher. The maxima
however were very broad.
In November 1931, Dimmi~kl~~ re-
ported the results of a series of deter-
minations of conditions for maximum
output from a 6000-cycle recording, using
Eastman positive 1301 for negatives and
prints, and
4,
6,
8
and 10 min in D-16
developer. The study covered an ade-
quate range of the four variables
-
nega-
tive (recording) exposure, negative de-
velopment or gamma, printing exposure
and print development. The results
showed that wide ranges in each
of the variables could be used with com-
paratively small loss of output, but for
any negative there was a print density
at which output was greatest.
It
made
Kellogg: History
of
Sound
Motion
Pictures
comparatively small difference (except
near the extremes) whether a given
density of either negative or print was
reached with small exposure and longer
development or more exposure and less
development, but in general the maxima
were broader with the higher values
of gamma, especially that of the print.
The two highest gamma values in the
series, 2 and 2.18 of both negative and
print, in general gave best results, with
negative densities (measured in the black
areas) in the range 1.5
to
2, and print
densities a little less in each case than
that
of
the negative.
While maximum high-frequency out-
put is of less consequence than avoidance
of cross-modulation (which is discussed
in the section on distortion) it is of in-
terest that recommended practices based
on the test just described come very
close to those found to be best in later
experience and after current testing
methods had become established. The
cross-modulation test did not come into
general use until 1938.176
For a number of years a print density
of
1.4
or
slightly higher, with appro-
priate corresponding negative density,
was taken as a practical objective. As
galvanometers and optical systems were
improved and finer grain films came into
use, the tendency was toward higher
densities for both negatives and prints,
especially for the negatives.
Evolution
in
a
Growing Industry
Greatb Expanded Deuelopmental Actiui-
ties.
The development work prior to com-
mercialization of sound was carried
on largely in laboratories supported by
manufacturers of supplies or equipment,
or in independent laboratories, and it
was done on the basis of hope for re-
turns which might be realized either
through patent royalties
or
through
sales of equipment
or
both.
Once sound pictures began to be made
and shown, developmental work was on
a different basis. Research and investiga-
tions of numerous incompletely solved
problems took on rather the character of
plowing in profits, with greatly increased
total expenditures for research and par-
ticipation by all the major picture-pro-
ducing organizations.
Of a11
of
the problems, the most fun-
damental and greatest in magnitude was
learning how to use sound pictures,
or
the evolution of a new art. This is dis-
cussed by
J.
E. Abbott.”? The expres-
sion “growing pains” aptly describes the
less successful phase of this evolution.
Capacity for readjustment is one of the
qualities of greatness in individuals and
in organizations, and the motion-picture
industry came through splendidly.
When any industry becomes large,
and especially if its requirements are as
diverse as those of sound pictures, it
provides a market for numerous special-
365