MTF EXPOSURE DEVICE

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04770A000800040002-9
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 11
• Document Creation Date: December 28, 2016
• Document Release Date: July 2, 2004
• Sequence Number: 2
• Publication Date: March 18, 1965
• Content Type: REPORT

Approved Release 2004/07/29: CIA-RDP78B0W0A000800040002-9
PAR 234
MTF Exposure Device
18 March 1965
Copy No. J.
Declass Review by NGA.
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 Approvedr Release 2004/07/29: CIA-RDP78B 70A000800040002-9
PAR 234
18 Mar 65
SUBJECT: MTF Exposure Device
TASK/PROBLEM
1. Develop and fabricate a device to expose sine-wave test patterns
upon photosensitive materials as a step in the process of measuring MTF
of the materials.
PROPOSAL
2. Introduction: We propose to develop subassembly designs and to
design, fabricate and test an instrument of moderate size to expose sine-
wave target patterns upon photosensitive materials for use in measuring
Modulation Transfer Function (MTF) of that material. The technique described
by R. L. Lambertsl will be the basis of the proposed instrument.
3. Approach:
a. The objective lens proposed for the instrument is a 40mm
E.F., f/2.8, narrow field Petzval-type lens with a field flattener and
fluorite elements. This lens design has shown very excellent performance
with nearly perfect chromatic correction. One sample of the lens has pro-
duced MTF in the aerial image, as scanned with a one micron wide slit in
a microphotometer, as shown in Table 1. The performance indicated should
not be considered as a guarantee but as indicative of the design capability.
As described by Mr. Lamberts, a high-quality cylindrical lens is to be
placed in front of this lens to convert the variable-area test objects into
a variable-intensity exposure on the film sample.
b. Consideration will be given to using the 40mm E.F. lens de-
scribed in paragraph 3 at 50:1 reduction from the test objects to the ex-
posed image to provide test exposures having 2.5 to 400 cycles/mm. With
this arrangement, it appears practical'to mount the test object holder,
the lens and the film sample holder upon a common rigid beam (about seven-
feet long). The beam can be shock mounted to isolate that critical portion
1R. L. Lamberts, Jl. Opt. Soc. Am., 49, 425 (1959): Reprint attached.
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 Approve 'Or Release 2004/07/29: CIA-RDP78B?OA000800040002-9
PAR 234
18 Mar 65
Performance' of 40mm E.F. Petzval/Fluorite Lens Sample
Spatial Frequency
MTF*
Blue
Green
50 cycles/mm
.82
.80
100 cycles/mm
.72
.68
200 cycles/mm
.56
.51
300 cycles/mm
.42
.36
400 cycles/mm
.30
.21
.79
.66
.48
.31
.12
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 ApprovedSr Release 2004/07/29: CIA-RDP78B70A000800040002-9
PAR 234
18 Mar 65
of the instrument from building vibration.
c. The proposed lamphouse for the instrument will provide a con-
denser lens to illuminate a test object area about 3.75 inches x .50 inches.
The light source will be a tungsten filament lamp with a high-diffusion
envelope, such as the General Electric Photoenlarger Lamps, Code PH/211,
Code PH/212, or Code PH/213. The PH/212 lamp will provide test object lu-
minance of about 6,000 foot-lamberts (envelope luminance). Means will be
provided to reduce the test object luminance by 2.5 to 3.0 decades below
that level. Insertion of diffusors, aperture plates, change of lamp position,
etc., will be considered for this control. A shutter system providing pre-
cise exposures times from 1/30 second to 110 seconds will be provided. Two
filter wheels and a position for the manual insertion of filters will be
provided. It is expected that one filter wheel will contain neutral density
"Inconel" filters of approximately 0.0, 0.3, 0.6 and 0.9 to permit convenient
production of an exposure series of the sine-wave test patterns. The second
filter wheel can be used for color filters to simulate various photographic
situations such as red, green, and blue separation filters for color photog-
raphy, simulation of Wratten 12 filter plus daylight illumination used for
black-and-white aerial photography, etc.
d. In the proposed camera arrangement, the various test patterns
are exposed upon the film sample sequentially. After each exposure, the
test object must be changed and the film sample advanced for the next exposure.
We propose to build and test breadboards of:
(1) Test object changer mechanism,
(2) Mechanism to hold and advance 35mm x 12-inch film samples,
(3) Adjustable luminance light source system,
before starting design of the complete instrument.
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 Approved#jr Release 2004/07/29: CIA-RDP78BW 0A000800040002-9
PAR 234
18 Mar 65
e. The instrument will have provision for making microphotometer
measurements of the modulation obtained in the aerial image formed by the
objective-cylinder lens assembly for each of the test targets delivered with
the instrument. These calibration measurements will be made before delivery
of the instrument and are expected to hold for an extended period of time.
f. The instrument controls and sample handling will be carefully
designed for convenient operation in darkroom conditions. In order to permit
the most convenient sample handling, it is proposed to completely enclose
the light source and optical path of the instrument and to leave the sample
open to room conditions during the exposure cycle. This has proven to be
the most convenient arrangement for similar film testing devices, such as
sensitometers and the like.
PROGRAM OBJECTIVE
4. Phase 1: Develop, fabricate and test one (each) breadboard model
a. Test object changer mechanism.
b. Mechanism to hold and advance 35mm x 12-inch film samples.
c. Adjustable luminance light source system.
5. Phase 2: Design, fabricate and test one prototype instrument to ex-
pose sine-wave test patterns on photosensitive materials as required for
measuring MTF values for those materials. Design drawings and sketches will
be prepajed as required by the contractor's model shop to complete fabrication,
assembly and test.
SCHEDULE
6. A tentative schedule covering major phases of effort is shown in
Figure 1. The time span indicated to complete Phase 1 of the subject program
is based on actual start of work. Upon approval to proceed and/or start of
work, schedule will be reviewed and necessary changes reported as required.
Phase 2 schedule will be reviewed at or near the completion of Phase 1 and
necessary changes will be reported.
Approved For Release 2004/07/29 : CIA-RDP78B04770A000800040002-9
 Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
MTF Exposure Device
1. PHASE I
a. Engineering
b. Breadboard Fab.
c. Reports
(1) Monthly
(2) Quarterly
2. PHASE III
a. Inst. Design Eng.
b. Calibration Eng.
c. FabricIation
d.. Test
e. Delivers
f. Reports
(1)
Quarterly
(2) Mnthly
(3)
(4)
TENTATIVE SCHEDULE
0 1 2 3
,MONTHS PAR 234E
18 Mar 65
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Figure 1
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
-5-
 Approved(or Release 2004/07/29: CIA-RDP78B'0A000800040002-9
Attachment to
PAR 2311.
Reprinted from JOURNAL OF THE OPTICAL SOCIETY OF AMERICA, Vol. 49, No. 5, 425-428, May, 1959
Printed in U. S. A.
Measurement of Sine-Wave Response of a Photographic Emulsion*
ROBERT L. LAMBERTS
Research Laboratories, Eastman Kodak Company, Rochester, New York
(Received November 13, 1958)
A method for experimentally determining the sine-wave response of a photographic emulsion is described.
It consists in determining the sine-wave response of a lens, using the lens to photograph a sinusoidal test
object on the emulsion under test, determining the response of the lens-emulsion combination, and dividing
out the response of the lens. Since sine-wave response is defined in terms of relative exposure in the emul-
sion, it should be substantially independent of development conditions and the exposure level when adja-
cency effects are absent. This expectation has been confirmed experimentally. It is also shown that the sine-
wave response of a lens and the response of an emulsion can be combined to predict the characteristics of
the lens-emulsion combination.
T HE application of communication theory to the
microstructure of optical images has provided
not only a better theoretical understanding of the proc-
ess of image formation but new means for the practical
measurement of images as well. This approach is very
useful in studying the photographic system because if
one knows the sine-wave response or transfer function
of each component of the system, the sine-wave re-
sponse function for the complete system is simply the
product of the individual response functions. In other
words, it becomes possible to "engineer" the micro-
structure of the photographic system.
However, until the present time, the majority of the
experimental work has been done with the response
characteristics of lenses and comparatively little work
has been reported for the sine-wave responses of emul-
sions. A single response curve for Kodak Panatomic-X
Film was reported by Ingelstam, Djurle, and Sjogren.t
Since the sine-wave response function is equivalent
to the Fourier transform of the spread function, the
sine-wave response and the spread function represent
to a considerable extent the same data-strictly so in
this instance, where the emulsion spread function can
be assumed to be symmetrical? It is then theoretically
possible3 to obtain the desired data by using either a
line image or an edge image to measure the spread
function, or by using sinusoidal patterns to measure the
sine-wave response. The selection of the method is a
matter of convenience, and there appear to be several
distinct advantages to using the latter method: (1)
The sinusoidal images can be made to cover an ex-
tended area of the film so that random fluctuations
arising from granularity can be averaged out, thereby
increasing the accuracy of measurement. (2) Since the
resulting data are in the form of a sine-wave response
function, the effect of the system used to put the images
* Communication No. 1988 from the Kodak Research Labora-
tories. Presented at the Washington Meeting of the Optical
Society of America, March 27-29, 1958, under the title, "Measure-
ment of Sine-Wave Response and Spread Function of a Photo-
graphic Emulsion."
Ingelstam, Djurle, and Sjogren, J. Opt. Soc. Am. 46, 707
(1956).
2 R. L. Lamberts, J. Opt. Soc. Am. 48, 490 (1958).
3 Lamberts, Higgins, and Wolfe, J. Opt. Soc. Am. 48, 487 (1958).
onto the film can be divided out of the system by a
simple procedure.
The principle of the convolution of images assumes
the simple additivity of spread functions. For this
reason, the spread function or the sine-wave response
of an emulsion must be expressed in terms of the light
falling onto the film or the exposure of the film rather
than in terms of the density or transmittance of the
developed image. The emulsion is then used as its own
photometer; a given density in the developed image is
taken to represent a certain exposure given to the
emulsion. The line spread function of an emulsion will
then be defined as the one-dimensional distribution of
light or exposure within the emulsion if an infinitely
narrow line of light were to be imaged onto the film.
The distance coordinate expresses the distance from
this line along the emulsion surface.. Since an emulsion
is almost always utilized as a plane, it is justifiable
from the viewpoint of application to disregard the
variation of exposure as a function of depth in the
emulsion.
Figure 1 shows a simplified sketch of the experimental
setup used for making sine-wave response exposures. A
test object of the form shown in the enlarged scale is
imaged by means of a high-quality objective lens. Since
the illuminance along the scanning slit must be con-
stant to give a trace like a variable-density motion-
picture sound track, a cylindrical lens is inserted in
front of the objective lens to smear the image in a
FIG. 1. Schematic drawing of apparatus used for recording
sinusoidal image on film or (alternatively) scanning it. A sketch
of the test-object pattern is shown on an enlarged scale.
-6-
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 Fio. 2. Test-object
patterns used to cal-
ibrate the microden-
sitometer. Each pat-
tern produces an
exposure of twice the
one above.
vertical direction. As was described in a previous paper,'
it is possible on the lens bench in these laboratories to
scan an aerial image photoelectrically with a narrow
slit, and also to replace this scanning unit with a camera
unit so that the photographic emulsion falls into exactly
the same position as was previously occupied by the
slit. This arrangement makes it possible to determine
the modulation of the sinusoidal patterns which fall
onto the emulsion.
The test object, as shown on an enlarged scale at the
bottom of Fig. 1, has a slit running parallel to the
length of the pattern. This slit is used to reduce the
contrast or modulation within the aerial image for the
lower spatial frequencies, because without it, the very
large exposure difference resulting between the light
and the dark portions of the image would exceed the
latitude of the film and make photographic photometry
impossible. For spatial frequencies higher than about
50 lines/mm, it was not found necessary to use such
slits since the lens itself reduced the exposure amplitude
by the necessary amount.
The objective lens was a 40-mm, f/2.0 lens of very
high quality. This lens was operated at an aperture of
f/2.8, where the quality was sufficiently high to make
possible imagery up to a spatial frequency of 400
lines/mm. Since this lens was not apochromatic, it was
necessary to use a restricted spectral region. All data
shown in this paper were determined with green light
(Kodak Wratten Filter No. 61).
A -4.0 diopter cylindrical lens was sufficient to
smear the image adequately. By reason of this smear-
ing, the width of tie clear portion in the test object
determines ,the illuminance in the image. With this
principle as a basis, a series of calibrating test objects
was prepared by printing a given line onto Kodalith
Plates to produce one slit, two slits, four slits, and the
equivalent of eight slits, as shown in Fig. 2. In like
Approved Release 2004/07/29: CIA-RDP78B'0A000800040002-9
Attachment to
PAR 231+
ROBERT L. LAMBERTS Vol.49
manner, a second series was also prepared with each
slit v2 times as wide as those in the first series. These
provided a calibration with a range of log E from 0 to
1.05 in increments of 0.15. These calibrating slits have
advantages over ordinary neutral densities in that they
have the same spectral and light-scattering character-
istics as the sinusoidal test objects.
Before the sinusoidal patterns were photographed
onto the emulsion, the system was analyzed photoelec-
trically to determine the modulation of the aerial image
for each spatial frequency. The point of best focus was
chosen to be the point of maximum amplitude for the
100-line/mm frequency in the aerial image. At this
focal setting, modulation values for each spatial fre-
quency were determined and tabulated.
To assure correct focus of the camera, a 100-line/mm
sinusoidal pattern was likewise photographed at a
series of focal positions on a very fine grained emulsion.
These images were then scanned with the microden-
sitometer to determine the focal position giving the
maximum amplitude in the developed image.
With the camera thus in correct focus, each member
of the series of sinusoidal test objects, as well as the
series of calibration slits, was individually photographed
on the film to be tested. The camera was made to space
exposures closely enough so that all the calibrating and
sinusoidal exposures could be made within about 1Z in.,
thereby minimizing any processing variations. Since all
the exposures were made for the same exposure time,
errors caused by reciprocity failure were eliminated.
The processed film was then scanned with a micro-
densitometer.4 The density patches obtained by pho-
tographing the calibration slits were first scanned to
0.3 0.6 0.9
Log relative E
FIG. 3. Microdensitometer trace of sinusoidal image and char-
acteristic curve of emulsion used to record the image, showing
how the trace is interpreted in exposure terms.
-7-
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
 Approved o of Release 2004/07/29: CIA-RDP78B'0A000800040002-9
Attachment to
PAR 231+
provide a calibration of the microdensitometer itself
in terms of exposure in the image. It was then possible
to draw up a calibration curve of the microdensitom-
eter chart reading as a function of log exposure in the
image, as is shown in Fig. 3.
The scan of each spatial frequency of sinusoidal
patterns was then examined to determine the maximum
and minimum values on the recording paper. Values of
logEmax and logEmin were then determined by using the
calibration curve just described. We can write
A logE= logEmax- logEmin
= log (Emax/Emin)
The modulation of the photographic system consisting
of the test object, the lens, and the film is, by definition,
Emax
-1
Emax-Emin
Emin
Emaa-f-Emin
Emax
-+1
Emin
This expression was evaluated readily from the anti-
logarithm of A log E.
To obtain the sine-wave response of the film alone,
the values so obtained were divided by the modulation
values measured for the aerial image by the photoelec-
tric scan described earlier.
50 100 150
Spatial frequency (lines/mm)
Fro. 4. Upper graph: Sine-wave response data for Eastman
Plus-X Panchromatic Negative Film, Type 4231, exposed behind
a Kodak Wratten Filter No. 61 and developed in Kodak Devel-
oper D-19 for (0) 2 min and (0)'10 min. Lower graph: Similar
data for a development time of 4 min and exposures to give net
densities of (0) 0.35 and (o) 1.70.
The sine-wave response of Eastman Plus-X Pan-
chromatic Negative Film, Type 4231, is shown in Fig. 4.
In the upper graph, the circles represent the response
derived from a sample developed in Kodak Developer
D-19 for 2 min at 68?F while the triangles represent the
response for a development time of 10 min.t The de-
veloper was agitated vigorously to reduce adjacency
effects. It is apparent that there is good agreement
between the two curves in spite of the differences in
development time. The slight rise of the curve above
100 percent for low frequencies can very probably be
attributed to adjacency effects.' Where high-exposure
regions are adjacent to low-exposure regions, there is
an accentuation of density differences compared with
those between large areas receiving comparable ex-
posures. This is exemplified by the Eberhard effect.
The present data show that curves for a given emulsion
with moderate differences in their low-frequency rise can
usually be matched closely for the entire region of higher
frequencies by a simple ratioing of the ordinates.
Likewise, it should be pointed out that the amount of
bromide in the developer will tend to control the amount
of adjacency effects. The higher the existing concen-
tration, the smaller will be the effect of the bromine
liberated in processing. Although a detailed study of
adjacency effects was not made, it is apparent that
sine-wave response techniques have possibilities for
appraising these effects.
Exposure time likewise has little effect on the sine-
wave response, as can be seen from the lower graph of
Fig. 4. In this case the exposures were such as to give
an average density in the sinusoidal-pattern exposures
of 0.35 for the circles and 1.70 for the triangles when
the films were developed in D-19 for 4 min.
The type of developer likewise appears to have little,
if any, effect upon the emulsion sine-wave response.
Measurements made for films developed in Kodak De-
veloper D-76 showed essentially the same curve as for
.D-19 except for a difference in rise attributable to
adjacency effects.$
Figure 5 shows a comparison of the sine-wave re-
sponse curves for several Kodak films. The extreme
condition is represented by the fine-grain experimental
documentary recording film (curve 4). This film is
almost as fine-grained as Kodak High Resolution Spec-
troscopic Plates, and the flatness of this curve confirms
t The data presented in this paper are representative of the
emulsions manufactured at the time of writing. However, it must
be recognized that the characteristics of products of the same
name may vary within manufacturing tolerances and may change
significantly as improvements are effected.
B C. E. K. Mees, The Theory of the Photographic Process (The
Macmillan Company, New York, 1954), revised edition, pp.
1031-1038.
t From the early data, reported in the oral paper, it appeared
that the form of the sine-wave response curve changed in going
to fine-grain developers such as D-76. A more thorough recheck
of these characteristics, however, showed no differences that
seemed to be experimentally significant.
Approved For Release 2004/071-2;-- CIA-RDP78BO477OA000800040002-9
 Approved
Release 2004/07/29: CIA-RDP78BW0A000800040002-9
Attachment to
PAR 231+
00 '50 200
Spatial frequency (hales/mm)
FIG. 5. Sine-wave response of the following Kodak films de-
veloped in the following Kodak developers: (1) Tri-X, DK-50,
5 min; (2) Panatomic-X, D-19,3 min; (3) Micro-File, D-11, 4
min; (4) Experimental document-recording, DK-50 (1:10), 10
min. Exposed through Kodak Wratten No. 61 Filter.
a general observation by users of such emulsions that
the image quality is much superior to practically any
optical system that can be used with them.
It should be noted that the errors arising from the
finite slit widths of both the microdensitometer and of
the photoelectric unit of the lens bench tend to com-
pensate for each other in the final result. The 1-1A slit,
as used for the photoelectric unit of the lens bench, is
of the same width as the slit generally used with the
microdensitometer and with which the present samples
were scanned. The slit in the photoelectric scanning
unit gives values of response that are lower than the
true values. Likewise, the microdensitometer has a
tendency to give values of response that are slightly
lower than the true values obtained with an infinitely
narrow slit. Since the microdensitometer data are di-
vided by the lens-bench data, these slit effects tend to
cancel each other.
COMBINATION OF LENS AND FILM
CHARACTERISTICS
To illustrate how the emulsion characteristics can be
combined with the lens characteristics and to show
how the combined data can be used to predict the
density-distance relationship in the image of an edge,
the following demonstration was prepared:
By using Plus-X emulsion and with the lens on the
lens bench placed a few thousandths of an inch away
from the position of best focus, an edge with an illumi-
nance ratio of about 4:1 was photographed and the
developed image was scanned on the microdensitometer.
The sine-wave response of the lens at this focal position
was then determined by scanning sinusoidal test objects
of 100% modulation. By making the exposures with
the lens slightly out of focus, the curve of sine-wave
response was made to have a different shape from that
used in obtaining the response of the emulsion. The
new response curve was multiplied by the response
curve of the emulsion obtained as described in the pre-
ceding section, and the Fourier transform was made of
the combined response to obtain the spread function of
the lens-film combination. This spread function was
then integrated into an edge distribution,' the required
amount of flare light was added to give the correct
contrast, and, finally, the characteristic curve of the
emulsion was applied to give the density-distance
function. A comparison of the measured and calculated
edge curves is shown in Fig. 6. The smooth curve was
T
C 1.5
0 50 100
Distance (,a)
FIG. 6. Comparison of computed edge trace for a certain lens-
film combination (smooth curve) with experimental trace
("grainy" curve).
calculated and the "grainy" curve was made with the
microdensitometer. The good agreement between the
measured and the calculated curves demonstrates that
the shape of the edge trace for a lens-film combination
can be predicted from the sine-wave response data.
CONCLUSIONS
It is possible to use the techniques described here for
measuring the sine-wave response of an emulsion with
good accuracy and reproducibility. Except for adjacency
effects, the sine-wave response of an emulsion tends to
be independent of exposure level and processing con-
ditions. The sine-wave response function of the emul-
sion can be used to predict the characteristics of a
lens-film combination if the sine-wave response of the
lens is also known.
Approved For Release 2004/07/29-- CIA-RDP78B04770A000800040002-9
 Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9
Approved For Release 2004/07/29 : CIA-RDP78BO477OA000800040002-9