PROJECT(Sanitized)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP67B00820R000300100001-1
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 24
• Document Creation Date: December 15, 2016
• Document Release Date: September 11, 2003
• Sequence Number: 1
• Publication Date: October 1, 1961
• Content Type: REPORT

Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
STATINTL
Supplementary Report
October, 1961
Prepared by:
STATINTL
STATIN L
Approved
USAF review(s) completed.
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
STATINTL
Supplement
11STATINTL
This report contains additional and supplemental material as requested
by the D/I and 544th RTGp for Project The following topics
are covered in this report:
The Control Extension Capabilities of the 544th RTGp
Communication between the Photo Producer and the Photo User
Calculation of the Printer Resolution for the SP/10/70 Printer
and SO-278 Film
General Procedure for Testing Duplicating Equipment
Edge Acutance and the Effect of Edge Degradation on Mensuration
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
II. The Control Extension Capabilities of the 544th RTGp
SiTATINTL During
investigation under Project
of STATINTL
the control extension capabilities of the 544th RTGp, an over-all flow
diagram was developed. This diagram included the operations of the
analysis center at SAC Headquarters, the operations at the Photo Records
Services Division (PRSD), ACIC, and AMS. This diagram indicated a command
function, a preliminary operational function, and a final operational
function. Between the preliminary and final operational phases, there
is a time delay while information is furnished to the 544th RTGp from
PRSD and ACIC.
As a part of the total analysis of the capabilities of the research
center, a detailed test program was conducted using simulated photographic
inputs. These inputs utilized information which duplicated, as nearly
as possible, the actual inputs to the research center. The results of STATINTL
this detailed investigation, as reported in the final Project
Report, proved a high degree of capability exists within the research
center for performing the required tasks of target location by photo-
grammetric control extension means.
The major points which perhaps were not stressed sufficiently in the
final report were the comparative capabilities and the time requirements
for all phases of the target location program. In evaluating the capa-
bilities of the research center, it is important to delineate those
functions wherein complete control is maintained within the unit and
those functions under which this unit has no command control. Such an
analysis indicates that the present capabilities of the center are suf-
ficient to perform the required task in this program. This sufficiency
includes all relevant factors--such as equipment, personnel, training,
space, time, and maintenance capability. The investigation pointed out
the fact that the research center is not the gating factor in this pro-
gram. That is, the services requested from PRSD and ACIC will control
the quantity, and to a certain extent, the quality of work that this group
is capable of performing. The selection of equipment, the training of
the personnel, and the overall capabilities of the group show a high de-
gree of proper planning in the building of such an effort.
This unit, as presently complemented, is entirely capable of keeping
abreast of present methods and of investigating new methods to determine
their suitability for use within this program.
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
III. Communication between the Photo Producer and the Photo User
Maximum intelligence yield from a given photographic product can be
achieved only if the analyst or interpreter is given the best print or
transparency for the particular job at hand. Depending on the require-
ments, the type of photo product best suited for his needs may vary widely.
This is most true when the analyst is struggling with small, low-contrast
detail, or when the detail in question is located at either extreme of
the exposure range. Frequently, an additional print or set of prints
made to meet a special requirement may provide substantial aid.
The initial transparencies or prints which the interpreter receives
must be made to present the best average picture. Having located a spe-
cific problem area, the interpreter can then request special service from
the Photo Lab. Here lies the pitfall. Unless the interpreter can state
in language the photo technician understands, what his needs are, he may
or may not receive the best product. Further, he may not know that it
is not the best product, unless he has at least a basic understanding of
photographic processes and materials, and is equipped to recognize defects
in his prints. His ability to wrest information from the picture can be
affected by a whole chain of events, starting with the planning of the
mission, and carrying through the illumination and viewing system employed
during his analysis.
In order to improve the special secondary prints, as well as securing
new prints when defects are found in the originals, the interpreter must
be able to recognize the defect, and to have a good idea of what he re-
quires to get improvement. Two examples may point this out. A slight
amount of slippage between original negative and duping film during the
continuous printing operation may produce a narrow, blurred band across
the print--just in a critical area. The analyst may accept this as a
vagary of the system--and fail to secure the maximum information from that
critical area. Or, if properly informed, he may immediately recognize
the defect, know that a new print will cure it, and within an hour or so
be working with a good print.
In another instance, critical information may lie in a bright, sunlit
sandy area. The average print may fail to show detail in this area--not
because it is missing in the original negative, but because the print ex-
posure was too light, or the paper too high in contrast. The informed
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
interpreter will ask for a new print--made particularly to show the cri-
tical detail at its best. It may not be sufficient merely to ask for ,a
darker print--best results will be mad if he actually indicates the par-
ticular detail that is of interest. Granted that it will take more time
and care to prepare his request, but the results will in most cases more
than justify it.
As an aid to the interpreter who must make his request through the
medium of written orders, it is feasible to prepare special order forms,
in which a number of pertinent facts or comments can be checked, and the
location of the particular detail or region of interest noted. It will
not be sufficient for the interpreter to request, "make a contrastyprint
of frame XYZ", or "a maximum enlargement of the airfield in frame XYZ".
It is better that he specify a print "to yield best detail contrast in
the forested area of frame XYZ", or "optimum enlargement to include the
entire airfield and supporting areas", or still better, "optimum enlarge-
ment and print contrast to accomplish an aircraft identification and
structure analysis".
The comments above apply also to the situation in which measurements
are to be made of small detail. Since the accuracy of measurement is
affected by the edge contrast, it follows that an improperly exposed
print can introduce an error into the measuring process. This is par-
ticularly true on a paper print, because the dynamic (density) range of
paper is appreciably lower than transparency material. For example, the
maximum reflection density a glossy paper can reach, with very heavy ex-
posure, is about 1.6; S0-278 duplicating film has a maximum density of
about 2.4, and the faster G-2 duplicating film has a maximum density of
over 3.0.
One course of action that could be taken to improve the ability of
the interpreter to communicate his needs to the photo technician would
be to prepare a manual for him which tabulates the defects he might en-
counter, and indicates the correction procedure, if any is possible. The
manual should also include original photo prints which illustrate these
defects, and the corrected appearance. The value of such a manual would
suffer if illustrations are not original prints, since much of the signifi-
cant detail would vanish in the half-tone screen reproduction.
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
Following is a table of common defects, separated into Pre-Printing,
Printing (Contact Printers), and Printing (Enlarging Printers).problems,
with suggested correction procedures.
A. Pre-Printing Defects
1. Underexposure Print lacks contrast markedly in the shadow
or dark areas. If taken under heavy haze,
entire print may be very flat, still lack-
ind shadow detail. Correction is difficult
--dark and light print may help.
2. Image Motion If due to excessive vehicle velocity for
shutter speed used, blur will be parallel
to flight direction. If due to vibration,
blur will be random, having appearance of
defocused image.
3. Image Defocused Image (in original negatives as well as
prints) will lack sharpness, especially
when viewed under magnification. Cannot
be corrected, except by sophisticated image
enhancement techniques.
4. Overexposure Negatives may be dense. Print may show ex-
cessive contrast; highlights blocked up.
Correction is difficult--use lower contrast
paper. In extreme case, chemical treatment
of negatives may be helpful.
5. Overdevelopment Negatives may be very dense. Prints may
show ppor contrast in dark areas, but ex-
cessive contrast in highlight areas. Cor-
rection. idfficult. Make selective prints
for dark and light areas.
6. Light Fogging Prints may show light streaks, with loss
in contrast in light areas. Definite pat-
tern may or may not appear.
7. Static Prints may show either very fine lightning--
like patterns, or diffuse round areas. No
correction except to minimize build-up in
static charges in camera system, or when
film is unwound, prior to processing.
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
8. Black Lines If in print, probably due to scratched
negatives. If parallel to flight
direction, occured in camera or during
processing. If random, occured from
careless handling of negatives, or
from rough handling of print prior
to development.
B. Printing Defects (Contact Printers)
1. Incorrect Exposure Print may appear generally too light
(Refer to Figure 2) or too dark. Note, however, that in
many cases, certain detail will be
better rendered.
2. Incorrect Contrast Highlights washed out, shadow detail
(Refer to Figure 3) black. Remedied by change in paper
grade, or use of lower contrast de-
veloper, or both. Again, contrasty
prints may be best for certain detail.
3. Poor Contact Prints may show random areas of poor,
(Refer to Figure 4) unsharp image. Most likely at edges.
Correction is to make new print, check
equipment for tension or pressure.
4. Slippage If from continuous printer, print will
(Refer to Figure 5) show band of blurred image, the width
dependent on the amount of slippage.
Often may be so slight detection is
difficult. Correction is to check
equipment, make new print.
5. Printing through Base Image will be laterally reversed, and
(Refer to Figure 6) there will be loss in sharpness. Amount
of degradation strongly dependent on
type of printer. Correction is to
make new print. Note: If original
negative was exposed in system with
odd number of mirrors, image will be
laterally reversed in negative, and
will appear correct when printed through
base.
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
6. Print or Transparency Print was fogged (exposed to unsafe
Background Grey room light., or too bright safelight),
or material was improperly stored.,
or is too old. New Print required.
7. Print Background Print received insufficient fixing,
Yellow or solution was exhausted, or re-
ceived insufficient washing, or was
dried at excessive temperature. Cor-
rection is to make new print.
C. Printing Defects (Enlarging Printers)
1. Lack of Sharpness Printer improperly focused. Can be
(over entire print)
separated from camera focus problem
by checking if silver grain structure
is sharp. If so, enlarger focus is
OK., original negative defective. If
grain appears blurred, new print
called for. May also be due to vi-
bration, particularly if magnification
is great or negative dense, requiring
long exposure time.
2. Lack of Sharpness May be due to non-flat negative., or
(localized) damaged optical system, or non-flat
paper or film in easel. Calls for
new print, attention called to defect.
3. Flat., Lacking in May be wrong paper or film grade., or
Detail Contrast may be due to dirty optical system.
Calls for new print, attention called
to optics.
4. Print Shows Excessive Magnification to high for negative.
Graininess
Excessive magnification can actually
impair ability to see fine detail.
Call for print at lower magnification.
5. Detail Too Fine to Insufficient magnification used.
be Seen Easily Calls for print made at higher magni-
fication--possibly photomicrograph
under critical situation.
-7-
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
D. A Manual of Photographic Examples
Experience has shown that nothing is as effective in pointing out
the deficiency in a photograph as showing the defective photograph itself.
Constructive comments are equally well effected by a corrected print.
The interpreter could profit from a manual which contains a series of
original prints, each one illustrating one type of defect and showing a
corrected print - if the defect is one that can be corrected in the dup-
licating process. This manual should contain original photo prints, not
screened offset reproductions, since the printing process may well destroy
the effectiveness of the presentation.
To illustrate what we propose, Figures 1 through 6 are included.
Each figure demonstrates a particular error - some of them made in the
original camera, such as image motion; others demonstrate the effect of
paper contrast, or printing exposure, or printer failures, such as slip-
page.
The Special Projects Lab is in an excellent position to adapt
this kind of presentation to the special requirements of the 544th TRGp,
and certainly has the capability to prepare a very useful manual of this
type. The examples shown here of course represent a small part of the
total spectrum of reference material that could be prepared for the inter-
preter. With such a reference, the communication between the photo pro-
ducer and the user should improve substantially. In many cases, the inter-
preter's order blank for additional photo aids could refer to specific
examples in the manual.
Approved For Release 2003/12/'1%': CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
qq~
s
"Q 1 ,9~ri~t ^~_ I p 0 0 1'
Figure I
Image Motion
Regsires Slight Magnification to be Readily Detected
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
-9-
 proved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
Approved For Release 2003/12/~gXpd(, -WDP67B00820R000300100001-1
Figure 2. Effect of Varying Print Exposure Time
Note that all prints contain areas of optimum exposure for certain detail.
 4XI
try, WN,
21 WA, .4w - t-I
High-Contrast Paper
Normal Paper
Figure 3. Effect of Choice of Paper Contrast
Approved Fo fWg@sp2j(a03//3Lt8 f qp? 9,9 gR000300100001-1
 Figure 4. Effect of Failure to Maintain Good Contrast between
Negative and Print during Exposure
Approved For Release 2003/12/1$1:2GIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
Localized areas of image degradation due to
mechanical slippage in continuous roll printer.
Gross Mechanica]_ Slippage
/. baa. t
Lesser Degree of Mechanical Slippage
Slippage Eliminated
Figure .5. The Effect of Gross and Slight Slippage between Original Negative
and eMIAo6dkOWOaTA
 Approved For Release. 2003/12/18-cIA-R
WAR; W!~,`
'r?r ~. '~3 ~t
M,
:Image Laterally Reversed
Corrected Image
PigZure 6. Effect of Printing through the Film Base
Note degrpy@4 Eqr f @$I;pe~t0L0 3/a ~ t At~i't ~71 ~$~~4~9A~ Q31gfl009i~n) .
-14-
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
IV. Calculation of Combined Resolution for SP 10/70 Printer and SO-278 Film
By making the assumption that the frequency response of each compo-
nent in a reproduction system is Gaussian in nature, it is possible to
derive mathematically the following expression for the combined resolu-
tion of a duplicating system:
12 12 + 12 + 12
Rc Ri Rp Rf
where R = combined resolution
c
Ri = resolution of original film image
Rp = resolution of printer
Rf = resolution of dupe film
If one can obtain values of Rc experimentally, as for example by the
procedure for printer testing outlined in another section of this report,
and if the individual values of Rf, R. are known, then the value of Rp
can be calculated.
STATINTL In the study, values of RC were obtained on several
printers, through several generations. In these tests, however, the
value of Ri was deliberately made low, by preparing a specially degraded
resolving power target. Two test films were used--one with a limiting
resolution of 48 1/mm, and one with a limit of 38 1/mm. The purpose of
these tests was not to obtain data from which to calculate numerically
the'krinter resolution", but to determine the effect of the printer on
test films which contained degraded targets simulating the limit of
typical reconnaissance photography at the time of the investigation. As
the Final Report showed, all the printers tested produced STATINTL
notable losses in resolution from these special test films.
In order to.determine the combined resolution for the SP 10/70 printer
with SO-278 film, the most accurate way would be to follow the procedure
described elsewhere in this report, using a high-quality, non-degraded
resolution test target with a resolution range which extends beyond the
capability of the duplicating film.
Frequency response, also known as sine-wave response, is a measure of
the ability of the lens, film, or printer, to reproduce the contrast in
the original image as the detail size decreases, or the frequency increases.
It is frequently expressed by plotting contrast (response) versus fre-
quency.
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
An alternate approach is to use the data obtained with the degraded
target, and to solve the equation shown earlier for R P . Referring to the
test data for the SP 10/70 printer, we find that the combined average
resolution in the first-generation printing was about 39 1/mm. Using
this value for R
c, and 100 1/mm for Rf (Type G-2 duplicating film), and
48 1/mm for Ri (the limit of the degraded test target), and substituting
in the reciprocal squared equation, one can calculate that the printer
resolution Rp is 90 1/mm. Having found this, and taking the high-contrast
resolution for SO-278 film as 335 1/mm, the same expression can be used
again to find a new film/printer combined resolution. This value is
calculated as follows:
1 __ 1 1
R2 R2 + R2
C p f
1 1
902 3352
R = 87 1/mm
c
In the following graph, the combined printer/film resolution is
plotted against printer resolution, for SO-278 and for G-2 film. This
data shows the combined printer/SO-278 resolution to be 87 1/mm, and
the combined printer/G-2 resolution to be 76 1/mm. This latter value
does not, and should not, agree with the experimental combined average
resolution of 39 1/mm mentioned above, because the experimental value
was determined from a degraded test target rather than a high-quality
test target. From the graph, note that the film choice becomes important
for printer resolutions higher than about 80 1/mm.
A useful set of values to bear in mind are the following: When one
component of a system is three times better than the other, the degrada-
tion caused by it will by 5 per cent or less. This is so small that it
is difficult to verify it experimentally. Remember that these numbers
are based on the reciprocal squared relationship, which is valid if one
assumes that the frequency response curves have a Gaussian shape.
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
PRINTER RESOLUTION (R )
Figure 7
Combined Printer-Film Resolution as a Function of Printer Resolution
for Two Duplicating Films, S0-278 and Military Type G-2, based on the
Reciprocal Squared Relationship
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
V. A General Procedure for Testing Duplicating Equipment
The procedure described below is intended primarily for testing of
continuous roll film printers. Modifications of the procedure for test-
ing step-and-repeat printers, or hand-operated, single-frame printers,
or enlarging printers will suggest themselves.
The procedure can be readily carried out by the photo lab personnel
who operate the 544th RTGp Special Projects Lab. The test will determine
the ability of the printer to reproduce a resolving power target pattern
over the full film width.
A. The Test Film
The test film shall consist of a length of the appropriate width
high-resolution film, consisting of a minimum of six feet of leader and
trailer with an array of high-contrast resolving power targets in the
center, dispersed over a typical frame area. (For example, 9 by 18 in-
ches, or 9-1/2 inch film, or 2-1/4 by 2-1/4 inches on 70 mm film.) The
target array should contain at least 13 targets, arranged according to
MIL STD-150A, with one major exception. Since the MIL STD-150A layout
is intended for lens testing, the off-axis patterns are rotated so that
the lines are radial and tangential with respect to the axis. For printer
testing, the targets should be arranged so that the groups of lines are
parallel and perpendicular to the direction of film travel. This will
simplify the analysis. The test film must be made on Eastman Kodak
High-Definition Duplicating Film (SO-105), or Eastman Kodak 649 Spectro-
scopic Film, or an equivalent film having a resolving power capability
of at least 1,000 1/mm.
B. Test Procedure
1. The test film should be spliced with additional leader, and
threaded through the printer so that when the target area reaches the
printing area, the instrument will be running under normal conditions.
2. The unexposed film type should be similar to that normally
used for duplication, if representative resolution values for normal
operation are desired. If maximum printer and film performance are
sought, then special slow duplicating films such as Eastman Kodak SO-278
or SO-105 should be used--provided the printer has sufficient printing
intensity.
Approved For Release 2003/12/18 : CIA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
3. A series of test exposures should be made first, with exposure
increasing in square-root-of-two (about 40 per cent) steps. This can be
accomplished by changing light intensity or by changing film transport
speed.
4. This test run should be processed to a gamma between 1.0 and
1.2, using standard developer and equipment. The film should not be
overdeveloped. Temperature should be 68? F ? 1/2? F. Drying should not
be forced.
5. Microscopic examination of the exposure series will be re-
quired to select the best exposure, or to determine if an in-between
setting is indicated. This examination should be conducted with a monocular
or binocular microscope, having magnification of at least 50X, and pref-
erably up to 100X. (The proposed ASA Standard calls for magnification
ranging from .3 to 1.0 times the maximum resolving power of the system.)
6. After setting the printer for the best exposure level, a new
test film should be exposed and processed as above.
7. The analysis of the test film should be conducted by at least
two, and preferably three, observers, who examine each of the thirteen
targets under the microscope and record the last resolved group in both
the parallel and perpendicular directions. Note that lines parallel to
the film travel direction will measure the resolving power across the film
width,-while lines perpendicular to the film travel direction measure
the resolution in the travel direction.
When the thirteen individual targets have been read, an average
resolution for the two directions can be computed. Maximum and minimum
readings should be noted, and any special degradation, such as the presence
of slippage or poor contact should be indicated.
The above test can be repeated with a test film containing low-
contrast test targets. The MIL STD-150A and the proposed ASA standard
agree on the following values for high, medium, and low contrast:
The decision about whether a given set of three lines is resolved or not
requires care. The criterion most workers use is that a density differ-
ence must be seen between each pair of the three lines along the entire
length. Exception may be made if a scratch or dirt has obviously im-
paired the reading, or if there are two or more smaller groups that are
resolved beyond the one in question.
Approved For Release 2003/12/18 16IA-RDP67B00820R000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
Contrast Description
Density Difference
Luminance Ratio
High
2.0
100:1
Medium
.8
? .05
6.3:1
Low
.2
? .02
1.6:1
Note that if several values of performance are obtained at dif-
ferent contrasts, it is possible to calculate the frequency response of
the printer. An alternate method for this would, of course, be to use
a special test film with microscopic sine wave test targets; however,
the analysis of these images requires a microdensitometer and a fairly
extensive mathematical analysis. At the present time, as far as this
author knows, no technique has been devised for making faithful sine
wave test targets on a production basis. We suggest that the USAF might
profitably support research in techniques for making such test patterns.
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
VI. Acutance and the Effect of Edge Degradation on Measurement
An aerial photograph:.presents to the observer a pattern of random
densities and edges. In some instances, particularly with cultural de-
tail in the scene, much information may depend upon the sharpness of the
photographic edge, which is the boundary between contiguous regions of
differing densities. Although the sharpness of the negative is a sub-
jective evaluation, it can be correlated with an objective quantity which
is determined from the photographic edge. This quantitative concept is
termed acutance, and is measured for a specific emulsion under given
conditions of illumination and development in the following manner.
Collimated light is used to expose a knife edge in close contact
with the emulsion under test, such that there exists a (perfectly)
sharp line of demarcation between the parallel rays of light incident
on the surface of the emulsion and the region of non-illumination blocked
by the knife edge. Diffusion of the light in the emulsion itself and
adjacency effects in the development process, however, render a smooth
rather than a sharp transition in density of the developed image, as
shown in Figure 1.
It is apparent that two factors are required to describe this
smoothened edge. On a microscopic basis, the mean square gradient, G2,
x
gives an indication of curve shape, and is calculated by first dividing
the curve into small equal increments, d x, from the beginning of the
curve at xa to the end of the curve at xb, and then measuring the cor-
responding density increment, D, for each division and squaring these
values. Finally, the average is found by dividing the sum of these
squares by the number of samples, N, i.e.:
G2 =
x
Secondly, on a microscopic basis, the sensation of acutance is also
affected by the dynamic density range of the edge, Dr, which is simply
the difference between maximum and minimum density, i.e.:
D = D -
r max D min
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
_.e
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
knife-edge input
xb
r----?---- Dmax
I
/It-\actual response
Figure 8
Typical Microdensitometer Trace of a Knife-Edge on Photographic Film
Showing the Degradation from a Perfectly Sharp Pulse
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
 Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1
The higher the mean square gradient, the more abrupt is the density
transition at the edge; conversely, for a given value of G2, the edge ap-
t
pears sharper as the density scale becomes smaller. The acutance factor
is therefore given by
G
X
D
r
It is apparent that as acutance increases, the decision concerning
the exact position of an edge becomes easier to make. The eye, which in
conventional mensuration devices is usually aided by optical magnifica-
tion, effectively broadens the edge and reduces the acutance. The mag-
nifier in a sense acts as a low-frequency pass filter, and as such re-
duces both contrast and acutance. However, by amplifying the low fre-
quencies, it aids the eye in seeing the edge.
The basic problem one is faced with in deciding where the edge lies
is one of relating the density gradient to the original optical image.
Since any stage in the reproduction process lowers the sharpness of
that edge (with the exception of specialized spatial filtering techniques),
it follows that the ability of the observer to make the decision about
the position of the edge will be best with the original negative.
An extensive series of experiments would be required to establish
numerical data concerning the accuracy of edge measurement as it relates
to the resolving power of the system components. However, the factors
which will affect that accuracy can be listed:
1. Imaging ability of camera system (frequency response)
2. Imaging ability of original negative film (frequency response)
3. Duplicating printer capability (frequency response)
4. Duplicating film capability (frequency response)
5. Number of generations between original negative and diplicate to
be measured
6. Contrast of original detail in aerial image
7. Negative film gamma
8. Developer type, and type of agitation
9. Printing film gamma and exposure level
10. Size of detail to be measured
11. Accuracy of measuring equipment
Approved For Release 2003/12/18 : CIA-RDP67B0082OR000300100001-1