TECHNICAL REPORT NO. 1 PROJECT 135 FORMAT 12 8 JUNE 1967

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04747A001500040051-6
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 9
• Document Creation Date: December 28, 2016
• Document Release Date: December 31, 2001
• Sequence Number: 51
• Publication Date: June 8, 1967
• Content Type: REPORT

le- -- /O'12- 'J'
Approved For RWse 2002/06/17: CIA-RDP78BO4747AM50 040051-6
TECHNICAL REPORT NO. 1
PROJECT 135
FORMAT 12
8 June 1967
Prepared By: J
Program Manager
DECLASS REVIEW by NIMA/DOD
Approved For Release 2002/06/17 : CIA-RDP78BO4747AO01500040051-6
 Approved For Rose 2002/06/17: CIA-RDP78B04747A?500040051-6
1 GENERAL
In the course of debugging the Multi-Purpose Data Block
Reader, two weeks (May 22 through June 2, 1967) were spent in an attempt
to read and analyze the film samples, containing format 12 recordings,
supplied by the Customer.
The purpose of this report is to inform the Customer of the
incompatibilities between the reader and the recorded material, their
possible reasons, and ways of correcting them.
2 TESTING PROCEDURE
2.1 DEFINITION OF THE TEST PHASE
The test phase for each format is defined as the effort
dedicated to reading actual film formats. This test period begins after
the logic for the format is successfully verified by use of simulated data
signals instead of live film. At this point, the tester is assured that an
-error indicated by the reader is truly an error in reading and not a
malfunction of the logic circuits.
2.2 TESTS PERFORMED
The following is a list of tests performed to determine the
readability of the test material and their results.
Approved For Release 2002/06/17 : CIA-RDP78B04747A001500040051-6
 Approved For Rose 2002/06/17 : CIA-RDP78B04747A`500040051-6
2. 2. 1 Film Density
The analog output of the read amplifiers was monitored, while
reading under different light intensities, to insure sufficient output for
density variations at different points of the test film. The result of this
test showed that the density differential between dots and background is
generally higher at the beginning of the film than at the end. No micro-
densitometer readings were taken; therefore, it is not known whether
the dot densities were lower or the background density higher. The
The variation, however, was tolerable, and the density differential seemed
to stay above the minimum required for reading.
2.2.2 Dot Size
This test was performed both visually, by using an optical
comparator, and electronically, by measuring the time lapse between the
leading and trailing edges of the dots as they were read by the reader.
Microdensitometer traces will be taken this week to confirm the results
of these tests.
The test showed a wide disparity in dot sizes from file to
file and within each file. In general, the index dots were smaller and the
data dots larger. Occasionally, and with no particular pattern, a small
Approved For Release 2002/06/17 : CIA-RDP78B04747A001500040051-6
 Approved For RdWse 2002/06/17 : CIA-RDP78B04747A=500040051-6
data dot appeared in a line of larger data dots. It is estimated that most
index dots are less than the required minimum of .00611 at their half density
points, while most data dots are within the allowed width. The micro-
densitometer tracing will supply more conclusive information.
2. 2. 3 Data Accuracy - Recording
Both visual and reading checks were performed to determine
the recording accuracy. The results showed that, out of 130 data blocks,
6 contained recording errors. All the errors consisted of extraneous
bits rather than missing bits, and all error blocks contained more than a
single error character. Some contained as many as ten error characters.
In these cases, some of the error bits had normal density, and some were
faded. The cause for these errors may be attributed either to noise bursts
in the recording logic or failure to inhibit data change during the recording
cycle.
2.2.4 Interference Test
This test verified that all interference dots and patterns on
the test film are rejected by the reader logic.
Approved For Release 2002/06/17 : CIA-RDP78BO4747AO01500040051-6
 Approved For RdWe 2002/06/17 : CIA-RDP78BO4747AW00040051-6
2. 2. 5 Film Wander
A series of tests were performed to determine the combined
wander of the film in the camera and reader and their effect on.the reading.
The result of these tests indicated that the camera has extremely low
wander characteristics.
2.2.5.1 Up Wander
The data block was placed with its index row approximately
005" below the upper limit of the wander correction band of the reader.
The roll of test film was then read in both forward and reverse directions.
In both cases, the only errors indicated by the reader's error printout
were in the blocks that had a recording error. All these blocks were
indicated as reading errors. None were indicated as "Excess Wander -
No Data".
2.2.5.2 Down Wander
.The data block was placed with its index row approximately
005" above the lower limit of the wander correction band. After reading
in both directions, still no excess wander errors were detected, but,
in addition to the 6 blocks recorded in error, 2 more blocks indicated
reading errors. This showed .a need for further testing after completion
of the film wander test. These were performed later under Reading Accuracy
Approved For Release 2002/06/17 : CI?RDP78B04747A001500040051-6
 Approved For REgse 2002/06/17: CIA-RDP78B04747A 500040051-6
2.2.5.3 Forced Wander
Tolverify that the previous results were correct, the film
was placed with the data block just outside of the upper limit of the wander
correction band. Upon reading, all blocks indicated "Excess Wander".
The test was repeated for the lower limit with the same result.
2. 2. 6 Reading Accuracy
The data block was placed in the center of the wander cor-
rection band, and the film was read in both directions. A total of about
20 errors were recorded. By increasing the light intensity, more errors
were detected. By decreasing the light intensity, "Incomplete Block and
No Data" errors were indicated. Moving the position of the block within
the wander correction band showed a variable number of errors with a
"preference" for certain block numbers, which varied with each setting.
The Up Wander Test was then repeated, and the result, once
-again, was a perfect readout. The Down Wander Test was then repeated
with a near perfect readout. An analysis was then made which resulted
in the conclusion that the active diode groups in the read head are not
selected correctly. A further analysis showed that this selection error
does not occur due to logic errors but because of the variable dot sizes in
the data block recording.
-5-
Approved For Release 2002/06/17 : CIA-RDP78B04747A001500040051-6
 Approved For Rose 2002/06/17: CIA-RDP78B04747A?500040051-6
The reason for a perfect readout during the Up Wander Test
-was because the index dot only was present within the wander correction
band. The data dots were outside of this band. Since the diode selection
is done by the diode in the wander correction band only, the decision for
each selection was done on the basis of one dot only, and, therefore, was
correct.
Similarly, during the Down Wander Test, all three dots of
the character were inside the band (the band is about 4 dots wide). In
this case, the leading edge of the bigger data dot was seen first by the
diodes, and the selection was again correct.
During the reading in mid-band, however, the index dots
were inside the band, while the data dots were partially inside and partially
outside the band. The data dot may be split by the band boundary, but
due to its larger size, a sufficient part of the dot stays inside the boundary
and is detected by the diode. Figure 1 shows such a case where the
shaded areas on the dots represent a minimum area required for detection
of a dot. Because of the larger size of the data dot, diodes A and C will be
selected simultaneously, while the proper selection is A only.
Approved For Release 2002/06/17 : CIA-RDP78BO4747AO01500040051-6
 Approved For Rose 2002/06/17: CIA-RDP78BO4747PS500040051-6
Approved For Release 2002/06/17 : CIA-RDP78BO4747AO01500040051-6
 Approved For Rose 2002/06/17 : CIA-RDP78B04747A(0500040051-6
3 CONCLUSION
The major cause of erroneous reading seems to come from
the variable size of the dots in the data block. The pattern in the dot
variation seems to indicate that, during the recording, no current regulation
is implemented in the diode firing. circuitry. Possibly, a current limiter
is applied to each file of dots so that, when a full file is fired (such as
an index file), the diodes are starved, and the dots are smaller. When a
data line with fewer dots is recorded, each diode receives a more generous
amount of current and the dots are larger. If this approach is employed,
a block with only one dot in the data line may cause too large a current
in the selected single diode and the junction may be damaged.
The proper way of firing the diodes is with a large current
source for each of the three files and 30 current regulators, one for each
row. This will insure an even amount of current to each selected diode
and, thereby, an even dot size and density throughout the data block.
The test results of this format show that the reading reli-
-ability increases appreciably when the dots are of the same size and density.
The dot characteristics will have a more pronounced effect on positives
where degradation of dot quality occurs almost invariably.
Approved For Release 2002/06/17 : CIA-RDP78BO4747AO01500040051-6