FINAL REPORT
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP79B00873A001400010011-7
Release Decision:
RIPPUB
Original Classification:
K
Document Page Count:
251
Document Creation Date:
December 28, 2016
Document Release Date:
August 24, 2012
Sequence Number:
11
Case Number:
Publication Date:
February 9, 1968
Content Type:
REPORT
File:
Attachment | Size |
---|---|
![]() | 9.48 MB |
Body:
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
FINAL REPORT
February 9, 1968
Progress Report for Period
23 November 1967 to February 9, 1968
File No. 11037
STAT
1 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
This.document is presented as the Final
Report under Contract to the U.S. Govern-
STAT
I
In addition, the report represented herein
covers the period 23 November 1967 to
February 9, 19 68 .
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
t Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
? VOLUME I
TABLE OF CONTENTS
PART I
Page
4
Introduction
I-1
System Configuration
1-3
Stage Control
I-10
Optics Control
.I-16
Automatic Stage Tracking
1-19
Automatic Optics Tracking
1-34
Summary of Automatic Tracking
1-35
Design Specifications
II-1
Task 1 - Statement of Work,
Specifications, Report Prep.
III-T1-1
Task 2 -Schedule and Plan
III-T2-1
Task 4 - Management, Admin-
istration and Supervision
III-T4-1
Task 7 - Main Frame and
Structural Elements
III-T7-1
Task 8 - Skin
III-T8-1
Task 9 - Granite & Ways
Assembly for Stages
III-T9-1
Task 10 - Air Bearings
III-T10-1
Task 11 - Stage Drives
III-Tl i-1
Task 12- Film Drive and
Transport System
III-T12-1.
Task 13 - Film Platen and Film
Clamping System
III-T13-I
Task 14 - Film Cooling'
III-T14-1
Task 15 - Optical Survey and
Specifications
III-T15-1
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Tasks 16,17,18 - Viewing Optics
Viewing Illumination, Reticle
Projector and Illumination
Task 20 - Platen Illumination
Task 21 - Optical Bridge and
Page
III-T16,17,18-1
III-T20-1
Supports
Task 22 - Interferometer As sy .
Task 23 - Optics Drive System
Task 24,25 - Scanning Device
and Correlation Logic
Task 26 - Digitizing Logic Sub-
Assembly
Task?28 - Output Logic, Inter-
faces and Systems
Task 29 - Cabling
Task 30 -- Control Console and
Chair
Task 32 - Computer Console
Task 33 - Electronic Racks and
Control Cabinets
Task 34 - Utilities, Vacuum and
Air Systems
Task 35 - Vibration Absorption
and Leveling
Task 36 - Overall Assembly
Task 37 - Radio Frequency
Noise Suppression
Task 38 - Environmental Control
Task 39 - Reliability Analysis
Task 41 - Stereocomparator
Mockup
Task 42 - Breadboards and
Test Services
Task 43 - Computer Programming
Bibliography of Task References
IIIT21-1
III-T22-1
III-T23-1
III-T24,25-1
III-T2 6-1
III-T28-1
III-T29-1
III-T30-1
III-T32-1
III-T36-1
III-T37-1
III-T38-1
III-T39-1
III-T41-1
:= Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
~, Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7 . I
PART IV
Page
Phase II Fabrication
IV-1
Statement of Work and
General Description
IV-4
Deliverable Items
IV-5
Performance Specifications
IV-6
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
LJ
VOLUME II
TABLE OF CONTENTS
APPENDICES
PART II Design Specifications
Effect of Pitch, Roll and Yaw on
Measuring Accuracy
Task 16, 17, 18 - Optical Design
Trip Report - SOPELEM
Task 24 - Scanning Device
Operating Instructions for the Image
Analysis System
Breadboard Tests and Components
of the Image Analysis System
Task 34 - Utilities, Vacuum &Air
Systems
Utilities Mechanical Schematic
Drawing E-6296
Tubing Assembly - Utilities Mechanical
Assembly - Drawing E5808
Electrical Diagram of Utilities Control
SK 405
Control Panel Schematic
Drawing D-6596.
Task 35 - Vibration Absorption & Level.
Dynamic Analysis of Barry Controls
Task 43 - Computer Programming
Figures T43-1 - 1.7 and Notes
Non-Real Time Computations
Appendix II-A
Appendix T16,17,18-A
Appendix T24-A
Appendix T24-B
Appendix T34-A
Appendix T34-B
Appendix T34-C
Appendix T34-D
Appendix T35-A
Appendix T43-A
Appendix T43-B
- . Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
c
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
i p
TABLE OF CONTENTS
COMPUTER PROGRAM - SPECIFICATIONS AND INSTRUCTIONS
Section 1 - Introduction and Summary
Section 2 - System Description
Section 3 - Program Specifications
Section 4 - Operator Interface
Section 5 - Timing and Storage Estimates
Page
1
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
U UNUAJMrILU U USE ONLY U LUNrIULNIIAL U ~oLLKtI
ROUTING AND RECORD SHEET
SUBJECT: (Optional)
FROM:
EXTENSION
NO.
C/PP&BS/NPIC
DATE
7 March 1968
TO: (Officer designation, room num?r; 'und
DATE
building)
OFFICER'S
COMMENTS (Number each comment to show from whom
RECEIVED
FORWARDED
INITIALS
to whom. Draw a line across column after each comment.)
t* ~' DDI P ,Ling Officer
2E-45 Headquarters
c m 2w-
S T'\Zy
Jay:
STA
Per your telephone request of
March
attached is a set of
,
h
3
fi
-volume
t
e
nal report
covering Phase I of the High
Precision Stereo Comparator
Program. The Statement of Work
for Phase II is contained in Part
IV of Volume I.
5.
It is my understanding that, prior
to your call, ca11ESTA
b
TS&SG directly. A co of the
attached was sent to ST A
by TS6SG on the afternoon of 6
March.
Accordin to STA
7
L
~ was particularly ST
8.
_
interested in the computer aspects
of the comparator. Volume III
9
deals with computer program
.
specifications and instructions.
TSgSG would appreciate the
10.
return of the three volumes when
you have completed your review.
11.
12
ESC
13.
14.
15.
T
T
T
T
FORM
F61~ USE EDITIONSPREVIOUS
a F-1 SECRET El CONFIDENTIAL F-1 INTERNAL n i1NCLASSIFIFf
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
-- Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
i
a 1,
STAT
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
G
c
.11
C'
C
C
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
DESCRIPTION AND OPERATION OF THE STEREOCOMPARATOR
C
INTRODUCTION
Ultra High Precision Stereocomparator is a highly
STAT
sophisticated tool for use in stereo analysis of aerial photographs..
The machine combines in one instrument facilities for overall viewing,
variable magnification, binocular viewing, and stereo presentation,
together. with capability for measurement with submicron accuracy. In
addition, the machine aids the operator's measuring task by automating
the job of stereo tracking. Every consideration has been given to
operator comfort, convenience, and speed; the resulting machine represents
the optimization of human-engineering in combination with state-of-the-
art accuracy.
The major functions of the Stereocomparator include:
1. Measuring and detecting image position to submicron
accuracy.
2. Simplified (and in some cases semiautomatic) accessing
"n I
of corresponding regions of the film to produce stereo pairs.
3. Providing ability to see detail on the film compatible
with measurement precision.
4. Seeing the converted equivalent regions in variable
magnification for best interpretability.
~,' Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
v
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
5. Converting these regions optically into stereo views.
6. Superimposing equivalent points on the photographs.
7. Providing data output for external processing into actual
ground. measurements and dimensions.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
fl Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
SYSTEM CONFIGURATION
The photograph included as the Frontispiece to this report
shows an overall view of the machine. The major assembly of the
machine is the unit containing the measuring engines, optics and
lighting systems, and operator console. In addition to the main
assembly, the system contains three double-bay Electronic Equipment
Cabinets, a double-bay Utilities Control Cabinet, and an auxiliary
Machinery Room in which are located the various compressors, vacuum
pumps, cooling equipment and other support functions for the machine.
In order to aid discussion of the system, each of these major assemblies
will be discussed separately.
A. Main Assembly
The main assembly is comprised of the following elements:
1. Measuring Engines: The measuring engines are
constructed of granite blocks for thermal stability. The base of each
engine is a monolithic block which has the top surface lapped flat
to within .000050 inch. A hole in the middle of the block allows light
from the illumination system to pass up to the film plane, and various
inserts are placed in the block to mount the engine drives, etc. A
granite tee is used as an intermediate stage and is driven along one axis
relative to the base. The top stage is guided by the tee and is driven
relative to it in a direction orthogonal to the driven axis of the tee. In
this manner a complete X-Y Cartesian system of motion is obtained. All
movable granite pieces float on extremely stiff air bearings for high
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
accuracy and low friction. The stages are driven by a special Threadless
Leadscrew (TLS) arrangement. The TLS offers excellent control and
negligible backlash. Non-cogging printed-circuit DC motors furnish
drive. power to allow a 16,000:1 range of engine speeds.
Attached to the measuring stages are the film drive units.
These drives are used to transport the film to the desired frame or to
rewind the film. The drive accommodates up to 500-foot reels of film
of any width from 70mm to 9.5 inches. Constant tension is maintained
on the film under all operating conditions, and a transport speed range
of from zero to 250 feet per to is provided.
The film platen is a special glass optical flat for low
distortion and uniform focus. Attached to the platen is a specially
made vacuum-clamping system which achieves extremely rapid pulldown
and release. Control over the clamping system is synchronized with.
film transport action to provide safe, convenient handling of the film
under all circumstances.
In order to provide overall viewing facilities, built in
light tables are included on the measuring engines. By merely pushing
a button, the operator can cause the measuring engines to travel to their
inboard forward limits, thereby placing the films adjacent to the
operator's chair. Console controls allow adjustment of the cold-cathode
tubes to secure a range of lighting levels.
The submicron measuring capabilities of the measuring
engines are made possible through the use of advanced laser-interferometer
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
fl Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
F
systems. A pair of servo-stabilized CW Neon-Helium lasers supply
the reference wavelength, and interferometers count interference fringe
pattern movements derived from a comparison of stage position with
a fixed reference mirror. Extreme precision, reliability, and simplicity
of adjustment make this linear measuring system most effective. The
least count of the interferometer system itself is O.16? , but by using
proprietary
count conversion equipment, a display and output
STAT
least count of 0.l? is obtained.
2. Optical Bridge: The optical bridge contains the
various lenses and drive assemblies which provide the controlled dis-
tortions necessary to rectify the photograph geometry for stereo. Addi-
tional equipment is included to allow injection of a floating reticle spot
for measuring purposes. Since the reticle is injected into the optical
path as close as possible to the film plane (for utmost measuring
accuracy), it is necessary to predistort the reticle image in a manner,
complementary to the distortion introduced by the main viewing optics
in order to maintain a uniform reticle shape and size at the eyepieces.
This is accomplished by means of servo-controlled follow-up systems in,
the reticle projector area in the optical bridge. Additional equipment in
the optical bridge includes the image dissector scanning assemblies
used for image analysis and correlation, and various photomultiplier tubes
and shutters which control the light levels in the system.
The optical bridge itself is. fabricated of heavy Meehanite
castings to provide an extremely rigid structure which is relatively stable
I-5
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
EI
with time and temperature. The cross-sectional areas of the members
are sufficiently large to keep deflections within extremely small limits.
The optical elements are fabricated into smaller sub-
assemblies which can be tested as integral units. These subassemblies
are doweled and bolted into place to assure precise alignment of the
optical axes.
The main viewing elements situated in the optics bridge
a) the viewing zoom, which has a 10:1 range
b) the objective changeover system, which has
1X and 2X lenses (to give the machine a total range of lOX to 200X)
and also contains the focusing mechanism
c) the anamorphic elements which-can be varied from
F
1:1 to 2:1 and stretch axis aligned over a continuous range
d) the image rotator system, which is also continuous.
The reticle projectors housed in the optics bridge
complement the main viewing optics to obtain round reticle floating
spots. A 75 watt arc lamp is housed in the optics bridge for the reticle
illumination source.
The binocular eyepiece assembly mounts onto the optics
bridge also. This assembly contains the various prismatic elements
used to accomplish reversal of views (left to right eye, etc.) and to
provide binocular viewing when desired. Also incorporated in the assembly
are the variable-density filters used to adjust the eyepiece brightness
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
fl Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
level and the high-speed safety shutters for eye protection. Micro-
switches activated by the prism changeover mechanism alter the
computer program and operator controls to.take into account the
various optical presentations at the eyepieces.
3: Operator Control and Display Console: This assembly
contains all of the normally-used operator controls. The upper portion
of the console is a display panel which contains the readout Nixie tubes
showing the X and Y coordinates of the two measuring engines (in . I?
units) and meters which show the settings of elements in the optical
trains. In addition, four sets of thumbwheel switches are incorporated
to allow presetting of the stage Nixie readouts to any desired numbers.
Pushbuttons are provided to allow the stage position
displays to be reset to zero, preset to the thumbwheel number, and to
have the count direction (sense of the axis) reversed.
Potentiometers are positioned on the display panel
to allow the image Analysis System to be zeroed to reference points.
The lower portion of the Operator Display and Control
Console contains the pushbuttons and control devices used to operate
the machine. A 180-button keyboard is centrally located to control the.
operation of the computer, data output, and miscellaneous functions.
Stage motion is controlled by a dual-range joystick
which is switchable to either or both stages and by a pair of trackballs
which are used for fine positioning. The optical elements can be
manually positioned by using dual-speed bi-directional velocity control
knobs placed conveniently around the trackball areas, Logic in the
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
U
machine switches the controls so that the operator is- not confused during
reverse stereo operation. The computer performs rotations on the track-
ball and joystick signals, so that in all cases the images appear to move
in the direction in which the operator moves the control element.
The platen illumination power and intensity controls
are also located on the operator console. A pair of joysticks used to
control film transport motion are located on the front surface of the console
for convenience.
4. Main Frame: The main frame is constructed of extremely
heavy box sections for utmost rigidity. The granite measuring engine bases
and optical bridge are mounted directly to this frame. Under and within the
frame are servo-controlled pneumatic shock mounts. The entire weight of,
the main assembly rests upon these mounts, so that the whole machine is
isolated from floor vibration. The Operator Control and Display Console
is not supported on the mounts, and is in fact isolated structurally from
the. main assembly. Since the operator's headrest at the eyepieces is
mounted to the Operator Console, it follows that the main assembly'is
entirely isolated from disturbances caused by the operator's manipulation
of the controls.
The high-intensity 450 watt arc lamps supplying the
main illumination and the dual-range variable condenser systems are- -
mounted to. the frame under the measuring engines also. This optical
arrangement allows illumination of only the area covered by the viewing
system for maximum efficiency and. constant illumination level. Control,
of this equipment is fully automatic.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Electronic Equipment Cabinets
There are three such cabinets in the system. Electronics
E
F
Cabinet Number One contains the servo and test equipment used for
the measuring engine drives and film drives. Electronics Cabinet Number
Two contains the servo equipment used in the optics drive systems and
illumination control systems. Electronic Cabinet Number Three contains
the internal computer, all digital logic, punch control, output data
link, counters, Image Analysis Equipment, laser power supplies, and
interferometer controls. All of the electronics equipment is designed t
for maximum reliability and serviceability.
C . Utilities Control Cabinet
This is a double-bay cabinet similar in appearance to the
Electronic Cabinets. In it are contained the pneumatic controls for
the stage air bearings, various air filters and regulators, pressure.
switches, and solenoids. Electrical controls and equipment contained
in the Utilities control Cabinet include the circuit breaker panels for
the machine, tally lights to indicate equipment status, arc lamp power
supplies, and various alarm and malfunction circuitry and interlocks.
D. Machine Room
This room contains various air compressors, vacuum
pumps, cooling systems, and support. equipment for the machine.
OI' Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
STAGE CONTROL
Position (motion) of the two measuring engines may be directed
by means of a joystick and two trackballs. Four pushbuttons marked -
JOYSTICK LEFT, JOYSTICK RIGHT, JS. /TB. BOTH, and TRACKBALLS
INDEPENDENT - allow selection of which stage(s) is controlled by
which control(s). The JS. /TB. BOTH may be selected by itself or in
conjunction with any one of the other three pushbuttons. In addition,
the TRACKBALLS INDEPENDENT may be selected in conjunction with
either JOYSTICK LEFT or JOYSTICK RIGHT, in which case the JS. /TB.
BOTH pushbutton is reset (not selected). In either the MANUAL mode
or the ENTER mode, direction of stage control by these buttons is
straightforward. With JOYSTICK LEFT (or JOYSTICK RIGHT) the left
(right) stage only moves in response to deflection of the joystick.
With JS. /TB. BOTH selected, and neither JOYSTICK LEFT nor JOYSTICK
RIGHT selected, the two stages move in unison as directed by the
joystick. With the TRACKBALLS INDEPENDENT pushbutton selected,
the left stage is controlled by the left trackball, and the right stage
is controlled by the right trackball. With JS./TB. BOTH selected and
TRACKBALLS INDEPENDENT not selected, both stages move in unison
in response to rotation of either trackball. In MANUAL mode or ENTER
mode, direction and velocity of the selected stage(s) are proportional
to deflection of the selected control(s). Thus in these two modes
neither eyepiece viewing mode or settings of the optical elements
causes any modification of stage control.
In AUTOMATIC mode and in AUTOMATIC WITHOUT ELECTRONIC
CORRELATION mode, direction and velocity of the selected stage(s), in
l Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
r Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
response to deflection of a stage control is modified in accordance with
settings of the optical elements in such a way that the image viewed in
the appropriate eyepiece appears to move as directed by the control.
Actual stage motion in this case may be quite different than deflection
of the control would otherwise indicate. In these two modes, also, the
eyepiece viewing mode affects the way in which the four pushbuttons
referred to above direct stage control.
In the two automatic modes, stage control direction is as
1 . JOYSTICK LEFT selected - Left stage only is controlled
by the joystick in all cases except that REVERSED STEREO selected
results in the right stage only being controlled by the joystick.
2. JOYSTICK RIGHT selected - Right stage only is con-
trolled by the joystick in all cases except that REVERSED STEREO
selected results in the left stage only being controlled by the joystick.
3. TRACKBALLS INDEPENDENT selected - Left stage is.
controlled by the left trackball and right stage is controlled by the
right trackball in all cases except that REVERSED STEREO selected
results in the left stage being controlled by the right trackball and
the right stage being controlled by the left trackball.
4. In cases where both stages are being simultaneously
controlled by the joystick or by either trackball the eyepiece viewing
modes have no effect on stage control. Thus, in either of the
binocular viewing modes, the stage which is not being viewed moves
along with the one which is, just as though the two were being viewed
l Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
7 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
E
El
in stereo. Stage tracking is exactly the same for REVERSED STEREO
as it is for NORMAL STEREO.
5. When both stages move simultaneously under direction
of one control, the modification of direction and velocity by settings
of the optical elements (once the latter have been adjusted so as to
establish a stereo model) is such that approximate stereo tracking of
the two stages will occur, at least over a short distance. If stage
tracking errors creep in, then the latter may be manually corrected by
using one of the individual stage control options. Thus a convenient
mode of operating would be as follows: With the JS./TB. BOTH push-
button and the TRACKBALLS INDEPENDENT pushbutton both selected,
common stage tracking may be directed by the joystick and stage
tracking errors may be corrected with the trackballs. An alternate
mode would be with the JS. /TB. BOTH pushbutton and the JOYSTICK
LEFT (or JOYSTICK RIGHT) pushbutton both selected. In this case,.
either trackball would control common stage tracking and the joystick
would permit correction of tracking errors.
6. In the AUTOMATIC WITHOUT ELECTRONIC CORRELATOR
mode, the computer repeatedly computes a stage to stage transforma-
tion whereby it attempts to correct (prevent) tracking errors. This
transformation type of stage control produces the effect of motion in
a stereo model of a geometric plane surface which represents the local
region of the ground surface. So long as the operator directs the stages
by either of the both-stages by ;one control. modes he may direct the
floating dot to any point in this plane surface. To move the floating
E Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
rl Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
ri
dot out of this plane surface, the operator may either use the track-
balls in the TRACKBALL INDEPENDENT mode or he may use the joystick
in the JOYSTICK LEFT or JOYSTICK RIGHT mode. Using the stage con-
trols in any of these independent stage control modes signals the
computer to discontinue tracking in the plane surface and allows the
operator to direct the floating dot to any point in the stereo model.
The operator may then, if he so wishes, direct resumption of tracking
in the previously established plane surface, simply by using the
joystick or either trackball in the. mode which calls for the selected
control. to drive both stages together. Note that switching in and
out of the tracking mode is controlled directly by the joystick and
trackballs without requiring that. any of the mode selecting pushbuttons
be changed.
While in the AUTOMATIC WITHOUT ELECTRONIC
CORRELATOR mode, the operator may initially establish a tracking
plane surface, or later establish a new (different) plane for tracking
as follows:
a. Using the joystick or trackballs in the respective
mode for independent stage control successively move to each of
three points through which it is desired to have the tracking plane pass.
b. While located at each of these three points operate
L
the pushbutton marked REORIENT to notify the computer that the floating
dot is at a point in the desired tracking surface. After three such
points have thus been established, the operator may direct tracking in
a plane through these three points by operating a stage control in its
respective common stage drive mode.
1-13
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
r
e
. After a tracking plane has been established, the
operator may also produce the effect of shifting the tracking plane
without changing its slope. This is done by operating either the
trackballs or the joystick in an independent stage mode to locate on
one desired point outside the tracking plane. While on this point,
the operator depresses the pushbutton marked REORIENT. The
operator now uses the joystick or a trackball in a common stage
control mode and the computer produces tracking in the plane through
the selected point, but parallel to the previous tracking plane.
It should be noted that computer control by means of
the transformation described above is possLble only in cases in which
complete information is available regarding the position, orientation,
velocity, etc., of both camera stations .
7. In the AUTOMATIC (with electronic correlator) mode
operation appears much as was described under 6, but tracking errors
are avoided by an electronic correlator (Image Analysis System)
instead of by a computed transformation. In this case, however, the
operator is not able to directly select the particular tracking surface
which is followed. The tracking surface is determined by the corre-
lator based on images formed on its two Image Dissector tubes. The
operator can sometimes exert an indirect influence on the tracking
surface followed by adjusting the scale factor control, thus limiting
the images being correlated to certain features within a restricted
field of view. Otherwise the operator may direct motion out of the
correlation surface by operating the trackballs or joystick in an
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
independent stage control mode. Similarly, the operator may direct
tracking in the correlation surface by operating the joystick or one
of the trackballs in a common stage control mode.
e
ri
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
E j Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Ell
r
OPTICS CONTROL
The Stereocomparator has 'two optical trains, each of which
may modify the image of its respective photograph in the following
four respects: magnification, rotation, anamorphic expansion, and
direction of anamorphic expansion. Eight five position switches on
the control console allow the operator to adjust the extent of each of
these optical transformations to any value in the available range. Each
switch operates as a velocity control of one particular element (zoom lens,
image rotator, etc.). The five positions are for no change (neutral) and
for high or low speed adjustment in either direction, with spring return
to the neutral position (center). Besides these major controls, there
are also a number of incidental controls (brightness, focus, size of
reticle, etc.). The latter will not be discussed here, however.
If either the MANUAL mode or the ENTER mode has been
selected, then all eight of these controls are operable at any time the
operator wishes to modify the setting of any optical element.
In either of the AUTOMATIC modes, however, all of the above
switches except the two controlling magnification are normally inoperative.
This.is because the computer (with or without help from the correlator)
is controlling the optics so as to maintain a stereo model automatically.
Nevertheless the operator may modify the scale factor of the stereo
model with either magnification control (i.e., either control causes both
zoom lens to increase or decrease at the proper relative rate so the
stereo model is not destroyed).
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
In the AUTOMATIC modes, the operator may temporarily take
manual control of the optics elements by selecting the OPTICS INDEPEND-
ENT pushbutton. Whenever this button is selected, the computer exerts
no control of the optical elements and the eight switches mentioned
above operate just as they do in the MANUAL and ENTER modes. If,
however, the operator remains in the AUTOMATIC mode, and simply
resets the OPTICS INDEPENDENT button, then the computer resumes
automatic control of the optics from the settings which existed at the
time this button was selected (i.e., the computer cancels out the
manual adjustments which were made in the interim).
In the AUTOMATIC modes, the computer (with or without
help from the correlator) is directing the various optical elements by
incremental drive commands. Hence, the actual settings of the various
elements are effectively the integrals of the past incremental commands.
The operator can at any time establish new additive constants for these
integrals. This is done by using the OPTICS INDEPENDENT button to
take manual control as was described in the preceding paragraph. Having
set the optics as he desires, the operator does not now reset the OPTICS
INDEPENDENT button. Instead, he selects the REORIENT button. The
computer then reads the settings of all the optics elements and resumes
incremental control. The incremental control this time, however, pro-
ceeds from the new settings. The computer itself resets both the REORIENT
and the OPTICS INDEPENDENT button in this case.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
e
F
Thus there is provision for the operator to control the optics
manually whenever he desires, and provision for the computer to take
control when directed to do so by the operator. The computer may be
directed either to resume control from an earlier group of settings
which it has remembered, or to resume control from the new settings
which the operator has established.
LI Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
71, Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
AUTOMATIC STAGE TRACKING
It is convenient to refer to one measuring stage as the
"master" stage and the other as the "slave" stage. In stereo tracking
the master stage is controlled so as to follow the operator's control
commands in a relatively direct fashion. The slave stage is controlled
so as to take positions corresponding to the master stage as required
to produce the stereo model. For each position of the master stage,
the proper corresponding position of the slave stage is determined by
the computer assisted by the Image Analysis System (electronic
scanning and correlation system).
At any instant the computer has values in some of its registers
for positions for both stages. For the master stage these values result
from integrating all past image motion commands from the operator
(via the joystick and/or trackballs), after first transforming these' in
accordance with the settings of the master optics. For the slave stage.
these values result from integrating the master stage motion commands
transformed as required to maintain stereo. The method of computing
these values is discussed in Part III under Task 43.
Periodically the computer compares its computed values for
intended stage positions with the actual stage positions (as read from
the stage position registers). The differences in these values are
output to the stage position servo systems which run so as to reduce
the amounts of these differences toward zero. Because the stage
speeds and accelerations cannot be infinite the actual stage positions
generally follow the computed positions with some time lag At this
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24 :CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
point it is well to examine the nature of this time lag for each stage.
For simplicity each stage will be treated like a one axis
servo system. Both stages actually have two similar axes with
inter-related position commands. Nevertheless, one axis approxi-
mations will give sufficient insight into the nature of the time lags.
Let Ow represent the image position increment read from
the joystick or one of the trackballs in some particular sampling
period. Let Oxi = Ow/M1 and Ax2 = Aw/M2 represent the corre-
sponding position increments for the master and slave stages, where
M 1 and M2 are the magnifications of the master and slave optical
trains. Since the computer integrates these position increments we
may represent the Laplace transforms by
sX1 = sW/M1
sX2=sW/M2.
Figure I shows a simplified block diagram of the two servo
systems, the electronic correlator and the computer. This diagram
is drawn according to the usual conventions for representing the
Laplace transforms of a set of differential equations. From Figure I
it is evident that
AI S1 W
1 1 - S + A 1 s I
M1
1-20
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
fl Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Computer
Correlator
L
1
M2 s
Master stage servo.
Al sl
2 Y
{ S +S1. s+A1sl
Master
optics
~- j Slave
M optics
X2 A2 s2 j
Y
sL + s2s + A2s2 2
Slave stage servo
Figure I. Block diagram of the stage tracking system.'
sW is the incremental input from the joystick or a trackball. Y1 and Y2
are the positions (one axis only considered) for the master and slave
stages. A,, A2, A3, sl, s2, and s3 are real constants, and. s is a
complex number corresponding to frequency.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A 001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
and Y2=
A2 s2
s+s2s +A2s2
W + A3 s3
[M2 M2s (s + s3) (M1Y1 - M2Y2)]
M
A2 s2 [s (s + s3)M2 + A3s3 M2 yl ]
s (s + s3)(-.2 + s2s + A2s2) + A2s2A3s3
The expressions
A 1 sl
s2+s1 s+A1s1
A2 s2
s2 + s2 s + A2s2
for the two servo systems would be identical if the two systems were
perfectly matched. They are shown different for generality but
numerical differences between the two are, in fact, quite small. From
tests on the stage mockup it has been found that reasonable numbers
are as follows:
s1 s2 100
A1= A2 50
corresponding to an 8 hertz bandwidth. The expression
A3 s3
S + $3
for the correlator includes an undetermined constant A3 and the angular
frequency response which is given in the Itek report* as about 1 over
50 milliseconds - i.e., 20 radians per second. Thus the denominator
* XI Final Report page 2 0
I-22
STAT
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
E
E
s (s + s3) (s` + s2s + A2 s2) + A2 A3 s2 s3
is s (s + 20) (s.22 + 100 s + 5000) + A2 A3 s2 s3'
Study of this shows that it is reasonable to make the constant
A2 A3 S2 s3 about 300 000 in which case the roots are -4.09, -14.5,
and -50.7 ? i49.8. The root -4.09 is close to zero, indicating slow
response. By making A2 A3 s2 s3 larger than 300 000, the corre-
sponding root is caused to move away from zero, i . e . , in the direction
of faster response. In the latter case the root corresponding to -14.5
moves toward zero and there is a limiting case with the two real roots
coinciding. If the constant A2 A3 s2 s3 is made still larger then the
e
two roots corresponding to -4.09 and -14.5 become complex and the
response is unstable for practical purposes. Thus the constant
A2 A3 s2 s3 should be as large as practical without having these two
roots become complex. Because the theory is only an approximate
representation of the actual physical situation it is desirable to allow
a substantial margin of safety. Practically speaking, 300 000 is
probably about as large as it's wise to go.
The foregoing discussion shows that the slave stage transfer
function (including the effect of the correlator) has, in addition to
a pair of complex poles close to those which apply for the master
stage transfer function, two real poles which are approximately
-4.09 and -14.5.
It will now be shown that the latter two poles essentially apply only to
the portion of Y, which is not given by the expression for the master
f
Id Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7 .
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
stage. Thus, let:.
Y 1 -L
s
Al s1 W
+ sl s + A1s1 M1
where Y0 corresponds to a difference (x-parallax) in the pictures on
the two stages. This difference must be adjusted primarily by the
correlator since the computer cannot predict such a difference. When
this value of Y1 is substituted in the expression for the slave stage,
the result becomes:
Y = A2 s2
s + s 1 s +A 1 s
s (s + s3) (s2 + sl s + Ais1) + A1A3s1s3 W
s (s+s3) (s + s2 s + A2s2) + A2 A3 s2 s3 M2
A2A3s2 s3
EI
Since Al = A2 and
2 s + sl s + Atsl M2 s (s + s3) (s + s2 s + A2s2) + A2A3s2s3
The first term on the right side of this expression is identical with the
expression for the master stage and represents the slave stage response
to operator commands. The second term shows the response to differences
in the two pictures as detected by the correlator. As stated above, the
second part contains time lags which are in addition to those contained
in the first part.. .
The foregoing analysis did not allow for possible time lags
directly in the optical systems. Figure II shows a generalized equivalent
of Figure I which provides for such possible time lags (insofar as they can
I-24
s + s 3) (s2 + s2 s + A2 s2) + A 2 A 3 s 2 s3 M2
s 1 = s2 this is approximately
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Master Stage
G 1(s)
e
Ge (s)
Y1
Master
Optics
Slave Stage
G2 (s)
Slave
Optics
Y2
Figure II. Generalized block diagram corresponding to Figure I.
G3(s)
G4(s)
0 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
71 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
111,
L
11
E.
11
L
be treated by linear theory). G 1 (s) and G2 (s) are the transfer functions
for the master and slave computer-stage servos. G3. (s) and G4 (s) are
hypothetical transfer functions for the master and slave computer-
optical systems. G5 (s) is the transfer function for the correlator. The:
functions G3 (s) and G4 (s) are not the basic transfer functions of the
optics servos per se but are functions which allow for the fact that
stage motion may generate signals which call for readjustment of the
optical system settings. Such readjustment may produce an indirect
effect on the stage position depending on the particular pictures which
are on the stages. Thus there may, be additional time lags due to the
readjustment of the optical system settings but the effect on the stage
servo response should be small.
From Figure II we write the following relations:
Y = G 1 (s) sW
and Y2 = G2 (s) [sW + G5 (s) (G3 (s) Y1 + G3 (s) Y0 - G4 (s) Y2)]
These may be written in the form
Y1 =G1 sW
Y2 =
1+G1G3G5
= G2 1 + G2 G4 G5 sW +
G2 G3 G5
1 + G2 G4 G
Thus, if G 1 (s) = G2 (s) and G3 (s) = G4 (s), the response
1-26
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
G2 [sW + G3 G5 (Y1 + YO)]
1 +G 2 G4 G5
Substituting the first in the second gives
_ G2 [sW + G3 G5 (G1 sW + Y0)]
Y2- 1+G2 G4 G5,
L Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
r
E,
C
r
the slave stage to operator commands (sW) is substantially the same
as the response of the master stage. On the other hand, the response
of the slave stage to differences in the two pictures (Y0) has additional
time lags which are now seen to result from the optical systems as well
as from the correlator. As stated before, however, the additional effect
of the optics systems on these time lags is not very great.
Some of the foregoing statements were made rather loosely and
were given without proof.. This was for the sake of readers who might
not care to go into _all the details. The following treatment covers the
same ground but with more detail.
On page 20 the following expression was given for the master
stage response
Y1 2
Al sl
s + sl s + A1sl
shown with the general operator command function (W(s). A particular
command function of considerable interest is that for a suddenly applied
constant velocity. In this case W (s) = v/s2, where v is the value of
suddenly applied velocity. Our primary nterest is in the response for
values of time greater than some very small value. Consequently we will
approximate the expression given, by its form for small values of s (the
latter correspond to large values''-of time, t). Then
A 1
7.
s +A1 M 1 s2
W
M
inverse Laplace transform of this is
1. )
yl =M (t _A +A e-A t
1 1
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Experiments with the test stage have given a value of Al about 50.
Let us, however, take the very conservative value of 30. v/iM1 has
a maximum rated value, for stereo tracking, of 100mm/sec. divided
by the overall magnification. Thus in what may be taken as the worst
possible case
y1 = 110 (t - 0.0333 + 0.0333e-30t)
This indicates a steady state velocity lag of (100/M) x 0.0333mm.
The field of view is greater than 150mm divided by the magnification,
Ell
hence the velocity lag as a proportion of the field of view is
3.33/150 = 0.0222.
The expression also shows that at t = 0, y1 = 0 and the velocity lag is
initially zero. Furthermore the velocity lag builds to 90% of its final
value in
2.3 t = 30 = 0.077 sec.
( since e-2.3 = 0. 1) .
The corresponding computation for the slave stage is as follows:
A2 s2 s (s + s3) (s2 + s1s + A1s1) +A1A3s1s3 v
Y2 2 + s1s + A1s1 s (s + s3) (s2 + s2s + A2s2)+ A2A3s2s3 M2s2
s
For smalls this is approximately
Y2
C
A2 [s (S + s3) + Al A3s3 (s + A1)-' ] v
(s + A2) [s (s + s3) + A2A3s3 (s + A2)-1] M2s2
A s
A2 [s2 + (s3 3 3 ) s + A3s3] v
A s 2
(s + A2) Ls2 + (S3 3A2 3) s + A3s3] M2s
1-28
E Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
E
L
In evaluating this expression it will be assumed that the slave
stage has a bandwidth which is 30% greater than that of the master stage.
The design of the stage servos is such that it should be easy to balance
the two more closely than this. Thus again the calculations are being
kept on the conservative side. The various values being assumed are
s1 =
60
s2 =
78
Al =
30
A2 =
39
Al s 1 = 1800
s3= 20
2 (s + 39) [s2 + 18.46s + 60] M2
A2s2 = 3042
A3s3 = 60
39 [s2 + 18.00s + 60]
39 (s + 4.417) (s + 13.58) v
(s + 39) (s + 4.210) (s + 14.25) M2 s2
The inverse Laplace transform of this is
v -4.210t 14.251 39t
Y2 = m [t - 0.0333 + 0.01224e - 0.005082e- + 0.02618e- 2
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
From the above we see that the steady state velocity lag for the
slave stage is precisely the same as that for the master stage - even
without assuming matched bandwidth. The time to reach 90% of the final
value of lag is longer, however, (0.31 sec.) for the slave stage. Thus the
two stages start together, move apart very slightly, and come back into
precise correspondence after a short period of time. The fast transients
will pretty much disappear in about the first 0. 1 seconds. Hence, for
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
times greater than this, the transient displacement between the two is
approximately
Y2 - yl M (0.01224e-4.21t)
At maximum rated tracking speed, as a proportion of the field of
view, this is
.0082e-4.21t
which decays to .005 in 0.14 seconds and to .001 in 0.21 seconds. It
appears doubtful that an operator can see such a small, brief displace-
ment between the two stages at maximum tracking speed.
Now consider the slave stage response to a difference between
the two pictures as detected by the correlator. For this purpose, the
artificial "assumption is made that, at some time, the two stages are
displaced apart by some definite amount and that the correlator suddenly
commences to command correction of any existing displacement. The
response to this suddenly applied correction of a definite displacement
is calculated. The appropriate formula for this calculation was given on
page 24 as
A 2 A 3 s 2 s 3 M 1
Y2 s (s + s3) (s + s2s + A2s2) + A2A3s2s3 M2 Y0
For the particular case described above
M1 _ d
M2 Y0 s
where d is the amount of initial displacement (scaled for the 2
cations). For small values of s the formula is approximately
I-30
magnifi-
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Y2
E
AZA3s3d
s (s + A2) [s (s + s3) + A2A3s3 (s +A 2)_ 11
A2A3s 3d
s
A3
S + A + A }
Putting in the values stated on page
Y2 s (s
(3W _(6 0)_d
+ 39) [s2 + 18.46s + 60]
39 60 d
s s+4.218 (s + 14.2s+39
The inverse Laplace transform of this is
y2 = d (1 - 1.591e-4.21Ot + 0.661e-14.25t-,0.0697e- 39t )
Thus the displacement is eventually corrected completely. Correction
becomes 90% complete when each of the transient terms is less than 0.1 .
Only the most slowly decaying one need be considered since, by the
time it's down to 0. 1 the others will be negligibly small. Hence:
0.1 = 0.0628
1.591
For e-4.210t = 0.0628,
t=2.77/4.210= 0.66 seconds.
Thus, as was stated earlier, the response of the slave stage to a suddenly
applied correction of a correlator detected difference is appreciably slower
than the response of both stages to an operator control command.
All of the preceding analysis assumed the correlator to be
functioning. If the correlator is turned off then the two stages respond
to operator control commands as interpreted by the digital computer. The
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
computer commands one stage as a master stage and the other as a slave
stage but there is negligible time difference in the computer commands to
the two stages. If, however, the two stages are not exactly balanced in
bandwidth then there will be a difference in the response of the two stages
to their respective computer commands. Figure I shows that if the corre-
lator circuit is opened then the slave stage has the same form of response
function as the master stage.
The master stage response was calculated on page 28 assuming
a bandwidth of 30/2ii hertz. The result was given as
= M (t - 0.0333 + 0.0333e-3Ot)
1
The corresponding slave stage response, assuming that it has a 30%
greater bandwidth is
y2 = AI (t - 0.02564 + 0.02564e-39t)
2
Thus the steady state velocity lag for the. master stage is 30% greater than
that for the slave stage. This produces a steady state displacement between
the two stages.at the maximum rated tracking speed* as a proportion of the
field of view
= 0.769/150 = .00513,
i.e., slightly over 1/2%. Hence in order to meet the specified maximum
value of 1/2% it is necessary either to make the bandwidth of the slower
stage somewhat greater than 30/2'rr hertz or else to make the difference
between the two bandwidths somewhat less than 30%. It can be calculated
that if the slower stage has a bandwidth over an even 5 hertz and the other
*It also produces a proportionately smaller displacement at lower stage speeds.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
j Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
I
i
stage exceeds this by up to 30%. then the velocity lag displacefnent
between the two at maximum rated tracking speed will be less than
1/2% of the field of view.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
E
AUTOMATIC OPTICS TRACKING
Insofar as the optics can be separated into equivalent one-axis .
systems, the analysis given above for stage tracking also applies at
least in principle to optics tracking. Thus Figure II applies to any single
optical axis (master and slave) if G1 (s) and G2 (s) now represent the optics
servos for that axis and G3(s) and G4(s) are included to allow for possible
additional time lags in the optics response due to stage motion. In fact,
G3(s) and G4(s) are probably simply unity gain (constant) transfer
functions. Nevertheless the input Y0 represents differences in the two
pictures and the latter are functions of stage position (different functions
for different pictures). Thus, as an approximate analytical device, the time
delays in Y0 may be absorbed into G3(s) and G4(s).
Thus the response time for the slave optical system, like that for
the slave stage motion, must be considered in two parts. The response of
the slave system to operator commands as interpreted by the computer
(sW) is substantially the same as that of the master system. The response
of the slave system in compensating for correlator detected differences in
the two photographs (Y0) contains additional time lags, however.
The servo system for any optical axis may be approximately
represented, as having principal poles consisting of a single pair of
complex poles. This is similar to the function used for the stage servos.
The optics servos, however, have a narrower bandwidth (about 3 hertz)
than the stage servos (about 8 hertz). Thus the principal poles for the
optics servos are approximately -20 ? i20. As a result the two real poles
which for the .stage servos were calculated as -4.09 and -14.5 are - for the
optics servos - somewhat nearer to zero. The difference is not very great,
however.
1-34
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
r
SUMMARY OF AUTOMATIC TRACKING
Although each stage has two axes and each optical train has
four primary axes (magnification, rotation, anamorphic stretch ratio,
rotation of anamorphic stretch direction) the design of the system is
such as to practically isolate each axis so far as internally produced
time lags are concerned. Hence, as a good approximation, each axis
(of the stage or of the optics) may be analyzed separately. The various
cross couplings which occur are treated as though confined to the inputs
to the several axes. Hence, the time response of the complete system
may be broken down to the time response of each axis to the inputs to
that axis .
Analyzing each axis separately shows that two inputs need to
be considered - one the operator's control signals, and the other differ-
ences in the two pictures which are not simply due to geometry but which
arise primarily from the relief in the object photographed. It is found
that the slave system response to the operator control commands is
essentially identical to the master system response, and is quite fast.
The slave system response to the photograph differences is, however,
appreciably slower.
To see how this works, in practice, imagine that the stages and
the optics are initially stationary with settings such that the operator,
the computer, and the correlator are satisfied that proper stereo corre-
spondence has been established. Now let the operator use the joystick
or one of the trackballs in a tracking mode to command motion of the
images in a certain direction and at a certain velocity.
E Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
El Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
The stages accelerate to 99% of their respective final velocities
in about 0. 1 seconds .(assuming an 8 hertz bandwidth). As the stages
move (assuming that the photographs change scale factor due to tilt
distortion and due to geometry) the computer generates changing position
commands to the optical systems.
Treating these changing motion commands, as suddenly applied
constant velocity inputs - the optics accelerate to 99% of the commanded
velocities in about .2 seconds (assuming a 3 hertz bandwidth). Insofar
as the photographs can be predicted by the computer and insofar as the
various master and slave systems have matched bandwidths, the two
systems stay in proper stereo correspondence while moving at the
commanded velocities (providing these do not exceed the rated values
for stereo tracking). If, however, some stereo non-correspondence begins
to develop, the correlator detects it and begins commanding corrective
action with a time lag not over . 05 seconds. Actual correction of the
non-correspondence is, however, somewhat slower.
The easiest way to state the time lag in correcting non-corre-
spondence (detected by the correlator) is to imagine that some definite
displacement has accumulated and that the correlator suddenly applies
a command for its correction. This is more pessimistic than the actual
situation wherein the correlator begins corrective action as soon as the
displacement starts to develope, but it is easier to analyze. Such a
non-correspondence displacement whose correction is suddenly com-
manded by the correlator will decrease to 10% of its initial value in
about .66 seconds for the slave stage and about .8 seconds for the
slave optics. Any displacement which is initially more than about 5%
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Ell
of the field of view is apt to be outside the pull-in range of the correlator.
Hence these times are ordinarily sufficient to reduce the displacement to
less than 1/2% of the field of view.
Maximum rated stage speeds which maintain correspondence on
identical photographs are rated at 10mm/sec. at l OX magnification varying
inversely with magnification to .5mm/sec. at 20OX magnification.
Maximum error in correspondence (on identical photographs) at these
speeds is rated as less than 1 /2% of the field of view at the selected
magnification. These values are believed to be more than adequate
for all practical applications. Since actual photographs are not identical,
tracking errors can be expected to develop if these maximum tracking
speeds are used. As seen above, there may be quite noticeable time
lags in correcting such tracking errors if they are allowed to accumulate.
It should be easy, however, to judge the maximum tracking speed for any
particular photographs since the first manifestation that a tracking error
is starting to develop will be departure of the floating dot from the surface
of the model. A slight reduction in tracking speed will then allow the
floating dot to settle back to the surface. Thus, it should be easy to
track in stereo without having tracking errors accumulate to the point
where they begin to produce eye strain and to do this at stage speeds
which are very respectable indeed.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
0 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
P
G
0 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
DESIGN SPECIFICATIONS
GENERAL SPECIFICATIONS*
1. The Stereocomparator has two optical trains. The following
specifications apply separately for each optical train.
Magnification with anamorphic stretch ratio at 1 /1: continuously
variable from 10X to 1 OOX, or from 2 OX to 2 00X; depending on selection
from two different objective lenses.
Anamorphic stretch ratio: continuously variable from 1 /1
to 2 /1: direction of maximum stretch continuously variable without limit
Image rotation: continuously variable without limit.
Brightness at each eyepiece: continuously adjustable from
0. 06 to 1 .2 stilbs (175 to 3500 foot lamberts). Set value is automatically
maintained for average film density variations over range 0 to 3. 0; high
speed shutter provides protection against sudden increase in brightness.
Type of reticle: floating dot principle; bright round dot projected
to center of each eyepiece.
Size of reticle: continuously adjustable from just over diffraction
limited** to 4 times diffraction limited; size and shape maintained auto-
matically for variations in magnification and anamorphic stretch ratio.
* Tentative until revised after completion of the Optical Design
** Equivalent object, size of Ares disc of an apparent point source at
the film plane.
Declassified in Part'- Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
L
Apparent position of reticle: superimposed, on the film plane;
position shift with respect to the? point at which the main optical axis
intersects the film plane is less than + 1/4 micron for changes of setting
in the zoom lens, anamorphic stretch ratio, image'rotation, fine focus,
size of reticle, and adjustment or switching of the eyepieces.
Range of coarse focusing control: 3 millimeters vertical
movement of objective lens.
Range of fine focusing control: 0.7 millimeters vertical,
movement of objective lens.
Low power objective lens: 80mm focal length, F/2. 1, operates
as a collimating. lens.
High power objective lens: 40mm focal length, F/1.25, operates
as a collimating lens.
Objective lens selection: via pushbuttons on control console;
accidental alteration of selection during a measurement sequence produces
automatic notification to the operator that measurements have been
invalidated and automatic provision for starting the sequence over.
Diameter of field of view at the film plane: inversely propor-
tional to overall magnification, greater than 15mm at 10X and greater than
0.75mm at 200X. Values are specified at 1/1 anamorphic ratio; at other
stretch ratios the field of view at the film plane in the direction of
maximum stretch is also inversely proportional to the stretch ratio.
Diameter of exit pupil: 1.2mm with low power objective and
1..0mm with high power objective.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
I L
C
Field of view at eyepiece: greater than 35? included angle.
Eye relief: 20mm + 2 mm.
Interpupillary distance: continuously adjustable from 50
Eyepiece line of sight: 15? below horizontal and 6? adjustable
e
E
convergence, with 2? vertical adjustment of one eyepiece relative to
the other.
Independent focusing of the two eyepieces.
Eyepiece modes: 4 selectable; normal stereo, reversed stereo,
binocular viewing of left stage, and binocular viewing of right stage.
r
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
E
E
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
2. The Stereocomparator has two measuring stages.' The following
specifications apply separately for each measuring stage.
Construction: Base block, top stage, and intermediate guide
stage are each a single piece of lapped granite. The top stage is
supported by 4 air bearings and the intermediate stage. is supported by
3.air bearings; both ;sets of support air bearings are with respect to the
top plane surface of the base block. Guidance of the intermediate stage
with respect to the base and of the top stage with respect to the inter-
mediate stage are by means of compensated air bearings. Film is vacuum
clamped to a glass platen mounted on the top stage. 'f'ilm spools and
drive system are also carried on the top stage. Film: llumination occurs,
from below, through appropriate openings in the base block and in each.
stage.
Measuring;; Range: 9-1/2' x 20-" rectangle.
Size of film, accommodated: any width from 70 mm to 9-1/2 inches,
any length from a cut chip to 500 ft., any thickness from 2 to 7 mils.
Maximum speed of top stage: 3 inches per second in any
(horizontal) direction.
Maximum. acceleration of top stage: 10 inches per second
squared in: any (horizontal) direction.
Maximum angular deviation of top stage from true rectilinear
translation: less than 1 arc second each in pitch, roll and yaw.
Maximum measurement error due to pitch, roll and yaw of
top stage: too small to measure (see Appendix II-A).
~, Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
E
Maximum vertical deviation of top of film platen: + 10 microns.
Type of measuring system: Twyman-Green interferometers on
each axis, powered by single mode gas laser light source.
Basic least count of measuring system: 1/4 wavelength of
He-Ne laser light (approximately 0. 1582 microns).
microns.
Converted least count of measurement read-out system: 0. 1
Maximum counting rate of measurement system: over 1 megahertz.
Maximum time required for reversing counting direction: less
than 2 micro-seconds.
Type of readout: BCD (1,2,4,8 code for each digit); sign -
magnitude representation; provision for setting count origin to zero or
to any preselected number.
Maximum deviation from straightness of interferometer travelling
mirrors: + 0.079 microns in any 2-inch length; + 0.3 microns over total
length.
Maximum non-perpendicularity of X and Y axis travelling mirrors
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
i
U
for interferometers: 1 arc second.
Room air conditioning:
Temperature: 72?F + 0.5?F
Humidity: 55% RH + 15% RH, -5% RH
Control digital computer: Honeywell DPD 516; 16-bit word
length, 16384 words of storage; high speed arithmetic option; teletype-
writer input - output; high speed parallel transfer to and from Stereo-
comparator interface electronics.
~,1 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
EI Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
a. White light from a Xenon arc
Linepairs/mm
Objective focal length 40mm 80mm
lOX Magnification 45
20X Magnification ,- S Q
10OX Magnification 40?
20OX Magnification
00
b. Yellow-green filtered Xenon arc light
Objective focal length 40mm 80mm
lOX Magnification { 6_0
2OX Magnification lop
10OX Magnification
20OX Magnification I ?00
c. The resolution degradation between the center of
the field of view and at one third of the distance
toward the edge of the field of view is less than 10%.
e
U
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
I E
c
C
E
Maximum RMS absolute error of coordinate measuring
system: 0.4 microns plus 10 parts per million, each axis, provided
the room environmental conditions are within specification (not includ-
ing operator pointing errors or errors in the film itself).
Maximum speed of stages while tracking corresponding points
in two identical pictures: 100mm/second divided by the selected over-
all magnification.
Maximum error in tracking corresponding points in two identical
pictures at maximum rated tracking speed: 1/2% of the diameter of the
field of view at the selected overall magnification. This specification
applies both when the Image Analysis System is on and when it is off.
If the Image Analysis System is operating, however, theory says there
should be no steady state tracking error when tracking on two identical
photographs.
At tracking speeds less.than the rated maximum, the tracking
error is proportionately smaller than 1/2% of the field of view.
Maximum time required for- the stages to accelerate to 90% of
the speed which is finally reached, after a command for constant
velocity (not in excess of the maximum rated tracking speed) is suddenly
applied through the control console: 0.1 second.
Maximum time required for the stages to accelerate to. 99% of
the speed which is finally reached, after a command for constant
velocity (not in excess of the maximum rated tracking speed) is suddenly
applied through the control console: 0.2 second.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
11
c
e
0
e
Maximum pull-in range of Image Analysis System (electronic
scanners and correlator): at least ? 5% of the field of view at the
selected magnification.
Maximum time required to reduce a suddenly released tracking
,error which is. detected through the Image Analysis System to less than
1/2% of the field of view (provided the two pictures contain sufficient
information for satisfactory correlation): 1.5 seconds if the error is
initially 5% of the field of view, proportionately less if the error is
initially less than 5% of the field of view.
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
e
e
0 Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
STATEMENT OF WORT{, SPECIFICATIONS, REPORT PREPARATIONS
Introduction
STAT
if
I
This report provides the technical summary of design
effort as performed during Phase I. Each task has been completed and
the technical results summarized, except for those items directly con-`
cerned with, or interfaced with, the Optical Subsystem. We anticipate
that the technical summary of those items, including the Optical Subsystem
will be completed and submitted during early March, 1968.
II. Summary
The report has been written in a manner which attempts to avoid
repetition and duplication of discussion which may have appeared in
previous monthly progress reports. Each task reported is augmented by a
reference index which appears as the last page of the task. T':-.:is reference
index indicates the volume and page numbers of previous reports where
further amplification of an item mentioned in the text can be found.
Figure TI-1 provides a convenient cross-index of volume numbers and dates
published. We feel that in this manner the reader can be spared repetition
of information of which he might already be aware and knowledgeable.
The reader will also note that there are three additional sections
in this report. We have included Part I - Description and Application;
Part II - Performance Parameters and Specifications; and Part IV - Report on
rr
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873A001400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
the Phase II Fabrication Effort. Part IV also includes the Statement
of Work and General Description (Appendix IV-A) and Specifications
(Appendix IV-B). This information is provided in accordance with
Phase I contract requirements.
III TI-2
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
Declassified in Part - Sanitized Copy Approved for Release 2012/08/24: CIA-RDP79B00873AO01400010011-7
El
Volume Period Covered
I January 9 through February 24, 1967
II February 24 through March 31, 1967
III April 1 through April 30, 1967
IV May 1 through May 31, 1967
V June 1 through June 30, 1967
VI July 1 through July 28, 1967
VII July 29 through August 25, 1967 - Appendices.
VIII August 26 through September 29, 1967
I