IMPROVED SCREEN FOR REAR-PROJECTION VIEWERS
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP79B00873A001900010119-3
Release Decision:
RIPPUB
Original Classification:
K
Document Page Count:
104
Document Creation Date:
December 28, 2016
Document Release Date:
September 6, 2012
Sequence Number:
119
Case Number:
Publication Date:
December 5, 1969
Content Type:
REPORT
File:
Attachment | Size |
---|---|
![]() | 4.07 MB |
Body:
-
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
IMPROVED SCREEN FOR
REAR-PROJECTION VIEWERS
Technical Reports N08. -- 47 and 48
December 5, 1969.
,
Electronics
Research
? \skkkvalkkm4.4
'5's?k*,k'? s;?.
'
Amz.WA*MOL
??
CORNING
ELECTRONICS
A DIVISION OF CORNING GLASS WORKS
RALEIGH, NORTH CAROLINA
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
I Declassified in Part - Sanitized Copy Approved for Release 2012/09/06 : CIA-RDP79B00873A001900010119-3
CORNING GLASS WORKS
ELECTRO-OPTICS DEPARTMENT
RALEIGH, NORTH CAROLINA
IMPROVED SCREEN FOR REAR-PROJECTION VIEWERS
Technical Reports Nos. -- 47 and 48
Date - December 5, 1969
Periods Covered - October 10 to November 7, 1969
November 7 to December 5, 1969
,
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
0 Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
lel
TECHNICAL REPORTS NOS. P-19-47 and 48
1. Introduction
? Eight experimental scattering-type screens and one
commercial scattering-type screen have been evaluated in
terms of observed resolution and judged quality by the
Aerospace Group of the Boeing Company. Their final report
is included in this report as Appendix CG3.
In the quality tests, each of the 12" x 15" screens
was compared side by side with every other screen in a
projector using standard imagery. Observations were made
by several experienced photointerpreters and a quality
scale factor Z was determined for each screen depending on
how many times it was chosen as the better screen. For the
resolution tests, a standard USAF tribar resolution chart
was projected onto the screens and the photointerpreters
recorded the highest resolvable spatial frequencies. The
correlation of these quality and resolution judgments with
measured screen properties such as axial gain, brightness
variations, MTF, substrate transmittance, etc., was then
investigated. In general, the differences among screens
were found to, be small, both in judged quality and in judged
resolution. This was true in spite of the fact that signi-
ficant differences existed in measured screen properties.
These results can be understood when the following factors
are taken into account:
1. Projector MTF
2: Projector brightness
3. Ambient light level
In many of the tests these factors had the effect of
diminishing observed differences among screens.
2. Projector MTF
The 'highest resolution reported in 'CG3 Is about 4 li/mm
(p. B9) for the unaided eye viewing, from a distance of about
7 inches, a high contrast target projected onto the rear-
projection screen Under acceptable ambient light conditions.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
2
On the other hand, typical square-wave MTF values obtained
by the contact method (P-19-41) for these screens were 0.97
at 5 lines/mm, 0.91 at 10 lines/mm, and 0.75 at 15 lines/mm.
If these contact MTF values are even approximately valid
for projected resolution targets, then the MTF of the pro-
jector must have been the controlling factor in the resolu-
tion determinations of CG3.
It is possible to estimate the projector MTF from the
limit-of-resolution determinations described in 2.7.3 of
CG3, in.conjunction with the square-wave response of the
eye. With screen removed, the target images in the screen
, plane were observed by use of a 7X magnifier. The in-
dependently-measured contrast CT and maximum resolvable
resolution number RN for each target contrast are reproduced
here from p. 15 of CG3. Included also are the corresponding
TABLE I
Limit-of-resolution data on targets of CG3
Resolution Spatial
Contrast Modulation Number Frequency (F)
(CT) (MT) (RN) (.mu-1)
4.45
0.69
43.5
13
0.86
0.30
42.0
11.2
0.38
0.16
40.0
9
0.073
0.035
22.0
1.12
-modulation of the target
and maximum resolvable spatial frequency for that target
RN/6
= 2
11.4
calculated from CT and RN. Square-wave modulation thres-
holds for the human eye are adapted from the data of DePalma.
and Lowryl/ and are plotted in Fig. 1. for a viewing distance
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r_Declassified in Part - Sanitized Copy Approved for Release 2012/09/96: CIA-RDP79B00873A001900010119-3
L;
of 7 inches. The 7X magnifier used at 7 inches effective
viewing distance provides a magnification of approximately
6. The effect of the magnifier is to reduce the spatial
frequency on the retina by a factor of 6. Thus, for a
target having modulation MT and a maximum resolvable spatial
frequency F, the appropriate point of the eye response curve
is at
6 in Fig. 1. The corresponding modulation threshold
is read from the curve. The product of the target Modulation
MT and the projector modulation M(F) must be equal to this
modulation threshold MTH at frequency F/6. Hence the pro-
jector square-wave MTF is
M(F) (F) - MTH (F/6)
(1)
MT
When these calculations are carried out for the four
: resolution targets listed in Table I, the results are as
shown in Table II and in Fig. I. The intersection of the
[:
_ TABLE II
Observed
Resolution Modulation:
Target with 6X Threshold Projector
: ,modulation Magnification of Eye Modulation
F
MT F F/6 MTH ( 6 ) M (F)
P
: (mm -1) (mm-1)
0.69 13 2.17 .010 0.014
[: 0.30
11.2
1.87
.0075
.0045 0.025
0.16 9 1.50.
0.028
: 0.036 1.12 0.187 .0022 0.063
eye modulation threshold curve and the projector modulation
Ecurve falls at 4.6 lines/mm. This implies that even with a
perfect rear-projection screen the maximum resolution would
: be 4.6 lines/mm. This low projector MTF largely explains
the 4 lines/mm limit to the observed resolution and also
explains the difficulty encountered in distinguishing sig-
nificant differences in resolution and quality among the
screens.
:
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
4
The above calculation is not highly accurate, because
of differences in experimental conditions for the eye re-
sponse measurements of DePalma and Lowry and the projector
resolution determinations. The eye modulation threshold
depends upon the observer, the nature of the test object,
the threshold criterion chosen,- the angular field covered
by the target, the luminance, and the condition of visual
1/
adaptation. The eye response data of Fig. 1. were adapted
from an experiment in which the target was square-wave over
a broad angular field, the luminance was 20 F.L., and the
criterion for threshold was ability to detect modulation.1/
Thus in the CG3 projector resolution measurements the
observer was different, the threshold criterion was more
stringent, and the angular field was smaller. For these
reasons, the projector MTF calculation must be considered
as an estimate.
While the above analysis shows that the projector MTF
was much lower than expected, it is also not clear from the
CG3 measurements that even the best screens did not degrade
the resolution. Direct viewing of the projected image with
a 7X magnifier gave a limit of resolution of 13 li/mm with-
out a screen. With a screen in place the limit of resolu-
tion with the 7X magnifier was about 7 li/mm for the average
screen, perhaps 8 li/mm for the best screens (CG3, p. B 16).
Thus it remains to be proved that contact square-wave MTF
values provide a realistic measure of resolution in the
projection situation.
3. Ambient Light
The ambient light level was 3 F.C. (CG3, p.7) and caused
little modulation degradation in the resolution measurements.
This was because the average film density was low for the
resolution targets and the minimum input illumination to the .
screens Was 10 F.C. (CG3, p..13). But in quality tests,
average film density was about 1.0 and the ambient-to-projector
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
iDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
5
illumination ratio was often greater than unity. The
approximate calculations below show a degradation of
modulation by this effect of as much as a factor of 5.
Trapped projector light was generally negligible compared
with ambient.
Reflected ambient light and trapped projector light
both have the effect of degrading the observed modulation
transfer by a constant factor for all spatial frequencies.
The ratio of the modulation, or contrast, y displayed by
the screen to the modulation yo projected onto the screen
can be calculated in an approximate fashion by reference
to Fig.
y
2. The displayed
IBmax + BT +B
- 1 D R
modulation
-
( min +
BD
is
BT
+BR
Bmax
D + BT + BR1 ( +
(
min
BD +
BT
+ BR
(2)
where Bmax min
D and BD - are the maximum and minimum brightnesses
directly transmitted through a local area of the screen, BT
is the brightness of the trapped projector light contributed
by all parts of the screen, and BR is the reflected ambient
brightness., Since the modulation projected onto the screen
-is.
Bmax- Bmin
Y 0 = max + min
BD B
(3)
the transfer of modulation by the screen can be written
1 (4)
YO 2B + 2B
1+ T
BMa + Bmin
The trapped light in Eq. (4) can be expressed in
terms of the measured trapped light ratio
BT
aT = BT/1-71.
D -N j max + Bmin)
2 BD
_
(5)
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
iLiDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Li 6
[.1
[:1
0
N = E-D
D
1 (Emax Emin
D + D 1 ( (8)B
max ,Qmin
2
- 1
YO 1 +N aT+RT
RD TS
where
N
D
1 (B max
2 D + Bi
D :
ml
(6)
is the ratio of the average brightness over the whole
screen to the local average brightness.
The reflected ambient light BR in Eq. (4) is ex-
pressible in terms of the measurable quantity RDTs.
Since the ambient reflected light suffers one diffuse
reflection and two traversals through the substrate,
the reflected brightness is proportional to RDTS2Eamb.
The local transmitted brightness makes a single pass
through the sub trate and is thus proportional to
(E + E
max min
S D The quantities E max min
D and ED
T
are the
incident illumination maxima and minima in the local area
corresponding to transmitted brightness B and BMin.
max
D ?
The ambient illumination is Emb. The reflected brightness
a
as a fraction of the incident local average brightness is
thus approximately
2
B R T E T E
D S amb RD TS amb
1 (Bmax min)
2 D D
B_ min 1 (E
max
2 S D + ED E.D/N
where
( 7 )
Equation (7) holds if the reflected and transmitted light
.have approximately the same angular distribution.
Equations (4) - (8) can now be combined to yield
ED
Declassified in Part- Sanitized Copy Approved for Release 2012/09/96: CIA-RDP79B00873A001900010119-3
LDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
:?
Quality test II, in which the open gate screen
brightness was limited to 10 F.L. for all screens, was
most strongly affected by ambient light. A sample calcu-
lation.of y/yo will.be made for the LS-60 screen. According
to CG3, p. 5, the projector provided a Maximum of about 30 F.C.
open gate to the screen under standard conditions. In order
to reduce the brightness of screen LS-60 to 10 F.L. it was
necessary to reduce this open gate illumination to
10 F.L.
30 F.C. x F.L. 3.75 F.C,
79
since under 30 F.C. illumination this screen produced a
brightness of 79 F..L. (LBRT-I = 1.82 from CG3, p. A2).
Because the average imagery density was about 1.0, the
average illumination projected onto this screen was
=.3.75 F.C./10. Then -Eamb/D = 3 F.C./0.375 F.C. = 8.
The assessment of image quality was made with emphasis on
dense, shadowed 'areas of the imagery where the transmission
was as low as 2% (CG3, p. 10). Then for an average film
,transmittance of 10%, the value of N was 5.- The product
RD TS was calculated from the values in Table II of P-19-40
.for all screens except LS-60, for which a separate measure-
ment was made. The value for LS-60 was RD TS = 4.4%. The
value a = 0.11% can be found in Table A-1 of CG3. The
quantity y/y0 can now be calculated for this screen under
the conditions of the test. The results of such calculations
for all the screens appear in Table III.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
LDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Table III
Parameters describing the effect of reflected ambient
light and trapped projector light. on the observed MTF.
(Quality Test II)
aT
RDTS E
/*E- am D
b
_1_
Screen
(%)
(%)
YO
Z
AQ-20
0.062
2.1
1.0
0.90
1.01
AQ-17
0.081
3.1
2.8
0.70
0.67
AQ-11
0.133
4.9
3.6
0.53
0
AR-27
0.086
4.4
4.7
0.49
0.24
AQ-18
0.630
6.6
4.4
0.41
-0.08
LS-60
0.110
4.4
8.0
0.36
0.40
AL-5
0.135
9.2
4.0
0.35
-0.67
AR-28
0.240
6.6
10.7
0.22
-0.58
AL-4
0.740
14.0
5.4
0.21
-1.01
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
LDeclassified in Part - Sanitized Copy Approved for Release 2012/09/96: CIA-RDP79B00873A001900010119-3
The quality factor Z is plotted against y/y0 in Fig: 3,
where it can be seen that the correlation isvery'good. The
effectiveness of substrate darkening in suppressing reflected
ambient light is well demonstrated. This is in excellent.
agreement with the correlation of -0.89 reported in CG3,
2 ?
Table C-10, between Z and RD T . Figure 4 shows this correla-
tion. The displacement of the LS-60 point from the others
prompted a remeasurement of RD T2 ' this time by a direct method.
S
The value of 2% obtained for LS-60 Should replace the earlier
value of 6..3%. This change causes LS-60 to fall in line with
the others.
When Vio is calculated fOr the Quality I and Quality II
tests, the results are not so clear cut as in test II because
the ambient light was not as large relative to the illumina-
tion provided by the projector. These results are plotted in
Figs. 5 and 6. In Quality test I, projector luminance was
held constant. Figure 5 Shows the quality factor increasing
as y/y0 increases, at low values of yho in test I. But at
high values of.y/y0,_the reduced screen luminance caused a
:rapid drop in judged quality. :LS-60 performed best here be-
cause of its high efficiency and adequate ambient light re-
jection. .
In the Quality III tests, screen luminance was maintained
constant, except for screen AQ-20. Figure 6 Shows a general
dependence on yho except for screen AQ-17 and AQ-20. The
reduced luminanceof AQ-20 explains its low judged quality,
but no good explanation for the performance of AQ-17 is ap-
parent.
As mentioned earlier, ambient light was of much less .
influence in the resolution determinations. The lowest value
of y/yo calculated by use of Eq. (9) for the constant--
luminance case Was 0.93. Nevertheless, for the low-contrast
targets a significant correlation was noted between RN and
RD T2 (CG3, pp C13 and C15).
S
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
10
4. Projector Brightness
The illumination produced on the screen by the projector
affected the tests through the ratio Eamb/ED as described
above. Also, in .some cases the screen luminance fell low
enough to cause decreased visual acuity, as in Fig. 5. If
projector power had been unlimited, it would have been of
great interest to see whether the highest resolution could'
be obtained by highly illuminating the very dark Substrate
screens.
5. Screen Parameters
The list of screen parameters in Table A-1 of CG3 was
purposely made redundant on the chance that some unexpected
correlations might be discovered. The following list .is
probably sufficient for interpreting the results:
T30
R T2 or R T
DS D s
V30
aT
MTF
DRTHI'
The correlations found between these parameters and
resolution and judged quality are found in CG3, 1011D; C10
C-15.
The last three parameters had negligible effect on the
outcome, although dry thickness DRTHI correlated extra-
ordinarily well with quality, in the Quality II test and
with resolution in the resolution tests. This must be
considered as fortuitous, arising largely because the in-
efficient screens AL-4 and AL-5 had very thin layers, and
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
frf
.m)
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
11
the screen which was given the greatest substrate darkening,
AQ-20, had the thickest layer. The low projector MTF pre-
cluded any significant dependence on MTF values of the
screens at 6 li/mm, which were not very different anyway.
The trapped light ratio aT would be important only if the
ambient light were quite low, which was not the case.
The parameter T30 TS is basically a measure of screen
efficiency and could equally well be replaced by LBRT,
B(0)TS' or T45TS' for which the correlations were very
similar. Not unexpectedly, at constant projector illumina-
tion T,30 TS correlates highly with quality and with resolution
for the high-contrast target. The correlation vanishes,
however, for low-contrast targets.
A significant correlation exists for RD T2 in Quality
S
test II for the reasons explained earlier. In all the other
tests the correlation is weak, although in the constant-
screen luminance resolution test with LS-60 excluded, the
correlation may reflect a. real ambient and trapped light
effect.
Large correlations were found for the brightness varia-
tion V30 in the resolution tests and in Quality Test II.
The latter is understandable in view of the strong dependence
of V30 on B(0). The surprisingly high correlation in the
resolution tests is at least partly fortuitous, since the
low efficiency screens AL-4 and AL-5, which nearly always
gave inferior performance, had very high brightness variation.
6. Variation of Resolution with viewing Angle
This phenomenon should be investigated further. Since
it occurs for all screens, it could be a property of the
projector. Also, if the screens were being used to best
advantage, i.e., in a high-MTF system, the effect might be
smaller or even more pronounced.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
0
01
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
12
7. Effect of Target Contrast
One unexpected result is that the darker screens
showed lower resolution than the lighter screens for high
contrast targets, but the reverse was true for low con-
trast targets. Figure 3-10 of CG3 illustrates this point
for constant projector output and corresponding. results
hold for constant screen luminance. The greater separation
of screens on the resolution scale for low contrast targets
can be explained by reference to the slope of the eye response
curve. At very low contrasts a given fractional change in
modulation produces a greater fractional change in detect-
able spatial frequency than at higher contrasts. However
this does not explain the observed interchange of rankings
of light and dark screens, as occurs most convincingly for
screens LS-60 and AQ-20. If this effect persists in a more
ideal projector arrangement, incorporation of heavier sub-
strate darkening may be justified.
8. Conclusions
Significant dependences on some screen parameters,
notably efficiency T30Ts and diffuse reflectance times
substrate transmittance RD TS' were established by the tests.
The more efficient screens performed best for a fixed pro-
jector output.
The projector MTF limited observed resolution to about
4.6 li/mm, whereas the screens should have been capable of
displaying considerably higher resolution.
Quality tests were dominated, by the projector. MTF and
by the ambient-to-projector illumination ratio. Calcula-
tions based on the known ambient light level revealed a
strong RDTs dependence, which was one of the principal aims
of the investigation. The importance of T was underscored
by an unexplained superiority of dark screens for the low
contrast resolution targets.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
'711
Fl
4.;.3
C-1
13
No physical justification is apparent for the large
negative correlation between brightness variation V3 and
resolution.- While it is partly fortuitous, it may be
significant.
Similarly, the reason for the observed increase of
resolution with viewing angle is obscure. This effect may
or may not be evident under ideal projection conditions.
Ambient light was generally high enough that the trapped
light ratio aT had little effect. Likewise, measured contact
MTF values for the screens were not sufficiently different
at 5 li/mm to have an observable- influence on the results.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
14
I. REFERENCES
1, J. J. DePalma and E. M. Lowry, J. Opt. Sod. Am. 52,
328 .(1962).
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
inDecl
' s- c 1069
Part - Sanitized Copy Approved for Release2012/09/06 : CIA-RDP79B00873A001900010119-3
_ _
....,
.....
, .
,__
8
-1-1-
--, -+-- H-
'-
-
-I
"117a
.w,
7.7_c_i
_-7 ,
--,
4?
6
1
-1_1-i_
,
.
-!
_I
L
5
mu
_L
_i I-
:
"
..IM
--1--'
!
-71
M
1 '
,
LI
i____
'MN=
_Li.
A
:III
17_1_ ....I
,
....
?
, ,I
....0.....?
,
a:
2
..
-
NM
1111155WEIII
sm
E--U,
r. w... ..=
.
IMO
---
MINIM."
MIME=
i
i
i___
-
MEM
-,-,--r,---
, -7-
girj-' MI
,
------
--
-
, I
-I-
"
MEM
MOMM
---,...-1 -r-Mill
]-7:)11
6111111111-11illa
' =
IN
!MEM
MOM
MEM
1
MUM
=
WM
M
MMTMM
M
M
MMIUMm
WM=
MMMMMMW
; !
?
1
9
WM
MEM
MILI
IMMIIIIIIM
MM
ILI J
I
UM
EA=?I
I I N
MEM
NNW=
=MEN
MIME'
MINE
?1
Mir?N?
r Imm
mumm?
IIIII
......r...
..
...
1 I ?
=MU
,
.......mm
MEI
????????????????
Ern.
?
?
MIME
MMINIMMEMOM
EIBMIMMEMMEM
MEM
MIR
WM
MMEMMIN
MOM
II
MEM=
UNWIND
MEM
OMMEMME
I IMMEIM
MEM Malls
iii.
OEM
UMW
7111111111
En
MO
II
OM
MOMMUMINIM
IMMO
M
OM
MPO
MAIM
ill
1111111
I
ml
flu
1411====l1R=i=
=
--`-
:
--PMI'S'
4,
>21=E='
8
= -_-_
-6
. '-
: :
---_,_
.
-MENNE
7......_-
-
''
IIIII
_
--!----i-
" "
6___
5
BMW
=
......
....
i___
J_
1_ t :
1 -
.
h
-
I 4
,
: ! :
f
--1-q-
=I=
..b..
'
...
mom
"
=MEM
...MMEMMEMMIMM
mum
-
mg=
iii
mis
4
3
11111111110:02111111
-1tH-
'
INEMH"..MMEEM
"POOmmiAMIMMEOLIOA
I
'
, MI
1111
, ,
, ,
_L__
,
" , :
,----t- ,---,
am--
==m-
--,
'
illli-4
EgEMPF
1111110-Ak
----h-t ,
?
'
-ASIII
---
i
mum
-
MEWL
'ffimoralibeit
--1-7'
A
11111-
,
,
---FH- ,
--i-
min=
,
magailjE
'.-.i...-4?==
.
Aril
M
u___
iiii
===inlinilm
I.
main
OMmirimmlinitirMEMEM
__m_ Arm'
'NEMMiall
-Ili-mEMM
MENE......
. m ..m."',
m
.
murilIMMIIIIMMII
mom:.
omms.
ARsoaire.
iim=1.1611..,
"
mom=
sm...
wm..0....m
...MO:
ummimm......
........016.8
-mug:
I
II
"NMI
I
unirlimmmmi
mum
um"
m
i
1
IN=
II
IIIIIIIII
ipu
II
I
1 I
II
1111
1
II
Iiii
Illi!!,,
I
h.
li
1
?
111041???1
???
?
.
gp
...
ri
Im
?
....
...
ini
_4,
m
in
iimm
imi
Immo
Imm
mii
m
m
INN=
000MEN11001
Figure
1.
Square-wave modulation
9
MWL"MMVEMEM11111
versus
spatial
frequency. Curve 1,
8
FAMES
-
-kJ-
square
-wave
modulation thresholds for
====,
-
ll
==
_ __ the
(adapted
eye at
from
7" viewing distance
DePalma and Lowry-').
lama__
...======....1
nUMMIII=1111
111111111W
BEEN
m.=0.
EUMNAmmmEM
ENEE-20-=.**
Immolkilmmmum
_ Curve
determined
2, modulation
of projector as
by limit-of-resolution
mum.=
...--lil
measurements
Immo.
MISIMM
0
MAIM
MMMimm
MINIM
Mil
-1IIIIMEMEI
M?
--`-
=um..
--x-=
===-----m=
-=.-------Em-
1-
ESSEM=ESS
-
______
Illii
Wi
I'M
,
___,
: _,...?,_
?.I..
m
.
me
Orm
.
lbw
k
__LA_H__,
H
41mEll
,
h
L___
,
,
,
.....
?.0.
,
LL,'
immilm
-
_
-4 :
,
--1
,
il
'--L?
mmelm
mm..0
?Iv
-
Mr
,
? ,
-[ '
PI --i
l
?
--, i
1
,
mi
HL
,
.
um
.
?
(
minommo
Li
1
mom
4
m
m
?
I
H
Eml.""
=II
, ,
1 ,
?
,_L
41
?
MO'
MO
E;
?
?
? '
4.
o 1
_LLIA,
1 I I w
__LLt
1 i
0.
/
?
MR
rilinani
MI
11
I
I I 1
O?
_L__
1-
II
1
ION
MOO
MOM
0
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
_Lb P-19-47 and 48
December 5, 1969
ViEWER
Li
SIDE
FK-Do.TF.-CTO
fl
.
Li
Li
LOCALi
L
,MAX
I LLUMIN4TION E.
)
Bg x 0:" LOCAL
OR'16,11-7"
TKA P.PF:
AV. BP 1601-,11,5-1 s
r
I L (May Art 0)4
PE.r-t.Eci p ?
ri w
5,e16 ?
Figure 2. Geometry and nomenclature for describing
trapped projector light and reflected ambient
light.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
11 P-19-47 and 48
December 5, 1969
f-
7
'1
fl
r-
-1
.QvilLrry resr EL
? Figure 3. Quality scale factor Z versus modulation
transfer factor Vy0 produced by reflected
ambient and trapped projector light. Numbers
beside points are abbreviated screen numbers.
Data for Quality Test.II. ?
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
6Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
J8 P-19-47 and 48
December 5, 1969.
20
Z7
f
6
QUALITY TESr
Rp1-5-4
a.41- 0 6 8 lo
16
2
Figure 4. Quality scale factor Z versus Rgs for Quality
Tes II. Screen Ls-6.0 has corrected value of:
R T
- D S'
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
0
19
P-19-47 and 48
December 5, 1969
QuALtr)( TEST
0
GO
0 0'1
v?-)
0
le
(
20
1,0
Figure 5. Quality scale factor versus y/yo for Quality
Test I.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Figure
-;20 P-19-47 and 48
December 5, 1969
quA-Lcry re_cr j-tr:
11
lb 27
0
.6 .8 LO
60
20
? Quality scale factor versus j,
Test III.
y for Quality
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
??eclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
FINAL REPORT
REAR PROJECTION SCREEN
EVALUATION STUDY
CG-3
29 AUGUST, 1969
CORNING P. O. 08209
THE BOEING COMPANY AEROSPACE GROUP SEATTLE, WASHINGTON
STAT
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
liDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
?
This report is prepared for the Corning Glass
Works in fulfillment of Purchase Order No. 08209
?
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3 _
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
CONTENTS
1.0 INTRODUCTION
2.0 METHOD
2.1 PROJECTION EQUIPMENT AND SCREENS
2.2 IMAGERY
2.3 TEST OBSERVERS
2.4 QUALITY TEST I
2.5 QUALITY TEST II
2.6 QUALITY TEST III
2.7 RESOLUTION TEST
2.7.1 PRIMARY RESOLUTION TEST
2.7.2 ADDITIONAL RESOLUTION TESTING
2.7.3 PROJECTOR RESOLUTION TESTING
3.0 RESULTS
3.1 SCREEN QUALITY JUDGEMENTS
3.2 SCREEN RESOLUTION
3.3 RELATIONSHIPS BETWEEN VARIABLES
3.3.1 SCREEN PARAMETERS
3.3.2 QUALITY JUDGEMENTS
3.3.3 RESOLUTION
3.3.4 QUALITY/SCREEN PARAMETERS
3.3.5 RESOLUTION/SCREEN PARAMETERS
3.3.6 QUALITY/RESOLUTION
3.4 REGRESSION EQUATIONS
4.0 DISCUSSION
5.0 CONCLUSIONS
6.0 REFERENCES
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D
PROJECTION SCREEN PHYSICAL PARAMETERS
RESOLUTION DATA .
CORRELATIONS BETWEEN VARIABLES
REGRESSION EQUATIONS
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
.1.- Declassified in Part- Sanitized Copy Approved forRelease2012/09/06 : CIA-RDP79B00873A001900010119-3
H. CG-3
ABSTRACT
Nine rear projection screens, eight experimental and one a standard
type, were evaluated. The evaluation was based on two measures,
judged screen quality and judged resolution. Screen quality judg-
ments were made by image interpreters while viewing operational
imagery on pairs of screens mounted side by side in a Richardson
rear screen projector. Resolution measurements were made by skilled
observers viewing tribar resolution charts at five contrasts and
three viewing angles with the screens mounted in the same viewer.
The differences among screens in both judged quality and judged
resolution were small. Quality judgments were strongly affected by
screen luminance - observers preferred the brighter screens. When
screen brightness was changed, the quality judgments also changed.
Screen parameters related to the distribution of luminance such as
axial gain and brightness variation were positively related to the
judged quality of the screens when large inter-screen brightness
differences existed and negatively correlated with judged quality
when inter-screen brightness differences were eliminated by match-
ii g their on-axis brightness. Screen resolution was considerably
lower than the visual capabilities of the observers. Low contrast
targets yielded lower resolution than high contrast targets. With
low contrast targets, resolution was worse with on-axis viewing. The
observers could distinguish higher spatial frequencies with the aid
of a 7X tube magnifier. The screen with the best resolution with low
resolution targets had the worst resolution with high contrast
targets.
iv
?
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
r"
1.0 INTRODUCTION
This report describes a study conducted to evaluate eight experimental
rear projection screens produced by the Electronic Research Laboratory
of the Corning Glass Works. To obtain a comparison with currently
available materials, a ninth screen manufactured by the Polacoat Company
and in regular use in rear projection viewing devices was included in
the study.
Many screen characteristics have been used as a basis for comparison and
evaluation, among which are: (1) spatial distribution of luminance;
(2) contact resolution; (3) breakup magnification; and (4) microphotometer-
measured brightness of a spot. The spatial distribution of luminous energy
is more important for group viewing than for use by an individual interpreter.
Spreading of the image over a large angle may actually lead to a reduction
of the resolution capability of a screen. Contact resolution is commonly
reported for screens (Klaiber, 1966, McHail and Soil, 1962), but there is
no indication it is related to the resolution of a projected image. Other
measures, for example breakup magnification, the maximum magnification
which can be used to view an image on the screen before it breaks up,
and the fidelity with which the microphotometer-measured brightness
characteristics of a small spot of light are maintained by the screen,
both appear to be valid indices of screen quality. Unfortunately, they
do not agree with each other (McHail and Soil, 1962).
Two measures were selected for use as indices of screen quality in the
present study, judged quality and judged resolution. Both involved
projected. images.. To collect the first,a potential rear screen
projector user, an image interpreter, judged the quality of the screens
as they displayed the same imagery he normally worked with. The second
measure, resolution judgements obtained with a range of target contrast
levels, was designed to simulate some of the critical information
elements of imagery.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r
Declassified in Part- Sanitized Copy Approved forRelease2012/09/06 : CIA-RDP79B00873A001900010119-3
CG-3
Initially two tests were planned, one involving judged quality and the
other involving judged resolution, with three .subjects to be included
in each, as described in the test proposal (DK-423, Firm Proposal -
Rear Projection Glass Screen Evaluation Study). These two tests were
carried out as planned. During the testing the projector brightness
was maintained at the maximum level normally available (referred to
hereafter as "normal brightness").
Preliminary analysis of data from the two tests initially planned
indicated that judgements were closely related to screen brightness,
so the study was expanded to include other brightness conditions. In
all, the following four tests were conducted:.
o Judged Resolution - After completing the initial test at normal,
unadjusted brightness, the test was partially repeated with the
'brightness of the screens matched by varying projector lamp
voltage.
o Screen Quality Test I - Three interpreters judged screen quality
with normal, unadjusted screen brightness, as initially planned.
o Screen Quality Test II - A fourth interpreter judged scKeen
quality with screen brightness matched by varying the projector
lamp voltage.
o Screen Quality Test III - Three additional interpreters judged
screen quality with screen brightness matched by means of
neutral density filters.
These tests are summarized in TABLE 1-1.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
U.
U.
TABLE 1-1
TESTS CONDUCTED DURING SCREEN STUDY
TEST
SUBJECTS
SCREEN BRIGHTNESS
CONTROL
METHOD
LUMINANCE
(FL)a
SCREENS
JUDGED
RESOLUTION
3
NONE
10-107
..
9
2
LAMP
VOLTAGE
- 10
9
QUALITY I
3
NONE
10-107
9
QUALITY II
1
LAMP
VOLTAGE
10
9
QUALITY III
3
FILTERS
30
a
Approximate open gate screen luminance in foot
lamberts. Appendix A describes the measurement
techniques used.
4
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
tAir?o
Li
Li
fl
2,0
METHOD
2.1 PROJECTION EQUIPMENT AND SCREENS
Nine rear projection screens,were included in the study, consisting of
eight experimental and
"Polacoat" and used in many
were 12 x 15 inches in size
glass. The diffusing layer
small particles of glass in
one standard type screen, manufactured as
rear projection viewers. All nine screens
and consisted of a diffusing layer on heavy
on the experimental screens consisted of
-
a binding medium. The diffusing surface
? was on the side away from the interpreter. The experimental screens
included an antireflection coating on the side toward the interpreter.
Physical parameters measured on each screen are listed in TABLES A-1
and A-3 of Appendix A. These tables include screen brightness
measurements made during the study.
The screens were viewed in a Richardson Model 705M rear screen projector.
For the quality studies, a frame was placed in the 30 by 30 inch viewing
area of the projector, which allowed two screens to be mounted side by
side, as illustrated in Figure 2-1. For the resolution study a different
frame was used to support a single screen in the center of the viewing
area. Projector magnification was fixed at 15x. Test subjects had
control over the projector focus and were encouraged to adjust it
whenever necessary. Using normal line voltage, the projector gave an
open gate brightness at the back of the screen of between 29 and 33
foot candles, as measured by a cosine receptor head. Control of the
illumination level is discussed in the sections below which describe
the individual tests.
Lamps were replaced several times during the testing. When possible,
screen brightness was measured for the different lamps. The typical
difference between a used bulb at the end of its normal life expectancy,
approximately one hour, and a new lamp was 5 per cent in either direction.
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
jDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
2 OF 9 SCREENS 7
UlIIlIdIJ
FIGURE 2-1
?
INTERPRETER VIEWING TWO SCREENS MOUNTED
IN THE REAR SCREEN PROJECTOR
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r Declassified in Part - Sanitized Copy Approved for Release 2012/09/06 : CIA-RDP79B00873A001900010119-3
The left-hand screen was generally 5 to 10 per cent brighter than the
right-hand screen because of lamp filament misalignment. The design
of the experiments eliminated any effect on the final results due to
.this difference.
Room illumination was provided by indirect flourescent lighting located
so there were no glare sources in the subject's line of sight. Illumination
near the screens was approximately three foot candles and the luminance of
the area adjacent to the screens was approximately two foot lamberts. This
illumination was adequate for reading test materials.
2.2 , IMAGERY
Imagery for the resolution test was prepared from a USAF tribar resolution
chart by varying exposure time to obtain copies at the five contrast levels
listed in TABLE 2-1. The microdensitometer measurements used to calculate
contrast were obtained on the smallest bars typically resolved on each
target. Brightness of the large square in the highest contrast target
was measured on a screen in the projector and yielded a contrast of 6.87
and a modulation of .77. The resolution target background density was
constant for all five levels. All five copies showed good definition
under magnification at higher spatial frequencies than could be resolved
under any viewing condition with the rear screen projector. The highest
contrast copy of the target showed a slight tendency for the bars to fill
into the spaces.
The spatial frequency of the. elements in the,tribar resolution chart was
specified in terms of a resolution number (RN), which was related to
spatial frequency on the screen as follows:
2(RN/6)
Frequency (linos/millimeter)
11.4
Frequency values for a range of resolution numbers are listed in TABLE 2-2.
The imagery for the screen quality testing consisted of 18 frames of large
scale *operational imagery on a 9-inch format. The frames were selected
.0
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
LDeclassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
TABLE 2-1
CONTRAST CHARACTERISTICS OF RESOLUTION TARGETS
/),e)
0!iii)
/
23.0
.92
4.45
.69
.86
.30
.38
.16
.073
.035
B1 -B2 = 2m
, where B1 and B2 are the luminance.
of the brighter and darker areas,
respectively.
bM = B1 - B2 = C
B1 + B2 2+C
Declassified in Part - Sanitized Copy Approved fo.r. Release 2012/09/06: CIA-RDP79B00873A001900010119-3
r i Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CU-:5
LJ
4.s1
TABLE 2-2
SPATIAL FREQUENCY ON PROJECTION SCREEN
FOR EACH RESOLUTION NUMBER
RESOLUTION
NUMBER
SPATIAL
9
FREQUENCY (1/mm)
19
.79
20
.89
21
1.00
22
1.12
23
1.25.
24
1.40
25
1.57
_ 26
1.78
27
2.00
28
2.24
29?
2.49
30
2.80
31
3.14
32
3.56
33
3.99
34
4.47
35
4.99
36
5.60
37
6.29
38
7.11
39
7.98
40
8.95
41
9.98
42
11.20
43
12.57
44
14.22
_
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
so that half had normal contrast and half had a lower than normal
contrast, usually because of ground haze. The selections were made
by personnel whose normal task was assessment and improvement of
image quality. On each frame a heavily developed ground area was
marked off for viewing by the interpreter when he was making a quality
judgement. Transmission of the imagery in these areas typically varied
from 2 to 25 per cent measured over a spot approximately 1/6 inch
wide on the imagery.
2.3 TEST OBSERVERS
A total of ten test observers were used. The three resolution test
observers normally worked on quality control .of operational imagery
and had used tribar resolution targets in the past. The seven screen
quality test observers were experienced image interpreters. All ten
observers stated they had 20/20 visual acuity (corrected) in clinical
tests. Most of the observers wore glasses and one of the resolution
observers had useful vision in only one eye.
2.4 QUALITY TEST I
Each interpreter in the first screen quality test judged 144 pairs of
screens, 72 with normal and 72 with low contrast imagery. The 72 pairs
provided that each of the 36 possible pairings would appear twice,
with the screen positions reversed to counterbalance the effect of any
tendency to left or right responses. Screen pairs and frames of imagery
were presented in a random sequence, with the restriction that the same
screen or frame did not appear in two consecutive trials. The three
interpreters made 'a total of 432 judgements.
The interpreters were told to use their own experience and judgement as
a basis for picking the best screen in each pair, but it was suggested
that they attempt to compare screens in terms of the amount of information
that could be extracted from the imagery. Specific features mentioned
as possibly providing a basis for choice included small, barely resolvable
objects and low contrast edges such as the base of a building in a shadow
area. To make a comparison, an interpreter would generally study one or
ri I 10
Th
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part- Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
0-7.1
4=i
more ground features on one screen and then on the other, adjusting
projector focus as he desired. This process might be repeated several
times during the 1 3/4 minutes allowed for a trial. A choice was
required on each trial, even though an interpreter would sometimes
complain that the two screens were identical.
The three interpreters were tested as a group, all three judging
one
of the screen pairs before the next pair was installed. While one
observer. was viewing a pair of screens, the other two were seated .
outside the test room. The necessity for obtaining independent
judgements was stressed and they were cautioned not to discuss their
choices with each other.
? Screen brightness was maintained at the
test. On-axis luminance of the screens
lamberts, depending on the transmission
maximum level during this
varied from 10 to 107 foot
characteristics of each
screen. The measurement technique and luminance data are included
in Appendix A.
2.5 QUALITY TEST II
The second quality test was like the first except that a single
interpreter served as an observer, and screen luminance was controlled
at an on-axis value of 10 foot lamberts. Control was achieved by
varying the voltage on the projection lamp, as described in Appendix A.
The interpreter viewed only one member of a screen pair at a time. The
other screen was covered with a sheet of cardboard hinged so it could
be swung aside quickly. When shifting from one screen to the other,
both were covered while the lamp voltage was adjusted to the proper
level for the screen to be viewed.
2.6 QUALITY TEST III
The third quality test was
first. Three interpreters
a time. Only seven of the
were excluded because they
conducted in much the same manner as the
were tested; but only one was present at
screens were tested; screens ADA and ALF-5
were generally poor in all the previous
11
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
? ?11,
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06 : CIA-RDP79B00873A001900010119-3
t=1
testing. Screen brightness was controlled with neutral density filters
to maintain a luminance level of approximately 30 foot lamberts for
six of the screens. The seventh, screen AQ-20, had a very low trans-
mission and a brightness of approximately 14 foot lamberts. Brightness
was matched over a 30 degree area, using the measurement technique and
data contained in Appendix A.
2.7 RESOLUTION TEST
2.7.1 Initial Resolution Test
Resolution judgements were obtained for each of the nine screens at
the five target contrast levels and three viewing angles, 0, 22, and
45 degrees, for a total 135 judgements per set. Three subjects each
completed three such sets of judgements under the maximum brightness
level normally available, for a total of 1215 judgements, thus completing
the resolution testing as described in the test proposal, DIC-423.
Additional judgements were made under other brightness and viewing
conditions, as described later in this section.
Screen resolution was measured by having the observers estimate the
smallest element in the tribar resolution chart they could distinguish.
Before testing started, the three subjects discussed and agreed upon a
single criterion for their judgements; the element named was to be the
smallest in which the space between both the horizontal and vertical
bars could be seen. The observers were allowed to change projector
focus freely and indicated different settings were sometimes required
for the .horizontal and vertical portions in a single element.
To enable the observers to maintain the proper viewing angle,sheets
of poster board were mounted chin high at angles of 0, 22, and 45 degrees
off the screen axis. Two of the subjects kept the proper sheet centered
between their eyes; the third aligned his one good eye. The experimenter
sat to one side of the projector and positioned the target vertically to
the height of the subject's eyes.
12
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06 : CIA-RDP79B00873A001900010119-3
,
A.,
Initially a fixed viewing distance of ten inches was planned. During,
the practice session, the subjects indicated they could resolve smaller
targets by moving closer So they were allowed to view the screens at
the nearest comfortable position. This was typically five to seven
Inches from the diffusing surface of the screen.
During testing, a single screen was mounted and judgements were obtained
at each of the 15 target contrast-viewing angle conditions. These
conditions were presented in random order, with no angle and no contrast
repeated on consecutive trials. Then another screen was mounted and the
15 viewing conditions were repeated following a new random sequence. A
single one-hour test session was usually sufficient to obtain the 15
judgements on each of the nine screens. Each subject received three
test sessions in this manner.
2.7.2 Additional Resolution Testing
Two facts emerged which led to an extension of the resolution testing.
The first screen quality test had identified screen brightness as an
important factor in judged quality, and the resolution test subjects
had indicated they felt that even at the closer viewing distance,
their responses were still partially limited by visual acuity.
Additional resolution judgements were obtained under the same controlled.
brightness conditions used in the second screen quality test; i.e., the
lamp voltage was adjusted to obtain an on-axis luminance of 10 foot
lamberts for each screen. Following this, resolution judgements were
obtained under both luminance conditions with a 7X Bausch and Lomb
tube magnifier used as a viewing aid. The amount of data obtained
under these conditions is listed in TABLE 2-3.
2.7.3 Projector Resolution Testing
The quality of the projector optical system was measured by viewing
the anal image with a tube magnifier in the plane of the screen.
The resolution values obtained with each target contrast were
13
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
U
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
TABLE 2-3
SCREEN RESOLUTION DATA
?
PRIMARY
TEST
ADDITIONAL TESTING
SCREEN
LUMINANCE
NORMAL
NORMAL
CONTROLLED
TO 10 FL
VIEWING
MAGNIFICATION
1X
7X .
1X
7X
TEST OBSERVERSa
1,2,3
2
2,3
2
SETS OF DATA FOR
EACH SUBJECT
3
1
1 .
1
a
The subject with useful vision in only one eye was number 2.
14
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
as follows:
CONTRAST
RESOLUTION NUMBER
23.0
41.5
4.45
43.5
.86
42.0
.38
40.0
.073
22.0
15
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part- Sanitized Copy Approved forRelease2012/09/06 : CIA-RDP79B00873A001900010119-3
CG-3
1=1
.4????
, 3.0 RESULTS
3.1 SCREEN QUALITY JUDGEMENTS
The quality data were analyzed by calculating the proportion of time
.each screen was chosen as best. These proportions were converted to
normal deviate Z scores corresponding to the proportions of a
dichotomous unit normal distribution. The effect of this was to
increase the weight given to extremely high or low proportions.
The quality scale values obtained in each of the three tests are
plotted in Figure 3-1 and listed in TABLE 3-1. Connecting lines
were used in the figure to show the scale values a screen received
on each test.
Some of the screens, for example AQ-20, varied
spectacularly between
consistently poor and
studies. Others, such as AL-4 and AL-5, were
some, such as AQ-11 and AR-27, were consistently
among the best liked screens.
smallest in Test III.
The range of screen scale values was
Quality scale values were calculated separately for each imagery
contrast level. These values are plotted in Figure 3-2 for each
quality test. The magnitude of the differences between imagery
contrast levels was generally small relative to the differences
? between screens.
To assess the effect of imagery density on screen quality judgements,
the 18 frames were divided into halves on the basis of density and
the data obtained in Test III were reananlyzed. The resulting scale
values can be found in TABLE 3-2 and Figure 3-3- Differences as a
function of imagery density were very small in comparison with the
differences between screens.
3.2 SCREEN RESOLUTION
.The resolution data obtained during the testing were described in
TABLE 2-3. The principle data were three replications by three
16
7
v.J
Declassified in Part- Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
t=0
41".?11
,
f;-?
SCALE VALUE (Z )
TEST I
.TEST II
TEST III
-1.5
PR-11
AR-2T
AQ-18
AQ-17
AQ-20
FIGURE 3-1 SCREEN QUALITY SCALE VALUES IN EACH TEST
17
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
-
\
!!
i
, 7
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
. CO
QUALITY SCALE VALUES
I-I
I-I
I-I
E-i
Cl)
ILI
El
COMBINED NORMAL LOW
---
,
?
P. 4 .`C:3 2
1
CD C\1 0 I???
Vi N Ctl 0)
I
c?-?-? e
r: ?1:1' -3 T.-
r-I 00 a) 00
I.0?.-INO3
? ? ? ?
I
'g3tral
? ?
I
I???? r-4
N N
? ?
I
? _.+ .....
- ...
co o
CNIv
? ?
I
4
I
In
r-i
?
I
--?
to
r-I
.
I
I-1
E-4
CO
Ix1
El
COMBINED NORMAL LOW
,,, ?
U)
?
r
1
t-
co
I
___0
ri
0
r-f
I
4)1 0 gO) 0 3
? ?
I
a) 0, ts) oo
or) o ?,14 ,-f CD
? ? ?
I I
\r- c, ..r- r-- z---_,.,
r- t- 00 ri
CID00000
I I r-I
0 t;;
I
cn cs)
V 444
I
V)
. ,
'1, 03
C\1111
I
`CI
?
?
(3)
V
.
-9
0
VI
TESTI
COMBINED NORMAL LOW
SCREEN P
AL-4 -.24 PO -.27 -.21
AL-5 -.16m -.16 7.16
AQ-11 .30 ,41- .16 .44
AQ-17 -.27,0 -.35 -.21
AQ-18 050
AQ-20 -1.25 4 -1.30 -1.19
AR-27 .35 AM .38 .32
AR-28 .24 ,(,? .33 .16
LS-.60G .86 p 1.03 .73
E-I
c0
ri M
0 E-4
? 8
18
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
SCALE VALUE (Z)
TEST I
TEST II TEST III
NORMAL LOW NORMAL LOW NORMAL LOW
CONTRAST CONTRAST CONTRAST CONTRAST CONTRAST CONTRAST
1.5
1.0
5
-1.0
-1.5
0
I 0
L- Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
^
LS-600
AQ-11
AR-27
AR-28
AQ-18
AL 5
AL-4
AQ-17
AQ-20
????
AQ -20
!VC11:211
AQ-18
AL-5
AL-4
77 AC2-11
4?1?11.101N.
FIGURE 3-2. hieFECT OF IMAGERY CONTRAST ON SCREEN QUALITY
SCALE VALUES IN EACH TEST
'19
AQ -18
AR-27
-17
Ls-6w
AR-28
Pia-20
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
CG-3
?
-,, TABLE 3-2
QUALITY SCALE VALUES FOR EACH SCREEN
AS A FUNCTION OF IMAGERY DENSITY
SCREEN
IMAGERY DENSITY
LOW
HI GH
AQ-11
.44
.74
AQ-17
.23
.13
AQ-18
.44
. .20
AQ-20
-.74
-1.04
_ AR-27
.25
.33
AR-28
- .36
- .44
LS-60G
-.10
.20
2d
-
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
i
CG-3
LOW DENSITY
IMAGERY
HIGH DENSITY
IMAGERY
FIGURE 3-3. PlikECT OF IMAGERY DENSITY ON SCREEN
QUALITY SCALE VALUES IN TEST III
21
L. Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
Declassified in Part - Sanitized Copy Approved for Release 2012/09/06: CIA-RDP79B00873A001900010119-3
t:=J
1
-n
Jo
observers,with normal screen luminanee and no magnification aids.
However, in order to assess the affect of screen luminance, analyses
were also performed on the remainder of the data. Four analyses of
variance, each involving a full factorial design, were performed to
identify which variables were associated with differences in resolution.
The basic data set of 135 resolution judgements on nine screens at
five target contrast levels and three viewing angles was included in
each analysis. The additional factors were as follows:
1. Three observers, three replications, normal screen luminance
and no magnification (this was the primary analysis);
2. Observers 2 and 3, one replication,
screen luminance, no magnification;
3. Observers 2 and 3, one replication,
luminance, no magnification; and
4. Observer 2, one replication, normal
luminance, 1X and 7X magnification.
normal and controlled
controlled screen
and controlled screen
The most important results from these analyses are reported in this
section. Summary tables for each analysis and the associated data-
are located in Appendix B. The summary tables include a statement
of the statistical significance of the differences in resolution
associated with each of the test variables. These will be indicated
in the text below as the probability (P) that the differences discussed
were due to chance.
At normal luminance levels, the screens differed in resolution (P(01).
As Figure 3-4 illustrates, the differences fell generally into two
groups. A Duncan's multiple range test indicated that the tWo groups
differed from each other, but within a group only screens LS-60G and
AQ-11 differed (P