825 
impared 
and Flight B 15,000 feet images (A 15,0 R1 vs B 15,0 Rl), and 
the flight B 6,000 feet image and the repeat measurement of the 
same image (B 6,0 Rl vs B 6,0 R2). Colors closely matching but 
significantly different were the comparisons of the red image of 
Flight B 6,000 feet as contained on two adjacent transparencies 
(B 6,0 Rl and B 6,0 R2 vs B 6,0 R3). The view angle, especially 
when one of the images is near the outer edges of the trans 
>up being 
parency as was the case in this study, can affect the color. 
Most of this change is in terms of visual transmittance given by 
the Y tristimulus value where a change from 68.92 to 63.72 is 
noted (Table 2). The CIE-UCS u,v coordinates, however, differ 
only at the fourth decimal place. Therefore, the chromaticness 
differences are negligible even though the extremely low variance 
values of this study provide a significant test difference. 
i a F- 
.ows: 
For the green target, all comparisons were significantly differ 
ent although the same tendencies exhibited by the red target were 
evident. The Flight A and Flight B 15,000 feet images (A 15,0 G1 
vs B 15,0 Gl) were quite similar as were the Flight B 6,000 feet 
independent color determinations of the same image on two adjacent 
transparencies (B 6,0 Gl vs B 6,0 G2). Also similar were the 
Flight A and Flight B 1,500 feet images (A 1,5 Gl vs B 1,5 Gl). 
observa- 
For the white target, the 1960 CIE-UCS u,v coordinates for all 
colors are so nearly the same that visually all images would be 
judged to be almost clear of color. However, the nearly nil 
variance terms associated with the image sample measurements 
make an analysis possible although possibly somewhat question 
able because of the unknown effects of slight equipment set-up 
and calibration differences which are included in the measure 
signific- 
ment data. That this concern is warranted is attested to be the 
fact that Flight B, 6,000 feet image and a repeat sample of the 
same image (B 6,0 W1 vs B 6,0 W2) are significantly different 
even though the difference was slight (Mahalanobis 1 D 2 = 2.55). 
Nevertheless, the Flight A and Flight B 1,500 feet images are 
not significantly different (A 1,5 W1 vs B 1,5 Wl) continuing 
the pattern for the images at 1,500 and 15,000 feet to be nearly 
the same color respectively on each of the flight dates. 
1 sample 
il value 
COLOR SHIFTS 
>ns were 
Imaged Red Target 
Reference to Table 2 and Figure 3 (2x enlargement insert) shows 
an altitude associated color shift toward blue-green for the red 
imaged target which appears as a saturated yellow color on the 
transparencies. The changes in the visual transmittance given 
by the tristimulus Y value and the analytical density of the 
ltly dif- 
Light A 
yellow, magenta and cyan dye layers on the different Flight 
dates-altitude combinations may be partially explained by the