TEILS WO TEN aT 
  
EE >>> 
  
  
  
y = re veu 
10. 
For this step the Modulation Sensitivity curve (MS) plotted on a 
transparent. overlay was set with the correct vertical position being 
determined by the granularity, and horisontal position by the appro- 
priate magnifications i.e. for 10X optical magnification the MS curve 
was shifted by a factor 10 towards the higher frequencies. Using the 
frequency limit target size and pointing precision were then derived 
from graphs of Trinder (1973). 
Variances of x-coordinate observations were deduced from the appro- 
priate pointing precisions, being subject to image movement, while 
those for the y-coordinate were subject to the enlargement factor of 
(1.3)? over the pointing precisions of x-coordinates, but not to image 
movement effects. 
Variance ratios for the low and high flying heights for each tar- 
get contrast and ray inclinations of 0? and 35? (Tangential) for the 
wide angle cameras, and 0°, 30° and 55° for the super wide angle 
camera are shown in Tables 1 and 2 for the optical magnifications of 
5, 10 and 20X for 3 target contrasts (excluding 1.2:1- for 20X nagni- 
fication which proved to be impossible to determine because of the 
very low FL). As the variation in precisions for wide angle cameras 
is approximately linear with respect to (ray angle)“, it was consi- 
dered unnecessary to present relative variances for all inclinations 
of the incident ray. This fact should be clear when viewing Figures 
2, 3, 4 and 5. The best and worst situations have been shown in 
Tables 1 and 2. No attempt has been made to compare precisions 
derived from MIF's for tangential and radial orientations of the test 
target. For 20° ray angle the differences would be negligible, and 
for 35° differences would become apparent only for low contrast, 
and/or high optical magnifications. 
2.7 "Discussion 
For all cameras, the most significant factor affecting pointing 
precisions is the contrast of the target. Clearly, low contrast 
targets will result in much higher coordinate variances than will the 
higher or medium contrast targets.  Premarking of targets is therefore 
essential for high accuracy photogrammetry, especially in the case of 
high altitude photography for which losses in contrast are greatest. 
In the case of low flying height (Table 1) there are only small 
differences between variances in the x and y-coordinate because of 
the separate effects of image movement on the x-coordinate observa- 
tions, and lower precision of pointing for the y-coordinate. As the 
ray angle increases the differences in variances are only marginal, 
for low optical magnifications, because image movement is the major 
factor affecting image quality. However the differences increase to 
ratios of 1:6 for the lower quality wide angle camera 188 for high op- 
tical magnifications. These variations of coordinate variance with the 
various parameters demonstrate the difficulties in presenting a simple