and with a focal length of 150 millimeters, an input scale of 
1/20,000 resulted. By using duplicate diapositives tilted 
identical amounts (three meters vertically for each 500 meters 
horizontally in both x and y) and producing three meter line 
drop contours, a value for sensitivity or "C" factor was obtained 
which was essentially free from all external errors such as 
those contributed by ground control, film distortion, etc. This 
value represents the cumulative errors of the UNAMACE itself. 
The output was a set of sharply defined bands running at 459, 
Measurement and analysis of this output showed an average height 
uncertainty of 0.25 meter or an instrument "C" factor sensitivity 
of 6,000 for vertical photography of this focal length and scale. 
b. Typical of single model tests are the following 
two. The first is a model of Gila Butte, Arizona flown at 33,000 
feet with a six inch focal length camera, giving an input scale 
of 1/66,000. Output scale was 1/62,500 with a 40 foot contour 
interval. This area is gradually sloping with approximately 
250 feet of relief in one corner. The model was compiled in 
one hour. The orthophoto with superimposed contours are shown 
in Figure 16. In the terrain shown there were only one or two 
stops requiring manual intervention. Based on interpolating 
elevations from the extracted contours for 44 spot elevations, 
there was a mean error or datum shift of 10 feet and an indicated 
"C" factor of approximately 1,500. The second model was of Fort 
Sill, Oklahoma and was flown at 25,000 feet with a six inch focal 
length camera, giving an input scale of 1/50,000. Output scale 
was 1/50,000 with a 40 foot contour interval. This model was 
about 400 feet of relief and was compiled in 85 minutes. A "C" 
factor of 2,300 with a mean error or datum shift of four feet 
was achieved based on interpolating elevations from the extracted 
output contours for the 37 ground points. Figures 17 and 18 
show this compilation as a contoured orthophotomap and the line 
drop output with the contour lines superimposed. 
c. Inspection of the line drop output reveals how 
the draftsman averages and interprets the dropped lines. The 
series of horizontal lines and symbols in the upper left indicate 
the boundaries where the "ADV-AR" was employed. The scrambled 
dropped lines in this area are over a large body of water where 
correlation does not occur. The UNAMACE failed to properly 
correlate at the edge of this lake and as a consequence approxi- 
mately fifty manual interventions were necessary in this area. 
This time is included in the overall compilation time of 85 minutes. 
The remainder of the model required no manual intervention. 
The drafted contours, grid lines, etc., are overprinted on the 
line drop output for comparison purposes. The indicated ''C" 
factor is sufficient to compile a 10 foot contour interval, but 
in practice this cannot be achíeved because of the line drop 
definition at this contour density. A 20 foot contour interval 
chart was drawn by shifting the datum by 20 feet, compiling 40 
foot contours and combining the two 40 foot contour charts. It 
is expected that when contours are compiled from the digital tape 
output this problem will no longer exist. This orthophoto, which 
is typical of a 1:1 product, represents a resolution of about 40 
line pair per millimeter as produced by a hígh contrast USAF type 
resolution target placed in the input diapositive holder. 
41