have proved invaluable in the large-scale computer-oriented design 
process. Although MIF has not yet entered into the merit function 
formulation to use in assessment of system quality, it typically 
outweighs other criteria. 
In the evaluation of prototype systems a correlation between 
paper design and hardware execution is essential to development 
process. As of this writing, modulation and phase measurements 
provide the best readily available measurements for the establishment 
of such correlation. 
It has not yet proved possible to utilize MTF and OTF 
evaluation in routine characterizations of wide-angle lenses 
such as those typically used for photogrammetric purposes. The 
difficulties encountered in attempts to do this lie both in physical 
contortions (the very precise adjustments and locations of equipment 
parts) necessary and the uncertainties in the results of subsequent 
analysis of the data obtained. 
It is envisioned, however, that routine contrast measurements 
with electro-optical equipment can be made to augment, or eventually 
to replace, other forms of testing in the screening, and even 
perhaps in the final evaluation of lenses having more modest angular 
coverage. 
5.9.6 University of Rochester and Trapel 
5.9.6.1 Slaymaker69 notes that in MTF measurements there are three 
sources of noise which can enter at almost any step which must be 
considered: Those introduced experimentally, those due to the 
quantum nature of light itself, and shot noise from the photocathode 
of the photodetector. Assuming that the first two can be reduced 
to negligible values, he deals with shot noise which appears both 
as a bias term added to the true value of the MIF and as a variance 
in the measurement. He shows that in making an OTF measurement by 
means of a knife-edge scanner, the effect of the shot noise can be 
reduced to an acceptable level by establishing a maximum value for 
the scan velocity. An upper bound for the noise bias can be calculated 
either from the variance in the experimental data or computed from known 
values for the electrical bandwidth, the time to make the scan, the 
gain of the photodetector and the anode current. To verify the noise 
calculation experimentally, Slaymaker measures the variance in the 
MTF data. He then calculates the mean and standard deviation for 
the data at specific spatial frequencies for a number of repeated 
scans. 
A comparison of the calculated and experimental results 
obtained with a 0.1 x Micro Tropar lens having a focal length of 
26 mm, an f/number of f/1.6, and a theoretical cutoff frequency some- 
what above 1200 cycles/mm, shows good agreement at spatial frequencies 
27