interferometrically, introducing these errors into the design, 
and computing the degraded MTF, Also a common practice at Itek 
is to adjust each lens for variations in glass melt indices, 
element radii, and central thicknesses from design values in 
order to maximize the MIF across the field of view. For very 
high quality systems, a "fine tuning" of the optics is sometimes | 
done based on lens system measured wavefront data from several 
field points. The lens MIF at these points are then optimized 
by tilts, decenters, and air space changes of key optical elements 
as determined by image analysis techniques. 
  
MIF analysis are further used to predict operational 
performance considering environmental effects such as thermal 
and structural changes and image motion effects due to vehicle i 
and camera dynamics. 
Another MTF evaluation is used on fiber optics. À 
fiber optics faceplate is composed of an ordered array of small 
individual light pipes which transfer images from input to output. 
Conventional methods of investigating the fiber-optics image-transfer 
process utilize the spread function and the edge response as 
a basis for transfer-function calculation. Unfortunately, a transfer 
function calculated from a fiber-optics spread function does 
not have clear meaning. A holographic method is, therefore, 
employed, photographic recordings being made of the summation 
of sinusoidal fringes at the input surface of the fiber optics. 
From these, the Fraunhofer diffraction patterns are produced. 
Because the irradiances of the narrow sinc light distributions 
in the Fraunhofer diffraction patterns are directly related to 
the amplitude of the sinusoidal irradiance variations, the pair 
yields a fiber-optics transfer factor, and a statistical MTF, 
the latter being a measure of the average performance. Tests 
were conducted with techniques that provided additional infor- $ 
mation on lateral and rotational distortion of fibers, missing 
or broken fibers, and resolution. 
5.9.7.2 Limations of Image Analyses 
A succinct description of use and reservations on MTF 
for image quality analysis or specifically resolution prediction 
follows: 
For proper MIF use the system on which the analysis 
is performed must be linear and a typical system (Skylab) was 
shown non-linear through the first generation duplication stage. 
If one uses MIF analysis for resolution prediction and good results 
are expected, then each component of the system process must 
be correctly modeled e.g. the duplication process would quite 
likely contain non-linear paramenters, and the proper threshold 
modulation curves must be used in the analysis. 
= 
30