38 
SYMPOSIUM PHOTO INTERPRETATION, DELFT 1962 
Discussion 
G. E. Olson (U.S.A.) : What effect would a change of contrast, from high to low, have on the 
validity of the transfer function and on the quality estimate discussed? Answer: The emphasis 
of the paper is on low contrast. Results obtained with low contrast images are different from 
those obtained with high contrast for any lens system. As high contrast tests give border effects 
that deteriorate results, low contrast tests are preferred. 
A. H. Birnbaum (U.S.A.) : If you were to compare two images differing in both contrast and 
resolution, how would you determine which of them is of higher and which of lower quality? 
Answer: Different qualities can be obtained and the transfer functions of each can be deter 
mined when they are analyzed in the way we have described. The quality - which has a 
relative value - depends on the details the observer wishes to distinguish. Correlation tests 
between quality and transfer functions still have to be carried out.