He regards OTF as a step forward from resolution, but not as a final answer. His company has, largely, 
dropped resolving power tests, because of disagreements in results. They had found that about half of their 
customers buy cameras using MTF specifications. His company buys lenses to MTF specifications, test to them, 
and use MTF for focusing. In one case they found OTF much more powerful in explaining focusing problems 
than was resolution. He was impressed with the speed and precision with which tests could be made, and felt 
it was a practical procedure for camera manufacture. However, the success of routine use in industry was a very 
different question from the establishment of an international standard. 
Mrs. Norton asked what variation from the norm is acceptable for an individual lens. 
G. Brock replied that any lens is rejected if it is more than 10% below the norm in resolution value as derived 
from MTF. He reported one example, when a poor lens just passed a resolution test, but failed on MTF and 
on picture quality. He mentioned that it is easier to check centering error by MTF than by resolving power. 
H. Meier said that we should not ask *MTF yes or no", but for which task: for system analysis, MTF is an 
important tool for finding bottlenecks which need improvement; for lens design, computers give us an 
assessment of quality by MTF or spread function. 
For production, his company have differentiated between photographic (pictorial) and photogramme- 
tric lenses. For the former, MTF gives good results, saves time, and is used extensively in nearly automatic 
testing devices. For photogrammetric lenses, distortion and symmetry are very sensitive to manufacturing 
defects, so MTF is not used except on prototypes. - 
Mrs. Norton asked R. Hopkins about the agreement in computed MTF between design and measurement. 
R. Hopkins reported a study showing that computer design programmes now agree well enough to 30°, but 
are doubtful at 45 For measured MTF there is very poor agreement in off-axis results in different laboratories. 
The differences are attributed to locational errors. 
W. Mandler remarked that he had grave misgivings about the possibility of standardizing, and that this 
meeting was confirming his view. He said that we should not make complex problems more complex by asking 
for a single figure of merit. If we can agree on a standard method of measurement for one point and one scan, 
all the rest is just mathematical processing. 
W. Pershing said they are obtaining equipment to measure MTF, but still depend on resolving power. 
Computed MTFs are currently converted to resolving power for judging lens suitability. They wish to see 
progress in MTF toward definitions and standards, and as a supplement for resolving power. 
Mrs. Norton called on several delegates for comments. 
W. Tayman said that the National Bureau Standards (USA) had been working with OTF and MTF for 15 
years, but feel that it is still 5 or 10 years to a set standard. Even then he does not expect one single method, 
but several. 
R.V. Shack felt that it was not yet time for a clear choice of merit function. However, one reason for the use 
of MTF was the ease of working with it by multiplication rather than by convolution. A merit function should 
retain this feature - perhaps by using the modulus at one or two selected spatial frequencies. He suggested 
seeking a figure of merit for an image at one point in the field. Then performance could be presented by 
plotting contours of constant merit in a graph of field position against focus position. 
P.D. Carman felt that resolving power was still most useful for some purposes, while MTF was most useful for 
others. He felt resolving power was at present the more useful to photogrammetrists, and suggested that 
Commission I needed to decide whether MTF was sufficiently useful to photogrammetrists at present to justify 
a working group, or if they should simply watch developments in the field. 
S. Hempenius described some applications of MTF pointing accuracy. He would favour, for photogrammetry, 
a figure of merit weighted for the corners, since they are critical in aerotriangulation. 
For photointerpretation, the centre is more important. MTF is particularly useful in photointerpretation 
because its mathematical approach is applied easily to matters of telemetry, and of correlation and 
auto-correlation. 
The meeting adjourned.