Two possibilities exist to overcome this difficulty which is not a 
special feature of OTF-measurements but is_characteristic for 
every exact sensitive image evaluation. Z^ Describing the imaging 
quality by OTF-values only without further information about the 
optical system means to know the OTF-values or curves of every 
field angle, every azimuth, every orientation and several image 
planes. [5] This large number of curves is inevitable, if an unknown 
system shall be evaluated which may suffer from rotational symmetric 
and non-rotational symmetric aberrations in addition to the aberra- 
tions based on the design. 
However, this large number of data can be reduced by combining some 
measurements which indicate the quality of adjustment and symmetry, 
for example the curvature of image field for different azimuthal 
directions. Then it is possible, at least for a reasonably well 
corrected, centered and adjusted optical system, to restrict the 
OTF-curves to some selected field angles and two azimuths in order 
to describe the image quality of the system. It should be emphasized. 
however, that badly centered lenses, constructions which are sensi- 
tive to adjustments and instabilities, are not only problematic for 
testing, but even more unsuitable for normal use. To exclude such 
undesired specimens, simpler tests than by means of the OTF are 
recommended. Today the imaging quality of certain optical systems 
is often limited more by accidental deviations due to fabrication 
than by the aberrations based on the design. 
Fortunately these cases are rare as far as aerial mapping cameras 
are concerned. They are mostly better centered and adjusted than 
lenses of photographic cameras. 
Summarizing one may say: 
/ s 
Agreement in interlaboratory OTF-tests is very good and we know 
the conditions which must be fulfilled for successful comparisons. 
In order to reduce the amount of data for describing the imaging 
quality, the deviations from the design must be known. 
-5-