(225)
satisfactory for grading binoculars. There seems to be an element of crispness
or sharpness of definition that is desirable and of which threshold resolving
power is not a satisfactory measure. Serious consideration is now being given
to the displacement of resolving power measurements by other data for the
specifications of photographic emulsions which will give more weight to the
density gradient at an edge. Furthermore, it should be noted that a resolving
power test is a threshold test which gives specific information concerning the
spatial frequencies that will not be recorded but gives no information, except
by inference, concerning the fidelity with which the resolved frequencies are
reproduced.
In the general field of photography some work has been done to determine
whether or not resolving power is a satisfactory criterion and there are indica
tions that at least for some purposes the more desirable photographs result,
not when the focus is that which gives the maximum resolving power, but for
an adjustment which tends to give better contrast. Again the determination of
this question for airplane photography requires an expensive program of flight
tests and such tests might well show that photogrammetrists using photographs
for mapping purposes and photo-interpreters interested in the military infor
mation in a photograph demand photographs of different quality. This differ
ence in requirements is mentioned as probable because for many phases of
map making, particularly for civilian mapping purposes for which the terrain
can be physically occupied, pointing accuracy is of great importance and
pointing accuracy or metrological power commensurate with the accuracy of
a negative can be attained with relatively low resolving power. In such a case
it might be desirable to trade resolving power for additional contrast. On the
other hand, for objects at the threshold of detection and perhaps for other
objects as well, resolving power or some equivalent characteristic may be of
great importance to the military photo-interpreter.
Although admitting the importance of a knowledge of the performance
of a lens, emulsion, airplane, and target it is desirable that the over-all perfor
mance be broken down into a series of contributions in order that each
element can be tested separately. This is in line with testing procedures in
general and will eliminate the need for the establishment of arbitrary conditions
which make the definition of resolving power for photogrammetric purposes
as embodied in the present specification so complicated.
As evidence of this it may be noted that the difficulty in arriving at a
standard procedure for testing resolving power as indicated by the final
decision to approve three types of charts arises because of different opinions
concerning the manner in which this simulation is to be achieved in the
laboratory. Furthermore, with the method of determining the resolving power
defined by specifying test chart, film, procedure for processing, and other
related details there is no simple definition of resolving power. Resolving power
is what one gets when a fully outlined procedure is followed. No room for
improvement is left so long as this type of definition holds because when one
changes anything one gets a different resolving power, not a better value of
what one originally set out to measure. Furthermore, if a new film which is
strikingly different from the present film comes into general use all the accu
mulated data on the old film become substantially worthless.
As additional advantages of substituting a series of measurements of fun-