han any other for
its undoubted dis-
by some new test,
tance”, the “effec-
methods continue
sasily? Will it not
f a camera's per-
y one involves as-
nd that frequency
e this is a contro-
: of the resolution
hods it is hard to
e circle containing
ction, especially if
results will not be
pear to be nearly
trate effort on the
concept of camera
an be used to pre-
onditions of deter-
ions of air photo-
our of the camera
arable with those
['herefore, in spite
'e in the study of
: undoubtedly that
ves about the best
ective impressions
a permanent self-
', especially in the
against the older
ration of the film
It is perhaps wise
Oo know — we may
. When a test has
s that gave vastly
eristic of a photo-
sirable. The ques-
d by the Eastman
graph to do. If it
THE PHOTOGRAPHIC IMAGE, BROCK 21
is to reveal the finest possible detail, then a resolution test is probably right, provided
that the modulation at lower frequencies does not fall below an easily visible level. But
the frequency response curve would need to have a very peculiar shape for this to
happen, and such shapes are rather unlikely in the lower frequency region corresponding
to the low contrast detail of air photography, well clear of the resolution limit. If,
however, the picture has to look sharp from a distance, the resolution test may well be
no guide, but this has nothing to do with air photography.
The writer is aware of the interesting experiments of the Eastman laboratories on
this subject, extended now to low contrast test objects, but would offer against that his
own experience, which indicates a rough correlation between low contrast resolution
tests and actual air photography. By “rough correlation” is meant that while the ex-
perimental conditions preclude the use of correlation coefficients, the air photographs
do not reverse an order of excellence determined from resolution tests. This experience
covers something like 100 lenses of different construction, tested by low contrast reso-
lution tests and air photography, over a period of some twenty years. Each new lens
was tested against a known lens. In addition, many new emulsions have been compared
with standard types. In some cases the tests have represented very different combina-
tions of emulsion and optics. Always the air photographs have given the same order of
merit as the resolution tests. One out-of-the-ordinary case may be quoted. An orthodox
camera with 36” lens, used at f/11 with a high-speed film of 27 lines/mm resolution,
was being compared with a 12” mirror camera used at f/2 with a slow film of 55
lines/mm resolution. The diffraction pattern of the mirror-camera image was broken
into three lobes by its internal supports so that the energy distribution was peculiar.
Resolution tests indicated that the ground resolution of the mirror camera was not
quite equal to that of the 36” refractor. The air photographs fully confirmed this result,
slightly more ground detail being shown by the 36” camera. This kind of confirmation,
repeated over many years, gives confidence that the low contrast resolution test is a
reasonably accurate guide to the practical excellence of a lens or film. On the other
hand, this same experience of testing has also left a sense of dissatisfaction with the
results, primarily because one number is felt to be inadequate as a description of the
“goodness” of an image. The testers sometimes comment that whereas two lenses may
give not very different resolutions, one has a “cleaner” image than the other, with more
contrast in the larger detail. Again one sometimes feels that the scale is not consistent,
e.g. that a given change of resolution due to different causes does not always produce
the same improvement in detail seen or in ease of seeing. This lack of correlation has
never extended to an inversion, as already stated; it is most noticeable in the presence
of very heavy aberration, and is less serious as lens correction improves.
4.2.2. High or low contrast targets.
For many years opinion has been divided on this point, with the British and Cana-
dians favouring low contrast, while the Americans and Europeans on the whole prefer
high contrast. There are, of course, arguments on both sides, and the situation is now
so complicated by considerations of expediency and continuity, with personal (or nation-
al?) interpretations and prejudices, that there seems no prospect of reaching agreement.
So far as the writer is aware, there has been no large scale experiment in which many
lenses are tested on high and low contrast charts with two or three different emulsions,
the results being then compared with actual air photographs.
Such an experiment would be very costly to carry out, but is about the only way to
settle the arguments on a basis of fact rather than opinion. Advocates of frequency
response tests would of course say that all the information needed could be obtained
from the response curves, which would also show that there is no unique answer to the
problem. The exercise would now be largely academic, but it might be worth doing if only