.d many lab- 
prime object- 
the long 
>al tools in this 
escribe the 
le the non- 
. The addition 
loped emulsion 
)hysics of 
il" has been 
greatly improv- 
mpenius, for 
jn.* From the 
lple test, 
able. The 
:nents and 
Laboratories which do measure transfer functions are emphatic about their super 
iority to resolving-power because of their "objectivity", but in itself this has no 
value except for routine testing. The importance of finding a link to psycho 
physical image quality in terms directly useful to the aerial photographer is still 
too little appreciated, though one major photogrammetric laboratory in West Germany 
has in fact worked on this problem. While such studies will throw much light on 
some obscure details, they cannot alter the fundamental fact that the imagery of 
every object is a special problem, soluble with full accuracy only after operation 
on its Fourier spectrum with the transfer function. Many approximate indices of 
imaging performance can be derived, such as re solving-power, single line contrast, 
or various integrations of the transfer function, but each of these will be fully 
accurate only within the limits of its own definitions, and some loss of precision 
is inevitable as soon as they are extended to include imagery in general. For the 
conditions of aerial photography, detailed study and experience with transfer func 
tions indicates that low contrast resolving-power is as good a compromise as any 
other criterion of performance, basically because (if the target contrast is low 
enough) it rules out of account regions of very low modulation transfer (say <20%) 
which make little contribution to micro contrast in the image. 
In general, the use of transfer functions, having taken image-evaluation into the 
spatial frequency domain and hence away from real images as we see them, has 
caused a good deal of confusion. This is particularly true when the interpretation 
, is an al- 
st, and is 
necessarily 
iirmation of 
anizations, 
esting lab- 
;ure a transfer 
ler information 
on of its 
of phase distortion has to be considered. Thus, although transfer functions are 
essential to the research worker and will probably be increasingly used in some 
form as a routine lens test, the situation in regard to their more general use is well 
summed up in a reply by Dr. Francis Washer. (U.S. Bureau of Standards) 
"Until methods can be evolved that substantially reduce the time and 
effort required to produce the data and interpret the results in a mean- 
ingful manner, it is likely that resolution will continue to 
be used as a quick convenient method for estimating the image 
forming properties of a lens." 
To the question "Has the use of transfer functions contributed to the improvement 
of photogrammetric cameras?", replies were received from two organizations in 
3 has been 
snce to 
.ation as a 
3 have been 
»rocal image 
e derived from 
irpose, i. e. , to 
i of targets in 
on by itself is 
ate indication 
:he general 
the best position to know. One reply was a strong positive, the other an emphatic 
negative. 
Though "image quality" is hard to define, it is easily recognized, and the question 
naire revealed a fairly general feeling that the average standard is still too low, 
though some consider it to be adequate. One factor in this, specifically quoted 
by some correspondents, is undoubtedly the continued use of the older wide-angle 
cameras; however opinions were sharply divided between camera, emulsion, image 
movements and printing losses as the prime cause. It seems likely that image move 
ments are more serious than is generally realized, which would account for some 
of the contradictory replies. To quote a Canadian reply "The vibration-absorbing 
characteristics of standard survey camera mounts appear to have fallen far behind 
the state of technological development in this field". One other cause of poor 
quality, emphasized by two replies, is improper exposure and development. These 
comments, indeed, are very much to the point, and make discussions of transfer 
function theory seem rather academic. In this connection the contributions of 
Professor Jackson to the 1956 and 1960 Congresses will be recalled; it is of interest 
that his recommended "low contrast" techniques are now in regular use by some