40 DISCUSSION ON PRESENTED PAPERS
My paper recognises two types of interpre-
tation, the photo interpreter and the photo-
grammetrist. Both require definition at some
time, the former, the photo interpreter may be
interested in tone rendering more particularly
than the photogrammetrist. By photogramme-
trist I mean the aerial triangulator, the contour
chaser, the man who wants to know the co-
ordinates of a single point rather than what kind
of vegetation that is.
Secondly, there are two groups of factors
that affect the interpretability of photographs,
the first pertaining to the production of the pho-
tographs and the second pertaining to the ob-
server and how he looks at it. These two, of
course, are very different. I propose to deal with
them separately and in a hurry.
First, speaking of the negative, at the last
Congress I showed briefly a diagram indicating
the results of our measurements relating reso-
lution and densitiy. I would now like to show
the first slide which includes that information
plus some measurements we have made since.
Here are four diagrams and the curves of these
films and a couple of different developers. The
strips were exposed through step negatives and
resolution targets made by Dr Howlett in N RC
at Ottawa. The densities were measured of line
and background, incidentally the resolution
targets were low contrast, 0.2, and the resolu-
tion was measured in each step in three separate
more or less top charts. These results are the
results of a great many measurements on as
many films as there are lines and indicate the
general distribution of resolution with respect
to density.
If you look particularly at the Kodak Plus X
developed in D 76 on the right-hand side, you
will notice we get a hill. These are contours of
resolution with its axis about parallel to the 0.5
density line. It does not make much difference
whether we get this density by long exposure
and short development or short exposure and
long development, and I think this is worth
noting. Consequently, we suggest that in order
to get the maximum definition in the negative
we must keep the density down and we suggest
a range of density between 0.2 or 0.3 to 1.2 and
no further because of the degradation of resolu-
tion in the higher densities. We will have a slide
[not reproduced here] a little later to substan-
tiate the findings of the laboratory work here.
A very simple cure to conventional proces-
sing at the moment would be to use one larger
stop and develop for half the time. This, of
course, is very crude and does not apply in any
case, but it does indicate roughly the direction
in which to move in order to get more definition
in the negative. I think we can take care of con-
trast in the positive at a later date.
The second point is on image motion, and
since my first paper which suggested a rather
smaller amount of tolerable image motion than
even Mr Brock's 0.6 — I suggested 0.3 — Mr
Trott, who is giving a paper at this Congress,
gave a very interesting paper on image motion
and resolution. In this slide I indicate the results
he showed at that time. Where the image motion
is divided by resolution the image motion of
resolution is 1 and loss of resolution is 30 per
cent. In the top figures there, we have the per-
centage loss in resolution with various pro-
portions of the resolvable distance in image
motion.
In the bottom I have indicated the tolerable
image motion for different films of different
lens film resolution, under various values; the
first was my original 30 per cent of the re-
solvable distance calling, for instance, for a film
with 20 lines per millimetre, a tolerable motion
of only 15 microns. This is the basis of our
specifications for the Lands and Forests Depart-
ment in Ontario, of which you will see some
pictures later.
I would relax and go a third of the way
towards Mr Brock and allow 40 per cent,
rather than his 60, and certainly not the 120 we
heard about a little while ago.
I would like to insert here a note on vertical
exaggeration. For purposes of interpretation I do
not think it is sensible to ask all the interpreters
to convert all the slopes and terrains they have
seen from what they see to what they know it
is, an exaggeration of probably 2.5 times for the
normal 60 per cent overlap for the 6-inch lens
or 9 by 9. I suggest that one sees a cube most
distinctly as a cube when one puts a cube of
sugar in one’s coffee. In other words, at the
minimum distance of distinct vision we have the
maximum disparity in the appearance of an
object, at a shorter distance it is indistinct, at
a greater distance the disparity is less and there-
fore less sensitive. Consequently, I suggest that
the zero exaggeration in aerial photography will
be obtained if we use a base to height ratio
similar to that natural one of the eye, base to
the minimum distance of distinct vision or 1 : 4.
In other words, for natural appearance of an
object, not as one would see it if one were on a
base at 23 feet but as one would imagine a
model at a distance of 10 inches from one's
eye, we would have a ratio of 1 : 4.
This means that in normal 6-inch or 9 by 9
photography, where we now use a 60 per cent
overlap we she
interpretation -
tures for mea:
you some of th
On the posi
that I am pro}
printing proble
control printer
have smaller e
continuously c
degree of conti
The choice
that it takes ca
within the spot
positive to beco
approach those
vicinities. Wha
the end points ¢
be avoided in tl
information ave
Now I woul
[not reproduce
duced not mere
but other things
several occasior
we have constr
cock-eyed ideas
for measuring t
ting. In the up;
view from the g
long, 16 feet wi
diagram you se
feet, and only a
larged indicates
target, with nine
smale, and on t
This is 16 feet
cells 4 inches b
painted glossy b
able to measure
This, we tho
tirely evaluating
the ground. Hav
ground, the only
tive at this point
from the haze.
[Professor Ja
mentary slides.]
This is the las
factors affecting
Archives 4
