62 THREE-DIMENSIONAL PHOTOGRAPHYBecause so many inquiries have been received about this re-
flector it should be repeated that it is not available commercially.
It can be made by altering a war surplus instrument, but it is
subject to the very real disadvantage that the field is extremely
narrow. It is more suitable for experimental purposes than for
the production of satisfactory pictures.
Keeping in mind the fact that the orthostereogram reproduces
the original in full life size and at full natural distance, we shall
now consider those variations in manipulation which result in the
stereoscopic distortion of either size or distance, or both. This will
serve to indicate those errors of technique of which the stereogra-
pher should not be guilty, and even more to the point, it will
indicate some violations of traditional stereo rules which may be
made to great advantage.
In practice, orthostereoscopy is rare because the average viewer
lenses exceed camera lenses in focal length by amounts varying
between 25 and 50 percent. A 25 percent degree of discrepancy
yields an image in which the distortion is not readily apparent,
but a 50 percent difference will produce a distortion of spatial
dimension, perceptible to the expert stenographer.
FIXED AND VARIABLE FAcToRs.-Although there is an infinite
variation in the interpupillary distance of human beings, these
variations are within such close limits that we may establish an
arbitrary standard such as 65mm and henceforth regard that as a
fixed factor.
Once the exposure has been made, the parallax of that stereo
image is fixed.
The perspective dimension of the image (dimensions in the
picture plane) is subject to alteration in making the negative by
changing the focal length of the lens. It is subject to further alter-
ation in the process of projection printing by either enlargement
or reduction, and it is subject to final variation by the focal length
of the viewer lenses.
To sum up, we may say that, generally speaking, parallax is
fixed while perspective dimensions are variable. Hence it shall be
our goal to control the perspective so that it will always match the
parallax correctly-or otherwise-as we desire. It is true that in
parastereoscopy we do this indirectly by deliberately introducing
an abnormal base in making the negative, but as this has already