264 THREE-DIMENSIONAL PHOTOGRAPHYtion has resulted in a space distortion which can be seen when the
stereogram is viewed.
As for the second argument, that of the nearby miniature, or
the giant's eye view, it, too, is far more important in theory than
in practice. As has been stated, our perception of absolute size is
as undependable as our perception of relative size is acute. The
optics of stereo will prove conclusively that the subject made by
hyperstereoscopic methods is shown as a miniature near at hand.
However, when the stereogram is viewed, this aspect is not seen
except by those stereo experts who know the stereogram is a
hyperstereo, and wh'o see what they have been taught to look for.
This is a common occurrence in many activities other than stereo,
so no more need be said about it.
When stereograms are shown to laymen who are familiar with
the original, and when the subject is such that hyperstereo is
warranted, the universal comment is that the hyperstereogram
best displays the realistic, natural aspect of the original. This
alone would warrant its use, even if the arguments against it
were not so wholly invalid.
ESTIMATING BASE SEPARATION.-The reason for using hyper-
stereo is that the maximum distance at which we have stereoscopic
perception is limited. Although the actual limits vary with indi-
viduals, it has been accepted that among those of highly developed
stereopsis, the limit of stereo differentiation is that due to the
parallax of 20 seconds of arc. Based upon the normal interpupil-
lary, this means that objects beyond 670 meters are not seen in
stereo relief. As a matter of fact, to most people stereo infinity lies
at about half that distance. Thus for the normal interpupillary
the maximum object distance is 670 meters, and as that is the
distance at which a round object appears flat, it is not a useful dis-
tance. At half that distance, a cylinder has the appearance of
having an oval section, one of whose diameters is half the length
of the other, the side-to-side dimension being the greater.
On the contrary, it is not at all uncommon to find among out-
door people a definite stereopsis two or three times this value,
occasionally more.
Thus, if an object is to be seen in readily appreciable stereo
relief, it must have at least that parallax of an object seen by