P3D The Third Deformation of 3D, last case and End. (6 of 6)

[abram klooswyk <abram.klooswyk@xxxxxx>]

Another and maybe the most interesting case of the Third or
convergence deformation is the viewing of aerial stereo
pictures, as used for cartography, mostly called vertical
photographs. Many books have been published with
pictures of this kind, some are still available, see
http://stereoscopy.com/bookshop/science.html.
In some countries you can order copies of aerial
stereopictures of specific areas at the topographic survey
authorities (don't try this if you are a spy). I have got
survey stereopictures of places where I have lived.

In viewing these pictures the 2d deformation (distance
effect) is easy to avoid. There remains a marked hyperstereo
effect (lilliputism), but also a marked stretch. Depth, or
rather height differences are greatly exaggerated.
The hyperstereo effect comes of course from the large aerial
stereobase. But in aerial survey stereophotography also an
extreme photographic convergence is used. Aerial base and
flight altitude vary, but a typical example would be a base
of some 2000 to 3000 m, with an altitude of some 5000 m
(1000 m = 0.6 mile), resulting in a photographic convergence
of more than twenty degrees.

Even Boris Starosta wouldn't recommend a base to distance
ratio of 1:2 for stereo portraits :-). For a Stereo Realist
this photographic convergence angle would mean a subject at
7 inch or less, which is of course impossible with that
camera.

Viewing convergence in looking at aerial stereopictures is
often small, and the mathematical, geometrical stretch cannot
be corrected by the visual system when the convergence
difference is so large.
This is the most conspicuous case of stretch due to the
Third Deformation.

Bruce Springsteen:
>To me the word stretch is a very specific "deformation"
>related to angle of view (...)
I hope that I have given evidence that stretch occurs in the
distance (or angle of view, or 2d) deformation as well as in
the convergence or 3d deformation. Therefore stretch is
"just another word" :-), not linked to a single deformation.
------

Why is the Third deformation mostly imperceptible but
sometimes so markedly present?

Note first that we deal with viewing versus photographic
convergence, so there is no direct perception of change,
it is a change with respect to a photography situation where
you may not even have been present, so you compare only with
preconceived ideas of how the shape of things is. You know
about the circular shape of flowers, the height of houses,
the form of human heads, and a few  other shapes.

I believe that the second factor is the fact that the visual
system is relatively insensitive for convergence changes
in the first place, as compared with the remarkable
sensitively for disparity, which is stereopsis.

I have postulated that there is a *threshold* of about three
degrees for the Third Deformation of the Third Dimension to
become visible.
(Classically, the threshold concept means that below it no
perception exists, but above it a larger stimulus gives a
larger effect.)

This threshold of course doesn't explain anything, but it
seems a workable concept.

(The End)

Abram Klooswyk