Mailinglist Archives:
Infrared
Panorama
Photo-3D
Tech-3D
Sell-3D
MF3D

Notice
This mailinglist archive is frozen since May 2001, i.e. it will stay online but will not be updated.
<-- Date Index --> <-- Thread Index --> [Author Index]

P3D Re: Scaling effects of hyperstereos


  • From: Bruce Springsteen <bsspringsteen@xxxxxxxxx>
  • Subject: P3D Re: Scaling effects of hyperstereos
  • Date: Mon, 11 Oct 1999 22:42:19 -0600

Dr. T has given a homework assignment:
> I wonder if Bruce (or anyone else) can give an explanation for the
> scaling effects of hyperstereos.  It has been said many times that
> increasing the stereo base makes the image look proportionally smaller
> and closer to the observer.  That does not seem intuitive since the
> size of the image remains constant.  How can you explain and 
> demonstrate this to a beginner?

Yes, George is right as usual.  Let's answer a related question instead: 
How can objects that seem the same size and distance when viewed with one
eye appear very different when examined in stereo?  Well, most people can
recognize that to ONE ground-level eye, a six foot tall man standing ten
feet away can appear to be the same size and distance as a 60 foot man
standing 100 feet away (ignoring different focusing).  But opening BOTH
eyes will break that illusion for two reasons:

1) The overall convergence changes needed when fixing your vision first on
the near man, then on the far man.

2) The different amounts of depth/form visible *within* the figure of each
man.  You can see "around" more of Mr. Small-and-near than you can of Mr.
Big-and-far, gathering more parallax information about his shape in the
stereo-rich near range. (Earlier this year I tried to make a nice
intuitive parable about this second effect which I called "Hugging the
Elephant" - try that phrase in searching the archives, or I'll post it
again if someone twists my arm.)

That stereo-induced difference in apparent size and distance of objects
that looked identical in monocular view is well illustrated in - of all
places - George T's NEW Realist book!!! ;-)  BUT in this case, the
differences in size and distance are illusions created by the
stereographer, using base changes!  On page 75, in the section on
size/space control in trick stereos, George has reproduced a stereo pair
by Ted Lambert.  Viewed without stereo, three toy-model cowboys standing
in a horizontal row appear to be the same size and distance.  Viewed in
stereo the image is found to be a composite of three shots of the same
figure, always at the same distance, but varying the stereo base between
each shot. 
When this composited scene is viewed in stereo, the three identical
cowboys suddenly appear to be at different distances, in three different
sizes, shrinking from the far cowboy to the near one, in proportion to the
increases in base.  If each cowboy had been left as a separate stereo
pair, the effect would still be present, but the helpful side-by-side
comparison would be missing, so we would probably be less aware of it.
 
My previous "boot camp drill" posting (which see) is an invitation for all
you clever folks to help me think of a more complete "standard
demonstration" of sample stereo pairs, to illustrate the difference
between base/scale (parallax) changes and focal length/viewing angle
(perspective) effects in a clear and memorable way.  I'm thinking about it
- you do too!

Bruce. 




__________________________________________________
Do You Yahoo!?
Bid and sell for free at http://auctions.yahoo.com