Re: 3-D phenomenon

[P3D Gregory J. Wageman <gjw@xxxxxxxxxx>]

Allan W. writes:

>Another important "autonomic" clue our brains use is the degree of
>divergence or convergence our eyes use to shift near/distant viewing.

I'd say, based on personal observation, that it is the primary cue
for a stationary observer and stationary scene.  And given that we're
talking about still photography of typically stationary subjects,
that explains why the Realist format works so well.

>So, we're looking at a projected stereo Realist slide where everything
>is as sharp as we like it and it all sort of looks right.  We can let
>out eyes wander over the image and no matter where we look, it appears
>to be "in focus" - BUT we are doing a lot of compensating because
>something - LOTS of things aren't quite right.  For one thing our
>eyes are always converged on THE SURFACE OF THE SCREEN where the virtual
>image is.  That doesn't happen when you just look at stuff.

That's not right.  If our eyes were always converged at one place there
would be no perception of depth.  There is still binocular disparity
between the two projected images, even if each eye is only seeing one
of the two images (just as in a viewer).  All one has to do is take off the
polarizing glasses to see that objects at different distances are
separated by varying amounts.  When we change our viewpoint of interest
in the projected scene, our eyes converge or diverge appropriately.  To
prove this to yourself, project a slide with something in front of the
window.  A real object placed at the apparent position of the virtual
object will be fused.  If you were converged on the screen, it would
not be.

>And if you move your head (carrier pilots maneouvre) perspective
>and parallax don't work either.

THAT is true, and it is a very spooky effect.  The image appears to
"point" at you, following your movement.  Very unlike a hologram.
Moving one's head breaks the illusion.

>So, convergence of eyeballs, out of focus periphery are more things
>which are part of seeing "normally."  In fact peripheral vision has
>major visual clues we don't think we notice.  Remember Cinerama? Or
>that round "See America" thing they used to have at Disneyland?  Or
>the 2-D OmniDome, MaxiDome things?

I'd never argue that the Realist format provides any kind of "immersive"
effect, which I think it's generally agreed requires involvement of the
peripheral vision and something approaching a 180 degree view.

But I think you misunderstand just how well the Realist format reproduces
a lot of the necessary cues.  If it didn't it wouldn't be as effective
as it is.

	-Greg


------------------------------