P3D Re: MagicEye Question Magic EyeQuestion

[Larry Berlin <lberlin@xxxxxxxxx>]

>Date: Sun, 18 Apr 1999
>From: Gabriel Jacob <jacob@xxxxxxxxxxxxxx>
>...............
>Okay, I know the magic-eye fad has faded, but today
>it struck me! I've never seen any magic-eye image
>protrude the surface. If there weren't any, was it
>because of some technical problem, or some strict
>adherence to some rule?
>


The rules of Geometry can be strict, I suppose. It has to do with the stereo
window and the geometry of freeviewing. In ANY parallel viewable mode, the
image is behind the stereo window. In ANY crossed viewable mode, the image
is in FRONT of the window. The window itself contains the 2D pattern.

What happens in the viewing zones close to the stereo window, is that the
eye locks onto the 2D information which is printed AT the window. This lock
is hard to break and results in the effect that one has to *jump* into the
image a ways to fuse the repetition of patterns in stereo. One loses the
stereo effect when you attempt to fuse too close to the window. 

As soon as you start to look *in front of the stereo window*, by definition
your viewing method has changed from parallel to crossed. ANY point in front
of the image requires crossed mode to see. These factors mean you can't have
an image floating at the window level, partially within each viewing space.

>...............(ron labbe) commented.............................
>Actually, it's a trick of perception:  "Magic Eye" images build from the
>back forward, not from the front to back... basically infinity is at the
>"window"! It's the ambiguity of the window, because of the "vistigal" images
>on either side that the window appears above the page!
>


*****  It's true that a typical stereogram is built in front of a background
formed by the tiled (repeating) pattern, but it's *not* true that infinity
is at the window! The background of a stereogram is usually considered to
define infinity for each image.  Infinity normally is defined by precisely
parallel lines extending outwards from each of the viewer's eyes, or the
average interocular distance. A parallel mode stereogram MUST have a maximum
pattern repetition distance that is equal to or smaller than the average
interocular because anything larger and most people wouldn't be able to view
it. Crossed mode stereograms can have large or small repetition rates.
Infinity for crossed viewing modes is harder to define, but would likely be
defined within each image and would be dictated by relativistic factors
relating to the average repetition of the chosen pattern.

The total depth for any scene is the difference between the maximum and
minimum pattern repetition rates. Due to perception factors, too great a
spread between max and min becomes hard to view. The limits for a single
image to be viewable all at once, are usually defined where the smallest
pattern would be > 1/2 of the largest pattern, or the largest is < twice the
distance of the smallest pattern. If your large pattern exceeds twice the
width of the smallest, you place the eyes where the smaller pattern is
doubled up offering a fused view other than the intended one, or I sometimes
think of it as going to the next octave... a realm of unique special effects.

Larry Berlin

Email: lberlin@xxxxxxxxx
http://3dzine.simplenet.com/


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