Re: SL3D per John B et al

[T3D John W Roberts <roberts@xxxxxxxxxxxxxxxxx>]

>Date: Thu, 29 May 1997 09:53:35 -0500
>From: T3D John Bercovitz <bercov@xxxxxxxxxxx>
>Subject: Re: SL3D per John B et al

>Paul Kline writes:

>> Apart from the shift, is there
>> any other difference between the two square images that are created?

>John R writes:

>> Yes.

>John B writes:

>What?

>John B

If the square is arbitrarily far away and/or small, I suppose you can get the
two images arbitrarily close to identical, but I believe that in a real-world
situation, at least one of those images is not going to be perfectly square.
(That's even *without* tilting the camera.)

I don't have a formula for the degree of keystoning, but I'm not prepared to
say that it would be an unimportant effect.

Come to think of it, I wonder whether the human visual system ever uses
keystoning as a depth cue.

John R


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