This will only work if your objects are close to the look_at point. If
there are objects much closer the viewer will not be able to fuse
the image. Also, is there a reason for the numbers you use? It
produces a separation more than twice the 1:30 ratio, which has
been widely recognized for decades as the normal separation for
humans. The results of that formula may often work but will result
in "stretch" and more people will have difficulty viewing the stereo
pair.

Harold

Phil Clute <pcl### [at] tiacnet> wrote in message
news:387FFA97.8C48E76B@tiac.net...
> This explaination was in regard to red/cyan 3D pair but it seems to
> work for cross-eyed stuff too.
> From my post back on 5/14/99 (I was making sure I had Grimdude's
> explaination right).
> file://---------------------------------------------------------
> I think I did it right, it seemed to work. But in case I
> just got lucky;
> let me make sure I have your directions straight. So it's:
>
> 1. Find the distance between the camera location and look_at.
>    ie. if the look_at is <0,0,0> and my camera location is
>    <0,.5,-4> then the result I want is along z which = 4 .
>
> 2. Divide the result of step 1. by 11.8(12.0 respectively).
>    I get 0.33893
>
> 3. Divide step 2. by 2(the number 2, Not step2).
>    which = 0.1694915
>
> 4. The result of step 3. is added to the x value in the camera
>    location for the right.tga. And subtracted for the left.tga.
>    left.tga = camera{ location<-0.1694915,.5,-4>
>    right.tga= camera{ location<0.1694915,.5,-4>
>
>
> --
> Phil
> ...coffee?...yes please! extra sugar,extra cream...Thank you.

Post a reply to this message