Re: Perspective simulation

New Message Reply Date view Thread view Subject view Author view

Angus Dorbie (dorbie++at++bitch.reading.sgi.com)
Mon, 8 Jan 1996 17:38:27 +0100

I'm not sure why your doing some of the things your are, what do
you mean by dividing the vertices by -z isn't your dividing the
coordinates by -z defeating the perspective correction (the opposite
of what I think your trying to do here)? The shading operation relies
on appropriate projected z information to apply perspective correction
to interpolated texel values.

Humble apologies if I've missunderstood your method, I've never been
masochistic enough to try this.

Rgds,
Angus.

On Jan 8, 5:58pm, Christophe DELEPINE (Tel 8663 ou 8357) wrote:
> Subject: Perspective simulation
>
> This is more a GL question. Here it goes:
>
> I would like to do the perspective computation in software (just as an
> exercise). To do this, I multiply each vertex coordinate by the perspective
> matrix then divide by -z. I send those new x,y,z values through the pipeline
> making sure the projection matrix is loaded with identity. This works fine.
> I get the same image than if I had sent original coordinates with a
> perspective matrix loaded onto the stack.
>
> The problem arises with texture coordinates.
>
> If i assign the same u,v coordinates to the new x,y,z values, the inter-
> polation along the polygon is incorrect.
>
> Is there a trick so that the interpolation takes perspective into account ?
>
>
> I can provide with a simple test case if needed.
>
> thank you
>
>
> Christophe DELEPINE
> delepine++at++tts.thomson.fr
>
>
>-- End of excerpt from Christophe DELEPINE (Tel 8663 ou 8357)

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:52:14 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.