fog on wildcat 4210

Im getting fog problems on a Wildcat 4210, meshes that have a lot of small polygons do not fog to the same intensity as that of larger polygons, I orignally thought it was a code problem but I do not get these problems at all on a GeForce and a GeForce2 Ultra.

Any ideas?

You don’t explain your problem very much, more than saying you get different results.

If the fog intensity differs alot from two different parts of your object, a low resolution mesh might look different than a high resolution mesh. If you use per-vertex fogging that is. Can be that the Wildcat performs per-vertex, instead of per-pixel, fogging as default?

Thats pretty much the reasoning that we came to, unfortunatly the Wildcat technical support is very minimal.


We do per-pixel fog calculations on GeForce.

  • Matt

If the Wildcat does per-vertex fogging as default, you might be able to tell it to do per-pixel fogging using glHint(GL_FOG_HINT, GL_NICEST). The driver can choose ignore the hint if it wants to, so you are not guaranteed to get better fog.

Yes, we tried GlHint but that had no effect,
its a bit odd that the Wildcat costs about 5000 dollars and the Geforce2 about 350 dollars and we are continually finding more and more reasons to use the GeForce, such as it has a faster fill rate. The only reason we are staying with the Wildcat at the moment is the excellent superscene anti aliasing, but when you compare the price differences you have to ask yourself whether it is worth it!

Hmm, it’s strange indeed…

(from 3Dlabs site, Wildcat specs:
“Hardware accelerated per-pixel fog with linear, exponential, exponential2, and user-defined fog models”)


Originally posted by David:
The only reason we are staying with the Wildcat at the moment is the excellent superscene anti aliasing

As I understend, such kind of multisampling (with irregular dinamic framebuffer memory allocation) should affect fillrate in a bad way…
What you can tall about it?
(How about some benchmarks?)

btw, more reasons: Wildcat has h/w support for imagine operations and for acc. buffer

[This message has been edited by Serge K (edited 02-02-2001).]