Supermodel Forum

Ian wrote:It might be a bug, make a save state and i'll try it out

here
https://drive.google.com/file/d/1Szl8c5 ... sp=sharing

yeah you are right it's broken
I think i will just give up and do it completely in the fragment shader

Right I fixed it ..
Can't believe it has taken so many attempts to get this right. The problem was caused by the fact the clamping to the range 0-1 must be done before fogIntensity is multiplied by the final value. Fog is supposed to be a simple per vertex attribute interpolated across the poly. But it turns out it's almost impossible to do it in the vertex shader because it's run before clipping happens. The model 3 must be calculating some of the attribs after clipping has been done. This can't be emulated directly in opengl, as the vertex shader always runs directly before clipping.

Ian
The specular new patch in November is not yet applied?

Not yet. Need some good comparisons to show the difference. The numbers Harry chose seem quite arbitrary. I would expect the coefficients to be something like power of two. Of course the ones Harry's chose could well be spot on, but without access to the specs it's very hard to reverse engineer.

Ian
R696
Virtual On
bug found

R685 screenshot (Normal screen)


R696 screenshot

Yoyo
can you try with svn 699 ?

Ian wrote:Yoyo
can you try with svn 699 ?

R699 normal
thank you Ian

Delete

This is off topic for the dev forum which is intended for development discussion, not feature requests.