3d backface culling

[SBM Productions]

As the subject says.. I'm having trouble with backface culling in 3d space. What I'm currently doing is finding the normal of my quad(using 3 of the 4 points), then using dot product theory, multiplied it with one of the quad's 3d points(ie x1 * normalx + y1 * normaly + z1 * normalz). If the result is less than 0 I reject the quad. I'm guessing a probable issue might be arising because the quads aren't all coplanar. But it completely fails with triangles. I'm reasonably sure the polys are all solid as I'm using the same objects from the nlife demo(it backface culled in 2d, objects appeared correct. For what I'm doing 2d backface culling isn't suited). I would've posted this in the fb forum, but they don't have a n00b forum..

[Anonymous]

nevertheless, more people look there, and, this is hardly a "newb" question like "what command can i use to clear the screen" :lol:

[SBM Productions]

true, but I'm pretty sure its simply solved(at least for the triangles...). by the way, are q2 md2 files solid or not?

[Oz]

well, if you want to do it cheaply, here are some steps
1) Check center of face to see if it's covered by any other faces of the current model.
2) CHeck the center of the object to see if it is covered by something else

Yes, this is cheap, but it'll do a 30% job...
There are more complicated ways...you may want to get in touch with Rel

Oz~