I know how to do it in OpenGL. The problem isn’t doing it, it’s WHY would you do it in screenspace? Doesn’t all culling happen before you draw it?
IF the triangle is in screen space haven’t you already culled it front or back facing and drawn it?
I don’t understand why you would want to cull it in Screen Space and not in the model space before drawing it. If i were building a software engine, I would first cull the visible polygons and then draw them in screen space, so I would never need to determine it’s winding in screen space. Unless I don’t understand something about software rendering.
Sorry this was supposed to be a reply to a thread. Probably hit the wrong damn button when I posted it.
The question was why would you want to determine the facing of a triangle in screen space. I know how to cull in OpenGL. I don’t understand why anyone would ever want to know it’s winding in screen space coordinates or how.
>>>I don’t understand why you would want to cull it in Screen Space and not in the model space before drawing it. If i were building a software engine, I would first cull the visible polygons and then draw them in screen space, so I would never need to determine it’s winding in screen space. Unless I don’t understand something about software rendering.<<<
Model space is not good enough, plus modelview matrix and projection matrices are concatenated on most modern & intelligently designed hardware.
The reason why its not good enough, is because people can do anything they want with the projection matrix.
Culling can be performed in clip coordinates or after with no serious performance loss.