Verts vs Textures

#1

Here is a simple element that im wondering how to optimize.

Asume I you have a house, and it will be on a plane of grass. Now, I could just make one large plane, and have a texture of grass tile all over it. Or, I could just make a large plane, and cut a hole in the center where the house will be, as this way the engine will not have to render extra textures that will not ever been seen. However, this process will obviously add a few more verts to the model. And I am a bit suspicious that it may take even more texture calculations, despite less texture will be mapped.

So, what’s the answere? I think I should just stick with simple geometry, despite more texture space will be used than needed?

#2

It would really depend what you where using to render the geometry. Mostly though (especially in the case of something like OpenGL or D3D) by cutting the ground polygon up you’re increasing the load on the transform section of the pipeline aswell as increasing general overhead in the texturing across the surfaces. Generally the texture stage will be highly optimised and you will only slow things down by adding more points to the mesh. I suppose it would depend on the size of the occluded section, but I would say just render the whole thing unless you’re seriously fillrate limited (mobile devices maybe?).

If this is a case of a couple of triangles difference in something that’s going to be rendered on a Geforce, Radeon, etc. then you’re not really going to see much of a difference.

It’s really a balancing act with triangle counts. Some techniques (particularly some shading methods) work better with highly tessellated meshes.

#3

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.