I'm not sure why D3DXComputeTangent gives downright strange results when calculating the tangent and binormal (aka bitangent) basis vectors for each vertex. With a standard bump-mapped mesh with diffuse and specular lighting, and 1 point light source in the direction of the camera, this is what I got with my own ComputeTangent fn: And this is what happens when I use D3DXComputeTangent: Note the incorrect shadowing and specular highlights on the circular bumps with D3D's fn. I was wondering whether anyone else has run into this same problem with D3DXComputeTangent?

This is kind of a wild guess, but it looks like the bumps in the second shot are inverted to being dips.

-Mezz

D3DXComputeTangent calls D3DXComputeTangentFrameEx with arguments that will never fail.

Try D3DXComputeTangentFrameEx() to get better results.

Also the WRAP_U, WRAP_V, WRAP_UV flags are VERY misleading....try it again without those flags.