Specular problem after lightbake
Hi all,
I keep having this issue:
Top image is before lightbake and bottom is after. There are 2 different materials on this mesh with different unwraps, but even if I change it to 1 material the problem persists. If I planar map the entire mesh in 3dsmax the problem goes away, but I guess I should be able to move/rotate/scale uv's without UDK messing up the reflection/spec. Think I tried everything I could come up with, re-exporting the mesh and make make sure all the normals are fine. Change textures (if I apply a simple gray material the problem still occurs) so I really hope someone can help me out here.
In exchange I can send you a pic of Peris' favorite mug. It has dino's on it.
I keep having this issue:
Top image is before lightbake and bottom is after. There are 2 different materials on this mesh with different unwraps, but even if I change it to 1 material the problem persists. If I planar map the entire mesh in 3dsmax the problem goes away, but I guess I should be able to move/rotate/scale uv's without UDK messing up the reflection/spec. Think I tried everything I could come up with, re-exporting the mesh and make make sure all the normals are fine. Change textures (if I apply a simple gray material the problem still occurs) so I really hope someone can help me out here.
In exchange I can send you a pic of Peris' favorite mug. It has dino's on it.
Replies
I have a few ideas that would be easier to look at rather than type out all the possible problems...
dons@envartist.com if you're interested in the assist
this is what i came up with... basically to solve this you'll need to re-uvwmap your object and not rotate any of your UVW clusters unless you rotate the lightmap uvw coordinates to be oriented the same way. You'll still have some lighting artifacts from the bleeding of pixels...
what happens is you create a break in the tangent when you do that because the vertex normal and the uvw normal are no longer oriented the same relative to the lightmap uvw coordinates...
I'm pretty sure if you broke the uvw coordinates of the lightmap in the same way and oriented them in the same way, they would be fine as well.
that probably doesn't sound right...
anyhow, the image will explain it better than my typing...
To solve problems like this, i generally try to find a case that works... then try to break it again. This gives me a lot of info about how to break it, then I can start messin around with it to get the thing to not break but give me the results i want...
The image is a slide show of that process.
Hope this helps brohammer