I can’t find any solid information on setting up proper UVs for lightmaps in blender. Some information for Maya says to take the lightmap size, divide it by 2, set the grid snap to that number, blah blah blah
How is this set up in blender? I know you can snap to pixels on a background image but what size does this image have to be? Would I just create a 64 x 64 imagine that is alternating white and black pixels and set that to the snap? I tried using blender’s lightmap project with not so good results. I’m not sure why this information is so hard to find as it seems like a very vital part of making models.
That technique is a way to get nice lightmaps by lining up the UV’s to match where the pixels in the lightmap will be. However, since pixels are square shaped the technique only works with box shaped meshes, so if your mesh isn’t box shaped then you can’t really do that. Because of that, I don’t really recommend going through the trouble of lining up the mesh to the pixels since there’s going to be a lot of cases where it can’t line up anyways.
So what do you suggest as far as making the UV? I did a test and made a really basic room, one single mesh, created my light map UVs. I bring it into unreal and I have seam issues (see attached picture).
Make sure your lightmap resolution is high enough, the default of 64 would not be high enough for that, and for your lightmap UV’s make sure that you try to have as few UV islands as possible, especially avoid small islands among large ones because you’d have to increase the lightmap UV’s to be high enough for detail on the small parts. You can use a lower lightmap for a mesh if there’s fewer lightmap seams.