Is it better for Hard Edges to keep the Lightmap UVs ongoing or to seperate them?
In this test-scene is a stationary Directional Light (I wanna have shadowmaps) and 2 static cubes.
A cylinder is casting shadows on the cubes. I don’t use Autogenerated Lightmap UVs.
One got an ongoing UV-Layout for the important front parts:
Pumping up the lightmap resolution just makes the "errors"smaller…
I read many stuff, like of course the documentation, and [this][6] here…
But i dont understand whats happening
Especially the offset in the ongoing UV’s example is not plausible for me…
Interestingly, with changing the lightmap resolution of the ongoing UV cube to 132 (only +/- 4 steps are possible) the bleeding disappears, but the spreed and offset getting worse…
I think the errors you are having are partially caused by UV’s. The mesh itself looks like a cube with equal edge lengths but not the UV’s. I’m guessing you scaled the mesh but didn’t remap UV’s afterwards. Also you should use power of 2 dimensions for lightmap(and textures too) 4, 16, 32, 128, 1024, etc.
Hi Jacky.
I think unlike with texture UV’s it’s should not be a problem to have non-uniformly scaled UV shells. As stated in the last section [here][1]
And the example in the documentation shows it also:
And yeah I was surprised, that its even possible to type in another values then power of 2…
I was just curious caused by those two threads:
[1 The MOST important thing about lightmaps !][3]
[2 Lightmap Density not matching UVs from Maya][4]
But I must say, that I’m uncertain about how important “UV-Grid Snapping” still is… Those threads are 4 years old…