koobas.hobune.stream
Should identical meshes in a scene have independent UVs for lightmaps? Should identical meshes be grouped into one mesh?
To better explain: I've got a scene with several identical mesh pieces reused, such as a chunk of a fence which is duplicated many times within my scene. From an optimization perspective, is it better to group each of these chunks together within my 3D modeling tool as one mesh, or should I keep them separate?
If I'm working with dynamic lighting, is it more efficient for each of these meshes to be identical, or would it be better to join all of these meshes (and give each one of them a unique UV map)? Are there any advantages or disadvantages to these techniques?
Please provide any information on the subject. I'm unsure of how these techniques differ on object culling or how lightmap information is baked, or if the difference between the two is substantial.
If the meshes are going to be static, grouping them together is a solid optimisation strategy, since you're only making one draw call rather than possibly multiple. Instancing should reduce the need to do this if they're identical though.
If they're spread out, keeping them separate is likely better, because only what you're seeing will be rendered if you set up occlusion culling.
