So I’ve been reading up on how UE4 deals with lighting and it appears modular results in less problems.
I’m a person who tries to keep everything modeled as one object, but now I’m wondering if that’s the right approach.
For example, I have a bathroom scene I want to build in UE4. Even though it’s simple, would it be better to construct the walls as separable planes instead of just having everything inside of a cube with the shading model set to unlit or forcing double sided?
In that example, it would be better to have the walls as a single unique object. Modular is a way to reuse assets which lowers the amount of memory that the game uses which improves performance. However, another thing that controls performance are draw calls, which increase with the number of items. In the case of walls, they’re usually very low poly so the added memory use isn’t a big deal. There’s a limit though, since each object can only have a single lightmap if there is too much surface area then you won’t get enough detail in your lightmap to get the results you want, in those cases you’d need to break it apart for that reason.
Lighting can actually be an issue for modular assets, because lightmap smoothing doesn’t go across meshes it can make the splits in meshes visible due to how the meshes are handled separately in lighting. So you wouldn’t want a split in the middle of a wall or the floor.
Yeah, lightmapping works better with one unified mesh.
But then what happens if I want DFAO/ray trace shadows? Apparently, this is where modular is more supported.