Can anybody give me some quick answers to the following questions…
If I crate a wall mesh in a 3D app (with UVs, colour and normal maps) and then use that multiple times in a scene inside UE4 (for creating 1 long wall for example…
Does this save resources such as memory because the engine only has to load 1 mesh which is then repeated?
Can I still bake lighting in to the scene or will the shared UV space be an issue?
An an the texture front what are the standard set of maps you would generally use on an environment mesh? I assume colour, normals, specular but I see a lot of talk about AO maps now.
1–it won’t really save memory because the full wall could be made using the same number of polygons as each section, however, if you have to use a lot of unique walls then it would use more memory that way than to use one mesh for everything.
But–this will not save a noticeable amount of memory, since the wall will be so low-poly it’s not worth the trouble. The main issue with this method is that each object increases the number of draw calls which will have a significant impact to performance, so for simple meshes it’s better to have more unique meshes than to use several meshes to build a wall.
2–each object gets its own lightmap so using a mesh multiple times will have multiple lightmaps. However, each object also gets processed individually for lighting, so if you have a flat surface that is made of multiple meshes then each one will have slight differences in the lighting which will look bad, it’s better to try and have a single mesh for a surface as much as possible. Sometimes it’s unavoidable because you might not be able to get enough detail in your lighting in one mesh even at the highest lightmap resolution and in that case then you need to split it up.