Gun as example. Ik it’s logical to just unwrap small meshes as one, but with a bit more complex meshes, it is better to unwrap them as one anyway, then incerase resolution if needed so the texture would still look decent, or separate on few different materials ? Would mesh like that function as game asset, or it’s prefered to be single texture for optimisation reasons or something ?
As far as I know, for game files you want to unwrap as one object, and keep it as low poly as you can. This is for optimisation as you say, but I also don’t think game engines particularly like multiple UV’s for 1 object. I could be wrong on that one though.
If this was for something else though, like an animation, or just a still render, then you could make as many materials and UV’s as you like.
It all depends on it’s purpose honestly…
Yes. fewer materials -> fewer drawcals = better performance
It is combined even further and it’s common to have multiple separate objects on single texture atlas