Hey guys,
I was wondering if it’s a correct modelling technique to combine multiple objects without connecting them geometrically. For example, if I were to make a simple button, would it be permissible/correct for me to just place a cylinder on the surface but not connect it?
There is no problem in modelling objects like that.
Games have been doing that a lot, and when you see a landscape scene, all those trees, vegetations, house etc… are usually not part of the same geometry, but different objects sticking together.
And in the end when you have a complex model, it can save a lot of triangles in the face count, by example this very simple object if everything is the same geometry :
And if it’s not same geometry :
18 triangles saved, but that’s a very simple object, now let’s imagine there are much more buttons/etc… on that object, there are a lot of object similar in the scene, + the land + every trees, rocks, houses etc… and you can see in the end how a scene can benefit in polycount from not everything being the same geometry
After that there are the texture/normal/bump map that can be used to make believe something is complex while it has very few poly :
Yes, if it looks right, it’s fine. But to add to Sanctuary’s reply, using intersecting geometry like that makes UV unwrapping easier too. It doesn’t need inside geometry so that part is always going to fold flat easier. If you’re going to unwrap it, don’t push it too much inside the other mesh because everything inside is also wasted texture space.
