optimization

Unity Engine
1

I plan to use mesh colliders for all of my bone animated models.

This means I can’t use mesh colliders for anything else that I plan to smash them into or bounce them off of, right?

does it make sense to have different geometry for colliders than I do for rendered meshes? (making a building out of individual box colliders and the render mesh out of a few objects (outside/inside) for instance.)

I know that this is an old trick. What I’m asking is will I take an appreciable performance hit by doing it this way?

Otherwise, I’d end up using mesh colliders for the buildings (so people can go inside) and (FAR less accurate) box, capsule or ball collision methods for my animated models (player, enemies…)

I know that there are as many ways to do this as there are people doing it. I have just never considered the performance hit until now and I want to plan carefully.

Any thoughts?
(I’m at work or else I’d probably be testing this and posting my findings).

2

Mesh colliders can collide with mesh colliders as long as at least one of them is a convex mesh collider (which has a limit of 255 polygons). Primitive colliders (boxes etc.) are faster than mesh colliders, so use them when you can. Yes, it makes sense to have a simple collider and a more complex mesh that’s actually rendered.

–Eric

3

that’s good to hear!

thanks!