In my racing game is it better for the track to be one giant mesh or lots of different road pieces?
What I’m assuming is that using the bounds of each sub piece it can decide which road piece to draw and then it wouldn’t have to draw the whole track each time.
On the other hand maybe it’s quicker if the entire track is just one mesh even though it will be about 10x bigger?
I’m having a bit of trouble with physics on the road where the car passes from one track piece to the next it kind of hits the end of it.
What do you think? Combine all the different road pieces into a giant mesh?
It’s not great if it’s one mesh, because then the entire thing is being processed every frame even though most of it wouldn’t be visible. The visible meshes don’t need to be related to the mesh you use for physics.
Yeah, thanks that’s a good point! So I could create a giant non-visible mesh just for physics.
Thanks! I might give that a go!
(Although… using bounding boxes for physics would be quicker too especially if the physics is not GPU accelerated. So I guess I’ll just have to do some tests. But I guess that might solve the smoothness issue of the road sections.)
Also, I wondered whether there would be any advantage using lots of submeshes in one object vs. lots of gameobjects with single meshes.