yes, it’s for optimizing render passes. you can combine in xsi or in unity. in the standard assets folder there are a couple scripts that will do it for you. see this thread:
there are much better threads on the topic but i was too lazy to search thru them, sorry. scroll down to aras’s comment and you’ll have a pretty good overview. couple things to note:
each material means a render pass (or more if it’s bumped etc). combining two objects with different materials gains you nothing.
you can’t exceed the 65k opengl poly count limit. not that you’d want to. the general thought is 4k per object is a good sweet spot. tho iirc, someone ran some tests a while ago and it showed 6-8k being optimal.
combining meshes that are far away from each other might not make sense depending on other factors like lighting and culling. ie something behind the camera that’s combined with an object in front will get rendered. if it has lighting on it the object in front will also be included in the lighting pass even if the light doesn’t reach it.
Yep, really. There are two levels to view frustum culling…object level and polygon level. Unity takes care of the object level, and OpenGL takes care of the polygon level. I assume there’s still some tiny performance penalty with having off-screen polygons (which polygons to ignore has to be calculated somehow), but it’s too small to make any difference that I’ve ever seen.
Have benn using the Unreal 3 engine/ed for the last few weeks. Most the the assets are single items and nothing to combined? is that because it renders different to Unity? or that each asset has a different material? so like mentioned above you gain nowt