|
I have read various articles on performance which mostly
deal with things like texture by reference etc.. etc. what I would like to know is whether performance is
hampered *much* by using many (eg 200 transform
groups to position objects in a scene) rather than defining the entire geometry
in one large QuadArray. I’m also considering
using 200 difference instances of appearance for each of those is this bad? Should
I just reduce spec and stick with just a few less appearances! Also will performance improve or decrease if I partition
these into sections using BranchGroups? Any comments welcome. ~~~~~~~~~~~~~~~~~~~~~~~~ |
