[stevek]

My experience in the past with collision objects, the more detail the objects the more accurate the collision will look- but that also increases the render time. So the more you want to cut back for rendering sake- the less realistic motion effects will look. (When I say detail, I mean segments and such- The computer will have more trouble calculating accurate collisions in a theoretical surface (ie the area between points). The accuracy comes when a vertex collides. I don't yet know the object creation parameters for the new Hair functions- but this is what its been like in the past doing - hair, toothbrush, grass, etc.

Net rendering is the way to go. If you have access to multiple computers, go for it. Of course it will render one frame per CPU! but that means it will simultaneously render that many frames as opposed to one at a time on one computer. When we first setup net rendering- it was taking us a few days to render our test animations- then we were able to run full resolution finals of our entire animation in one night (that included working late and starting early the next morning.) So we were able to run the entire thing in several hours instead of days.