the principle of XGen is that you setup the scene description in maya - expressions, ptex property maps, etc - and then the instancing/interpolation/etc engine will run uring rendering, in the xgen geometry shader. So it has the upper hand with very large data sets because the data is not calculated unless needed, and will never be created on the DCC side. It's about having a system that can scale to milions of hair/instances, and there are some definite tradeoff to that - xgen's engine is simpler than ICE.
On Thu, Aug 8, 2013 at 9:01 PM, Ahmidou Lyazidi <[email protected]> wrote: > > So finally, this is how it look like: > http://www.youtube.com/watch?v=uPZvG5H8MoY > > The funny thing is it's not that far from what we can do in ICE, actually I > made a set of compound > that can do pretty much the same thing: > https://vimeo.com/19323411 > > It's just missing the vector paint tool(on my TODO list), a better viewport > integration, and a nice > caned UI, which is problematic with ICE.
