[Sidefx-houdini-list] reducing dops simulation time
dogboy_l at yahoo.com
Thu Apr 12 21:03:29 EDT 2007
That's an nxn search which is exhaustive. Try using point clouds, which would
give you a log(n) worst case search. (n is the number of spheres) This would be
pretty easy to code up in vex, I'm not too dops savvy, so I don't know
specifucally how you could hook this into your sim.
--- "Velazquez, Cesar" <Cesar.Velazquez at disney.com> wrote:
> I was wondering if anybody had any useful tips for reducing dops
> simulation times. My dops simulation is actually pretty straight forward.
> I have a bunch on spheres that move around and interact with one
> another. They never stop moving. One optimization I found was the
> "implicit sphere" for the
> sdf representation. Another is "sphere" for the culling method in the
> rbd options. Other than that, I'm kinda stuck. Dops doesn't
> seem to do sdf->sdf collision detection that I know of. So I'm stuck
> each point in the sphere with the sdf of nearby spheres. With the
> potential of thousands of spheres, my simulation times are becoming
> unworkable. Seems like there should be a built-in optimization for
> using spheres since collision detection between spheres is pretty
> thanks for any suggestions.
> Cesar Velazquez Walt Disney Feature Animation
> FX Animator 500 S. Buena Vista Street
> cesar.velazquez at disney.com Mailcode:4870
> 818-460-8797 Burbank, CA 91521-8965
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
God, how much more profitable would it be,
if the Thinker would learn from the Drunk?
Be a PS3 game guru.
Get your game face on with the latest PS3 news and previews at Yahoo! Games.
More information about the Sidefx-houdini-list