djwarder Posted October 8, 2012 Share Posted October 8, 2012 Hey, just been looking into simple sphere to sphere collision & response, and wondered if this is possible to do in VOPs. There's all the pointcloud access which would be fast and the algorithm doesn't look too bad - is this possible & has anyone tried it at all? ps. Sample C++ code ... http://studiofreya.c...ision-response/ Quote Link to comment Share on other sites More sharing options...
anupamd Posted October 8, 2012 Share Posted October 8, 2012 you could just do simple sphere collision in bullet.. its probably about the same speed as if you did it in POPs.,.. Im guessing.. not that I actually tried.. but I probably should Quote Link to comment Share on other sites More sharing options...
ikarus Posted October 9, 2012 Share Posted October 9, 2012 That solution really only works when you have two objects colliding. In a complex collision you could have hundreds or thousands of collisions which need to be resolved. The Gas Particle Integrate dop does this for particles and collision geometry. But if you want a sphere based collision, your best bet would probably be the bullet solver. Quote Link to comment Share on other sites More sharing options...
djwarder Posted October 9, 2012 Author Share Posted October 9, 2012 Hmm ok, I'll have a look the GasParticleIntegrate microsolver, cheers for the heads-up. Bullet could be an option, but I need to have ALOT of spheres colliding and Bullet in Houdini seems to use tons of memory and is not as quick as I remember in being in Maya. Might have to look into if there's any settings on the solver I can change to speed things up ... Quote Link to comment Share on other sites More sharing options...
cubiccube Posted October 10, 2012 Share Posted October 10, 2012 I'm pretty sure this can technically be done using point clouds and VOPs but Bullet might end up saving you more time in the end. You can use PC Open's search radius to represent the sphere, max points = 2, then use Num Found/Import by Index and a For loop to access and transfer velocity/position info between the two points. As Ikarus mentioned, you're only ever calculating collisions between two points/spheres at a time but you're doing for every point. Not a very robust system but it could maybe work for debris in the background or something similar that doesn't necessarily call for a great deal of accuracy. Quote Link to comment Share on other sites More sharing options...
papicrunch Posted October 10, 2012 Share Posted October 10, 2012 you can try to use some technic expose on cmivfx flock system, especialy "separation behaviour" look here is an interesting paper. have fun, thomas 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.