Small object jitter: approaches summary?

User avatar
Erwin Coumans
Site Admin
Posts: 4221
Joined: Sun Jun 26, 2005 6:43 pm
Location: California, USA

Re: Small object jitter: approaches summary?

Post by Erwin Coumans »

Only polyhedral convex shapes are affected by 'setPlaneConvexMultipointIterations', see the check in the code. I should add the same check in for the convex-convex case.

So spheres/cylinders are not affected. If cylinders/capsules are automatically rolling, it is best to include angular damping until they automatically settle down.

Thanks,
Erwin
User avatar
ejtttje
Posts: 96
Joined: Mon Nov 03, 2008 9:57 pm

Re: Small object jitter: approaches summary?

Post by ejtttje »

Thanks for all the help so far, I've started integrating what I've learned into my main code.

I know you probably want to be done with this topic, but there's still a couple loose ends, bear with me a little longer :)

In the attached BasicDemo, I've applied a scaling factor on dimensions, positioning, gravity, camera, and collision margins. The objects are stable when SCALING==1, but as we increase SCALING to 100, objects become impaled in the ground (box or plane) and jitter around. By 'conventional wisdom', increasing the scaling should have made things more stable since this puts all the dimensions closer to 1... so I guess there's something wrong with the high gravity per simulation step time.

What's surprising is that this issue is not restricted to the gravity acceleration per object dimensions, since we're scaling them all together. Any thoughts on what's going on? Maybe there's another threshold parameter that I should be scaling with gravity? Otherwise, the result of this would be to advise people to keep gravity ~10 instead of scaling the world to keep object dimensions ~1...

Thanks again!
You do not have the required permissions to view the files attached to this post.