I’m running into a problem in particular rigs. They are using fk chain with subdivided joints, meaning that each controller drives 2 or 3 consecutive joints in chain, until the next controller.
I did a ragdoll setup connected to each bone, but didn’t work at all, the next option was to connect ragdoll markers only to controllers, which worked, but the baking result is unusable.
Is there a solution for this problem, have anyone came across this issue ?
Hi @Jose_Carlos_Poeiras, welcome back.
Fundamentally, this is a problem of resolution. The joints driving that mesh have a higher resolution than your controls.
In an ideal world, you would pass the full resolution to Ragdoll, and extract the full resolution back onto the controls. But in the case of IK or Spline IK or any kind of control that drives more than a single joint, you are effectively downsamping your simulation to fit the limited resolution of your controls.
You can either:
You almost always benefit from simulating at full resolution than the other way around. At least then you have something to compare the recorded results with.
Recommended approach
At no point will you get the full resolution of your simulation though, since the controls are what ultimately drives those joints. But by passing the full resolution into the simulation, you will at least have a high resolution to downsample from.
The most common alternative is to have “extra” controls to drive those middle joints, or an “FK control” option that bypasses the IK controls, that you then simulate and record onto. IK in general is an approximation to aid in the animation of too much resolution. Simulation is also an aid in animating too much resolution, so you typically only want one or the other.
Thanks Marcus,
Yeah , I tried you use a ragdoll to match joints resolution, meaning that I match ragdoll markers to joints, it worked out better, but not great.
The solution was to use a rig with matching controllers to joints.
But thank for your reply