What is your experience on cleaning up the opticaldata on the optical nodes versus after it is plottet to the rig? Of course swaping and missing markers must be fixed on the actor, but I’m thinking about noisy and jittery curves and stuff like that.

It looks easyer to do it on the opticals because you dont have to wory about rotation. But if I do it on the rig; I can do the cleanup after several takes has been cut into the final shot. So I wont have to cleanup on more than what is actualy going to be used.

I have thought about clipping the takes together directly on the opticals. This will work sometimes, but the problem is if I have several actors on the same shot on different takes :|

What is your experience on this issue?

For me it is faster to clean the final move than all the data leading in to it, after I have cropped and clipped it for what I need.

I worked in a system for a long time that all we had was skeleton data to clean so I still prefer it that way.  If you want to clean the data and jitters on the skeleton , set the reach to the 100% for the IK ctrls and then butterworth filter and edit them instead of letting the FK skeleton jitters magnify down the chain.

I think both have their place but I favor doing it in the edit pass on the rig.

Brad Clark
Co-Founder: Rigging Dojo-Teaching the art and science of character rigging
Author: Inspired 3D Advanced Rigging
The Character Animator Toolkit for Motionbuilder :available now: