90s-early 2000s: smart white guys and asianslate 2000s-2020: poor people show up2020-presetnt: smart white guys and asianswe're going home bros
>>978177Now we just need smart white guys and asians to take over Hip-Hop.
>>978177are you talking about the cg industry? because it's full of indians.
So, somebody in another thread wanted to see mocap with shoulder movement. I was thinking about starting a new thread but this one will do. Thanks Cris for making it.
With the armature visible.
This >>980058 >>980060 is Microsoft Rocketbox, which you can get from here:https://github.com/microsoft/Microsoft-RocketboxJust in case you want to make something, have fun and not have to listen to some self-styled "artist" for a while.
This is another action. But the problem is the same as with the other set. Due to bugs and missing features in the NLA editor, it's (almost) impossible to re-target, combine and modify the actions effectively.I'm glad the Blender developers may be doing something about it in 2025 after they're done with Grease Pencil which is obviously a priority.
>>980058what's the point? I always delete the shoulder movement and when I'm done re-animating hands I constrain the shoulders to elbow (elbows and hands are on IK) with aim at about 40%. what you posted looks terrible
>>980068>what you posted looks terribleCan you show me something good of your own? Don't answer. I know the answer.
>>980069why make the post then, you fucking retard. what's even the point of your other posts ITT? you wanted to discuss mocap techniques or just shill your github project?
>>980071>shill your github projectIf you don't want to visit the link, at least read it. What does it say? You know there are those letters things in there.
>>980072okay an avatar library..cool. then what's up with the mocap? you want to discuss it or what
>>980077The main issue is that the library provides building blocks (in term of animations) but it's really hard to use those in Blender. So, if you have concrete ideas or techniques to do so, that would he interesting.
>>980078I'm not familiar with freetard software but as I mentioned previously, I don't use shoulders from mocap at all most of the time because I reanimate the arms anyway and then the original movement doesn't correspond (I reanimate with IKs in a kinda retarded way but it's much faster than classic workflow)
>>980159I'm curious to try to do that but I don't think it's going to work because Blender's IK system relies on joints being a little pre-bent in the right direction and I don't think the CMU skeleton has that feature.
>>980159I'm surprised. It kinda works. However not with the default solver. The itasc solver has an option to copy the rotation of the target instead of using a pole and that seems to work. It should also allow to set rotation limits, stiffness...
For years I've tried to understand what the iTaSC solver thing purpose was. Because the documentation is cryptic and it works kinda like a cut down version of a rigid body simulator.https://docs.blender.org/manual/en/2.79/rigging/armatures/posing/bone_constraints/inverse_kinematics/introduction.html#rigging-armatures-posing-bone-constraints-ik-model-itascTurns out it's for advanced re-targeting. When it's enabled, the "Copy Pose" mode inside the IK constraint can be used instead of the Copy Transform constraint. It's outside the scope of what I'm trying to do now but I'll get back to it
>>980192yeah, the trick is that 80% of the important motion comes from the ankles, chest and wrists (all on IK).. you can just grab the poses you need (from the original mocap) and delete all other keys and you will still get sort of cleaned up original motion that you can restore via constrains and playing with IK pulls. of course you need to know what you are doing as an animator. sometimes reanimating things yourself is faster than trying to clean up shitty mocap
