>>2829678It can be caused by models having a different spatial relationship (relative positions) between leg IK and foot IK (no real workaround to this, aside from maybe multiplying all key values in the motion to match), foot IK (ankle) angle limits (what I currently use is -50/70, -20/40, -30/20 for the left foot, -50/70, -40/20, -20/30 for right), incorrect leg IK—foot IK relationship (leg IK must target the ankle, foot IK must be parented to leg IK and follow it in the deformation order) foot IK chain makeup (in practice, I cannot recommend non-standard IK foot chains at all. I included the 足つま先1-2 bones in it in early May, only to abandon that in early November because after LATATA I was fed up with how much extra adaptation work they had introduced due to videorelated). I doubt there is a solution to this, however, that doesn't break the ability to rotate the foot by rotating leg IKs, which is used in way too many motions to abandon.
>>2829784Based.
>>2829759This must be the most lightweight stage I've seen.
