HOME | DD

Description
Not going to any work to make this pretty, because I don't expect very many people to be interested. This is intended for people that are already very comfortable with IK bones and appends.

Division of bones into individual axes is important for direct control over the solutions found by IK bones, for translation of motion or rotation in one axis into a another axis, and for the rigging of muscles. Note that these measurements are not perfect. It's probably unfair of me to say this, since I've never made an IK solver, but MMD's is crap.

Finding the average angle or mid-position between two bones is useful in a few places. I designed this technique in order to address animators applying rotation redundantly to both UB and LB, rather than at the center, groove, or waist. However, finding the midpoint is useful for pseudo-physics, and I have a feeling that you could make a bone that agreed with any particular BDEF2 vertex, basically eliminating the need for BDEF4, permitting the equivalent of BDEF anything, and allowing collision bodies in full agreement with the deformed mesh. Can't find the midpoint yet. This would be useful though.

If you use IK solutions to drive other IK systems, you need to increment deformation tier on each additional IK layer, or MMD won't work right. PMXE will, but MMD won't. (Children, including append-children, also need to have a deformation tier at least as high as their parents.)

Edit: I was wrong. Half the angle is part of the solution-- the midpoint lies along that line. But the midpoint between two bones is not the endpoint of that line. Changed the picture because I was embarrassed enough about my mistake.

Description
Not going to any work to make this pretty, because I don't expect very many people to be interested. This is intended for people that are already very comfortable with IK bones and appends.

Division of bones into individual axes is important for direct control over the solutions found by IK bones, for translation of motion or rotation in one axis into a another axis, and for the rigging of muscles. Note that these measurements are not perfect. It's probably unfair of me to say this, since I've never made an IK solver, but MMD's is crap.

Finding the average angle or mid-position between two bones is useful in a few places. I designed this technique in order to address animators applying rotation redundantly to both UB and LB, rather than at the center, groove, or waist. However, finding the midpoint is useful for pseudo-physics, and I have a feeling that you could make a bone that agreed with any particular BDEF2 vertex, basically eliminating the need for BDEF4, permitting the equivalent of BDEF anything, and allowing collision bodies in full agreement with the deformed mesh. Can't find the midpoint yet. This would be useful though.

If you use IK solutions to drive other IK systems, you need to increment deformation tier on each additional IK layer, or MMD won't work right. PMXE will, but MMD won't. (Children, including append-children, also need to have a deformation tier at least as high as their parents.)

Edit: I was wrong. Half the angle is part of the solution-- the midpoint lies along that line. But the midpoint between two bones is not the endpoint of that line. Changed the picture because I was embarrassed enough about my mistake.