Blending scaling between bones

[Stuart Rogers]

A model I'm working on has an IK spine and an FK spine. A 'geometry spine' is to blend between the two according to a pose slider. Blending the position and orientation of the spine bones works fine. However, the IK spine stretches some bones, and I'm at a loss for how to determine the blend bone scaling.

 

For example, to blend the orientation of a bone I use two 'orient like' constraints - one to the FK bone, the other to the IK bone, with the enforcement scaling from 100% to 0% or 0% to 100% accordingly over the slider range.

 

This doesn't work for 'scale like' constraints - it looks like the results are multiplied, so that at each end of the pose range the scaling factor is 0% x 100% = 0%, and that at mid range the scaling factor is 50% x 50% = 25%. (By default, the bones in all three spines are of identical size.)

 

Does anyone have any good ideas how I can blend my bone lengths? (It would be a lot easier if I had access to absolute bone lengths within a pose... (I must experiment with that when I get home - is a property available to an expression even if it's not available as a pointy-clicky item?))

[Steven Cleary]

Just a thought off the top of my head:

 

You could try making a property where you create two separate spines (with one extra bone at the top on each whose origin is located at the tip of the top bone), locating them in the same positions as the geometry bones and constrain one to react to the FK functionality and the other to IK. Then create a percentage pose where one end is FK and the other IK. At the FK end, give all the geometry bones ‘aim at’ constrains with the ‘scale to reach’ option turned on (only in the Z axis) then make the targets the next bone up in the bone sequence of the FK spine (e.g. make geometry Bone1 aim at FK Bone2). The extra bone is there so that the top bone has something to aim at. Then do the same for the IK side of the slider.

 

That might work. Apologies if what I’ve typed is hard to read; I’m knackered and my brain has gone to sleep.

 

Steve.

[Steven Cleary]

Ah no, I've just tried my idea this morning and it still has the same problem and might introduce some new ones.

 

Sorry about that, I'll keep thinking about it today now that I'm more awake.

 

Steve.

[Stuart Rogers]

Ah no, I've just tried my idea this morning and it still has the same problem and might introduce some new ones.

Thanks for the idea. I tried something similar, using blended 'translate to' and 'aim at' constraints, but that wasn't too good - if the FK and IK spines were noticeably different, the bones can end up quite small in mid-transition, as the 'translate to' blending is a linear interpolation between positions, rather than the arc a rotating parent bone follows.

 

I suspect you're suggesting something slightly different, so I'll play with it a bit more. I'm also beginning to suspect that I'm worrying too much - in practice, if a want to blend between FK and IK, I would do it from similar start and end states, so a slight shortening of bones probably wouldn't show.

 

This will have to wait a couple of days, as Real Life is getting in the way at the moment.