I have a character with legs that are not parented to a torso bone (they end at the top of the thigh - each thigh is a highest level parent). Other than that, I rigged the legs normally.

I wanted my character to be able to retract his legs upwards into his body. To do this, I scale each thigh bone down in pose mode.

This works for each leg, but each foot remains the same size, because each foot bone is parented to its heel IK bone.

To solve this, I tried setting each thigh bone as a parent of its respective heel IK bone. This worked fine in the left leg. However, in the right leg, this causes the bones to move wildly all about in pose mode.

To see what was wrong, I tried separating the right leg with all its bones into a different armature. When I did this, the right leg behaved normally.

My paint weights are fine, and influence the mesh properly; this problem is only with the armature. Both legs appear to have identical bone hierarchies, and identical constraints.

I've been able to solve my problem by setting a bone constraint to match each heel IK's scale to its thigh bone's scale. However, I'd still like to learn why parenting the bones the way I did caused the strange behavior, and I'd really appreciate help. Thanks!

I can post more information or screenshots if that would help.

Edit: https://drive.google.com/open?id=1-bWzTfC_T1BdgUn1wnHGU0S8ra1XRreM

  • $\begingroup$ If the left leg is OK while the right leg has the problem, probably there's a little hidden difference, so I think you should upload your file for inspecting. $\endgroup$ Oct 14, 2019 at 19:36
  • $\begingroup$ I think you're right. Thanks, I uploaded the project. $\endgroup$ Oct 14, 2019 at 19:57

1 Answer 1


To have an IK controller which is child of a bone which is part of the IK chain must be considered a rigging error, because it creates a circular dedendency, leading to unpredictable behaviours (and jitterings).

Instead you can create a new bone (which I called "scaler") that will be parent of the thigh and the HeelIK. In my picture scaling the scaler bone caused the whole left leg to scale proportionally, maintaining full functionality of the IK leg chain, without any need of constraints.

enter image description here

  • $\begingroup$ Awesome! Thanks for the answer. Any idea why it worked for the left leg? I got the sense that there was something invalid about the chain of operations I was performing, but I couldn't tell for sure because I only experienced problems in one leg. $\endgroup$ Oct 16, 2019 at 5:10
  • 1
    $\begingroup$ Circular dependencies are impredictables. On my computer your file had jittering in both legs. $\endgroup$ Oct 16, 2019 at 5:51
  • $\begingroup$ Gotcha! That makes the most sense. I was suspecting that the left leg might actually be the anomaly. Thank you! $\endgroup$ Oct 16, 2019 at 17:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .