If IK handle is under a transform node , moving the last joint in X translates it’s parent, not itself.
Can someone explain why that is?
When an ikHandle is parented under a group, that group takes over the space of the ikHandle, in a sense. You can even rotate the group and it will rotate the entire ikPlane. (If you don’t have a pole vector constraint.)
What you are seeing happen is that the ikHandle is no longer free to move. It sticks in place.
When you move the joint on the unparented version, the ikHandle moves along with the end joint. Try moving that ikHandle back into place, and you’ll see the bones are in the identical position as each other.
Interesting. I’ve never tried doing this before.
It looks like if the handle is directly under the world it’ll update its translation to match the new location of the end effector. If it’s parented to something other than world, it won’t update translation. Instead, it’ll rotate the parent bones to reach the new end effector. I have no idea why it would behave differently like that.
In general, I like to make sure that I’ve got the bones in the right place before I set up IK. If I need to adjust anything, I’ll delete the IK and recreate it after moving joints. That way I can make sure I’ve got the joint orientation I want.
Whoops! Looks like @clesage already answered. Probably should have checked for replies before writing mine.
The interesting part is, why would it not be free to move if it’s just a child of the group node. How is this different from other child parent relationships?