favorite
Development Fund
Support Blender Core Development with a monthly contribution.
theaters
Blender Studio
The creators who share. Access production content and knowledge from the Open Movies.
code
Blender Developer Blog
Latest news and updates on Blender development.
people
Get Involved
Join the community and help with design, development, docs and more.
bar_chart
Open Data
A platform to collect and display the Blender Benchmark results.
menu_book
Blender Docs
Documentation on Blender's features, tools and API.
Blender Conference
The yearly event that brings together the Blender community in one place.
download
Get Blender
Download the latest Blender version, or try the beta!
6 Comments
Join to comment publicly.
Philipp Schoch
14th November 2017 - 23:10
This is cool, but very complicated to set up. I was wondering if this isn't actually unnecessarily complicated. And I think I have actually found a far easier solution now to achive the same behaviour:
With my solution you do not need an inverse spine at all. Instead you need only one additional mechanism bone let's call it "MCH-pivot". (But you do still need both the original chain and the control chain.) MCH-pivot should be a copy of the first spine bone. (So they are aligned the same way in edit mode, such that they have the same local space. And they should probably have the same parent, too.)
Instead of sliding the original chain along the inverse spine, just set it up so that MCH-pivot slides along the control chain. Now add a copy location constraint on the first bone of the original chain, which copies the location of MCH-pivot. In this constraint check the Invert-checkboxes for all three axes. also set this constraint to local space.
That's it. :D
Besides thanks for this tutorial series. I learned a lot. I think this inverse spine rig is the only video in this series that is probably unnecessarily confusing in my opinion. :D
Philipp Schoch
16th November 2017 - 19:19
Ok, probably there is a little problem with my approach. I don't know if this is important. But at 12:30 he alignes the rig in a way such that it does not move inititially when going in and out of edit mode. That is quite tricky with my solution. because it messes up the copy location constraint, since the local spaces don't align properly anymore.
Philipp Schoch
16th November 2017 - 19:38
I think we can fix this by adding an additional intermediate parent to the original spine. Instead of applying the copy location constraint to the spine we can apply it to this parent.
Philipp Schoch
30th November 2017 - 22:05
@philippschoch: Ok, wait in the original solution in the next video the two chains are parented to the root bone directly not to the body bone. but for my solution I think they need to be parented to the body bone. Besides I think I encountered a few other issues with my solution. I think my solution is actually a bit trickier to implement correctly than I thought.
Fin O'Riordan
22nd March 2018 - 11:36
Ingenious!
Tighe Racicot
17th February 2022 - 09:15
This is an incredible lecture. I remade this rig in Blender, but I honestly don't understand how the pivot is effectively changing. Can anyone explain further? Edit: Ah, I'm starting to get it. Not quite 100%, but close. The key principal is the parallelogram he demonstrated. In the rig, the reverse chain is parented to the control chain, so it's the reverse chain that pivots, which takes the original chain with it.