2022-11-10 Animation & Rigging module meeting

The meeting will be open for everybody interested to join on Google Meet (link below).

Present: 3d Illusions, Brad Clark, Christoph Lendenfeld, Colin Basnett, Daniel Salazar, Demeter Dzadik, Fani-Eleni Pandi, Hans Goudey, Ingo Clemens, James Ledger, Julien Duroure, Léo Depoix, Marc, Marion Stalke, MohammadHossein Jamshidi, Morgan Loomis, Nate Rupsis, Nathan Vegdahl, Rik Schutte, Sybren Stüvel

People present are referred to by first name for brevity. Others are referred to by full name.

Links

New Faces

  • Ingo Clemens, half of Brave Rabbit, saving the Maya & Blender communities with cool scripts. Been in the industry for a long time, usually as generalist, but is actually interested in rigging & animation.
  • Morgan Loomis. From New Zealand, used to work at Weta. Rigging supervisor at Steamroller Studios in Florida. Also mentor at Rigging Dojo.
  • Léo Depoix, pipeline developer at a VFX studio in Montreal. Using Blender since he was 8 year old, later contributed to the Geometry Nodes project, on the subdivision surface node. Worked on student project as technical artist, rigged & developed custom nodes in Maya to help rigging. Wants to port work from Maya to Blender.
  • 3D Illusions, just here to watch.

Since the Last Meeting / Announcements

  • Christoph signed a Development Grant contract. He’ll be working Wednesdays and Thursdays, starting 23 November.
  • Results of the workshop have been published on the wiki: Presentation and Results. This includes a PDF of the Mural workboard.

Landed

Short-term goals

  • T101130: Regression: Scaling of NLA Strip Via S Hotkey Not Working
    • Nate is taking over from Thibault de Villèle, he’s investigating the issue.
  • T87548: Propagate Pose, Next Keyframe
    • Sybren: This task needs a decision on desired behaviour. Has been open for a while, and fixing this likely makes an existing tool (Propagate Pose to Next Frame) work better / more intuitive.
    • Sybren: can we just look at the first key in time, and create keys on the other channels for that if necessary?
    • The module agrees this is the right approach.
    • Christoph: we could have a “frame” field that is set by the operator to the frame it propagates to. That way the exact frame can be adjusted in the redo panel.
  • T87413: Driving object scale also scales its particle instances
    • Sybren: do we break people’s file when they use this?
    • Rest of the module: people shouldn’t use this, it’s clearly broken.
    • Hans: we could also close or mark as known issue, because particles are on their way out.

Constraint Nodes design by Léo Depoix

Léo (@PiloeGAO) has been working on a design for a node-based constraint system, and presented his ideas.

He wants to work together with the geometry nodes team to get the required socket types implemented, so that he can continue working on a prototype. The actual implementation of constraint nodes will likely be separate from geometry nodes, though.

Brad is worried about feature parity, like the “Remove constraint without moving” button, the visual keying, blending, animation of targets. Can we get away with all of them in one graph or need to stack them like now to have many weights/targets?

Hans: the geometry nodes evaluator can certainly be used for this. And even having things like a Bezier Curve node in your constraint setup can be very powerful. Could be good to keep the integration with geometry nodes in mind from the start, not not close doors from the get-go.

Léo is developing on a separate Blender branch on GitHub for now. He’s open to migrate to git.blender.org when it’s the right time.

Patch Review & Decision Time

  • T88489: Animation: Interpolation Ease In and Ease Out names are fliped
    • Marion and Daniel already noticed it’s broken.
    • Sybren: if we change this around, it likely should be a Blender 4.0 change.
    • Daniel: do we know the scope of the change? Might only be in a few places.
    • Demeter: bendy bones also have ease in/out, but it’s unclear whether this has the same issue or not.
    • Marion checks the bendy bones, seems to do the right thing, but isn’t sure either.
    • Nathan: long-term it makes sense to address this.
    • Ingo: CSS (mentioned in task as consistent with Blender) is for coders, not animators, so not a good reference to use here.
    • Christoph: maybe then also rename it to “slow out” and “slow in”? People shake “No!”.

MohammadHossein’s Physics Based Poser

MohammadHossein gives a quick demo of his physics-based poser system.

Next Meeting

The next meeting will be a somewhat experimental “sync-up” meeting. It’s intentionally on a different day and time than the normal meetings, and will be held once per month.

It’s set for Monday 14 November, 12:00-13 CET (Your local time: 2022-11-14T11:00:00Z). Again it will be open for everybody who’s interested. The provisionary meeting agenda will be linked in the #animation-module channel before the meeting.

11 Likes

Exiting stuff! Regarding this part of the constraint nodes document:

Can we add a new addon with “helper” node groups that replace constraints that didn’t need to be translated to nodes ? - Could useful for users with not a technical background and don’t take much time to develop.

Would it really need to be an addon? As far as I know the plan, at least for geometry nodes, is to ship these kinds of nodegroups with blender as assets

1 Like

It’s usually easier to start with new features as an add-on. That way it’s easier to publish new versions independent of Blender releases, and iterate much faster. Of course the final implementation of constraint/rigging nodes will have to be well-integrated into Blender itself.

1 Like

MohammadHossein’s Physics Based Poser

That sounds really interesting! Anything to look at?

3 Likes

Physics Based Poser? like Cascadeur?

3 Likes

it really does sound interesting, ¿is there a link to some extra info?

There’s a glimpse of that on my portfolio. please take a look at Physics-Based Rigging (Physirig)

6 Likes

Yes, just like that. it provides collision detection and physics-based constraints

2 Likes