• Editor

  • [Request] Extra mesh deformation animation track?

    inglese

Sorry if this had been asked here before...
Would it be possible to allow users to enable an extra animation track for mesh deformation?
for example.. if I had a meshed face, I could have a track for eyelids blinking and then a separate, independent track for mouth and lip movements, all on the same mesh.. (or even a 2nd mesh, if that was the way to go about it).

currently I use bones to animate facial expressions and mouth/eye lids (like any good animator should!).
but my work involves many simple character portraits, and the slow part for me is always the face rig. sometimes I can save a lot of time by just manually deforming the mesh for blinking or mouth movements. but usually the mesh deform track is reserved for perspective tricks on the mesh.

On an almost related note... what would be amazing for my workflow would be to allow a 2nd deforming set of weights on a single mesh.. sometimes I rig my character with bones to cheat perspective... does works very well.. however if I'm using this technique on a rigged face.. then I have to use a bunch of transform constraints on the bones for eyebrows, eyes, eyelids, lips, etc. to make sure they also move with the bone I use to push the perspective. it works great but setting up the rig is a very tedious process, especially with the constraints are currently not kept when I duplicate a bone or rig.

It's probably not trivial, I can guess.. just thought I'd throw the idea out there anyway.
thanks for your time.

Related Discussions
...

  • Ddc_all

    • Messaggio #2
    • Modificato
    • Questo post è in inglese

    • Erika

    • Messaggio #3
    • Questo post è in inglese

    Hello dc_all,
    As you can read on this thread:

    Nate ha scritto

    Every deform key has to start all mesh verts. Also, each vertex (a float) is stored for every bone that affects a vertex. So a mesh with 100 verts where each vert is affected by 3 bones is 300 verts, then someone goes and keys that mesh a dozen times in 20 animations = 12 * 20 * 300 = 72,000 verts stored! This affects data file size and memory consumption

    And more here:
    Meshes - Spine User Guide: Vertex count
    Metrics - Spine User Guide: Vertex transforms

    Using deform keys for animations simply because it's quicker, is also unfortunately exponentially heavy under various aspects.
    It's true that you may take longer to set the rig up, but the advantage of the number of animations you can later create effortlessly pays the effort immediately.

    Regarding the fact that you have to redo the transform constraints, you could consider changing approach, by adding the new images to a copy of the already set controls, instead of doing the opposite and adding the bones in the new skeleton.

    What is truly long in this process is creating a good mesh and weighing the bones correctly to gain a good 2.5D movement, but if you ever tried to animate only by direct deformations, you should know how not only heavy for performance, but also tedious it is to keep correcting parts, while changing the weights that influence a bone will fix a problem you might have on all of your animations without having to triple-check. I suspect you may have already seen this, but I'll leave this video here for people who might be curious about how to create the face controls you mentioned on a single mesh:

    Hi Erikari,
    Thank you for your informed response and taking the time to offer a solution, as well as some tips to assist my workflow.
    Now that it's been pointed out, I can definitely understand how keyframed mesh points can be such a performance hit.
    I agree rigging the characters is a better practice, because as you say it makes it easier when attempting multiple animations from a single rig.
    relying on a keyframed mesh can turn to a tedious mess very quickly when it comes to fine tuning the animation.
    That being said, I'll still be using animated mesh points on my simpler rigs. Most of them only have a single, 30 frame animation per skeleton, and I'm yet to have any complaints from our programmers during implementation.
    I think it's acceptable in our situation.

    Parenting the facerig bones with the perspective offset bone is a much better idea, I have no idea why that didn't come to me sooner.
    previously, I had each facial bone with it's own constraint, so they could be tuned individually. I had more (unnecessary?) control over my rig this way, but things would always slow down when it came to adjusting the weights
    I had scrubbed through that video before, but playing it in it's entirety allowed me to pick up on some things I had missed.
    I sure hope her voice recovered.