20 replies to this topic

[Zanozza]

Hi!

I've been very impressed by Weta's "Digital Tissues" ( http://www.fxguide.c...-tissue-system/ ) so decides to try to create simplified muscle rig.

The target:

1. Muscles can be created procedurally.
2. Muscles can be created manually (sculpted) but placed procedurally.
3. Muscles not intersected with bones or another muscle layer.
4. Muscles can move by active or passive way (joint bended by muscle itself or by another force).
5. Muscle can be affected by gravity, acceleration, in-between muscle layer or skin.

So, this is the first, very simple iteration:



How this works for now? (simplifier version):

We have a shulder and sub-shoulder:



Place the main points of muscle and tendon attachment:



Generate the points normals, perpendicular to surfaces:



Project points to surfaces:



Left muscle points only:



Generate the spline from one muscle point to another on bone surface:



Sure, we can do the same including the tendon' attachment:



Now generate the muscle based on this spline and given muscle profile:



The parameters of muscle:



The form of muscle is based on "form" curve. The form changes with muscle contraction is controlled by "width scale" and "height scale" curves. They depend from joint angle.



So, to be continued!

Any questions or suggestions is welcome!

[abvfx]

This is very cool, i will be looking forward to more updates. Great work!

[Sifis]

Looks promising, I 'm looking forward to see what it might bring.

[edward]

Now generate the muscle based on this spline and given muscle profile:

Interesting. If you generate a "center" spline in between, then I wonder how well the Inflate SOP will work with this on some skin geometry.

[Netvudu]

oooh...keep at it, please. It looks like a nice tip of the iceberg

[Zanozza]

Thank you, my friends, for a warm feedback!

I haven't so much time as I would to spend for this wip, but, anywa, there is the new week and I have something to show you again!

Today is "Fat Skin Day"

Please take a look on video of sliding skin:

This sphere' skin have most of "freedom" of "fat under" sphere' Equator and less freedom on sphere' Poles.

Bonus is - skin is polygonal object and can have ANY topology structure. On this example I have very low-poly polygonal object. Sure, with more polygonal dense we can get more "smoothed" topology deformation (or "flow"). Dynamic itself will stay same.

Sure, every elasticity attribute can be painted or settled procedurally:

- "freedom";
- layer of "fat thickness";
- fat linear constant;
- fat linear damp;
- fat angular constant;
- fat angular constant;

So, this is original sphere:



Painted "freedom" attribute:



Finally, this is visuar representation of "freedom" as a normals length (for your visual consideration):

[Zanozza]

Interesting. If you generate a "center" spline in between, then I wonder how well the Inflate SOP will work with this on some skin geometry.

This is not muscle (yet), but some hard obstacle, like knee (sure, imagine a Pretty Girl' knee! ), ribs, clavicle etc movement under skin:

[Zanozza]

Another small update:



video:

As you can see, the muscles can slide one on another and working completely independently.

Sure, every "bottom" layer affect to "upper".

[Fabiano Berlim]

That's a very nice tool! Keep it going please!

[goldleaf]

Very cool!

Does the system remain pretty responsive when applied to a character? (if you've had the chance, of course; if not, still very, very nice!)

Kudos!

[Zanozza]

Today this system in "prototype" condition, so I don't worry (yet) about performance. Sure, this isn't realtime, but approx 10-15 fps.

I'm planning to write own solvers later, for muscles, inter-organs(muscles) layers, ribs and skins. But I want to create a working solution by standard Houdini operators first.

Thank you for your warm feedback, my friends!

[Zanozza]

Another short update (sorry, this is my hobby).

As you can see, I've got control on int-muscle sliding.

1 - muscles sliding without affect;
2 - under layer is "pulling" higher muscle layer in full affect;
3 - "pulling" is variable and depends from higher muscle' tension.