Vellum muscle simulation - target shape

[chovasie]

Hi guys,

I am doing some tests with vellum muscle setup to decide if it's ready to replace my current Ziva workflow.

So far i find it pretty good, and with help from the Rex leg example scene im getting pretty nice results so far. I managed to wrap my head around it and getting nice results.

It's still not FEM, but pretty close, and compared to Ziva setup, this is at least for me much better workflow. Cant wait to see FEM integration of this setup, as they mentioned in Houdini 19 lounch reveal, that Vellum is first integration, and thean they will do same thing for FEM.

Only thing that i am still not to sure how to replicate from Ziva setup is "targetShape". where you kind of create shape, which will act as target for your muscle shape when it's contracted. It's not 100% as with blendShape, but it will try to get shape of what you modeled. In Trex leg scene i see notes in Muscle Solver section where it says that is possible.

"Similar to the way fiber contraction is handled, you can also have activators for muscle shape and muscle bulge. By default, all three qualities are driven by the same @muscletension attribute, but you can specify a different attribute for each and thereby affecting shape independently from bulge response (volume boost) and fiber contraction.
Shape activation will look for a @targetshape position attribute. This static attribute supplies a rest shape to the solver that a muscle will then attempt to conform to when activated."

 

But am not really sure how this setup would look like. Anyone tried to do this? Any suggestions?

Thanx a lot,
C

[kushaldas]

Take a look at the Vellum Rest Blend nodes (Paul Esteves explains them well here). As long as you are able to create an approximation in SOPs of what you'd like the muscle contraction to look like, you can drive the Rest Blend to match that contracted muscle when the muscle contracts with an attribute.