Muscle and Tissue - Jiggle

   1334   13   2
User Avatar
Member
12 posts
Joined: Oct. 2015
Offline
Hello,

I am wondering if people have had success with adding convincing jiggle using the new muscle and tissue system.
Overall I am happy with the results and behaviour of the muscles and tissue passes but one thing I am struggling for both is adding convincing jiggle.
When I reduce the stiffness parameter, the muscles do get more jiggle, but it's a bit too flabby and floppy, do they don't look like muscles anymore. I am looking for more of a springy jiggle, that is more high frequency but doesn't flop around as much.
With the tissue I get a slightly different issue, where I do get some jiggle, but if I want to ramp it up, the geometry starts to collapse.
If anyone has some tips on how to achieve more convincing jiggle or which parameters to look and consider for this, that will be much appreciated

Many Thanks
User Avatar
Member
12 posts
Joined: Oct. 2015
Offline
Aside from adding jiggle using the stiffness parameter, I am curious if there is any way to add jiggle on impact.
Edited by Milen Piskuliyski - Aug. 5, 2024 16:37:42
User Avatar
Member
172 posts
Joined: May 2021
Offline
Here are some resources:

https://www.sidefx.com/docs/houdini/muscles/frankenmuscle.html [www.sidefx.com]

https://www.sidefx.com/contentlibrary/capy-ripple-solver/ [www.sidefx.com]

https://www.sidefx.com/contentlibrary/muscle-target-shape/ [www.sidefx.com]

https://www.sidefx.com/docs/houdini//vellum/wrinkledeformer.html [www.sidefx.com]

Do these align with your "high frequency jiggle" concept?
Can you provide reference motion context, part of body, and visual etc?
PHENOM(enological) DESIGN;
Experimental phenomenology (study of experience) is a category of philosophy evidencing intentional variations of subjective human experiencing where both the independent and dependent variable are phenomenological. Lundh 2020
User Avatar
Member
12 posts
Joined: Oct. 2015
Offline
Hey phenomdesign, thank you very much for the links. These are not exactly what I am looking for, but still super useful to see. Maybe an additional sim pass on the skin for some extra jiggle and wrinkles.
But what I am looking for is some general tips on achieving realistic muscle and fat behaviour, using the newly released 'muscle and tissue' system.
The franken muscle link that you sent above is actually part of that system, but it doesn't say much about how to control the look and behaviour of muscles. The examples provided with the reindeer actually look quite convincing and realistic, but I am struggling to achieve the same level of realism.
Unfortunately I can't share any examples as the project is under NDA, but it is for a quadruped animal with some fatty parts and loose skin. So will need to read some muscle definition, but also some nice jiggle on the fatty parts.
User Avatar
Member
6 posts
Joined: July 2016
Offline
Milen Piskuliyski I have the same question. The Houdini documentation has a note saying that the information that they provide is sparse because it's in beta (ie the docs are out of date). The muscle/tissue example files in the content library are from the old H19.5 beta and have a note saying the input values have changed since then and you need to guess what they are now.
All the other info I have found seems to be based on the H19.5 beta and if you load them in H20.5 they break. What we need is an explanation of "if you want this look in the simulation you need to modify and combine these values on these nodes". At the moment you need to read each of the breadcrumbs on every parameter and try and figure out which ones to modify and combine to get a certain look.
@SideFX a masterclass based on the latest muscle/tissue workflow in Houdini 20.5 would be great explaining how to configure for the look of certain types of muscles/tissue. Also up to date demo files on the content library that can be run without trying to work out why they are now broken in the latest Houdini release would be good. Trying to work out what has changed since the old beta based files is not ideal.
I've looked at and read everything I can find and there's still limited information out there (even reading every sticky note on every .hip file).
User Avatar
Member
172 posts
Joined: May 2021
Offline




PHENOM(enological) DESIGN;
Experimental phenomenology (study of experience) is a category of philosophy evidencing intentional variations of subjective human experiencing where both the independent and dependent variable are phenomenological. Lundh 2020
User Avatar
Staff
21 posts
Joined: June 2024
Offline
Hi,

@Milen, have you tried lowering your stiffness damping values instead of your stiffness to increase the jiggle? To increase the jiggle on impact I would try to increase the mass density so that the muscles have more momentum. Also if you have areas of your tissue that are very fatty, say a big belly, I would try to make a "muscle" for the fat volume and give that more jiggly properties, rather than trying to achieve it all with the tissue sim. Let me know if that helps.

@etwiss we are working on all the items you suggested, thank you.
Liesbeth Levick
Technical Director: CFX
SideFX
User Avatar
Member
6 posts
Joined: July 2016
Offline
@Liesbeth, thanks!
Having a reference table that is based on simulating real world muscle behaviour would be really handy. A lot of the time I'm trying to get similar behaviour for different parts of the anatomy on different models that should be within a range of values. Even going from biped to quadruped has a lot of overlaps in anatomy definition so being able to quickly setup upper and lower arms for example means you can spend time refining rather than running endless wedges playing whackamole to try to track down the right combination of settings.
User Avatar
Member
172 posts
Joined: May 2021
Offline
@etwiss
I am a graduate student working on some tool-making approaches in this area and wanted to know a bit more about:

1.
etwiss
Having a reference table that is based on simulating real world muscle behaviour
2.
etwiss
being able to quickly setup upper and lower arms

How has this been done elsewhere in the past?
In the future, how do you envision interacting with this while making?
(viewport, parm panel, pop-up window, online, etc.)
PHENOM(enological) DESIGN;
Experimental phenomenology (study of experience) is a category of philosophy evidencing intentional variations of subjective human experiencing where both the independent and dependent variable are phenomenological. Lundh 2020
User Avatar
Member
6 posts
Joined: July 2016
Offline
@PHENOMDESIGN in response to your requests:
1) I'm thinking of the types of tables you see online in the Houdini documentation for the different nodes but relating to artist recreation of real world muscle and tissue behaviour. Sort of like the jello, rubber, solid etc settings in the new MPM solver. A good example is the tables for real world based IOR or PBR settings for shaders that are online and readily accessible.
Something like a very cut down version of this data as it relates to value ranges for the settings for the muscle nodes in Houdini would be helpful https://academic.oup.com/view-large/333867009 [academic.oup.com]
I'm thinking more as an online reference to keep open when configuring different muscle settings so you know a combination that should work already to give a certain look. Presets to start from in the nodes already configured would be ideal then a table reference of ranges to stay within so your sims don't blow up.
2) I'm not sure if you've seen it but the Unreal ML deformation Houdini .hip demo file has a good example of splitting the different body parts up for use in the tissue workflow. Splitting the muscles then merging them back together seems to work better than having a lot of tabs in a node for a lot of grouped muscles to keep track of. If you group the muscles by body parts you can then add tabs to separate out the settings for specific muscles within that group.

Personally I'm a fan of staying in vanilla Houdini if possible but if you want some ideas on tools that would be useful please send me a direct message and we can discuss. I've already put together a script to populate the muscle flex links for all muscles and procedurally generated activation lines for every muscle that I'm using a lot if you want that.
Also have you looked at Procedural Anatomy? There's a ton of stuff in there to learn from: https://www.procedural-anatomy.com/ [www.procedural-anatomy.com]
PA_Chris is a good guy to talk to more about this type of thing.

Cheers,
Edward
User Avatar
Staff
21 posts
Joined: June 2024
Offline
Milen Piskuliyski
Aside from adding jiggle using the stiffness parameter, I am curious if there is any way to add jiggle on impact.

The Muscle to Bone constraint is also limiting a fair amount of the jiggle, especially the jiggle due to impact. As an experiment try toggling off "Enable Muscle to Bone" under the Constraints dropdown of the Muscle Solver and see if that gives you the jiggle you're expecting. If it does, you can then turn it back on and reduce the muscle to bone stiffness and/or damping to dial in the exact look you want. You could also try reducing the muscle to bone distance threshold, but I would be wary of doing that too much with high stiffness values still in place as it could end up distorting the shape of your muscles.

I hope that helps!

Cheers,
Liesbeth
Liesbeth Levick
Technical Director: CFX
SideFX
User Avatar
Member
12 posts
Joined: Oct. 2015
Offline
@Liesbeth_Levick

Adding a volume for the belly was really helpful.
Also disabling the muscle to bone contains was super helpful too. The constrains were limiting the muscle movement quite a lot so was great to see without any muscle to bone constrains. In fact a lot of the muscles were behaving much better without the constrain at all.
Thank you very much for all the great tips, these're really valuable!
User Avatar
Staff
21 posts
Joined: June 2024
Offline
Hi Milen,

Great to hear, happy I could help!

Cheers,
Liesbeth
Liesbeth Levick
Technical Director: CFX
SideFX
User Avatar
Member
6 posts
Joined: July 2016
Offline
@Liesbeth thanks! That seems to be working better for me too. I was trying to change the properties of the muscles instead of focusing on the constraints.
  • Quick Links