Optimizing Ragdoll Simulations for Complex Character Animations

Hello

I have been exploring ways to make ragdoll simulations more efficient when working with characters that have complex rigs / multiple layers of animation. One approach I found helpful is tweaking the solver iteration settings to balance simulation accuracy and performance. :slightly_smiling_face:

Additionally; using low-poly proxy geometry for collision calculations significantly improved simulation speed without compromising quality.

Another tip I discovered is fine-tuning the damping values for joints to prevent unnatural movements in the ragdoll state. I have also been experimenting with custom gravity settings; which can help create more realistic falls & reactions depending on the environment. :innocent: I have checked https://forums.ragdolldynamics.com/c/tips-and-tricks/6- Selenium reference guide but still need advice .

I would love to hear how others optimize their setups! Do you have any specific solver / rigging tricks that you use to get the best results? :thinking: Let’s share ideas!

Thank you ! :slightly_smiling_face:

Hi @segidas, welcome to the forum and thanks for sharing your thoughts! :partying_face:

Some video examples would work wonders here; can you post some typical complex setup you find needing optimisation? If it is proprietary, maybe use one of the default assets to demonstrate, maybe connect 3-4 manikins together to form some kind of mega-manikin? :sweat_smile:

How do you find iterations impacting the performance/accuracy? What does a fast proxy geometry look like, compared to the default (256 vertex limit)? What does custom gravity mean, more moon-like or saturn-like? And how does it impact performance?

The link you shared isn’t working, did you mean to include -Selenium?