Hello, this is my first time posting on the DevForum so please do correct me if I am incorrect.
I have been scratching my head off for weeks on possible ways I could somehow implement an Active Ragdoll system or Dynamic Motion Synthesis Simulation for a more realistic feel and better immersion when viewing player models.
Here is a video demonstrating Euphoria Engine Physics ^^^
So far, I have tested out some methods such as utilising an IKPF system to sort of fake this engine’s behaviour, however this has resulted in a not so authentic experience in comparison to its original. I have also tried to use inverse kinematics to move and exert a force using the player model’s legs to sort of physically simulate the walking process of the Active Ragdoll, however that did not work as my bipedal player model turned into the greatest Michael Jackson revive and pulled a depressing moon walk.
Here is another great video showcasing the same Euphoria Engine by visualising it’s capability of creating smooth procedurally animated ragdoll animations in real-time.
It would mean a lot to me if anybody is able to explain some concepts or methods I could utilise to recreate this flawless work of scripting on roblox studio.
You could reproduce this effect by manipulating bones or Motor6Ds (which you have to add to the player character yourself) and once hit, release a ragdoll effect with some stiffness, you could even add some procedural animations based on which part of the player was hit to further add realism
Yes, this is the method which I have been using, however, my main problem with this is figuring out how to make the character balance itself while utilising the procedural animation system.
As shown in the videos provided, when the ragdoll is shoved, it attempts to balance itself and push whatever is shoving it away from itself, displaying a sort of self-preservation behaviour, all the while maintaining the aspects of a realistic Euphoria-like ragdoll.
This is especially great for reproducing immersion when tackling other humanoids or tazing/lassoing them.
Could anyone give me some advice on achieving this sort of ragdoll behaviour?
Nobody can really give you ‘advice’ on how to code AI, a bunch of math is involved, I would suggest just playing around with it like I said. You can make it balance itself by literally making it balance itself
Maybe you should show us a video of what you have now?
Sadly I am unable to use my PC to record a clip right now, however, just to summarise my system as of current, I basically have an R6 IK FootPlanting system which procedurally animates the walking part using sine and cosine math as well as trigonometry, this FootPlanting system also utilises an inverse kinematics module(CCDIK) for accurate solving in calculating foot movement, hence, I experimented with this module to adjust to multiple nodes which move around my character to simulate dynamic motion synthesis as shown in this short demo video I made here(pardon the horrible editing):
Also, I know nobody can give me entire chunks of code for free what I was referencing by “advice” is some sort of guidance through practical methods or some knowledge on how a system may work in logical terms, such as the demo video I made above ^ showcasing how my system works under the hood
Ah, thank you for linking this, however, I already have a method of calculating Inverse Kinematics, my real problem is calculating the end position (e.g. where a realistic step would ideally be placed when stumbling, how the body would balance itself when tumbling or stumbling, etc.)
what I was referencing by “advice” is some sort of guidance through practical methods or some knowledge on how a system may work in logical terms, such as the demo video I made above ^ showcasing how my system works under the hood
