Switch to maya... it may take 3 to 6 months time to learn maya but you will not regret... it may be one of the best decisions you have ever made in your career.
lol try Houdini, steeper learning curve but much better. they have 15fps on simple rig like this? there is 120+ fps on similar rig with apex in Houdini. Maya is dinosaur that needs to be let to die.
My question is: Is it worth to continue studying rigging and animation basics? Do these tools come configured? Or are we going to have to be not only artists, but also technicians, engineers of all this? Is this feasible on a normal computer? And when will we have these new features running on our camera machines?
Always. What people call AI (its really machine learning or LLM not "AI" but hey, its sounds better) will always make mistakes. You will not be able to correct them if you dont know how things work.
It's totally worth it. Somebody needs to use those tools. Automation is also not something new, it's been around since maya became a thing. Being a rigger is more about how to automate stuff and become faster at it, rather than actually rigging. Regarding animation, there is a lot of work in capturing motion capture and if it's stylized there are still no tools that can do that proper. We can maybe get some of the way, but to get good animations it needs a human touch or a lot of help from animators. It's basically like motion capture, sure it's great, but it looks like crap out of the box. It needs a lot of work.
yes. Imagine having to rig a cartoon umbrella. None of these fancy AI tricks are of any use. Or if you need to rig a cartoon character. Even Human IK is completely useless in that case
The animation research was impressive, the ML deformer not so much. The input rig was way too simple/stylized to impress anybody here who actually works in the field
I would say the Animation part was 'Ok', it works for simple tasks, and budget projects, don't see them working for 'hero ' projects where you need to have that artistic feeling for it, it felt like sketching mocap mostly. Still is good, imagine for previz and animatic, will do wonders. for rigging is always depend on the pipeline, training your rigs to achieve poses at least is helpful to correct the mistakes and enhance the final result. This is what actually where I can see AI thriving , just like testing your code in many situations with generated tasks and configurations, it helps a lot in detecting unexpected errors and situations.
![[1hr Talk] Intro to Large Language Models](http://i.ytimg.com/vi/zjkBMFhNj_g/mqdefault.jpg)
I agree with the similar comment - bringing a similar toolset to 3dsMax would be appreciated.
Switch to maya... it may take 3 to 6 months time to learn maya but you will not regret... it may be one of the best decisions you have ever made in your career.
Amazing progress for animator' s control, tools, and all with the benefit of getting things done much faster, which is key, congrats!
This is crazy, I love maya ❤️
lol try Houdini, steeper learning curve but much better. they have 15fps on simple rig like this? there is 120+ fps on similar rig with apex in Houdini. Maya is dinosaur that needs to be let to die.
impressive
3ds Max update coming? A ton of expert generalists using 3ds Max would benefit a ton from having a tool like this to augment animation pipelines.
My question is:
Is it worth to continue studying rigging and animation basics?
Do these tools come configured?
Or are we going to have to be not only artists, but also technicians, engineers of all this?
Is this feasible on a normal computer?
And when will we have these new features running on our camera machines?
I think learning the basics will help a bit. You have to know the rules to know how to break them
You first need to make an amazing Riggs to teach the model how to deform amazingly. Right? This is just to make it faster
Always. What people call AI (its really machine learning or LLM not "AI" but hey, its sounds better) will always make mistakes. You will not be able to correct them if you dont know how things work.
It's totally worth it. Somebody needs to use those tools. Automation is also not something new, it's been around since maya became a thing. Being a rigger is more about how to automate stuff and become faster at it, rather than actually rigging. Regarding animation, there is a lot of work in capturing motion capture and if it's stylized there are still no tools that can do that proper. We can maybe get some of the way, but to get good animations it needs a human touch or a lot of help from animators.
It's basically like motion capture, sure it's great, but it looks like crap out of the box. It needs a lot of work.
yes. Imagine having to rig a cartoon umbrella. None of these fancy AI tricks are of any use. Or if you need to rig a cartoon character. Even Human IK is completely useless in that case
those initially locomotions seem to be warming up for training 🙄💪🤙
Amazing work!
NIce, but I want this for 3DsMax. Thank you.
Wow this is a game changer!!
Great stuff, looks very useful :) Can you add a "make cartoony" button? XD
Please release this
❤❤❤
The animation research was impressive, the ML deformer not so much. The input rig was way too simple/stylized to impress anybody here who actually works in the field
I would say the Animation part was 'Ok', it works for simple tasks, and budget projects, don't see them working for 'hero ' projects where you need to have that artistic feeling for it, it felt like sketching mocap mostly. Still is good, imagine for previz and animatic, will do wonders.
for rigging is always depend on the pipeline, training your rigs to achieve poses at least is helpful to correct the mistakes and enhance the final result. This is what actually where I can see AI thriving , just like testing your code in many situations with generated tasks and configurations, it helps a lot in detecting unexpected errors and situations.