Audio2Face Nvidia Omniverse and iClone8

[@Markom3D]

What do you think of the initial animation?

[@josepht1575]

this is a great "go to tutorial" because there's so many steps. Thanks for making it quick! btw if you download nucleus and set that up, under the "localhost: omniverse://localhost/NVIDIA/Assets/Audio2Face/Samples/reallusion/ there's the viking present and debra that you can load in

[@DataJuggler]

I finally have this working, I think. One thing I like about this versus Acculips, is you can export long audio files with Audio2Face, it just takes forever. My file is 50% done after 15 minutes. I have a 14 hour audio book, so my computer will be busy for a few months while I sleep. Testing the introduction now.

[@Kmankeg]

Most helpful vid on this topic so far! Nice one!

[@incrediblesarath]

Hi, Lower teeth is not moving much, what do I did wrong?

[@davidvideostuff]

Thx dude !!! Really helpful when your mother tongue isn't english. Does a better job in french for my part than acculips does.

[@iclone3d]

how do i find the a2f folder? thanks

[@Skyrider87]

Dear people, I have 2 questions, I have been trying to get it done for days, no idea how to proceed. I have watched the video a hundred times but can't figure it out. I used a sample character in character creator 4 (talking Kevin) When I export it it already goes wrong, because I see in the folder at audio2face and then the folder A2F only 1 file. What am I doing wrong here? Furthermore I do not succeed to export the file again to Iclone8 I see as if the field in I clone 8 remains empty. In addition, I also do not see motion files in folders that do appear in the video. How is this possible? Greetings Martin Netherlands.

[@sinanimacionnn-q3z]

Is it possible to bring the eyebrow movement generated in Audio2Face to Iclone?

[@Soojinmiao]

There is not anymore a cavity by the export.. the result isn't so good yet with the jaw.. any idea how to solve this?

[@johnmrs]

i have the same experience in iclone 8.1. it doesn't work now

[@himagrawal4903]

cant find audio2face plugin

[@MrSubinantony]

In the iclone and character creator i could see only the export menu of omniverse. There is no import menu. I exported the character from character creator now how to import it back to character creator.Should i need to install anything apart from omniverse connector

[@Markom3D]

I need to update this tutorial. I will remake it soon

[@christiandebney1989]

Just getting into iclone what are the advantages of this method over what Iclone8 offers with face puppet and the audio parts? is it extra emotional expression in one controller?

[@StevenCookFX]

That option at :33 "include Omniverse Audio3face mesh is disabled on my end for a character I have that was rigged using autorig in CC4

[@StevenCookFX]

Just tested this more and found that unless the character is a CC3+ or better it will not work. Humanoid characters will have to be converted.

[@gregorybennings8718]

Do you HAVE to use liveface? What if you have a windows phone?

[@leodegrance5609]

After watching the Audio2Face animation, it seems that something gets lost when it's transferred back into iClone. Especially the character's bottom lip? Or maybe the lower jaw didn't get copied back into iClone? IDK. Lips seem more expressive in A2F than in the final animation. Maybe the plug-in to bring everything back in needs to be updated. I remember it was around for iClone7. And it still says 'beta', so there is that I suppose. Gonna do some of my own tests now. Thanks for posting this. The process seemed daunting before, especially configuring everything manually in the original A2F a couple of years ago now.

[@Markom3D]

So, the issue is that it is running of blendshapes/shapekeys and it does it’s best to convert that information from a fluid mesh. Also the iclone character jaw normally works off a bone control and not a shapekeys. I’ll be sending an email to nvidia with some suggestions for improvements

[@leodegrance5609]

@@Markom3D Thanks for the info. Just doing a quick test now. I think the embedded auto-emotion of A2F can make the case for this workflow over accu-lips in that you can get both the lip movement and the emotions applied at the same time. Going to take it back into iClone8 now and do a side by side comparison and see the differences. Just a random thought out loud as it were.

[@leodegrance5609]

Okay, I couldn't resist. Did my own test following along with your tutorial. It really is a shame that a lot of the emotion blendshapes seem to be left behind. Still, it's a lot easier than manually imputing the expressions. Link to my humble attempt: kzbin.info/www/bejne/epCbcmmDd7xoisU

[@Markom3D]

@@leodegrance5609 it just adds to it with acculips. The other thing aswell is that you can bulk export the facial animations from A2F. You only really have to set it up one then export like crazy

[@WillyCash69]

@@Markom3D Ive been noticing this in video demos as well. The gray face is amazing looking, full of expressions.. but once it gets converted and loaded into other platforms it loses nearly all that made it worthwhile. So, sort of useless if you still wind up spending a ton of time trying to recreate the emotions manually

[@yahbin77]

That was sick..

[@AVISIONMUSICTV]

Can you just export Audio2Face data to Iclone? And just apply it to a character

[@Aghashujaakhan]

Thank you for your video. You were using iClone 8, but in iClone v8.4 they introduced to select a solver. Kindly update to resolve this problem. I am getting error No solver selected.

[@AVISIONMUSICTV]

The Audio2Face Ai tech is intriguing but Tbh honest acculips is absolutely amazing. Audio2face is an alternative but I would recommend diving into acculips it's fantastic.

[@DataJuggler]

Acculips is also much easier to setup. I still have high hopes for Audio2Face.

[@blendertutorialforme9746]

can it done it iclone 7?

[@andreasnoutschmidt9964]

Interesting, but why? Seems like an aweful lot of complicated steps to practically outsource AccuLips, or did i miss something?

[@Markom3D]

Just showcasing other ways of doing things. Might be some people out there that don’t want to use acculips.

[@DataJuggler]

It seems the teeth and some other properties have changed. Instead of Mesh I see GeomSubset, which Audio2Face doesn't like.

[@greenmonstercast8193]

Pretty cool

[@DataJuggler]

3:40 My eyes are all messed up on my character after I click Yes Attach. Edit: I fixed this by adjusting the eye post processing offsets and lowing the Saccade Strength to about .14.

[@DataJuggler]

I am glad I left this note, because it happened again and I forgot.

[@viktorshtam2104]

что то тут не так. Головы, которая по умолчанию, не появляется. Нигде ее нет

[@DataJuggler]

You should remake this video with the latest Audio2Face. My teeth end up way above my character, and the Post Wrap steps didn't fix it.

[@gennaroschiano8085]

thanks!!!

[@KwasiMedia]

I'm still on Win7 so I can't even use Omniverse yet. 😫

[@brechdt]

It requires RTX

[@stevensteverly]

How the hell did you figure all this out?

[@Markom3D]

lol, with alot of punching myself in the face. Lets just say they rewrote Audio2Face based off my videos lol

[@TearDownTeddy]

I dont think this works with humanoid characters..

[@MrRac2010]

Que culiiaaauuuu que sos! 😅👌🙌👏👏 Yo no lo puedo llevar a iClone todavía

[@DataJuggler]

I've tried this too many times. I give up.