Exporting USD files

Hi everyone,

I have my Ready Player Me avatar in Blender with glb/fbx format. I want to export it with a USD format & import it in OV Audio2Face to work on my RPM avatars lip syncing.
But I am having problem with exporting my charater & it doesn’t work.
Do you have any suggestions?

OS: Windows 11 Pro
GPU: RTX 3050
App version: Audio2Face 2022.2.1


Hello @NedzZone! Could I get some more information on the issues you are having during import/export? It sounds like you managed to get your RPM avatar into Blender. Are you using the Blender Connector to open your GLB/FBX file?

I did find this article on Medium that might be able to offer you some assistance: Made With Unity | Using Blender to Convert Ready Player Me Avatars to FBX | by Rehtse Studio | Medium. You should be able to change the export from Blender file format from FBX to USD, but I have not tried this myself yet.

Any details you can provide would help!

Hi there!

If you could be more specific on the errors we might be able to help more. Some things to think about:

  • The Audio2Face Add-on that ships with Omniverse Blender is only supported by Omniverse Blender. Omniverse Blender supports many more USD features for both import and export.

  • The Audio2Face Add-on was written to streamline this workflow, specifically because there are a lot of confusing parts to getting a rigged character transferred.

  • The RTX 3050, I believe, doesn’t have enough video memory. I think that for most Omniverse Apps the minimum requirement for video RAM is 8 gb-- does your graphics card have that much video RAM?

If you followed the videos, are there specific spots you found confusing? I could help clarify.

Dear @WendyGram ,

My apologies for the delayed reply. I overlooked your previous response and I am truly thankful for the document you shared with me, it has proven to be invaluable.

Primarily, my interest in exporting the RPM avatars in USD format is for their application within the A2F. However, it’s not entirely clear yet how to facilitate their use within this framework, or at least, I haven’t found extensive documentation on it yet.

I greatly appreciate your assistance, and once again, I regret the tardiness of my response.

Best Regards,

Hello @cwardlaw

Sorry for the belated response!

I’m grateful for your thorough advice and the wonderful tutorial you’ve shared.

My primary goal is to successfully utilize RPM avatars in the A2F application, and your insights have been tremendously helpful in this pursuit.

Here are the specifications of my laptop:
System Memory (RAM): 32 GB, DDR4
CPU: I7-12700h 12th Generation
Graphics Card: Nvidia GeForce RTX 3050, Memory: 4 GB, GDDR6

With respect to the Audio2Face Add-on, I’ve been getting acquainted with Omniverse Blender, so your detailed explanation certainly comes in handy. I understand the intricacies involved in transferring rigged characters. In fact, I attempted to rig and import the RPM avatars via the connector but without success. I was following this tutorial: Rig a Face Blender Addon for FREE - Nvidia Omniverse Audio2Face - YouTube.

It appears the main obstacle is that RPM avatars require additional adjustments in their modeling and design before they can be effectively integrated with A2F. Could you provide some pointers on how to approach this?

Thank you once again for your invaluable guidance.

Best Regards,