I’ve been learning how to work with Audio2Face lately, but I’ve run into trouble exporting my results back to Blender. I initially exported my skin mesh from Blender into a2f, then used its mesh separate function and setup the character for audio2face. I was successfully able to do mesh fitting, and get the animation playing in audio2face. I previewed and exported the 52 arkit blendshapes in audio2face successfully as well. However, when I go back to Blender and try to import the blendshapes, I have trouble. I’m able to import my rig file back into Blender, but when I hit “Transfer Shape Data” I get this error:
Python: Traceback (most recent call last):
File "C:\Users\James\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\omni_audio2face\operators.py", line 870, in execute
result = self._transfer_shapes(context, source, target, mapping_object)
File "C:\Users\James\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\omni_audio2face\operators.py", line 767, in _transfer_shapes
attr.data.foreach_get("value", mapping_indices)
TypeError: foreach_get(attr, sequence) sequence length mismatch given 14062, needed 25711
Up to this point I’ve been unable to figure out what is causing it and would appreciate any help diagnosing/solving so that I can ultimately get my animation out from a2f and into Blender. Thanks!
Yes, and I’m quite sure I’ve followed everything in the first video properly, but I still have this issue when I go to transfer the shapes over.
In the tutorial, it doesn’t explain how he has separated the head mesh from body mesh. But I think he does it in Blender rather than using Audio2Face to separate them. Can you do a test where you cut the mesh in Blender before sending it to A2F?
My mesh isn’t attached to a body going into a2f. I’m using the separate mesh function in a2f to separate the eyes/teeth from the skin mesh for the a2f character setup. To be honest, I’m not sure how I would separate them in Blender beforehand, so I can’t test it right now though it is something I can look into learning/trying if you think that’ll make a difference.
Any further help on this? Still haven’t had any luck getting the exported blendshapes to transfer onto the original mesh.
i am just another user, but is it possible to do a clip of you going through the A2F process? at the very least upload some relevant screenshots (of your setting, properties, etc). i am just thinking it would allow others to know whether it’s mesh related or process.
another thing that might help is to include more of the log, just in case it’s related to something else. you can upload the most recent A2F .log file here:
C:\Users\James\.nvidia-omniverse\logs\Kit\Audio2Face\2023.2
Problem is, if you separate meshes in A2F, their vertex order will probably be different from the one in Blender. To have matching verts, you will probably need to re-export separated mesh from A2F into Blender and use that as your base. It’s just a guess though.
1 Like