We have collected some of the issues people are seeing into a single place. We hope this will help others find the information they are looking for.
04/21/21 - These are the most frequent questions or issues we are seeing from users
05/07/21 - Change this to general FAQ with different sections, updated with usage FAQ
=============================================================================
System Requirements
Hardware and OS requirements
Minimum Specifications
OS Supported Windows 10 64-bit, 1909 and higher
CPU Intel I7, AMD Ryzen
2.5GHz or greater
CPU Cores 4 or higher
RAM 16 GB or higher
Storage 500 Gb SSD or higher
GPU Any RTX GPU
VRAM 6 GB or higher
Min. Video Driver Version 455.28 (Linux), 456.71 (Windows)
Note: Omniverse is built to run on any RTX-powered machine. For ideal performance, we recommend using GeForce RTX 2080, Quadro RTX 5000, or higher.
For latest drivers, visit here.
=============================================================================
INSTALLATION FAQ:
Most frequent install issue - Outdated Windows Version:
(NOTE: From Audio2Face2021.2.4 and on - this issue has been addressed. The app now will error out if your windows version is wrong with a more clear message - instead of not letting you install)
Pre Audio2Face2021.2.4 - below error can still apply
"Error occurred during installation of Audio2Face: Command failed: “…....\pre_install.bat”
- This error happens when you have unsupported Windows Version (English Windows)
- For non-English windows of the same error, read below
- Windows have to be at least: Windows 10 64-bit, 1909 and higher
Solution:
- Update your Windows 10 and rerun install
Corrupted Shader Cache
Example: " 2021-04-19 16:46:49 [9,446ms] [Error] [gpu.foundation.plugin] Failed to compile compute shader: rtx/system/GenerateMipMap.comp.hlsl "
Log example:
2021 - 04 - 19 16 : 46 : 44 [ 4 ,744ms] [Error] [gpu.foundation.plugin] PsoCompute::createPipeline and shader loading failed. RtxResult: 3
2021 - 04 - 19 16 : 46 : 44 [ 4 ,745ms] [Info] [carb] file 'C:/Users/conta/AppData/Local/ov/cache/Kit/101.0/d34aba4f/shadercache/ff1c16dd5215b8b11dead59fa46698ba.d.desc' does not exist
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Info] [carb] file 'C:/Users/conta/AppData/Local/ov/cache/Kit/101.0/d34aba4f/shadercache/dfc06f72089e694021524bc28a3ec7ec.d.desc' does not exist
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Error] [gpu.foundation.plugin] Cannot load shader file 'rtx/scenedb/NormalGeneration.comp.hlsl' .
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Error] [gpu.foundation.plugin] buildShaderAsync() failed: rtx/scenedb/NormalGeneration.comp.hlsl
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Error] [gpu.foundation.plugin] Cannot load shader file 'rtx/rasterizing/Point.geometry.hlsl' .
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Error] [gpu.foundation.plugin] PsoCompute::createPipeline and shader loading failed. RtxResult: 3
2021 - 04 - 19 16 : 46 : 44 [ 4 ,746ms] [Error] [gpu.foundation.plugin] buildShaderAsync() failed: rtx/rasterizing/Point.geometry.hlsl
2021 - 04 - 19 16 : 46 : 49 [ 9 ,412ms] [Error] [gpu.foundation.plugin] Cannot load shader file 'rtx/system/GenerateMipMap.comp.hlsl' .
2021 - 04 - 19 16 : 46 : 49 [ 9 ,412ms] [Error] [gpu.foundation.plugin] buildShaderAsync() failed: rtx/system/GenerateMipMap.comp.hlsl
2021 - 04 - 19 16 : 46 : 49 [ 9 ,412ms] [Error] [gpu.foundation.plugin] Failed to async compile compute shader: rtx/system/GenerateMipMap.comp.hlsl
2021 - 04 - 19 16 : 46 : 49 [ 9 ,412ms] [Info] [carb] Initializing plugin: carb.blockcompression.plugin (interfaces: [carb::blockcompression::BlockCompression v0. 1 ]) (impl: carb.blockcompression.plugin)
2021 - 04 - 19 16 : 46 : 49 [ 9 ,446ms] [Error] [gpu.foundation.plugin] Failed to compile compute shader: rtx/system/GenerateMipMap.comp.hlsl
- Basically, sometimes the copy of the ‘shadercache’ files does not complete on the first run and you end up with a broken shadercache directory, and subsequent runs will error out due to this broken shader cache.
Solution:
- You will have to delete shadercache with path below and start the app again.
“C:/Users/%USERNAME%/AppData/Local/ov/cache/Kit/101.0/d34aba4f”
Non English Windows Installation Issue - FIXED
(NOTE: This issue has been fixed from Audio2Face2021.2.4 and on)
Windows OS version being older than 1909 + Non English based Windows OS
- Right now pre_install.bat approach to checking Windows OS only works for English version
- Non English Windows are failing to install due to this
Solution:
- A new patch is being worked on to fix this, will announce once this is released.
More user experience here:
Windows running into this same Assertion Failed issue
[ 4 .704s] [ext: omni.audio2face.player- 0.1 . 1 ] startup
A s s e r t i o n f a i l e d !
P r o g r a m : C : \ U s e r s \ James \ A p p D a t a \ L o c a l \ o v \ p k g \ a u d i o 2 f a c e - 2 0 2 1 . 2 . 3 \ k i t \ k i t . e x e
F i l e : s r c / h o s t a p i / w d m k s / p a _ w i n _ w d m k s . c , L i n e 1 0 6 1
E x p r e s s i o n : F A L S E
2021 - 04 - 19 18 : 29 : 33 [ 16 ,097ms] [Warning] [carb.audio.context] 1 contexts were leaked
We are seeing some people with this error when running Audio2Face 2021.2.3
The solution below has not been tested by Nvidia. Try at your own risk.
Forum User solution from tymaho
However, I was able to fix it by replacing the files:
C:\Users%USERNAME%\AppData\Local\ov\pkg\audio2face-2021.2.3\exts\omni.audio2face.core_deps\deps\py_sounddevice_data\portaudio-binaries\libportaudio64bit.dll
and
C:\Users%USERNAME%\AppData\Local\ov\pkg\audio2face-2021.2.3\exts\omni.audio2face.player_deps\deps\py_sounddevice_data\portaudio-binaries\libportaudio64bit.dll
With the file portaudio-r1891-build.zip\portaudio-r1891-build\lib\x64\Release\portaudio_x64.dll (Rename this file to “libportaudio64bit.dll”) in this zip file:
=============================================================================
GENERAL FAQ
Mesh Fitting Panel missing after Mesh Selection
Solution:
the Mesh Fitting panel is not expanding because it silently failed to find the Default Prim in your Usd Stage. (We will avoid this error in the future release).
Meanwhile, you can set the Default Prim on the top node of your mesh by Right clicking on it on the Stage view and choose Set as Default Prim .
Once you set this, try reselect the mesh again in the Meshes panel.
And Mesh Fitting panel should show up again
2 Likes
Where is the Maya file for male_bs_46_start.ma? Why is this file not included in the Omniverse Nucleus Navigator download?
Script for maya 2023 dont work?
@yseol Can you reply the above questions?
Hi @1259741322 ,
Sorry for the very late answer, but you can find the maya file is on the nucleus server. We uploaded this file along with our recent release.
localhost/NVIDIA/Assets/Audio2Face/Samples/blendshape_solve/maya/male_bs_46.mb
Hi @k.attila3d,
Which script do you mean? can you be more specific please?
I tested “load_Bs_weights_maya.py”, and it works good on maya 2023.
i open load_bs_weights_maya.py or “load_bs_weights_mh_rig_maya.py” with good path folder for load .json… it works in maya 2020, but not in maya 2023, I have its error for the two scripts loaded…
Hi @k.attila3d
I found you’re running python script in the Mel script editor. Could you make a python script editor by clicking + button and try it again?
Your error says blendShape1.browLowerL does not exist, can you check if you have the blendshape node with the correct name?
yes i check blendShape1.browLowerL , i’m not sure because work on maya 2020 . On python i have this error : # Error: ValueError: file line 15: No object matches name: blendShape1.browLowerL" I even deactivated in audio2face the browLowerL and browLowerR nodes and the error still continues with other lines… Have you a tutorial for maya 2023?
Hi @k.attila3d
Not sure why it is not working on your maya2023. The same workflow still works for 2020 and 2023.
From your screen capture, I don’t see browLowerL in the list. Is that a reason why it fails?
If you can attach or send the maya file and the .json file you’re using, I can test it here too.
i use exactly the same method in maya 2020…maybe maya 2020 bypasses the problem… this is my .Json
a2f_cache_nathan.json (310.3 KB)
male_bs_46.usd (13.2 MB)
it almost looks like a reading problem between python 2 and python 3, I suck at python.
I used the same export tutorial to Maya following all the steps in audio2face with the duplicated male_bs_46.mb, add blendshape solver and export in .json… nothing more
maybe a conflict between your script that works in python 2, while maya 2023 is python 3?
The script is compatible to both python2 and python3.
How did you use male_bs_46.usd in Maya? If you just import it, Maya wouldn’t give you the blendshape node, but just the neutral mesh.
Can you download this file from the localhost and apply the script?
localhost/NVIDIA/Assets/Audio2Face/Samples/blendshape_solve/maya/male_bs_46.mb
