Audio2Face 2022.1.1 Hotfix released

AI & Data Science Digital Humans (closed) Audio2Face (closed)
1

Audio2Face 2022.1.1 introduces hotfixes for the following issues.

  • OM-60167 Character Setup - Eye Field selections not saved / lost on load.
  • OM-59915 CharTransfer test failed due to deprecation on last warp v0.4.0.
  • OM-59341 It takes more than 3 minutes for Audio2Face to launch.
  • OM-59224 When Character Setup duplicates the eyes some geometry settings are lost.
  • OM-58440 Set new kit version in title bar.
  • OM-58448 Eyelid offset parameter is loaded with a negated value.
  • OM-58451 Riva TTS produce bad stream audio.
  • OM-58485 Fail to attach Eyes xform prim to A2F Instance.
  • OM-5444 Issue with loading audio from OV server.
  • OM-59465 CharTransfer Axis arrow doesn’t update when user changes camera.
  • OM-59464 Ability to handle different blendshape orders.
  • OM-58442 new a2f launcher description update.
  • OM-57780 Fixed issue that prevented Audio2Face from loading on vGPU A40.
4 Likes
3

Can you add an [optional] auxiliary time-stamped text that can be used to assist in automatically cutting the audio to avoid false animations caused by background noise?
Because some long audio has non-speaking parts, there may be some noise, audiotoface will be animated in such cases, timestamp text can cut and mask the audio.
If you have timestamped text, you can omit the following steps:
Cut the audio manually into small segments, then enter each segment into audioToFace, then set the parameters one by one, then export them one by one, and finally merge them together

4

I can´t find the audio2face tool window, can you help me please?

5

It is own app. it should open like this:

image
image2143×1399 338 KB

If you start two A2F apps at the same time, the other might not have that audio2face -tab.

6

We can consider that for future.

For now you do have a range setting that you can use to limit the in and out times of your clip.