v0.4.0 LipSync

v0.4.0 LipSync

This version brings the real lip-sync feature.

We implemented the issue #38, replacing the Live2dView dependence package guansss/pixi-live2d-display with RaSan147/pixi-live2d-display which supports the lip-sync based on audio analyzing. Thanks @RaSan147 for his work.
MuvtuberDriver's sayer module is rewritten, joining with changes on Live2dDriver, to supports this new feature.

This feature also makes it easier to supports any live2d model. Latter versions (v0.4.x) will test & support the out-of-box live2d model replacement.

✨ Feature

• lip-sync based on audio analyzing
  • cdfmlr/live2dview@f8d3769
  • cdfmlr/live2ddriver@64cb639
  • cdfmlr/muvtuberdriver@5fe057d

🐛 Fix

• driver can positively ask audioview to reset if there are too many successive errors.
  • cdfmlr/muvtuberdriver@bad21f5
  • cdfmlr/audioview@fbc89dc

🛠️ Engineering

• sayer has a shorter playback timeout
  • cdfmlr/muvtuberdriver@dc9997a
• logs changes
  • cdfmlr/muvtuberdriver@83e900d
  • cdfmlr/muvtuberdriver@e656691
  • cdfmlr/live2ddriver@fdd44dc

📄 Documents

• Put versions changelogs into /docs

⚠️ BREAKS

This version focus on the new feature and slightly neglect the stability as well as the compatibility. We have found the following important feature is broken:

• "悲欢形于色": the new lip-sync based on audio analyzing is not capable to works with the original live2d expression & motion control based on the text emotion analyzing .

Later versions (v0.4.x) will fix it.