AVoc: Real-Time Voice Changer with Linux Optimizations

[left finch]

Local (non-web) application, similar to W-Okada, deiteris, tg-develop voice changers. Looks like W-Okada voice changer. Open Source. Compatible with your usual RVC models.

Uses PipeWire multimedia framework in Linux for lower latency. Doesn't need virtual devices / virtual cable setups.

Has global hotkeys and popup notifications.

Split into PySide6 GUI and a voice conversion library for the ease of development of different GUIs or CLIs.

Distributed as python packages (PyPi) or an AUR package.

Linux developers are welcome to test, check GPU compatibility, desktop environment compatibility, etc. It's fine to use with EasyEffects: put "Noise Reduction" and "Autogain" as the input effects there.

The voice changer is theoretically cross-platform, but other platform weren't tested at all. It will definitively snag on something. 🙂

https://github.com/develOseven/avoc

https://github.com/develOseven/avoc/raw/master/doc/screenshot.png

[edgy hemlock]

Interesting and nice job!
unfortunately I’m not a Linux user and there aren’t many linux users here (mostly windows)

I might think about adding it in the AI HUB Docs, but i can’t really test it other than checking the code

[left finch]

windows users can test too, it's theoretically cross-platform. It will definitively snag on something, but it's up to them to fix it then.🙂

[iron heron]

This is so cool for voice artists

[edgy hemlock]

Btw is it based on W-Okada (forked) or made from scratch?
Currently are these the only perks compared to Wokada, deiteris and tg-develop?:

• PipeWire multimedia framework in Linux for lower latency without needing VACs
• easier GUI and CLI development
• easier installation via PiPY and AUR packages
• not really built in but could be said in the AI HUB Docs that it could be used with EasyEffects