[Feature]: Realtime Decode MDX for Online Karaoke Plugin

[GenzPN]

Description

Feature Request:
Support for Realtime Decode MDX (Online Audio Separation)

I came across an inspiring project called python-audio-separator-live, which provides audio separation in real-time. I believe integrating a similar capability into the project would open up new possibilities, particularly for those of us interested in building applications like online karaoke plugins.

Use Case:
By enabling real-time audio decoding and separation, users could livestream music tracks while isolating vocals and instrumentals on-the-fly. This would be a game-changer for online karaoke setups, allowing seamless performance and playback directly within a live environment.

--realtime

[beveradb]

This is a duplicate of #3, closing.
If someone wants to raise a PR implementing this feature I'd love to test and merge it!