Make a realtime infer to virtual microphone

[arigandores]

Is your feature request related to a problem? Please describe.

It's only possible to hear realtime infer at output device (speakers) for this moment. So i can't use trained model as 'voice changer' at discord, teamspeak etc.

Describe alternatives you've considered

It would be great to make a realtime infer to virtual microphone which be installed at 'so-vits-svc-fork' installation.

Additional context

As an example you can see how 'clownfish' (voice changer program) works

Code of Conduct

• I agree to follow this project's Code of Conduct

Are you willing to resolve this issue by submitting a Pull Request?

• Yes, I have the time, and I know how to start.
• Yes, I have the time, but I don't know how to start. I would need guidance.
• No, I don't have the time, although I believe I could do it if I had the time...
• No, I don't have the time and I wouldn't even know how to start.