background noise and music in output

#1
by Diksha2001 - opened

I’ve been testing your TTS model and noticed a recurring audio quality issue. Whenever I generate speech using the voices provided by vibevoice-community , I hear noticeable background noise and sometimes even music artifacts mixed into the output. The same issue appears when I use my own custom voice as well.

Could you please check if this is a known problem with the current model release, or let me know if there are recommended settings to avoid these artifacts?

Increasing the CFG sometimes works.

Sign up or log in to comment