Text-to-Speech
Safetensors

Question about the reference audio duration limit

#1
by Bluebomber182 - opened

I tried using a 23 second reference input audio file but the ai voice always start speaking in gibberish. What is the duration limit for a reference audio file?

Hi, the total length of the input speech and the target speech should not exceed 30 seconds. I think a speech within 10 seconds is enough as a prompt.

Bluebomber182 changed discussion status to closed

Sign up or log in to comment