Produce different voices for the same sentence

[JRMeyer]

>>> ritesh.nitjsr01
[August 15, 2019, 7:34pm]

Hi, slash
I wanted to know if there is a way to generate different voice samples
for the same sentence, using any of the pre-trained models ?

Thanks

[This is an archived TTS discussion thread from discourse.mozilla.org/t/produce-different-voices-for-the-same-sentence]

[JRMeyer]

>>> erogol
[August 16, 2019, 9:51am]

different speakers you mean ? If so, no there is not.

[Archived Post]

[JRMeyer]

>>> nmstoker
[August 22, 2019, 7:06am]

Out of interest, for the same speaker, would the same text always
produce the exact same audio output?

Is there a way to produce slight variations, in the same way that people
don't say a phrase in precisely the same way each time?

[Archived Post]

[JRMeyer]

>>> erogol
[August 22, 2019, 1:15pm]

you can enable the prenet dropout at inference

[Archived Post]