piper Docker - Multiple Voices?

Is it possible to configure the docker image for multiple voices?

I have tried adding more than one --voice in the docker run command, but it only seems to respect the last one.

I have also tried comma and semicolon separated lists, which result in errors.

Thanks

May 15 '23 16:05 rsbwhass

Not yet, but this is planned.

May 15 '23 16:05 synesthesiam

@synesthesiam Would the plan allow for multiple voice selection within the Home Assistant Add-On with non-Docker installs?

May 28 '23 19:05 snarlingllama

I need to figure out a good way for this to work. Some voice models also contain multiple speakers, for example the LibriTTS voice has 900! This will quickly become unmanageable in the UI with drop-downs or checkboxes.

Jun 21 '23 21:06 synesthesiam

Can the language field of the tts action be used to select a matching voice if the docker image did allow multiple languages?

Oct 05 '24 03:10 reijop

Unintentionally got halfway there trying to answer my own question. If you redeploy the docker container with another language, it will download the new one (And assuming you've persisted the volume where) side by side to the first one.

Using developer tools, action tts.speak, specify the language and it will at least switch. Mine are just reversed. More debugging.

Oct 05 '24 03:10 reijop