I would like to integrate TTS engine Cepstral: cmd line tool swift.
Swift is able to get a text or text file to generate audio stream or file.
Is there an interface (or a finite number of places) to configure a handover?
Thank you for support, Uwe.
I would like to integrate TTS engine Cepstral: cmd line tool swift.
Swift is able to get a text or text file to generate audio stream or file.
Is there an interface (or a finite number of places) to configure a handover?
Thank you for support, Uwe.
Look at the code in mycroft-core/mycroft/tts
That is where different TTS that Mycroft current supports. There isnt Cepstral/swift support yet, so you or someone have to make that first. You can look into the others TTS to se who they are made and get inspiration.
As I see the Cepstral/swift have a API that can be used if one gets a API key so adding support for that in Mycroft seems posible.
If you can get a wav file in return from anything, it could be used. See
but instead of the web call (lines 36-37), have it call the binary for cepstral, and return the file name.
A really motivated sort would also address the need for the config bits and the name of the engine.
i’ve added a few TTS and STT engines to mycroft, feel free to reach out to me in http://chat.mycroft.ai @Jarbas
mycroft is super modular, you need to subclass the TTS class and then it’s just a matter of editing your config