AceCentre / pasco

Phrase Auditory Scanning COmmunicator - AAC App for iOS and the Web
https://app.pasco.chat
GNU General Public License v3.0
14 stars 6 forks source link

Add support for Azure TTS #86

Open willwade opened 6 years ago

willwade commented 6 years ago

Support Azure TTS

Original Issue

Possible TTS Engines to integrate:

Need to investigate Acapela - even if possible for FLOSS apps

willwade commented 5 years ago

We would need:

Note the workflow is usually to pick a voice, purchase it with in-app purchase and store it locally

willwade commented 5 years ago

Some of this gives us some more features - like adding in the ability to change pitch and intonation mid-phrase (check out https://www.acapela-group.com/voices/voice-tuning/)

willwade commented 5 years ago

This maybe useful: https://github.com/CoughDrop/extra-tts (NB: Would need to also add in in-app purchasing of voice)

willwade commented 4 years ago

Cepstral:" A Personal Use license is $31.41 - this allows you to use one voice on one device. " - https://www.cepstral.com/en/personal/linux

Acapela: https://www.acapela-group.com/solutions/acapela-tts-for-linux-embedded/ - Ace would have to negotiate a price. But all possible

willwade commented 2 years ago

Just an update to this..

azure cloud and Google cloud voices support and more languages.

For example we desperately need AAC that supports Persian/Fashto (Which Azure supports) or Urdu (Azure/Google)

I wonder if she would rehash an idea for these online only voices to support offline only use - wouldn't work for spelling but fixed phrases #188

gavinhenderson commented 1 year ago

@willwade Can we narrow this down to something more specific?

If you could pick one TTS system to add integration for what would it be?

willwade commented 1 year ago

Haha.

I'd plump for Azure. Its got more voices / languages , we already run a Azure instance. But yes.. its offline. Other than that I guess I'd plump for Acapella

gavinhenderson commented 1 year ago

@willwade Got you, im not saying never, but just for the sake of the issue I think its good to have an outcome of a single TTS system. Ill switch this issue to specifically be azure