search

sukeesh / Jarvis

Personal Assistant for Linux and macOS
MIT License
3.03k stars 1.04k forks source link

Changing the voice of assistant #663

Closed chekoduadarsh closed 4 years ago

chekoduadarsh commented 4 years ago

Hey, When I was checking up the voice mode of Jarvis, it wasn't pleasant to listen it was super robotic and non-human like. Proposing a migration from pyttsx to a combination of gtts and playsound which gives a way better humanitarian approach!!

pnhofmann commented 4 years ago

Sure! Just, would prefer to let the pyttsx code in Jarvis and add gtts as an alternative backend (and make the user choice if he wants to send stuff to google or not).

jarviscli/utilities/voice.py does implement 3 backend anyway - one for Windows, one for Mac and one for Linux. On Mac pyttsx isn't even used ;).

chekoduadarsh commented 4 years ago

Just, would prefer to let the pyttsx code in Jarvis and add gtts as an alternative backend (and make the user choice if he wants to send stuff to google or not).

Will do it

OneHighCloud commented 4 years ago

I am working on Win10 with the code. In voice.py, I made some changes regarding voice (VoiceWin()). Might help you with sound.

self.engine = pyttsx3.init("sapi5") voices = self.engine.getProperty("voices") self.engine.setProperty("voice", voices[1].id) self.engine.setProperty("rate", 140)

Methenitis commented 4 years ago

Hey!

After communicating with @chekoduadarsh , my colleague @annakotsa and I have worked on adding GTTS as an option for voice generation and fixed the parameters of pyttsx3 in order to make the voice sound less robotic.

Would love to get your feedback on this @pnhofmann

Thank you in advance!

pnhofmann commented 4 years ago

686