Closed Xanthos-Code closed 8 years ago
https://pythonspot.com/speech-engines-with-python-tutorial/
Tried googleTTS - it is throttled so you need a key and you get a couple hundred requests a day. Tried voiceRss - robot like speech on their web site duo page. I got a 404 using the API
google_speech looks nice - just requires installing a bunch of stuff
Another option - have users sign up and get an api key for google - then they can use their own service plan.
You could use https://github.com/marytts/marytts. Easy to install and easy to use crossplatform TTS.
Excellent suggestion! I will give it a go.
On Tue, May 31, 2016 at 1:51 AM, PageArkanis notifications@github.com wrote:
You could use https://github.com/marytts/marytts. Easy to install and easy to use TTS.
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Xanthos-Eve/vintel/issues/6#issuecomment-222629203, or mute the thread https://github.com/notifications/unsubscribe/APr2Bvo7XVjc1mGHXK-MP_-tlzhW0GDkks5qG_Z0gaJpZM4G9Fgq .
What do you think about pyFestival? https://pypi.python.org/pypi/pyfestival/?
Interesting, though the manual has a broken link. We are moving to python 3 and QT 5 now - may have more options with that mix.
On Wed, Jun 29, 2016 at 10:41 AM, Jesse Litton notifications@github.com wrote:
What do you think about pyFestival? https://pypi.python.org/pypi/pyfestival/?
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Xanthos-Eve/vintel/issues/6#issuecomment-229432161, or mute the thread https://github.com/notifications/unsubscribe/APr2BsdlKFCyVM19J7Yhlpd_9rk_9zQvks5qQq5WgaJpZM4G9Fgq .
I haven't really had time to look at the api, but I have played with festival in the past so it was the first thing that came to mind. I'm willing to help out from the Linux side of things - I'll try to familiarize myself with what you're already doing on OSX when I get the time.
I thought I found an easy solution in the pyttsx package, but as it turns out... not only is the speech quality much like a speak n' spell (like Festival if you don't install a higher quality voice), but it only works with Python2. Too bad, because it was literally 2 minutes to implement.
Let me continue looking at other options. I will try pyfestival if I don't find something better first.
nm.... with the high quality voice I have in festival, pyfestival sounds great! I'll make a develop-based patch and send it your way tomorrow. We can apply it to master if you want to push it out sooner. Since I can only use PyQt5 on Linux, I actually do all my dev on the develop branch.
Hey, am I crazy or is the voice not wired up to anything yet? I guess I was expecting there to be code somewhere that said things when on OSX, but maybe you just haven't added it until you had a cross-platform solution?
I've added a "Test Voice" button to the sound dialog to test it for now. I left the code such that it will prefer darwinSpeak if you're on OSX - but we can change that if you like. Will send you a pull request to look over when I get home tonight.
OSX has native support so its easy there - once we have a cross platform module we can refactor the code as needed. I looked at festival and thought it was overkill for what we are doing.
On Wed, Jul 6, 2016 at 1:18 PM, Jesse Litton notifications@github.com wrote:
110 https://github.com/Xanthos-Eve/vintel/pull/110
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Xanthos-Eve/vintel/issues/6#issuecomment-230893648, or mute the thread https://github.com/notifications/unsubscribe/APr2BmK8UZO_Dr-oZNQfegAhMYggn6qbks5qTA2wgaJpZM4G9Fgq .
gTTS looks great but google locks down the api from time-to-time so its not reliable.