Comments on: Synthesizing Speech with pyTTS

By: Richard

Richard — Wed, 07 Jul 2010 20:50:33 +0000

Alright, I finally got it to work! MSSam should be Microsoft Sam. Other voices have their own particular names. Obviously, you have to get the names right, but there appears to be a discrepancy between the names given by
tts.SetVoiceByName(‘MSMary’)
vs. in XML

Why the difference?

By: Richard

Richard — Wed, 07 Jul 2010 16:38:25 +0000

Sorry, I didn’t realize that the tags would be taken out of my previous post. The XML tried (and didn’t work) was supposed to read:

I am Sam. Sam I am.>
Hi, I’m Mary.>

By: Richard

Richard — Wed, 07 Jul 2010 16:35:48 +0000

I tried the XML:
I am Sam. Sam I am.
Hi, I’m Mary.
but it doesn’t change voices. I’m on Windows XP with SAPI 5.1 and python 2.5.4 installed.

Same goes for this:
I am Sam. Sam I am.
Hi, I’m Mary.

Am I doing something wrong, or does pyTTS not support changing voices?

Also, I find that if there’s any typos in the tags, pyTTS will throw and exception. That’s pretty harsh. I think it might be better if it just ignored the tag and gave a warning, but finished the TTS anyway.

By: Python Test-to-Speech: Making Your PC Talk « The Mouse Vs. The Python

Python Test-to-Speech: Making Your PC Talk « The Mouse Vs. The Python — Fri, 02 Apr 2010 22:57:19 +0000

[...] after getting hired at my current job, my boss sent me a script (which I think was based on this article) about Python and a certain text-to-speech module called pyTTS. This was right after Python 2.5 had [...]

By: Speech with Python « There is no security on this earth. Only opportunity.

Speech with Python « There is no security on this earth. Only opportunity. — Sun, 17 Jan 2010 09:45:35 +0000

[...] http://mindtrove.info/articles/synthesizing-speech-with-pytts/ [...]

By: Peter Parente

Peter Parente — Wed, 13 Jan 2010 21:50:35 +0000

I’m not sure if that’s part of the info provided by SAPI. You should get reasonable results by registering listeners for utterance and word callbacks, starting a timer when you see the start of an utterance, and calculating the elapsed time each time you get a word utterance. Is that accurate enough for your needs?

You might also want to look at http://mindtrove.info/pyttsx which is meant to be a cross platform replacement for pyTTS.

By: Dave

Dave — Wed, 06 Jan 2010 20:35:27 +0000

Is it possible to get the timing marks for each word spoken?