Dave (or anyone else), can you comment on whether an enhancement can be added to P2A to send the correct speed parameter for those of us who are using Polly?

 

Here is the documentation on Polly's speed parameter. Right now, the 0-10 Speed setting in Config-->Voices has no effect for Polly voices, I assume this is because it is not making a request that includes the '<prosody rate=...' parameter that Polly accepts.

Is this possible to implement in P2A?

 

 

P2A currently treats Polly Voices as SAPI 5 voices.  Amazon's Polly for Windows is supposed to provide their voices as if they were SAPI5 voices, which I don't believe would have the parameter you mention.

I will have a look to see what might be done with the Polly for Windows voices, if anything.

Dave

 

6 hours ago, Dave-Pilot2ATC said:

P2A currently treats Polly Voices as SAPI 5 voices.  Amazon's Polly for Windows is supposed to provide their voices as if they were SAPI5 voices, which I don't believe would have the parameter you mention.

I will have a look to see what might be done with the Polly for Windows voices, if anything.

Dave

 

Thanks for taking a look Dave. It seems as though all the documentation for Polly indicates it accepts SSML tags, but even within windows itself, polly voices don't seem to respond to the "Voice Speed" slider in Text to Speech settings. I opened an issue on their github, since technically Polly is open sourced.

I had a look and don't believe there is a way to change the speech rate directly on the SAPI 5 AWS voices.  There is a whole other architecture for using Polly voices directly, but P2A would have to be modified significantly to accommodate that.  This is something we will look at for a future version, but not something that will be done in the short term.

Hopefully there's something that the Polly for Windows software can do to create SAPI voices that behave the same as native SAPI voices.

If you find something please let me know.  You can email me at [email protected].

Dave