[DECtalk] Formant-based Neural Text to Speech Voices from Blastbay Studios

[Jayson Smith]

The following was posted to the https://audiogames.net forum on Friday 
morning:


Hello all,

Blastbay Studios is proud to announce the release of our first US 
English text to speech voices for Windows. These voices work with both 
32 and 64 bit versions of SAPI 5, and they are responsive enough to be 
used comfortably with screen readers and games.

When you download a voice, you can either install it as a trial, or 
unlock the full version if you own a serial number for it.

In trial mode, the voice will function normally for the first minute of 
audio that it generates. After this, the audio quality will be degraded 
and you will hear background noise. When you restart the application 
with which you are using the voice, the clock will reset and you will 
hear the normal output for an additional minute before it reverts back 
to the degraded quality mode.

Please note that the serial number that you receive when purchasing a 
voice is not tied to your hardware. This means that you can use it to 
activate the voice on as many of your own machines as you like, even 
machines without an Internet connection.

A common request from users has been support for a custom lexicon. The 
voices do have support for this, but we will be publishing a separate 
download with a user lexicon that will apply to all US English voices in 
the next few days.

Please visit www.blastbay.com to listen to audio samples and/or download 
the voices.

On a personal note, I would like to thank all of you who have given me 
feedback and encouragement throughout the development of these voices. 
It has meant a lot. I hope you enjoy testing the voices, and I look 
forward to your feedback!

Kind regards,

Philip Bennefall


I have purchased both voices, and am happy with them. As always, there's 
room for improvement since this is an initial release, but it seems he 
has a good thing going.


Jayson