[DECtalk] an article about speech synthesis methods

[Nikita]

Hi Blake,
There are two main types of parametric speech synthesis: statistical and
neural.
RHVoice is a statistical parametric synthesizer. It is capable of running in
real time on a CPU.
Neural parametric synthesizers run on GPU and require large servers.
Thus, a statistical parametric synthesizer can run on a personal computer
and integrate with screen readers, while a neural parametric synthesizer
requires constant data transfer between a personal computer and a server,
which is not very suitable for screen readers, although theoretically it can
also be used.
For example, on Android, the synthesizer of TalkBack can work over the
Internet when switched on for the first time, if you activate a language for
which the voice package is not preinstalled.
Best regards, Nikita.

-----Original Message-----
From: Dectalk <dectalk-bounces at bluegrasspals.com> On Behalf Of Blake Roberts
via Dectalk
Sent: Sunday, September 12, 2021 5:58 PM
To: DECtalk <dectalk at bluegrasspals.com>
Cc: Blake Roberts <BEarlRoberts at aol.com>
Subject: [DECtalk] an article about speech synthesis methods

Hi all,

While browsing the most recent Top Tech Tidbits newsletter, I came 
across this blog post which explains different methods for creating 
speech synthesis.

https://tink.uk/notes-on-synthetic-speech/


The author claims, at the time this message is written, that parametric 
synthesis is not available for screen readers. I wrote to her already 
about the parametric synthesis solution RHVoice. Some of you might find 
the article explaining various TTS methods of interest. I definitely did!


Blake

_______________________________________________
Dectalk mailing list
Dectalk at bluegrasspals.com
https://bluegrasspals.com/mailman/listinfo/dectalk