speechd-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [orca-list] making Linux a hospitable place for TTS engines like Vox


From: Bill Cox
Subject: Re: [orca-list] making Linux a hospitable place for TTS engines like Voxin
Date: Tue, 22 Dec 2020 11:14:56 -0800

I've installed RHVoice, and it is working with Orca on my machine.  The English voice sounds OK, though I hear more of a Scottish accent rather than American.  It is a nice voice, though.  The maximum rate is waaay too low, possibly no more than 2X speedup from default.  Some blind folks find rates up to 6X faster than default useful for work.  I usually use 4X speedup.

I see that libsonic is linked in, which I wrote, and I believe Samuel Thibault initially packaged for Debian.  It is also linked into espeak, but now days most commercial TTS engines incorporate the core algorithm used in libsonic for speeds > 2.0 directly into the vocoder.  This produces less noise in the audio and mostly eliminates the CPU overhead of libsonic.  So, actions to consider:

- Assuming libsonic is used for speech speedup, just increase the high-speed limit in sd_rhvoice to something far higher, like 6.0X faster than default.
- Longer term, consider upgrading the vocoder to produce cleaner audio without the overhead of libsonic.

If you or anyone on the RHVoice team would like to chat about how to upgrade the vocoder to generate smooth high speed speech, just send me an email.

Thanks for writing RHVoice!  I hope to see it all over the place, just like Espeak.  In the meantime, until it gets rolled out onto popular stable releases of Linux, I may be able to help you write a portable binary version of your sd_module executable, so folks on just about any Linux distro could use it.

Bill

On Tue, Dec 22, 2020 at 2:42 AM Sergei V. Fleytin <fleytin@gmail.com> wrote:
Hi Kyle and others.

Just to let You and others know that address for the official github RHVoice
repository now is:

http://github.com/rhvoice/rhvoice



Stay well!

Sergei.


-----Original Message-----
From: orca-list <orca-list-bounces@gnome.org> On Behalf Of Kyle via
orca-list
Sent: Tuesday, December 22, 2020 5:48 AM
To: speechd-discuss@nongnu.org; Orca-list <orca-list@gnome.org>
Subject: Re: [orca-list] making Linux a hospitable place for TTS engines
like Voxin

I messed around with MaryTTS some years back. I even wrote a little
interface at one point that would allow the generic sd module to talk
through it. But I found it to be very cumbersome on most hardware, as it had
to talk to MaryTTS by sending an http request to the speech synthesizer and
playing the wav output that was returned. I also have found Java to be very
very slow overall. There was also a strange tendency for MaryTTS voices to
suddenly raise their pitch to a very squeaky high, although I didn't hear
many of the English voices do that.
I did hear bdl get rather tinny and oldschool at random times however,
sounding autotuned or like packet loss was occurring in an extreme
low-bitrate speech recording.


Perhaps work could be done to integrate the already existing RHVoice module
into sd. I use RHVoice here every day all the time, and it sounds smoother
and runs with less resources than MaryTTS. It also doesn't get tinny or
scratchy at random times. I'm not fully sure, but I don't think new voices
require proprietary tools in order to train them either, though I could be
wrong about that. In any case, have a look at the excellent work over at

https://github.com/olga-yakovleva/rhvoice

~Kyle

_______________________________________________
orca-list mailing list
orca-list@gnome.org
https://mail.gnome.org/mailman/listinfo/orca-list
Orca wiki: https://wiki.gnome.org/Projects/Orca
Orca documentation: https://help.gnome.org/users/orca/stable/
GNOME Universal Access guide:
https://help.gnome.org/users/gnome-help/stable/a11y.html



reply via email to

[Prev in Thread] Current Thread [Next in Thread]