[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[VINUX-DEVELOPMENT] Next speech project?
From: |
Bill Cox |
Subject: |
[VINUX-DEVELOPMENT] Next speech project? |
Date: |
Fri, 5 Nov 2010 22:23:15 -0400 |
On Fri, Nov 5, 2010 at 9:01 PM, Trevor Saunders <trev.saunders at gmail.com>
wrote:
> I'm not sure I'm comfortable with two different ways to change rate and
> two sets of rate setting options and commands. ?I'd sugguest getting
> this merged in to the tts systems.
Hi, Trev. Most TTS synths in my experience don't perform well at high
speed. There are some, like Dec Talk and Eloquence that do, but we
have to face reality: Formant synthesisers are going away.
Concatenative speech engines are winning, because only people with
vision impairments prefer formant synths.
That's why I was so focused on speeding up speech with low distortion.
Now, any TTS can be played fairly clearly at high speed, including
all those "natural" voices we hate so much. This means we wont have
to lose productivity as support for Eloquence, voxin, and Dec-Talk
fades. This trend scares the heck out of me, but now I think we have
a decent open-source solution.
I think the most natural place to integrate sonic support is in the
speech switch - speech-dispatcher. That way, we only have to do it
once, and it immediately works for all synths, both open-source and
proprietary.
I don't know how to best integrate into SD. Having global options
simplify things for users who only use one synth, but synth specific
settings would be better. Also, this feature needs to have a way to
be disabled, and for the near future, it should be disabled by
default. That means we'll have two ways to control speed. I don't
see a way around it.
Bill