speechd-discuss
[Top][All Lists]
Advanced

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

Re: piper output module?


From: Kyle
Subject: Re: piper output module?
Date: Wed, 14 Aug 2024 16:21:13 -0400
User-agent: Mozilla Thunderbird

The only thing I am aware of currently that should work is modification
of the generic module configuration. I was able to get this to work for
MaryTTS some time back, as well as for Mimic3 more recently. The problem
I have found with all of these, and the reason I don't use Mimic3 with
Orca even though I have it working in speech-dispatcher, is because
these are all very slow. Most of the time, a message takes a second or
more to begin speaking, and pauses between sentences or clauses can last
from one to two seconds. This was my experience with MaryTTS on a laptop
some years back, and has been my experience with Mimic3 as recently as a
month ago on a fairly fast machine. I know Piper has been designed from
the ground up to be usable on a Raspberry Pi, but I don't think it has a
packaged version yet, nor any other easy way to get it installed,
although I haven't had a lot of time recently to play with it. I could
try pip to see whether it's included there, and then I could modify my
mimic3-generic.conf to see what happens with it. I did hear of some work
possibly being done with a non-generic module, but I don't know the
status of that work.

~Kyle





reply via email to

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