Re: Adding native audio support to ports

[adr <adr%sdf.org@localhost>]

On Sun, 28 Jan 2024, Brian Buhrow wrote:
> 	hello.  How is the aaudio wave form getting passed from Espeak-ng to the audio device?  Is
> it going through pcaudiolib?

Pcaudiolib is just a banch of wrapper functions, nothing sophisticated.

Although it is a good idea to try to achieve the latency like in
the alsa code in pcaudiolib (I did it already), I'm afraid this is
not the issue. Also, latency should be configurable, as a said
pcaudiolib is very simple. I'm not interested in this software, I
just want to make espeak-ng and emacspeak's espeak server to work
ok. After doing some testing using directly the espeak server, the
problem seems to be in the threaded mechanism that allows to call
flush while it is already processing a synth command. Let's see
what I find.

Eflite+Flite works ok. The problem with this project is that only
English and Indic voices are distributed. There are instructions to
build voices from bible recordings from a web site, they made
indexes, &c:

http://www.festvox.org/cmu_wilderness

The process seems tedious and frankly, although I respect whatever
people want to believe and do with their lives I just got literally
sick having to navigate through myths and Late Bronce Era superstitions
exposed like facts while searching for copyright information.

The raspberry pi 4 (And surely older models too) are more than
capable of running these synthesizers, emacspeak runs perfectly on
linux.

I tryed yasr with your patches and I got frustrated like with the
others screen readers I've tryed, including the last release of
yasr in linux. They all end up screwing the screen. For a complete
blind person this is not noticeable but for me without any visual
impairment it's problematic. Also there are people who are not
complete blind and may want to use the screen reader as an aid.

Thanks for the advices, let's se if I can put all this together.

adr