[Top][All Lists]

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

Re: Package proposal: greader, an audio emacs reader for blind and disle

From: Tim Cross
Subject: Re: Package proposal: greader, an audio emacs reader for blind and dislexic people
Date: Thu, 31 Jan 2019 11:54:16 +1100

Speech-dispatcher is software installed on most Linux systems to provide TTS services (as used by things like Orca and other accessibility tools). It provides a unified interface to many TTS engines, such as espeak, but many others as well. It has nothing to do with Emacs - it is just the backend TTS service you would need (rather than implementing an espeak specific interface). There is a speech-dispatcher lib for emacs which provides the functions necessary for Emacs to send data to speech-dispather. The speechd.el module uses this to send text to the TTS. 

The advantage of using speech-dispatcher for your TTS service is that it would provide support for many different TTS engines (quality and features differ greatly between TTS engines - as too does support for different languages).  The other advantage is that it provides a unified interface for doing things like setting speech rate, volume, pitch, pausing/restarting etc. I *thiink* speech-dispatcher will also work on OSX, but probably not windows. 

The speechd.el package will provide you with examples of how to take text that is in a buffer and send it to speech-dispatcher to have it read out. The library is small and well written (though suffering a little from 'bit rot', so can probably be improved upon - but the general idea is sound and will likely provide some clear examples. You will also see how you can enhance the speech by adding support for different voices - both speechd.el and Emacspeak use this technique - text is spoken in a different voice based on the font-lock properties of a word. So, you could, for example, have a eww buffer of a web page spoken such that headings are spoken in a different voice to paragraph text etc. This can make the spoken text easier to understand/follow for the listener. While speechd has this in a very basic way, Emacspeak has a far more sophisticated version. 

speechd.el is not part of core emacs and therefore is not in the GNU ELPA repository. It is GPL'd. With something like the package you are suggesting, you are probably best off developing it as a separate project and once it becomes mature, see what interest there is in having it moved into becoming part of the Emacs project. I suspect this is unlikely as it isn't core Emacs functionality, but you never know. Of course, that doesn't mean it cannot be a GNU project.

BTW, you may want to choose a different name from greader - there have been packages in the past called greader, which were interfaces to the old Google Reader RSS interface. 

On Thu, 31 Jan 2019 at 09:09, Michael Heerdegen <address@hidden> wrote:
Tim Cross <address@hidden> writes:

> nIn addition to emacspeak, there is also a package called speechd.el,
> which also does a similar thing, but uses speech-dispatcher.

But speechd is not part of the Emacs project, right?  I see it is
distributed in Melpa, but not in Gnu Elpa.  So I guess this is not a
reason to reject the suggested package, or do I miss something?




Tim Cross

reply via email to

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