Speechd-el provides very powerful features that are not always immediate to understand and make work. Most of this documentation describes my own experiences with Speechd-el settings. I have been using at most Speechd-el features even if I was sometimes lead to implement a few lines of code to enhance certain features. Anyway you should always keep in mind that Speechd-el has a very wide set of features that you first should try to use before thinking that you have something to implement.
I think that what is required by the user is that speech characteristics fit the particular situation/context where the user find itself. The goal of this small documentation is to explain how to build a mapping
|situation/context –> speech characteristics .|
speech characteristics are kept in Speechd-el implementation in an object called a voice. The mapping above hence becomes a mapping
|situation/context –> voice .|
A voice contains all characteristics of the speaking flu i.e.language, speedth, punctuation pronounciation … voice definition is controlled by the speechd-voices variable that can be customized. You should then first define voices fitting your needs by customizing this variable through the standard Emacs customization mechanism (cf. 2 .)
The situation/context has not a so precise definition and is not implemented in Speechd-el by an identified object. In fact it is represented by a batch of variables that can be checked when trying to deliver a voice output. Certain mapping mechanism are already implemented by Speechd-el like the
|face –> voice|
mapping implmented by the speechd-face-voices variable customization (cf. 3 .)
One of the most important concept in Emacs is that of mode. The goal of this small documentation dedicated to Speechd-el is not to explain in detail what is an Emacs mode but we refer the reader to Emacs documentation to discover this very powerful concept. However it almost implements the notions of situation/context we used above. Anyway as we will see in 4, mode might not be sufficient to take care of all situation/context parameters.
Moreover it is unfortunately not possible to directly map modes onto voices. Two successive mapping should be used: first mapping a mode onto a connection (cf. 4 ,) then mapping a connection to a voice (cf. 5 .)
If you encounter problems with this package content please send an email to devel.
|Apache/2.4.38 (Unix) PHP/7.3.2 SVN/1.11.1|