6.4.3 Miscellaneous Speech Commands

Speech can be stopped using command dtk-stop — though in normal use, the action of moving the cursor will stop ongoing speech. The speech server can be stopped and restarted for cases where the user wants to switch to a different server — or in the rare case to nuke a runaway speech server.

C-e s


Stop speech now.

C-e d q


Toggle state of the speech device between being quiet and talkative. Useful if you want to continue using an Emacs session that has emacspeak loaded but wish to make the speech shut up. Optional argument PREFIX specifies whether speech is turned off in the current buffer or in all buffers.

C-e C-s


Use this to nuke the currently running dtk server and restart it. Useful if you want to switch to another synthesizer while emacspeak is running. Also useful for emergency stopping of speech.

Finally, here are the remaining commands available via the TTS related keymap C-e d.

C-e d a


Add this pattern to the list of repeating patterns that are cleaned up. Optional interactive prefix arg deletes this pattern if previously added. Cleaning up repeated patterns results in emacspeak speaking the pattern followed by a repeat count instead of speaking all the characters making up the pattern. Thus, by adding the repeating pattern ‘.’ (this is already added by default) emacspeak will say “aw fifteen dot” when speaking the string “...............” instead of “period period period period”.

C-e d d


Select a speech server interactively. This will be the server that is used when you next call either M-x dtk-initialize or C-e C-s. Argument PROGRAM specifies the speech server program.



Toggle splitting of speech on white space. This affects the internal state of emacspeak that decides if we split text purely by clause boundaries, or also include whitespace. By default, emacspeak sends a clause at a time to the speech device. This produces fluent speech for normal use. However in modes such as ‘shell-mode’ and some programming language modes, clause markers appear infrequently, and this can result in large amounts of text being sent to the speech device at once, making the system unresponsive when asked to stop talking. Splitting on white space makes emacspeak’s stop command responsive. However, when splitting on white space, the speech sounds choppy since the synthesizer is getting a word at a time.



Interactively set how text is split in chunks. See the Emacs documentation on syntax tables for details on how characters are classified into various syntactic classes. Argument S specifies the syntax class.

C-e d t


Prompt for and dial a phone NUMBER with the Dectalk.

C-e d cap V


Use this to find out which version of the TTS firmware you are running.

C-e d z


Send this command to the TTS engine directly.