Say client to Speech Dispatcher

This manual documents Say for Speech Dispatcher, version 0.4.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation License".

Say for Speech Dispatcher

1 General

spd-say is a very simple client for Speech Dispatcher that takes some text from the command line and some simple parameters and passes them to Speech Dispatcher for synthesis. The resulting synthesized speech is then played on your speakers. It can also be used for stopping and cancelling speech on Speech Dispatcher.

Usage: spd-say [options] ["some text"]

-r or --rate: Set the rate of the speech (between -100 and +100, default see below).
-p or --pitch: Set the pitch of the speech (between -100 and +100, default see below).
-R or --pitch-range: Set the pitch range of the speech (between -100 and +100, default see below).
-i or --volume: Set the volume (intensity) of the speech (between -100 and +100, default see below).
-o or --output-module: Set the output module. Default see below.
-O or --list-output-modules: List the output modules that are available with the current server.
-I or --sound-icon: Play the sound icon.
-l or --language: Set the language. The parameter is an iso language code, like en or cs, possibly with a region qualification, like en-US or en-GB. Default see below.
-t or --voice-type: Set the preferred voice type (male1, male2, male3, female1, female2, female3, child_male, child_female).
-L or --list-synthesis-voices: List the ”synthesis voices” supported by an output module. If no -o option is supplied, -L displays the synthesis voices provided by the default output module. Otherwise, it displays the synthesis voices provided by the chosen module.
-y or --synthesis-voice: Set the synthesis voice. The synthesis voice is an attribute of the chosen synthesizer. Most users will probably not wish to set it, but this setting is provided for completeness.
-m or --punctuation-mode: Set the punctuation mode (none, some, most, all). Default see below.
-s or --spelling: Spell the message.
-x or --ssml: Set SSML mode on (default: off). The text contains SSML markup.
-e or --pipe-mode: Set pipe mode on (default: off). Reads text from stdin, writes to stdout, and also sends to Speech Dispatcher.
-w or --wait: Wait till the end of speaking the message. In this mode, spd-say returns only after the message is fully spoken on the speakers or after it gets discarded.
-S or --stop: Stop the previous message in Speech Dispatcher (regardless of the client who sent it and it’s priority) before saying the specified message. If no message is specified, it just performs the stop function.
-C or --cancel: Similar to -S, but it stops (cancels) all the messages in all the queues in Speech Dispatcher. Use this for total stop of speech.
-v or --version: Print version and copyright info.
-h or --help: Print a short help message.

Defaults for rate, pitch, and volume are determined by DefaultRate, DefaultPitch, and DefaultVolume settings in the speechd.conf file. Factory defaults are 0, 0, and 100 respectively.

The default language code is determined by the DefaultLanguage setting in the speechd.conf file. Factory default is en-US (American English).

The default output module is determined by the LanguageDefaultModule setting in the speechd.conf file. Factory default is flite.

The default punctuation mode is determined by the DefaultPunctuationMode setting in the speechd.conf file. Factory default is none.

2 Pipe Mode

In pipe mode (option --pipe-mode or -e), spd-say reads text from stdin, writes to stdout, and also sends the text to Speech Dispatcher. Example:

 $ cat README | spd-say -e

If a line begins with the characters !-!, the line is sent to Speech Dispatcher as a SSIP command. Example:

 $ spd-say --pipe-mode
 The quality of mercy is not strained.
 !-!SET SELF VOICE FEMALE1
 It droppeth as the gentle rain from heaven
 Upon the place beneath: it is twice blest;
 !-!SET SELF RATE 50
 It blesseth him that gives and him that takes.
 ^D

Note that each line of input is individually sent to Speech Dispatcher, therefore in the example above, the word heaven is taken as the end of a sentence. It should be possible to use other text filtering utilities, such as tr or fmt to compensate. For example, the following command would clean up most cases:

 $ cat INSTALL | fmt -u -w 1500 | spd-say -e

If the text itself contains !-!, just make sure it is not at the start of the line to prevent it from being interpreted as a SSIP command. The following sed command will do the job:

 $ cat myfile | sed -e 's/^!-!/ !-!/g' - | spd-say -e

The input is sent to Speech Dispatcher as fast as it will accept and queue it. To wait for each line to be spoken, use the --wait or -w option. Example:

 $ spd-say --pipe-mode -w
 This line is spoken and processing waits for speaking to finish
 before this line is sent to Speech Dispatcher.
 ^D

You can abort a long speech narative with the --cancel or -C option.

 $ spd-say -C