The Centre for Speech Technology Research, The university of Edinburgh


Project Summary

The Sable Consortium

Project Details

Text-to-speech synthesizers must process text, and therefore require some knowledge of text structure. While many TTS systems allow for user control by means of ad hoc escape sequences, there remains to date no adequate and generally agreed upon system-independent standard for marking up text for the purposes of synthesis.

Sable is a recently formed consortium aimed at providing a single standard for speech synthesis markup. The consortium's principle aim is to merge the two existing proposed standard, namely STML developed by Bell Labs and Edinburgh, and JSML, developed by Sun.

The present groups that are actively involved, or who have expressed an interest in this project include:

Release of the Sable v0.2 spec

Version 0.2 of the Sable spec was released in March 1998. This is the second draft spec agreed upon by the Sable coordinators and is now available for general discussion. The primary purpose is not to present Sable as a fait accompli, but rather to interest other TTS research groups in collaborating and contributing to the development of this standard.

The draft specification for Sable version 0.2

Sample Implementations

Festival-1.4.1 contains a basic implementation of Sable 0.2 in its standard distribution.

A more detailed and documented example implementation for the Bell Labs synthesizer is available here: blsable.tgz. This implements Sable by converting the tags to synthesizer specific escape sequences. This example is a good place to start if you wish to implement a sable interpreter for another synthesis system.

The Sable Mailing List is mailing list for discussing Sable and tts markup languages in general. You can subscribe to this be mailing with the following line in the body of the message subscribe tts-markup Archives and information is also available at

Bibliography of papers related to STML and JSML include