Introduction to SSML

The VoiceXML 2.0 specification adopted Speech Synthesis Markup Language (SSML) as the standard markup language for speech synthesis.

SSML provides developers of speech applications a standard way in which to control speech synthesis and text processing parameters. SSML enables developers to specify pronunciations, volume, pitch, speed, and so on.

The SSML standard

This section describes SSML support in IBM Text-to-Speech System. The implementation is based on the Speech Synthesis Markup Language Version 1.0, recommended by W3C on September 7, 2004, which can be found at http://www.w3.org/TR/speech-synthesis/

The IBM Text-to-Speech System implements this specification, with the following exception:

SSML parsing

The SSML processor silently ignores unsupported tags. The text contained inside an unsupported <say-as> tag is synthesized as-is; that is, only the tag is ignored.

If the syntax of the input text is not legal, the SSML processor returns and logs an error.

SSML language support

The level of SSML support is language dependent. The topic "SSML language support" provides detailed information about what is supported for each language.

Related information
SSML language support