As far as speech mark-up languages go, ssml seems to be the one that is being currently used. Speech-dispatcher documentation mentions it in the section about making output modules. I think if you were to make use of ssml from speech-dispatcher, it would need to come from a specific module, the documentation seems to say this, and it seems to indicate that there is a file with the basic parts of the module written, so that a developer will only have to add synth specific parts. More information about speech-dispatcher output modules for developers is in the speech-dispatcher documentation chapter 4.2 and the use of the generic module configuration file is explained in chapter 2. If you would like to see some examples of what a mark-up language can do with speech, I could send you a sample file from the ETI eloquence synth I have in windows, and the sound it produces (as mp3. ogg vorbis, etc). Before anyone says eloquence does not use ssml, it might not, but it will give an idea of what mark-up can do. From Michael whapples
