Advanced TTS voice tags work in EDDI. Havn't tried them in Voiceattack but might work their to. The milli-second pause tag is great for getting voices to sound better when reading sentances with proper pacing. You can work wonders with bad sounding phrases just by adding pauses in the right places.
Eg
{Pause(00000)}
<prosody rate="x-slow">
<prosody pitch="-50Hz"> I'm</prosody>
<prosody pitch="-30Hz"> flow</prosody>
<prosody pitch="+60Hz"> ting </prosody>
<prosody pitch="+80Hz"> in </prosody>
<prosody pitch="+60Hz"> the </prosody>
<prosody pitch="-20Hz"> most </prosody>
<prosody rate="slow">
<prosody pitch="+30Hz"> paculiar</prosody>
</prosody>
<prosody pitch="+90Hz"> way.</prosody>
<prosody pitch="-50Hz">my</prosody>
<prosody rate="slow">
<prosody pitch="-30Hz"> ship is</prosody>
<prosody pitch="-10Hz"> blown </prosody>
</prosody>
<prosody pitch="-30Hz"> to </prosody>
<prosody pitch="-50Hz"> pieces </prosody>
<prosody pitch="-20Hz">and I</prosody>
<prosody rate="slow">
<prosody pitch="+20Hz"> hope my insurance</prosody>
</prosody>
<prosody pitch="+90Hz"> pay!</prosody>
I get wierd results when altering the pitch of the voices. I wanted the voice to sing, but even when trying to use proper notes, it don't work right.
I got the tags from here.
http://www.ibm.com/support/knowledgecenter/SSMQSV_6.1.1/com.ibm.voicetools.ssml.doc/ssml_tags.html
Interesting, where does it go ?.