brycehowitson / ssml-prosody-library

A collection of pre-built speech synthesis settings used to convey emotion

SSML-prosody-library

A collection of pre-built speech synthesis settings used to convey emotion.

Prosody "presets" (Volume, Pitch, Rate) matched to theatrical voice descriptors, including terms like these:

• grumble
• ambivalent
• harsh
• sarcastic
• scornful
• arrogant
• impassioned
• indignant
• inquisitive
• sincere
• jovial
• solemn
• thoughtful
• elated
• threatening

Here's a little background on the project and rationale on Medium

---

Example of "Excited" Prosody

<prosody rate="fast" pitch="high" volume="loud">This voice is very excited!</prosody>

SSML can be tested using IBM's interactive demo

---

How to contribute

Any help would be much appreciated! I'm no expert in the human voice but this library may prove extremely useful to anyone developing a VUI tool.

Simply add an xml document to the repository containing the prosody tag and attributes to emulate a style of voice. Make sure to name the document with your term. For example the above lives in excited.xml.