About Text-to-Speech
Text-to-speech (also known as Speech Synthesis) is
great for IVR systems that need to speak people's names,
street addresses, and things that you cannot sensibly
pre-record as prompts. It’s also useful if you
use pre-recorded prompts that you expect to change regularly.
SQLphone uses SAPI (the Microsoft Speech standard)
to carry out text-to-speech. SQLphone supports SAPI
version 4 and version 5 standards.
To use text-to-speech in your IVR system, you need
a Speech Engine. If you are running Windows 2000/XP/2003,
you already have the Microsoft speech engine installed.
Just select Text-to-speech from the SQLphone Connect
menu, then select the Microsoft voice -- you are now
ready to use text-to-speech in your IVR system.
Speech Engines
The free Microsoft speech engine is not high quality,
and it speaks only US English.
You can buy high quality speech engines from the suppliers
below. Most suppliers have online demos.
| AT&T Natural Voices |
US English, UK English, French, German, Spanish |
Buy from Regsoft
Buy from Wizzard |
| Acapela |
Many languages |
Buy from Acapela
Buy from NextUp |
| Cepstral |
US English, UK English, French, German, Spanish |
Buy from Cepstral |
| Realspeak |
21 languages |
Buy from Nuance |
| NeoSpeech
|
Korean, Japanese, Chinese |
Buy from NeoSpeech
|
|