SVOX demonstrations


Audio files synthesized by SVOX


Demos at Eurospeech 95

Audio files synthesized by a preliminary text-to-speech system using a neural-network-based model for duration control (presented at Eurospeech 95). For these sentences the output of the text analysis unit was manually corrected.

Data and file format: 16KHz, linear, 16bit, mono, big endian, no header

Data and file format: 8KHz, ulaw, 8bit, mono, big endian, with sun header


Swiss weather forecast read by SVOX

Call +41 31 934 05 50 for a forecast of the weather in Switzerland (updated 5 times daily). This application uses the SVOX system for text-to-speech conversion.
Go to the SVOX Project description.


Last modified Dec 17, 1997
riedi@tik.ee.ethz.ch