Dhvani is a Text to Speech System for Indian Languages. It includes language specific synthesizer modules and an Indian voice database. The system was designed by keeping the linguistic features of Indian languages in mind.
Dhvani started as a subproject of simputer project, a low cost handhelpd PDA like device. The project was started and designed by Dr. Ramesh Hariharan, Indian Instritute of Science , Banglore in 2001. The simputer project was not successfull and stopped, But dhvani was taken by Santhosn Thottingal in 2006 and he rewrote many parts of the system and added other language modules too. Currently Santhosh Thottingal maintains the project. The dhvani project won "FOSS India award 2008" for its relavance in Indian IT field for visually challenged people.
Dhvani is written in C and it is written from scratch. It has a command line interface as well as a C/C++ API Interface. It includes a speech database and total size of the application is less than 2 MB. Pitch, Temp modification is possible. Intonation and Prosody is not available as of now. When integrated with Speech Dispatcher, it can be used in Python with python-speechd binding. It can be used a sreenreader with Orca using a generic driver too.
Instructions on how to use this software follow. Please let the authors know your comments on the software and of any improvements that you would like to suggest.