| @@ -22,11 +22,16 @@ | |||
| ---------- | |||
| The [eSpeak Text-to-Speech](http://espeak.sourceforge.net/) program is an open | |||
| source speech synthesizer that supports over 70 languages. It uses spectral | |||
| formant synthesis by default which sounds robotic, but can be configured to | |||
| use Klatt formant synthesis or MBROLA to give it a more natural sound. | |||
| This is a mirror of the eSpeak sources with the following improvements on the | |||
| *master* branch: | |||
| 1. An autotools-based build system that supports building the phoneme data and | |||
| language dictionaries from source. | |||
| language dictionaries from source on POSIX systems. | |||
| 2. Detection of the supported audio APIs via the `configure` script — no need to | |||
| modify the build files. | |||