Text-To-Speech on the Web

This demo is 100% clientside JavaScript. It uses eSpeak, an open source speech synthesizer, which was compiled from C++ to JavaScript using Emscripten. Source code for this demo can be found here.

Browser requirements:

Typed arrays. The eSpeak code is not portable to the extent that would be necessary to avoid using typed arrays. (It should however be possible to rewrite small bits of eSpeak to fix that.) This is present in Firefox and Chrome, but not IE, Safari or Opera.
Support for WAV audio in data URIs. eSpeak's simplest form of output is a WAV file, and the easiest way to use that is via a data URI. (It should however be possible to extract the raw audio directly by hacking eSpeak.) This is present in Firefox, Safari and Opera, but not IE or Chrome.

So currently this demo will only work in Firefox. Help is welcome regarding the workarounds mentioned above that will let it work elsewhere, we would like to develop this demo into a useful project for people to use around the web.