This release focuses on speed, and getting the unspoken sounds to play as close to the associated speech as possible. To do that, we've stopped using NVWave, and begun managing the sound device ourselves through OpenAl. If you experience issues with device switching or audio ducking, please file an issue.
On top of a massive speed increase, we've also:
- introduced support for adding themes (place them in the folder under the addon)
- introduced support for localization (though no translations exist yet...submit a pull request!)
- fixed the default volume of the sounds being too low
- done some back-end work to make future updates easier
- removed and simplified multiple settings that were confusing users, and not working as intended
This release wouldn't have happened without @akj and his laser-like focus on getting unspoken-ng to be as fast as possible.
