- Add support for models with multiple subvoices
- Add all different qualities of voice models
- Add automatic voice data generation script (thanks to @lainedfles)
- Persist download managers through list updates (Fixes #15)
- Fix output detection when sox is available (Fixes #14 and #21)