Tamas GFeb 25
New Android TGSpeechBox build (ha, beta 8 or beta 3, depending on how you see it.) This one adds a lot of small things: Punctuation look-ahead, ability to choose a speech rate and make it global in engine settings, pitch mode selection comes to Android. So, solid update. To those who had the debug APK from earlier you might need to uninstall it to get this one installed, but good news: Signing keys will not change from this release to the Playstore version, so eventually your app will just get updated over Playstore as it rolls out there.
https://eurpod.com/synths/tgspeechbox-v300b8.apk
Show threadSalih Kunduz
@Tamasg Guess I've missed a lot of buzz here. Congratulations. What I'm curious is that how do you add new languages?
Feb 25 at 6:25amTweeseCake
Show threadTamas GFeb 25
@salihkn so new languages are a bit tricky, but not impossible. Easiest entry point to test them is the NVDA driver honestly, or SAPI. With SAPI you can drop the language YAML file in and the engine picks that up as a selectable voice token when you next restart or SAPI does so. NVDA driver means adding that language and Espeak voice-code (usually the same) in constants.py, adding the YAML to the packs\lang folder, and modifying it. You can always start from default.yaml as a base, change normalization rules or settings. The Phoneme Editor (Windows for now, but actively exploring this to go cross-platform) lets you load that pack file and edit rules in an easier way.