>
Synthesize the given Text into speech.
Engine:Voice specifies the speech engine and voice/language to use. default:en specifies the default engine and English voice.
Stream/File specifies where the sound should go. Stream specifies a sound channel, File specifies a location on the SD card; the synthesized speech is stored in WAV format.
Network: use network synthesis, if supported by the current engine/voice. This may result in better quality.
Continue Task Immediately: if selected, Tasker will continue with the next queued action while the speech is still ongoing. You must select this in order to do a Shut Up action later.
Android settings may override locale, speed and pitch. Please see the 'Always use my settings option' in Android's Text-To-Speech settings section.
If Reduce Resource Usage in Tasker Prefs / Misc is checked, the speech engine will be shutdown after each utterance, meaning the speech will always begin later for subsequent Say actions.
WikiD Instuctables is a not-for-profit venture by CommWi Projects. CommWi Projects or any person affiliated with, are not in any official way, affiliated, endorsed, sponsored or licensed by any of the software, hardware, companies or persons documented on this site. Please address any Copyright or IP concerns to myself so they may be dealt with appropriately. Content hosted on this site, unless otherwise stated, is provided under the license Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).