currently, when using swift TTS engine with app_cepstral, generated audio is streamed to the channel. This means that a call to ceptsral operates like app_playback. I need the functionality of app_background. I'm thinking I have two options... 1.) use system() to call swift engine, create a temporary file, background() the temp file, and then delete the temp file. option 2.) write this functionality into app_cepstral and use a flag in the call. Has anyone already solved this problem, or does anyone have another suggestion? Jason Jason