I have a voice PE and a rpi satalite. The RPI plays back responses from piper just fine but the voice PE does not. I checked the Voice Assistant Debug for both and notice the big difference is that the rpi PIPER produces a WAV file but for voice PE it produces a FLAC. If i put the link to the wav file into a browser or VLC it plays back, however if i put the link for the FLAC into VLC it errors and if i put it into a browser I get what appears to be the ASCII resprenstation of the binary data. I suspect it is an issue with my HA server not streaming the FLAC correctly but not sure how to correct that.
#Issue with voice PE and response playback
1 messages · Page 1 of 1 (latest)
Voice PE doesn't produce any files. The TTS file, created by Piper, gets converted by HA to FLAC format, so it's easier for PE to play, and more standardized.
i understand that, but that
the FLAC file produced by HA is not playable apparently. even in VLC
If I take the URL from here, prepend of course with my HA url
And use either a browser or VLC it will not play back. However if i take one from my RPI request that is a WAV it works fine.
Try restarting Piper, and make sure you have correct voice set up.
no change
pretty sure piper is fine since like i said my rpi satellite works fine. I think the issue is with HA doing the conversion from wav to flac for the Voice PE
actually i was able to get the produced FLAC file by quicktime after manually downloading it via the direct URL above. So now more confused why Voice PE is not working