You just updated the Media Player with an Announce option to allow the sent media to interrupt what is currently playing to make an “announcement” and then return to what was playing. That function belongs in the TTS service as well.
Indeed now you have to play the tts first and then you can use the right .mp3 from the tts folder…
Ok, you’ve completely lost me. How is this a solution? What exactly are you saying I have to do?
I use a lot of the same sentences to announce people at home. When you play them trough tts they are stored in the /config/tts map
Then you can figure out which one is which sentence. At last you can use the announce option for the media player.
But this isn’t a real solution.
Just like you said
That function belongs in the TTS service as well.
I don’t use this “solution” as well, I still use the snapshot and restore service…
+1 I hope this will be seen
FWIW, I recently learned from TheFes that when you use a TTS service call, like tts.google_translate_say, it is automatically converted to a media_player.play_media service call with announce automatically set to true.
In other words, the TTS integration already (transparently) uses announce: true. Therefore the TTS integration is already using the announce option so there’s no need to “add it”.
What’s important to keep in mind is that few media_player integrations support the announce feature. Currently only the Google Home/Chromecast integration supports it.
Is that a recent update, because I never was able to do it before.
Also, I no longer use google translate, I use tts.cloud_say
+1
When playing some music, any tts.cloud_say breaks it (correct) and then stops playing all (wrong).
+1
Any update on announce: true with tts.cloud_say ?
Ok, how about TTS through piper 
the service call “speak” does not have an announce option. Yet.
Any news on this?