This is on the iOS app. I don't use the autoplay mode for a slideshow. It has done this from cards from different stacks. It is doing it today as well, so nothing particular to yesterday's session.
Normally I have turned the sound down low because I'm using the app while in public, waiting on something and having time to kill. I didn't even know until yesterday that I could turn it on or off.
I took screenshots of cards that were affected and I'll take a closer look at them later, if you are not noticing this problem on your end. Thought occurred to me that maybe it has to do with the text used to create the card, like extra carriage returns after the the text or if it was text imported from Google spreadsheets vs created on the mac app (I don't create cards on the iOS app). I don't know if that would make sense or not, just a thought. Or if it has to do with whether the stack falls inside multiple subdirectories (cases)??
Well, after going through my study session today I see that the issue is triggered by moving from one card to the next quickly before the audio has completed on the initial card.
Best I can tell so far, with question/answer decks like I have, if while listening to the audio of a prompt facet, you move to the next card (the prompt facet of the next card) while on the audio is still in progress, it will voice the audio of that next card's prompt *twice*.
If you listen to a prompt and then check the answer, but then go to the next card's prompt before that answer audio has completed, it will voice that next card's prompt facet *three* times.
That said, I thought I had it happen once where an answer facet repeated its audio, but I couldn't reproduce this so I'm guessing maybe I just accidentally tapped on the audio button to cause it to repeat (perhaps I was intending to tap for a default grade and advance but tapped too high).
This issue is no big deal to me since I don't make use of it. Actually, if I were using it, it might be nice to have a setting where you could purposely have it pause and repeat the audio prompt.