I am trying out https://cloud.google.com/speech-to-text/ but every attempt has failed. I have a couple of .amr files with recorded phone calls, that I want to convert to text. Uploading them directly fails with the following error message:
We could not decode your audio. Try a different file.
So I converted files to AAC / MP3, but it still fails. Now with a different error message:
Invalid recognition 'config': bad sample rate hertz.
I have tried a few different sample rates, but I always get the same error. What sample rate is supported? The page doesn't give any hints about supported formats or other details.
Googling this exact phrase didn't return any (!) results.