Hello,
When I call the Google Speech API with gRPC for sync requests, only first part of the audio data is recognized. Tested witn Google sample audio file "How old is the Brooklyn Bridge", for which only "how old" is returned. Other audio data shows the same behaviour, only first 1 or 2 words are returned. In all cases, no errors are returned, and confidence factor is also returned.
Calling the Speech API via REST works fine for the same audio.
Please note I do not use Google's client libarries. Calls are made using plain http/2.
Any ideas are highly appreciated.
Best regards, Ronald
PS For streaming gRPC request, the same issue seems to be present, depending on audio buffer size. To be investigated later.