OK, but how do I do that? SpeechRecognizer has only start and stop recording function and no callback for what you say, I dont see any options to change what is sent to the recognition engine or catch it before sending, if you could please explain what functions or callback can I us, there is nothing related to it in the code or the documentations