I have speech-to-text working fine from microphone input, but is there some way to feed a speech audio file to the speech recognition plugin?