Hello, I am using the audio_common package and trying to create a node that performs speech recognition on captured audio. I therefore have a node subscribed to the audio/audio topic published by the audiocapturenode.cpp node that recieves arrays of bytes (uni8[] in the message) aka chunk. How should this chunk be interpreted if the sample format is S16LE (signed 16 bit little endian)? These bytes do represent the the captured audio in a raw/byte like form correct?