How would you explain the concept of audio codecs and container formats to a developer who is struggling to format their payload for Deepgram’s API?