It would be great to get the audio volume from users to be relevant to the how far or near your are to the user. At the moment, if everyone speaks you would hear everybody even if you are far or near to that user.