I was reading this great guide on doom9 on Dolby encoding and Dialog Normalization (found here http://forum.doom9.org/showthread.php?t=56020), but as well written as it was, it still left me with a few unanswered questions. The guide stresses on finding the RMS level of the selected audio for the proper Dialog Normalization setting. It also states to only find the RMS level of a portion of the audio representative of dialog (no music or sound effects). My problem is that the parts of my audio that have dialog have music in the background. How am I supposed to find the RMS level if the only parts that have dialog are mixed with music? Would I just select the portion where music is least audible with the dialog?