To correct for the non-linear response of our ears at different listening levels I assume we would need to have a level-calibrated playback system and some digital circuit applying the normalised ratio of the loudness curves at playback level vs reference recording level. At least for movies...