I can imagine that—if the music has more extended frequency response than the dialogue, as it inevitably will, then full range speakers will play the music at a relatively louder volume as compared to the voice—speakers that would be more centered on the vocal range with less extended frequency response would have relatively louder voice playback as it would attenuate the lower and higher frequency parts of the music. Makes sense to me. I think a center channel speaker with a surround sound receiver could possibly help—some of the available surround modes are meant to emphasize dialogue during TV watching, e.g., “TV-Logic” on my receiver (though as I look at the manual online that requires a 5.1 configuration on my receiver). Also there are settings to reduce or broaden stereo width that you can play around with with a center channel speaker and stereo speakers. I can also turn up the relative volume of the center channel speaker or separately tweak the EQ on it. Although as I just sit here and flip through the modes I have the best luck with just plain Dolby 5.1 surround. The headphone situation seems like a puzzler. I don’t know how you’d work around that.