OK I'll have a punt at this, which may end up being way off base and I'm only referencing the plot you posted. The temporal pre- and post-masking of human hearing are reckoned to be 20 msecs and 100 msecs respectively. That would make them between 8 and 40 times longer than the main events in the plot., (2.59 msecs). Also, the simultaneous masking threshold is usually quoted as being about 60 dB. See this Wiki: http://en.m.wikipedia.org/wiki/Auditory_masking So my guess is that nothing in that plot would be audible on music, but maybe, just maybe on test signals. The peaks in treble are imo, frequency response anomalies and there's no evidence that they'd cause audible ringing. In any case, there's not too much energy at 15 kHz in recorded music. The bass anomalies are 90 dB or more down anyway. IMO a pretty clean trace, I've seen much worse, even from pretty expensive phones and magnitudes better than you'd get from a speaker in an untreated/uncorrected room. Now whether anyone would actually like these phones, (I have no idea what they are), is another question entirely.