I don't really like headroom's method of normalizing waveforms. That's why the measurements I did recently take a different approach. Headroom obviously matches volume at a 1 Khz, disregarding any dips or bumps a particular headphone might have in that area (relative to other frequencies). With my measurements, I normalized the recorded waveforms to a uniform average RMS (Root Mean Square) over the entire spectrum. In other words: the average volume of each recorded file is the same, not only the 1 KHz output is the same.
My conviction is that this allows for better comparability even though it still doesn't say everything about how headphones really sound compared to each other. Psychoacoustics and HRTFs are hard to show in graphs

My method has other shortcomings but at least I know that the results mostly conform to what I'm hearing.
http://www.head-fi.org/forums/f4/hea...phones-388711/
I'm also wondering why so many AT in the new headroom measurements have severely rolled-off deep bass. My measurements show something totally different (my AT are pad modded, but still) and my hearing certainly confirms that they're not
that bass-light.
AT comparison
The Balanced K701 measurement seems odd to me. Can balancing turn them upside down in terms of Frequency Response? I doubt it. Something must be wrong there. Oh and notice the 1 KHz volume level doesn't even match...
K701 / HD650
This is really interesting by the way:
AD300 / HD650
AD300 and HD650 sounding somewhat similar? Haven't seen that coming

But I'm suspicious given the K701 measurements.