Headroom's graphs show some significant higher order distortion which I think accounts for a lot for their cold sound, even after some EQ.

http://www.headphone.com/learning-center/build-a-graph.php?graphID[0]=963&graphID[1]=573&graphID[2]=&graphID[3]=&graphType=1&buttonSelection=Compare+Headphones

I just noticed it ate my link...

Anyway, sometimes I'm a little suspicious of those distortion plots.  The HD650 for example is scary clean.  I don't remember the midrange being that clean.  Probably the cleanest on the whole graph generator.  At least it was the last time I went though them all anyway.

OTOH some other ones match what I hear really well.  The 770s and 990s have less overall distortion and the harmonics are lower order as well which matches with me liking them more than the 880s.

The AKGs don't come off so hot either but I think don't think the K601s sound as bad as that graph looks.

Is the distortion a tangible problem though? Headroom seem to have posted another graph, exactly the same labelling - assume it's another measurement of the 250 ohm...

Thanks for the reply, but I'm not entirely sure what is meant by low and high order. I understand that what they've done here is played a test tone and see what frequencies that are multiples of the test tone have been produced also, and that's called harmonic distortion, but not too sure on higher and lower. It all seems to be below 70db peak so I'd assume it's inaudible, or am I missing something?

Higher order = the little distortion spikes in the treble. 4th, 5th, 6th, etc.

Possibly inaudible b/c -70db down. Hard to say tho. I never ran my own distortion tests on them.

Exactly,

Will elaborate on the plot.
The big one ar 0dB is a pure sinewave that SHOULD have little to no distortion but in any case should not have any harmonics at least not higher than say -120dB for THESE measurements.
The HP (or amplifier if a crappy one is used, but is unlikely) will add frequencies that are multiples of the one applied.
So say a 400Hz tone is applied (which is the first order) the 2nd order harmonics (that is thus a distortion from the original tone) is 800Hz, The third one (a higher order) 1200Hz.
Add 400Hz for each order.
Non-linear distortion can also have other frequencies popping up b.t.w.
higher order harmonics are thus higher frequencies  and thus more to the right in the curve.

Thing is the lower the amplitude (-70 dB and -80 dB is already pretty low for a HP) the less audible.
Harmonics are present in any instrument and most of the natural sounds and are in most case higher than the ones generated by the HP.
Also the spectrum of music is rather wide and thus all those frequencies happening at the same time will add their own harmonics.
If there is too much of those and too high an amplitude this will affect the SQ.
a 'nice' distortion is when the amplitude gets lower at each higher order harmonic.
This seems to be the case as the FR plot shown only goes to 3kHz (and not 20kHz as in most plots)

Tube amps for instance in general have higher amplitude 2nd and 3rd order amplitudes BUT the amplitude drops fast when the frequency increases.
Most SS designs have lower amplitude harmonics (and less audible) BUT the amplitude of the highest order harmonics doesn't drop that much compared to the 2nd harmonic which is less desirable.

High amplitude harmonics generated by X CAN sound very pleasant (and thus is distortion but in practice a nice 'addition' to the signal) IF the higher harmonics get lower in amplitude fast.

This means amps with HIGH distortion figures but a fast decaying spectrum (higher order harmonics dropping fast in amplitude) may look bad but sound pleasant.
It is NOT more accurate sounding but more pleasant.

I agree about the audibility of harmonics (especially in music) when harmonics are below -80dB and lower in amplitude fast at each higher order.
Those harmonics will probably be 'masked' by higher amplitudes of harmonics present in music signals anyway.