(I'm sorry for intruding with English here, hope it's not a prob.) Really love the measurements you're doing, but I was wondering: If you look at the px100 ii graph you can see that the volume levels are very high. I think this is close to the limit the headphone is able to produce without taking damage. The problem is that distortion increases a lot at such levels, which also distorts the accuracy of the frequency response graph. In case of the px200 ii it looks like the speakers are limited mechanically (!) and cannot produce bass at such high levels - while it might be perfectly fine at lower levels. In other words, the balance between bass, mids and treble might look quite different at more reasonable volume levels. Nobody's listening at 120 dB anyway (I hope!). Therefore, I would like to suggest to add a second FR line to the existing measurements, but with much lower volume level (94 dB on average or at 1 kHz would be nice, instead of the fixed 1 Vrms. I certainly do not listen above this level with headphones.). In fact, it would be interesting to see how the frequency response curve changes with rising volume. I wouldn't expect much change with an HD650, but with something like the px200 the change should be quite big. As a quick test, you'd just need to play a 100 Hz sine wave at 85 dB and 120 dB and plot the frequency response. Harmonic distortion should increase a lot with the px200, but also noise, because I think the small speakers approach their mechanical limits.
