Just out of interest, and since I had time to run by the lab this morning, I ran some brief, gain-matched/calibrated measurements on my HD800S and HD820. I'll have one of the guys that does this professionally validate the results before I share any actual plots - since I am NOT practiced at doing audio measurements, but three things stand out. The basic shape of the plots are similar to those in the linked images. By which I mean that there are dips and peaks in essentially the same places. No matter what I did, I couldn't get an excursion anywhere near as large for that 300 Hz dip, either relative to rest of their plot or to the HD800S. It's almost 20 dB in the linked graphs. In mine it's shallower and ranges from 2.4 dB to about 4.6 dB. The mid-range bump is a similar width, but it's amplitude is, similarly, much lower than the linked graphs suggest. 4 dB to about 8.5 dB, spending on what I'm comparing to. I am not saying my measurements are better, or more representative. Though I will say that they're a lot closer to what I'd expect, based on what I'm hearing, than the linked ones. Particularly compared to similar comparisons with my other closed-back cans.