It's the expression of dynamic range, how much difference there is between soft and loud, particularly as it deals with the leading edges of notes, the transients. This manifests as bass slam as well as punchiness and liveliness throughout the frequency range, like how notes "pop" in the music. This is also referred to as "crest factor" in music production. Dan Clark has mentioned this in interviews as an aspect of his design and tuning. How much a headphone overshoots the crest factor in transients is effectively how dynamic it is. Dan found that most headphones overshoot the transients present in the signal to varying degrees. His headphones (at least during the Aeon 2 and Stealth time frame) were designed to not overshoot at all, hence the huge amount of damping in the designs. If a signal has a 6 dB crest on the leading edge of a note, then a DCA headphone will only give you a 6 dB increase in SPL, no more. Thus it will only provide that sense of punch and slam if the music itself is mastered with huge dynamics.
You can give any headphone more slam and dynamics by using dynamic EQ. The issue with parametric EQ is that any filter affects all levels of the signal equally. Both soft and loud parts are boosted in equal measure, so the difference between them is the same no matter how much you boost. Dynamic EQ is nonlinear, so it will look at the signal and adjust how much to boost or cut depending on what's happening in the signal. Hence, it can act as a dynamic expander and increase the difference between soft and loud by boosting only when transients occur. This has a transformative effect on DCA headphones in my experience (mainly with the Aeon 2 Noire). All of the bluntedness and constriction that I heard is lessened by expanding dynamics across the whole frequency range and I found it far more enjoyable as a result. But then I'm forced to use a computer for listening, as dynamic EQ only exists as VST/AAX plugins as far as I can tell, and I feel like I'm undoing Dan's work in tuning his headphones. Why go through all of this effort to undo what is an intentional part of the tuning?
Since dynamic EQ is nonlinear, I suspect that it will not show any difference in frequency response graphs, which measure the linear variations in SPL. I think it would show up as some form of distortion. Maybe it's encapsulated in THD and we just need to pick it out, or it might be something else like TIM (temporal intermodular distortion), I don't know. DCA headphones measure with extremely low distortion, particularly in the bass, while headphones that are known for slam, like Focal and Abyss headphones, usually measure with very high distortion in the bass.
Thus my suspicion about reports of improved dynamics. If the E3 has better dynamics, either Dan Clark has reversed/softened his position on avoiding overshoot, or he has somehow figured out a secret sauce to allow more perceived dynamics without generating actual overshoot. Or that he was wrong earlier about his headphones being accurate to the signal and they've actually been undershooting (though that doesn't make much sense as undershoot would imply insufficient bandwidth). I note that the E3 has more venting in the back of the earcups, which should allow for greater diaphragm excursion as the air behind the diaphragm can more easily escape rather than compressing, and thus should not pose as much resistance to diaphragm movement.
All of this theorizing aside, I'll give the E3 a shot. I actually quite enjoyed the sound of the Noire after I expanded its dynamics with full-spectrum DEQ and no other parametric EQ changes to its tonality, so if the E3 has more dynamics without any DEQ necessary, I might be interested. Though I'm a bit wary of its midrange tuning; I've heard on some reviews that it sounds similar to the LCD-5, and I think all of the new Audezes (LCD-5, MM-500, CRBN) are a bit too shouty for my tastes.
