@chocomel167 The [obvious] problem with measuring with a loose couple is the measurements are no longer objective. It's more an experiment than a measurement.
Such experiments varying coupling to the test fixture would favor driver designs that are damped (or electronically controlled) for free air, and in as such most likely sound worse in a headphone application even though they vary less with coupling to a microphone. Again, coming at it from what humans hear while listening to music with headphones, testing loosely coupled does not move the measurement bar forward. All it does is favor worse sounding headphones and sh.tty measurements, where the driver could care less about damping, or the person making the test cares less about results. Some 50 cent dynamic driver would look way more consistent uncoupled vs coupled, while a planar or estat, some of the current best sounding headphone driver designs available, would contrast far more.
If the headphone is decoupled, more explanation and information needs to be given, such as photos of the decoupling, or more specifically the circumferential area of decouple. Not only that, a measurement with proper coupling to the test fixture is necessary as a baseline, something lacking from ASR. Otherwise how would anyone know the difference? Amir knows this, yet he's talked a fair number of you into believing that what he's doing is in some way forward thinking. How does this headphone actually measure when used as intended, he doesn't know, does't care, all of the above, from his data we can't tell. In all cases, to be impartial to headphones that have complex shaped ear pads designed to more closely follow head shape, the test fixture must at least attempt to mimic the shape of a human. A flat plate test fixture would only exasperate the decouple from any real life scenario, as is the case with ASR. If his intentions were truly to accurately contrast and compare the affect of head coupling on any given headphone, he would have opted for the test fixture shaped as a head, not a flat plate (photo of both below).
In the end, from my perspective as a headphone designer, and end user of such, looking into the crystal ball shows that making measurements that vary wildly from test to test, experiment to experiment, test set-up to test set-up, subjective desires to subjective desires, will only reduce trust in such to a point where they're nothing but a joke, nothing more than an excuse for poor results and lower sound quality. You guys are most definitely being misdirected on this one, and you don't even see the train coming.