these were measured on a 45CC
the 45CC is not a full head, just two metal plates with microphones in the middle
measurements with the 45CC DO NOT show the sound as it is perceived by the human head+ear
at least at frequencies above 1 kHz. But above 1 kHz nothing noteworthy changed anyway, so I'm only plotting up to 1 kHz in this graph
the big upside of the 45CC is that it is very precise. You can put the same headphone on the 45CC twenty times and you will get twenty times the exact same result.
This isn't possible with other (more accurate) measurement setups. Accuracy often comes at the cost of precision.
WIth a "real" head like the 45BC I can measure sound just like it is perceived by a human ear, but due to the shape of the measurement rig, it's not possible to put the headphone on two times in a row and hitting the exact same position again.
With the 45CC however this is very easy, as there is exactly ONE possible position that you can put the headphone on. And you will hit that exact same position every time you put the headphone on. This means that measurement variation is negligible.
Which also means that whenever I see a difference between two measurements made on the 45CC, I can be sure that this difference is caused by the headphone itself and not by e.g. a different position of the headphone on the measurement rig.
now
what I measured here was the right earcup of the Avantone Planar that I was sent.
- I measured it in stock configuration ("take it out of the bag and put it on").
- Then I tightened all screws until they were, well, tight.
- then I losened all screws lightly (about a quarter turn with the screwdriver)
- then I losened them again by another quarter turn
- then I losened them AGAIN
- then I tightened them all as best as I could
- then I took the whole headphone apart and reassembled it
- then I tightened all screws again
after every one of the above steps, I performed a measurement (three measurements actually, and calculated the avg, although those three basically gave me the exact same result due to the afore mentioned precision of the 45CC setup).
you see that the drop-off below 100 Hz varies quite a bit, but it never reaches full linear bass response, even with all screws completely tightened.
you also see that the response from 100 Hz to 1 kHz (the midrange) shows various peaks and dips, and their spectral position ("at which frequency do they occur") changes everytime I touch the screws.
This is what makes me believe that the mechanical tension/stress onto the baffle affects magnet geometry - this could explain such behaviour.
since the seal is obviously also affected by this, I'm thinking that warping (stressing) the baffle affects the driver itself, and how it is mounted into the baffle.
If you were to mod this headphone, one thing I'd do is to strengthen the baffle, maybe replace it with a metal part instead of the plastic.
Another thing you could do is to make sure there is a clearly defined end position for every screw.
Maybe replace them with metal threading too, because the plastic appeared to get quickly worn out by the screws.