A lot of headphones resonate around 5kHz

[Rasboo]

Like the title says. Measuring a bunch of headphones I have noticed this phenomenon. It is common between many over-ear and on ear from what i can see. I am beginning to wonder if it is perhaps a poorly documented characteristic in ear simulator couplers? I have used different coupler models (IEC 60318 variants) from different brands but the behavior is basically the same..

I have been messing around quite a lot with dampening and removing parts of the acoustic assemblies in search of the root cause of this specific resonance but nothing i have tried has really had any meaningful impact.
Note that this doesn't necessarily show up clearly on frequency response plots, especially when pinna simulators are used (since they are resonance like heck, just like our ears...). But even if the headphones are equalized to be practically "flat", this frequency is still ringing (still visible on impulse response measurements).

On the attached frequency plots I measured a handful of popular headphones in the $100-500 bracket using a 318-4 coupler with a custom attachment to reduce all "external" influences that might cause resonances (It is not my custom attachment that is ringing. Source: Trust me bro ).

Would love to hear your thoughts

 

[TurtlePaul]

5 khz has a wavelength of 2.7 inches, which corresponds to the height and width of speaker cups. For normal speakers, the room modes appear below 100 hz. For headphones, the problematic 'room modes' resonances appear at several khz because the 'room' is a couple of inches.

 

[Rasboo]

5 khz has a wavelength of 2.7 inches, which corresponds to the height and width of speaker cups. For normal speakers, the room modes appear below 100 hz. For headphones, the problematic 'room modes' resonances appear at several khz because the 'room' is a couple of inches.

That is a good hypothesis! I did an experiment where i removed the cushion from one headphone and measured, same resonance present. I removed the back cover, same resonance. So I doubt that is the reason in this case.. My whole experiment is basically about killing all modal resonances in a pair of headphones, so i have gone through a lot of possible causes

Typing out my message here actually helped me get my thoughts together a bit. I am now suspecting that what I am picking up is indeed a "feature" of the 318 couplers meant to simulate the ear canal and eardrum.
I remembered this old diagram:

So I did a "free field" measurement of one the headphones that (previously) showed a clear 5kHz ringing, and there was no detectable ringing present! (Headphone cup flat on a table with a GRAS reference microphone right in front of the cushion opening).
So right now I'm leaning towards this being either a "coupler thing" or an interaction between coupler and headphone. I have to follow this up by doing some measurements of the coupler in an anechoic chamber to see if it is inherent or stemming from an interaction between coupler and headphone.
Thanks for letting me vent my thoughts

 

[respice finem]

Could it be a "self-resonance" of the membrane?