• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

New IEM Threads & IEM Design Principles

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,211
Likes
1,632
I don't know about everyone else, but I have very low interest in new IEMs.

I've said many times that the only way I've had consistent (subjectively speaking) FR is by using deep-insert IEMs by Etymotic. Other IEMs, although cheap and good (or expensive and good), don't fit well and I can hear FR shift as they jiggle and move while I wear them. Don't say "try different eartips", because I have, and apart from foam-based ones like Comply, they are, IMO, a tweaker accessory trying to patch a more fundamental issue.

My main question is what aspects of IEM design are important to ensure consistent FR across many people. Outside of deep insertion (or offering custom fitting), what else should we look for?
 
  • Like
Reactions: MCH
Ultrarrelaxed FRs with very little pinna gain seem to be the new fad: Mega5est, Dusk, Supermix 4...

They seem to work for most people.

Quite funny how after all the research, people are gradually gravitating towards a flat response.

1000064867.png

1000064868.png


1000064869.png
 
There's no magic bullet, only trying different stuff until you find something that suits you. My best IEM is Truthear Red with stock silicone tips, this goes against traditional ideas and I've owned Etymotic sets to know it's better for me. Generally extra shallow IEMs like many TWS or 7Hz Zero and Truthear Hola does give people more issue. And smaller shell and bullet styles IEMs have a floaty fit which may also be a concern. Even more concerning if unvented. But I can't say this must be true for everyone. You will have to deal with some level of uncertainty, take it or leave it, even custom IEMs come with bad experiences to some people.
 
Last edited:
We don't really know what works or what doesn't, other than based on somewhat unreliable personal anecdotes, because there's no comfortable way to measure response at the eardrum.

The whole response is skewed individually based on leakage and canal shape. Same thing in headphones too, but at least in-ear microphones help.
 
My main question is what aspects of IEM design are important to ensure consistent FR across many people. Outside of deep insertion (or offering custom fitting), what else should we look for?

Look into active systems exploiting microphone(s) in the front volume. At low frequencies the feedback system from ANC headphones (and IEMs) can already lower the in situ variability caused by varying impedance or leakage. Most feedback control loops seem to be effective to control FR up to around 300-800Hz depending on the design. These "classic" feedback systems can be a double edge sword as it can make the inter-individual variation worse than passive IEMs in the mids in particular.

At higher frequencies (most severely above 1kHz or so) - and perhaps in some leakage scenarios for IEMs - the front volume microphone can't directly measure the frequency response at the eardrum (if the feedback control loop was active it would lead to errors, whether in terms of FR or ANC), so quite clever systems have to be employed to indirectly estimate the response at the eardrum and individualise the filters, whether the goal is to individualise the filters to reduce the "undesirable individual variation", to deliver a constant response at the eardrum, or perhaps more ambitiously to individualise the filters to actually introduce some form of desirable individual variation (above 1kHz same FR at the eardrum is not necessarily a desirable design goal).

The APP2 or Bose CustomTune's TWS IEMs are two such IEMs that use active systems to adjust the response above the range where typical "classic" feedback systems operate, typically up to around 3-5kHz for the APP2 or 5-7kHz or so with the Bose, but with quite a different approach and different goals. Neither of them seem capable of adjusting the response above that range, and I am not certain that we currently have the means to test on our end whether they're successful or not in achieving their design goals.
 
Last edited:
Back
Top Bottom