Zaylli Lyrö. These on-ear headphones have promising measurements.

[harvey.zhx]

Cool, so it’s going to be a Bluetooth headband with battery that plugs into the drive units?, sounds ideal.

You can think of it that way — it does connect directly to the drive units, but the functionality will be a bit more interesting.

 

[harvey.zhx]

Agreed there... for lower listening levels a bit more 'warmth' and 'presence' (is what I call the 2-6kHz region) does not heard.

I prefer a little bit more 'presence' myself (as if listening to near-field monitors) but don't like it 'peaking' there and prefer a 'wider' and more gentle 'bump' if it has to be there.
Especially when listening at studio ref levels (80-85dB average so to speak) in that case a light depression there (lightly reduced 'brightness') is more desirable at those levels.
It also turns out that most people seem to prefer some mild recession in that area rather than peaking as it is known that the test fixtures 'over-report' that area compared to human perception and peaks are more detrimental to SQ than mild recessions (dips). BBC dip effect in speakers is an example.

Are there time domain plots available (like CSD) ? and is that peak visible as ringing ?
Some ringing in that part of the frequency range is not problematic (as the ear 'rings' there too and the brain ignores it) but should not be too long.

Anyways... I like the modular design, simple acoustic tuning for the bass, low weight, easy (consumer) serviceable at least when parts are available for > 10 years.
The acoustic bass tuning is often only seen in closed headphones and can actually work there to compensate for seal issues which is a major issue in closed headphones.

I'd like to see you succeed in your efforts.

A low weight also could mean low clamping force which would be good news for people wearing glasses.

Another desirable, but not easy to achieve, issue is cable microphony. Is there attention for this issue ?
Would be solved wireless (in an active headband).
Nothing as annoying (to me) when a cable is microphonic.

Another (overlooked and often not even recognized) issue is the cable being 4 wire all the way up to the TRS jack. This is not feasible with mic/remote anyway.
How is the cable that comes with the Lyro ? 3-wire from split down or 4-wire all the way to the plug ?

Another question .... how is the impedance curve? Are there plots available for that ?

Totally agree with you here. Especially in the presence region, a sharp peak (and a noticeable spike around ~5 kHz) is something many people simply can’t tolerate — I’ve heard similar feedback from listeners who are sensitive to 5–6 kHz peaks.


On time-domain plots: I’m posting a CSD/waterfall plot here from our R&D stage. It wasn’t captured in a perfectly silent environment, so you may notice some low-frequency “hangover” patterns, but it still shows what matters: there’s no obvious high-frequency ringing tied to that area.

Regarding cable microphony: I agree it can be extremely annoying. We tested multiple cable options and made optimizations for handling noise. The jacket material is intentionally very soft while remaining stable over long-term use, and the 3.5 mm cable uses a tightly constructed coax-style geometry.


And good catch on the wiring: for the cable with a mic/remote, even with shielding, there are typically additional conductors allocated for the mic/controls. I’ll confirm the exact conductor layout with the team so I don’t state anything incorrectly.

 

[Ropeburn]

While it does indeed look promising, and also very interesting because that's the usual price range for headphones I buy, the exaggerated peak (+5dB!) way beyond Harman curve is an absolute no-go. What good are headphones for what's already good money when they're unbearable without EQ? A personal sensitivity exactly there 3-4kHz, which I've heard isn't all that rare either, certainly doesn't help.

Weird design choice. A slight dip there relative to Harman would probably work much better for many people. I don't get how sharp peaks like that can be acceptable.

 

[harvey.zhx]

While it does indeed look promising, and also very interesting because that's the usual price range for headphones I buy, the exaggerated peak (+5dB!) way beyond Harman curve is an absolute no-go. What good are headphones for what's already good money when they're unbearable without EQ? A personal sensitivity exactly there 3-4kHz, which I've heard isn't all that rare either, certainly doesn't help.

Weird design choice. A slight dip there relative to Harman would probably work much better for many people. I don't get how sharp peaks like that can be acceptable.

Haha, I get the concern. Just to clarify first: we haven’t observed anything like a consistent “+5 dB above Harman” on our side. On-ear headphones are extremely sensitive to placement and clamp/seal — even small shifts can noticeably move the 3–6 kHz region — so some spread between different rigs and setups is unfortunately common.

We’re also looking forward to Amirm’s measurements once he receives the unit, since that will be a very useful independent reference. In the meantime, external community impressions (especially comments about tonal balance and whether something feels sharp or fatiguing) are very helpful for us to iterate, and internally we rely on multi-listener blind comparisons as the main anchor for final tuning decisions. For on-ears, measurements are best treated as guidance, and their “precision” in this form factor really benefits from more blind-test data and more samples — in a sense, this is where on-ears sit at the edge of what the standard Harman targets were originally designed to cover.
One more measurement detail from our side: in our lab we typically use aged-to-stable, serial-numbered pad sets for comparative measurements. Because our pads are highly breathable and quite compliant, their shape and conformity can continue to settle during the first period of use, which can shift coupling (especially in the bass and presence region). As a result, the response may only converge toward our “stable/ideal” curve after some use.
Similarly, when pads are brand new, measurements usually need a clamping/pressing condition closer to real wearing force to get closer to our lab-repeatable results. This is another reason why on-ear measurements can vary more across rigs, samples, and pad states.

 

[Ropeburn]

Haha, I get the concern. Just to clarify first: we haven’t observed anything like a consistent “+5 dB above Harman” on our side. On-ear headphones are extremely sensitive to placement and clamp/seal — even small shifts can noticeably move the 3–6 kHz region — so some spread between different rigs and setups is unfortunately common.
View attachment 505641

View attachment 505645We’re also looking forward to Amirm’s measurements once he receives the unit, since that will be a very useful independent reference. In the meantime, external community impressions (especially comments about tonal balance and whether something feels sharp or fatiguing) are very helpful for us to iterate, and

Thank you for the detailed reply. It's certainly appreciated. Very interesting to hear about the care you are taking during the design process, particularly this:

internally we rely on multi-listener blind comparisons as the main anchor for final tuning decisions.

Very good.

Of course, about the +5dB peak, I was referring to this, from the original post:

That Harman peak there is something my personal ears have a problem with due to a natural sensitivity further emphasised by old hearing damage, and I'm not the only one with that problem. It's a theoretical curve that naturally doesn't fit everyone's ears, and that peak I see there, even beyond the target curve, would make it literally painful and unusable for me.

Not sure how universally valid that measurement is; if you say it's way less pronounced or even absent in your own data, then that sounds good.

I must admit I overlooked a crucial detail: these are on-ear, not over-ear headphones. Sadly that rules them out for me because on-ears are impossible to use for me due to physical pressure pain on ears after a few minutes (wearing glasses only makes it worse). So I'm sad to inform you I definitely won't be a future customer for that reason alone.

But still, very interesting. I'll follow the development, and maybe recommend it to friends as I see fit, who don't have that problem. I appreciate the technically detailed reply, and I'm sure others will too. I wish you the best, and good effort!

 

[Somafunk]

I must admit I overlooked a crucial detail: these are on-ear, not over-ear headphones. Sadly that rules them out for me because on-ears are impossible to use for me due to physical pressure pain on ears after a few minutes (wearing glasses only makes it worse). So I'm sad to inform you I definitely won't be a future customer for that reason alone.

Not all on ears produce discomfort

 

[berjine]

Not all on ears produce discomfort

I wanna add that I especially like on-ears with glasses because I can set the arms of my glasses just above my ears/earpads, so that there is actually no pression at all compared to over-ears

 

[solderdude]

I wanna add that I especially like on-ears with glasses because I can set the arms of my glasses just above my ears/earpads, so that there is actually no pression at all compared to over-ears

That's how I do that too.
Clamping force is also a thing though.
Given the light weight clamping force could be low which also helps with discomfort.

Instead of only having the option to use regular or neck headband one could explore ear-hooks... then again... not everyone is totally happy with those too.
An advantage is that the angle the ear has does not matter anymore (swivel wise).

On time-domain plots: I’m posting a CSD/waterfall plot here from our R&D stage.

That plot only shows the decay of the filters.
It would be more interesting to set the CSD plots with more appropriate settings for measuring above 500Hz.
Try: 5ms window, 0.2ms risetime, 50us interval and 50dB range.
Graph limit 500Hz -20kHz and use CSD mode.

Note for thread followers... the plot below is an example and NOT a plot of the Lyro but just an example to show the settings in REW in this case.

Still seems like a promising product.

Is there an impedance plot ?
This also can show if there are resonances or that it is just an acoustical event.

 

[harvey.zhx]

Haha — that’s exactly how I do it too.
And I’m genuinely happy you asked these questions in such detail… it’s rare to run into someone who’s this interested in headphone measurements and the on‑ear edge cases.

1. Clamping force & comfort (LYRÖ)​

You’re absolutely right: clamping force matters, and with a lightweight on‑ear it can be a major comfort win (especially for glasses wearers).


On Zaylli LYRÖ, besides the very soft & breathable pads, the clamp is intentionally low.
The clamp force below is measured at a 15 cm “opening” (pad‑to‑pad distance). (Numbers shown in N / gf are clamp force, not headphone weight.)

• HD800: ~6.17 N (≈630 gf) — not directly comparable (over‑ear), but a reference point
• AKG K420: ~3.8 N (≈390 gf)
• Koss KPH30i: ~3.68 N (≈375 gf)
  (All of the above are headphones I personally find comfortable.)
• LYRÖ: ~2.1 N (≈215 gf)

That low clamp has two downsides:

1. It won’t feel “tight.” We used a composite ergonomic approach (contact geometry + pad compliance + structure) to keep it reasonably secure. Some YouTube reviewers also show real‑use stability. If someone needs more security, a sports support / ear‑hook style accessory is the practical answer — and I agree with you: ear‑hooks can also reduce swivel/angle sensitivity.
2. Bass seal can vary and some low‑frequency leakage can happen — that’s basically the nature of on‑ears with very breathable pads + low clamp. We tuned with that real‑world behavior in mind, and there’s enough bass headroom for typical use (you can also see the overall trend in the response plot).

And there is a benefit: low clamp + breathable pads can make the presentation feel more open and can reduce some pad‑cavity standing‑wave effects that tend to show up with tighter clamp / less breathable pads.

2. Waterfall/CSD + “extreme settings” (more engineering than listening)​

Just to set expectations: the part I consider most directly related to listening is the CSD/waterfall I posted above — it suggests there isn’t an obvious “ringing” issue in perception. What I’m sharing below is more in the direction of engineering/mechanics discussion.


Because these ultra‑zoomed plots can be influenced by many factors (placement repeatability, pad settling, background noise, noise floor, etc.), most manufacturers don’t publish them. But since you’re clearly interested, let’s treat this as a small “measurement‑geek experiment” for fun.
Everything below is measured at home, purely as a personal hobby and not an official brand/company release.


Also: this is not a whitepaper. Different tools/methods can differ in absolute numbers, but for side‑by‑side comparisons on the same chain, the relative differences are consistent — that’s the only way I’m using these here.


Measurement chain (home):

• GRAS RA0402 + KB5001 pinna (SN 602464), regularly calibrated
• DAC+amp THD < 0.00007%, DR ~131 dB
• ADC THD (20–22 kHz BW) < 0.00003%, SNR ~132 dB
• Battery powered where possible; USB isolation on DAC/ADC

To keep myself honest, I also measured HD800 and my beloved STAX SR‑507 as references.


Level / alignment note:
All sweeps are level‑matched to 94 dB SPL at 500 Hz (GRAS calibration point). SPL plots use the same compensation (Harman 2015 over‑ear target on GRAS).

THD note (published spec & factory requirement):Our published LYRÖ THD figure is intentionally conservative: across a larger sample set, for each frequency “node,” we keep the highest (worst‑case) THD. Factory QC is conservative too: LYRÖ @ 94 dB SPL must be < 0.05% THD.


With the sweep set to 94 dB SPL at 500 Hz, in this measurement I got roughly:

• HD800: THD < 0.061%
• LYRÖ: THD < 0.026%
• SR‑507: around 0.0030%, and dipping to about 0.002% around 1 kHz (it’s wavy, but still extremely impressive)

What I envy most about electrostats is the low‑frequency distortion behavior. And SR‑507’s LF THD should actually look even better here, because it seals very poorly on my coupler (lots of leakage). On my head it doesn’t behave like that, and one time I “cheated” with tape + foam to reduce the leak and the low end became basically ruler‑flat. It also doesn’t mate perfectly with BK5128 because the 5128 pinna is quite stiff and can lift it — but since we’re mainly looking at >500 Hz high‑frequency waterfalls here, those seal quirks shouldn’t affect the main takeaway too much.
The main goal of these measurements is simply to confirm that the headphones are “healthy” and behaving normally, broadly in line with what we see across most samples, and to give me a system-level reference baseline for comparisons.

For the CSD/waterfalls: I’ll post the comparisons using the “fast” settings you suggested, and I’ll also include a more extreme view with “Normalise to peak at each frequency” enabled, shown in 3D. I’ll set it to a 100 dB range/scale — this is a method I personally use to help spot material- or acoustic-related issues (e.g., hidden resonances / stored energy).
HD800

SR‑507

LYRÖ

Other plots (impedance, impulse response, etc.):
Some independent reviewers have already published very comprehensive measurements online — if you’re interested, searching the product name (Zaylli LYRÖ) should bring them up. Otherwise, you can also wait and see if an official set gets released later.
Either way, what I posted here is only my personal measurement share.
One last note: in my personal experience, these kinds of “extreme-scale” measurements don’t map tightly or one-to-one to the immediate listening impression. I see them more as engineering/reference comparison material than something to draw direct tonal conclusions from.

 

[solderdude]

• HD800: ~6.17 N (≈630 gf) — not directly comparable (over‑ear), but a reference point
• Koss KPH30i: ~3.68 N (≈375 gf)
  (All of the above are headphones I personally find comfortable.)
• LYRÖ: ~2.1 N (≈215 gf)

My HD800 measures 2-2.5N (13-17cm) 6N is clamping like a vise and my HD800 even moves when I tilt my head forward.
I measured the KPH30i at 1.5N.

2N is low and a good value for on-ears. the lowest one I encountered was the Snowski Wind (0.5N) but the thing only weighed 46 gram so stays on the head.

Question:
In your SPL plot which setting was the bass adjustment in ... halfway ?

I really like the bass adjustment setting especially for an open headphone as all the other passive bass adjustments are closed headphones and all they do is change the port to the outside world.
The HD 630VB is different though and does it using a variable high shelf (see below).

The upside is that there is one easy to adjust large control on one side that controls both sides.
This is a downside for passive controls as one must set both sides to the same setting.

That headphone has other issues though such as poor driver matching, not only bass is affected and the headphone is very seal dependent. Also the pads (nor headband padding) are not available as spare parts any more (10 years after release).

I always wondered why manufacturers did not make use of passive tuning of the driver damping itself. It could be the change in distortion is a possible reason and figuring out a method to make it 'logarithmic' in nature.
Having played around with tuning drivers of headphone in the bass showed that it is possible but sometimes requires minuscule changes.


CSD looks great.
The impedance plot suggests a resonance at 1.5kHz but it is just barely visible in the FR and not so much in CSD.

Easy to use on higher output impedance sources. The bass level would still vary depending on the setting but as that is always done by ear it does not matter.

I found that there is quite a substantial difference in recordings when it comes to bass levels.
The HD630VB single sides easy adjustment is a great tool for that.

The Lyro looks like a well thought design that is likely to sound good to a lot of people.
While not really Harman bass (who came up with those bass filters ?) this headphone can easily be adjusted to preference which others don't.

Like to see you try with a comfortable over ear.

One last note: in my personal experience, these kinds of “extreme-scale” measurements don’t map tightly or one-to-one to the immediate listening impression. I see them more as engineering/reference comparison material than something to draw direct tonal conclusions from.

Absolutely.
Measurements can show issues with a design but whether or not they are serious enough to be audible is a totally different matter.

 

[harvey.zhx]

Thank you so much for doing that clamp‑force re-check — having someone independently re-measure and compare results across different setups is really important. I honestly underestimated how tricky this test can be for other headphones… I spent about 3–4 hours today trying to find what went wrong on my side (and I’m not exactly “free” either, since I’m in the middle of preparing LYRÖ’s first showcase event ). I re-ran the test 2–3 times and the results were very consistent, so it clearly wasn’t random.

1 Clamp force — my footnote mistake + corrected 15 cm data (apologies)​

My clamp rig is a pressure-style setup: one side is held by a spare GRAS 711 fixture frame, and on the other side I mount a remote digital scale, which I tare/zero every time. In principle, this avoids some issues of handheld pull-force gauges (especially for on‑ears), because it doesn’t “ignore” the pinna side and it captures some of the small instabilities that happen in real wear. (Also, pull gauges are honestly inconvenient.)

But I finally realized the key problem: about two years ago, when I was testing LYRÖ at extreme clamp, I changed my “opening” from 15 cm to roughly 18–20 cm. LYRÖ is not very sensitive to that change, so it didn’t ring any alarms. For other headphones, clamp can change a lot with opening — and I mistakenly wrote 15 cm in the footnote out of habit, because our factory requirement is defined at 15 cm. Given my relationship with Zaylli, I understand how that could look misleading, so I want to apologize for the incorrect footnote.

Here are the re-tested clamp values at a true 15 cm opening

• HD800: 351 gf ≈ 3.44 N
  Head/slider position changes it noticeably — I saw up to 380 gf ≈ 3.73 N between smallest vs. largest setting.
  I later checked third‑party measurements and they’re in the same ballpark
  
  
• AKG K420: 330 gf ≈ 3.24 N
  Close to my earlier number; the steel band has a large elastic travel so it’s more stable.
• Koss KPH30i: 289.7 gf ≈ 2.84 N
  About ~100 gf lower than what I previously wrote — plastic behaves differently than metal and is more sensitive to measurement conditions.
• LYRÖ: 208 gf ≈ 2.04 N
  Almost unchanged, thanks to its very long spring‑steel flex range (this is exactly why I tried so hard to fit such a long steel band into such a small headphone).

Factory spec / QC (15 cm opening):
Our pass requirement is 200 gf ± 30 gf, i.e. 1.96 N ± 0.29 N.

One more takeaway from this: clamp force alone doesn’t fully predict comfort. Head-shape adaptability matters a lot. Some headphones don’t measure “high” but still feel like a clip, while others measure higher yet feel comfortable because the spring steel has a long, smooth compliance range across different head sizes.

2 Bass adjustment setting (SPL plot)​

On your question: the SPL plot I posted was measured in the Zaylli setting (maximum bass / minimum treble — the warmest setting). Under the DF setting, the headphone sounds noticeably “faster”, but around the mid/treble region the difference is relatively small; most of the change is in the low end (and you can also see it reflected in the impedance behavior). In practice, I don’t expect many people to use the fully DF-oriented setting all the time, so most of our development work is based on the Zaylli setting.

3 HD630VB, passive tuning, and driver damping​

On the HD630VB: from what I remember, it’s a variable RC network / internal filter approach and it’s closed-back. I don’t own one myself, and personally I prefer headphones where ~1 kHz doesn’t rise above Harman or diffuse-field. And yes, it seems it has been discontinued.

On passive tuning by driver damping: I agree with you — it’s extremely difficult. As you said, it can significantly increase THD and the effect is hard to control. For the current LYRÖ we ran thousands of finite-element iterations and printed hundreds of physical prototypes.

4 Over-ear attempts (and what we’re trying to solve)​

LYRÖ is modular. I did design an over‑ear cup module and the curves look very good on both BK5128 and GRAS. The only thing “missing” is a bit of upper‑mid energy around the ~1.5 kHz impedance feature — which isn’t necessarily a bad thing, because less 1.5 kHz can increase perceived space/atmosphere.

However, in my own listening and in blind comparisons, that over‑ear module currently ends up with extremely strong treble (much more than expected from the curves). It’s basically the opposite behavior compared to on‑ear LYRÖ. I’m still hunting down the cause, but I think it’s solvable.

Longer term, we do have many concepts for a true over‑ear product, but we don’t want to make a “traditional” over‑ear that traps heat and becomes uncomfortable over long sessions. Our goal is to address the ventilation/comfort limitations of conventional over‑ears. LYRÖ is one small step in that direction.

And lastly, thank you for the kind wishes and encouragement — I really appreciate it. We’ll keep working even harder!

 

[solderdude]

In case you were considering an ear-hook....
I don't think you need a suggestion but if you do have a good look at the KSC75 ear-hook.
It is silicone with a bendable steel wire in it.
This can be bent in the correct shape and is very important.
The rigid plastic KSC35 ear-hook for instance can be very uncomfortable and kind of 'cuts' in the outer ear.

Yep HD630VB is discontinued and wasn't a success. I got it mine to sound reasonably well.
It is very handy and easy to quickly adjust the bass.

The headband construction/tooling is used in the Neumann NDH-20 and NDH-30 and the HD630 designator is kind of re-used in the HDB630.

My clamping force measurement method:
put the headphone on a scale.
hit tara button (so set weight to 0 gram).
Put one cup on the scale and pull the other cup under the table and stretch the headphone 13 to 17 cm and check the values.
This way I end up with a range in Newton as heads can be different sizes and clamping force usually differs within that range.

It gives the same values as when you hold a scale vertically and put a headphone on the side and pull it outwards on the other cup alongside a scale.

Easy to convert grams to N.
I have no idea why my method gives lower values than headphone manufacturers usually quote.

Longer term, we do have many concepts for a true over‑ear product, but we don’t want to make a “traditional” over‑ear that traps heat and becomes uncomfortable over long sessions. Our goal is to address the ventilation/comfort limitations of conventional over‑ears. LYRÖ is one small step in that direction.

Good idea, I hope the low bass won't suffer.

 

[oratory1990]

Sai did not measure the peak though. At least not nearly as high.

We get pretty much the same result actually:

Differences below 10 kHz are easily explained by a difference in position and/or unit variation (we're talking about less than 1 dB for most of the frequency range).
Above 10 kHz we see a difference that can either be unit variation (speakers tend to vary more above 10 kHz), or the effect of the coupler (unheardlab use an RA0045 coupler, I use an RA0402 Hi-Frequency coupler)

 

[solderdude]

@oratory1990

So it seems for raw data ... targets aren't the same though making it appear as though the peak is not as high.

That peak is also bass tuning dependent which is also in the same 2dB difference range.
Were these plots made with the same bass tuning (given different bass response) ?

Of course product variance could be a thing as well as placement etc.

Question... were you troubled by that peak and was it audible as much as it measured ?
This seems to be the rationale of the peak, that it is a measurement artefact caused by on-ear measurements on certain fixtures.

I looked at the other on-ears you measured and those all did not have similar peaks, which one would expect IF it were indeed pinna/coupler not being that suited for on-ears, see post #12.

 

[oratory1990]

@oratory1990
That peak is also bass tuning dependent which is also in the same 2dB difference range.
Were these plots made with the same bass tuning (given different bass response) ?

Both measurements are done with the knob set to maximum, as per unheardlab's annotation

Placement has a big effect on on-ear headphones, given that with just a minimal change of position (1mm) the coupling to the ear can change drastically. This is especially pronounced on closed on-ear headphones (the Lyrö is an open headphone), but it's still visible on the Lyrö too.

Below is my measurement result for spatial variation, that is: moving the headphone up/down/forward/back by only a couple millimeters (all positions being reasonably, nothing that wouldn't realistically happen on a human head).
I took 16 measurements in total (2 each for up/down/forward/back, and 8 measurements more or less centered)
The thick line shows the average of all positions, the shaded area shows the range within which 90% of results fell.

It's a lot of variation, yes, but that's inherent to supra-aural headphones. It's normal.

You can see how a 1-2 dB difference in the bass compared to Unheardlab's measurement results of the same bass knob position aren't troubling at all, and are easily explained by a slight difference in position.

As far as audibility of the 3 kHz peak, yes, it was audible to me.
I didn't publish those results (because they're even more prone to misinterpretation than standard head simulator results), but I did do MIRE measurements as well (microphone in real ear, open meatus), and the results generally showed the same trend.

@oratory1990
Question... were you troubled by that peak and was it audible as much as it measured ?
This seems to be the rationale of the peak, that it is a measurement artefact caused by on-ear measurements on certain fixtures.

I looked at the other on-ears you measured and those all did not have similar peaks, which one would expect IF it were indeed pinna/coupler not being that suited for on-ears, see post #12.

I'm using standardized equipment (ITU T.Rec P57 Type 3.4 ear simulators) for on-ear and over-ear headphones. They're well suited to measuring on-ear and over-ear headphones.

 

[solderdude]

Thanks for this extra info.

What can be seen is that the mids and lows vary a lot with positioning yet the 3kHz peak stays the same on the fixture.
This basically means that when listening to it, considering how humans evaluate tonality, the 3kHz peak can be anything between audible and very audible depending on the positioning.
That is all with the same bass setting.

Then there is the difference between noticing a peak at 3kHz and being bothered/annoyed by it.
In any case ... when you can hear it, it is too loud. There are solutions for it though (when bothered by it).

Perhaps for customers the right positioning and playing around with the adjustment it would be possible to get 'balanced' sound with it.

Noticeable in Koss thingies as well as DT1350/T51p as well (not the 3kHz but position dependency).
Different pinnae will also make changes, certainly in real life.

In any case I applaud all efforts of (new and old) headphone manufacturers trying to push boundaries and trying to improve them.

 

[oratory1990]

I'll try your proposed circuit when I'm back in the lab!

 

[oratory1990]

If only they would have gotten rid of that 3.5kHz peak then it would not have desperately needed EQ.

Oh wait.... there would be an easy fix for that.

View attachment 503300

Before I get to soldering, I ran a simulation to see the effect on the frequency response:

This shows the voltage going into the headphone (the LT-macro has the same impedance as the headphone, I can feed it directly with measurement results, so I don't need to simulate the headphone as a resistor but can take its actual complex impedance)

Reading the voltage and multiplying it with the measured voltage sensitivity of the headphone gives me this (simulated) result:

A bit of overcompensation at 3.5k.

Cutting the resistor in half (to 34 Ω) gives us something that's pretty spot on though:

This is with the bass knob set to the middle position (notch 4/7).

I have to say, this looks promising. Time to get to soldering I think!

 

[solderdude]

The filter design was just a 'first attempt' which I often build and test in practice.
More often than not the attenuation is adjusted a little (the resistor value) by measurements and listening tests.

Interested to read/see the results of the passive filter by lack of owning a sample.