Dan Clark E3 Headphone Review

[Chagall]

I was bored...made a sticker

Get in stock, already!

 

[Dan Clark]

Hahaha! Awesome!

 

[Mr. Haelscheir]

Posting the step responses derived from the impulse response measurements (horizontal scale from -200 us to 5 ms) might provide a prediction of the subjective "speed" or "impact" (namely objectively showing whether or not DCA's AMTS and damping actually dulls transients) as I have lately found with measurements of my own headphones, showing my HiFiMan Arya Stealth as indeed by default being sharper than my two previous dynamic headphones, though I did also demonstrate that a headphone's step response can be sharpened through the phase corrections of minimum-phase EQ. You can find these step responses in https://www.head-fi.org/threads/stax-sr-x9000.959852/post-17850191 (post #2,738; for the sake of not duplicating that information here). I believe the sharpness or overshoot within solderdude's (if I am remembering the right person) own step response measurements in https://diyaudioheaven.wordpress.com/headphones/measurements/ plausibly correlate with subjective reports of "impact", "speed", or transient sharpness. Do note that an accurate phase response must be captured alongside the magnitude response, which requires proper sample synchronization which I guess is bound to be supplied by Amir's test setup.

 

[MayaTlab]

That should be identified as a fitment issue, not a general assessment of a headphone. What I read frequently is of latter nature.

At this point I am not certain that this is of much use given that you seem to me like the sort of person who would still be arguing that the Trojan horse wasn’t hollow long after the greeks were ravaging the city, but here’s in more details how the K701 and Stealth behaved on two subjects from Harman’s latest known study on that subject :

From a presentation given at https://pubs.aip.org/asa/jasa/artic.../Selecting-a-reference-headphone-for-3D-audio

Regardless of the seating, whether it's the left or right ears, subject SO consistently experienced less bass than subject FB with the Stealth (headphone B), while he experienced a similar amount of bass with the K701 (headphone A).

Now this slide also is a good example of some of the difficulties in performing in situ measurements. On one side subject FB seems to experience a rather sharp dip around 5-7kHz that could have been caused by the mic placement, and not necessarily representative of the coupling behaviour of these headphones. As a general rule I'd be quite careful of making direct comparison of the absolute levels past around 5kHz with that method (and if the mics didn't seal the canal well enough, below that actually, but that should be rather easy to spot for a particular reason that would be too long to explain here).

Another way of illustrating the data is to assess the relative difference between the headphones. For example by plotting the Stealth's response over the K701. To some degree this mitigates issues related to mic placement when comparing absolute levels, and with some caveats makes comparisons easier between fixtures measurements performed at DRP and measurements performed at the blocked ear canal entrance (particularly below 1kHz, and with some potential for moderate errors up to several kHz as well).

From the slide above I had already digitised a while ago the "L4" trace for all four graphs. This isn't the average of the 5 traces for the left channel, but this is just to make a point, and besides the magnitudes involved are dramatic enough that it doesn't matter I think. As I'd encourage people to not look too closely at the data past 5khz with that method, I've elected to grey it out.

I also had already digitised the measurements at DRP with a GRAS fixture, for these headphones, that were shared on twitter by Sean :

If we plot the Stealth over the K701 for all three data points, we get this :

For subject FB, the relative difference at low frequencies measured on the 711 fixture translated reasonably well to his own head, not so for subject SO.
There's also a rather interesting discrepancy in the 2-5kHz region between both human subjects and the 711 fixture. This is in the range where it would be preferable to have some idea on how the load of the ear canal affects the relative difference between the headphones, but that's something that I've frequently done in situ for my own headphones and I have never seen an error of that magnitude, so my suspicion, enlightened by the circled region on the graph below, is that this could to a good extent be an illustration of something representative of what was experienced (ie a coupling issue).

Interestingly, subject SO is Sean Olive :

https://x.com/seanolive/status/1550173465545560067?s=20

Now you could very well suggest to Sean that he should perform plastic surgery on his head so that the Stealth fit him better, or perhaps less ambitiously wear an elastic strap around his head to press the ear cups more firmly against it, but I think that it's rather easier to suggest that these headphones have a shortcoming in that regard, whether because their acoustic design is too sensitive to coupling issues to begin with, or because their physical design can't compensate for it and ensure as good a seal as it requires on enough individuals. In other words, that the capacity of a pair of headphones to deliver the target it aims at is an intrinsic characteristic, which, luckily, can be quantified - albeit quite impractically I'll admit, particularly in ASR's context.

I'm using that data simply because it includes two headphones from Dan Clark, one of which performs quite a good deal better than the other closed backs, but I'll reiterate that this is nothing new and has been known for years, and there are several articles on that subject published on AES or other journals, including from Harman. Which is why I had a good chuckle when I read that self-aggrandising nonsense :

I read all the literature

Maybe instead of literally making things up and misunderstanding the graphs that you quote, you should just... actually read it ?

Now what's quite interesting is that Dan Clark says that attention was given to how sensitive to leakage the E3 is. And given that as we can expect these days from the company, it otherwise measures really, really well, it would be quite brillant if it performed more predictably and consistently than the Stealth when worn by people.

And considering that it merely is a "fitment issue" is doing a disservice to Dan's efforts in that regard if they've proven successful, or to all of the engineers who are actively trying to make their headphones perform better for real humans and not consider that their job ends when their headphones measure nicely on a particular fixture.

 

[El_Arte]

And considering that it merely is a "fitment issue" is doing a disservice to Dan's efforts in that regard if they've proven successful, or to all of the engineers who are actively trying to make their headphones perform better for real humans and not consider that their job ends when their headphones measure nicely on a particular fixture.

Yes, it should work similarly for all humans. Analyzers and rigs are very nice as long as they tell you something reliable about that.

But, since I have found changing pads or position of cups on my head can make a dramatic difference on what I hear, I believe it’s probably impossible for a manufacturer to cover all scenarios. (not referring to E3 specifically here, just in general)

I would really like to try these E3, but I should not spend $2K on headphones at this time, as I have not made much money in months.

 

[MayaTlab]

But, since I have found changing pads or position of cups on my head can make a dramatic difference on what I hear, I believe it’s probably impossible for a manufacturer to cover all scenarios. (not referring to E3 specifically here, just in general)

I also share that sentiment to some degree, at higher frequencies in particular - at least for now. I think that there are interesting questions to ask in terms of how much positional or inter-individual variation is audible, how much can be tolerated, perhaps by habituation, until the latter can no longer be effective, either because the time interval between variation is too short, or because the magnitude is too important, and perhaps which parts of the spectrum are more crucial to maintain stable and predictable - and I am not going to pretend that I have read it all in that regard, let alone understand it .

However, the sort of inter-individual or positional variation that can be seen in some headphones is in my opinion rather unarguably beyond these questions, and there are some interesting solutions to these issues. For example, while closed headphones seem to generally struggle more so than open headphones to ensure a consistent response at lower frequencies, a side benefit of the feedback system of some ANC headphones is that it compensates for leakage, up to a certain point, and maintains the response constant across positions or individuals, up to a certain frequency (typically around 500-800Hz). Thomas_A has an interesting thread here on ASR on the subject : https://audiosciencereview.com/foru...250-ohm-bose-qc25-measured.32622/post-1164980
Rtings has long measured over-ear headphones at low frequencies on real humans and ANC headphones with a feedback system usually perform more consistently than passive closed headphones in that part of the spectrum.

As the E3 are passive and closed, DCA probably has it the hardest - I hope that they've made a successful design here, and if so, I'd prefer to see it recognised.

 

[Cuckoo Studio]

The most impressive thing is that it is a closed design.

 

[Benesyed]

At this point I am not certain that this is of much use given that you seem to me like the sort of person who would still be arguing that the Trojan horse wasn’t hollow long after the greeks were ravaging the city, but here’s in more details how the K701 and Stealth behaved on two subjects from Harman’s latest known study on that subject :

View attachment 332451
From a presentation given at https://pubs.aip.org/asa/jasa/artic.../Selecting-a-reference-headphone-for-3D-audio

Regardless of the seating, whether it's the left or right ears, subject SO consistently experienced less bass than subject FB with the Stealth (headphone B), while he experienced a similar amount of bass with the K701 (headphone A).

Now this slide also is a good example of some of the difficulties in performing in situ measurements. On one side subject FB seems to experience a rather sharp dip around 5-7kHz that could have been caused by the mic placement, and not necessarily representative of the coupling behaviour of these headphones. As a general rule I'd be quite careful of making direct comparison of the absolute levels past around 5kHz with that method (and if the mics didn't seal the canal well enough, below that actually, but that should be rather easy to spot for a particular reason that would be too long to explain here).

Another way of illustrating the data is to assess the relative difference between the headphones. For example by plotting the Stealth's response over the K701. To some degree this mitigates issues related to mic placement when comparing absolute levels, and with some caveats makes comparisons easier between fixtures measurements performed at DRP and measurements performed at the blocked ear canal entrance (particularly below 1kHz, and with some potential for moderate errors up to several kHz as well).

From the slide above I had already digitised a while ago the "L4" trace for all four graphs. This isn't the average of the 5 traces for the left channel, but this is just to make a point, and besides the magnitudes involved are dramatic enough that it doesn't matter I think. As I'd encourage people to not look too closely at the data past 5khz with that method, I've elected to grey it out.

View attachment 332467

I also had already digitised the measurements at DRP with a GRAS fixture, for these headphones, that were shared on twitter by Sean :
View attachment 332468

If we plot the Stealth over the K701 for all three data points, we get this :
View attachment 332469

For subject FB, the relative difference at low frequencies measured on the 711 fixture translated reasonably well to his own head, not so for subject SO.
There's also a rather interesting discrepancy in the 2-5kHz region between both human subjects and the 711 fixture. This is in the range where it would be preferable to have some idea on how the load of the ear canal affects the relative difference between the headphones, but that's something that I've frequently done in situ for my own headphones and I have never seen an error of that magnitude, so my suspicion, enlightened by the circled region on the graph below, is that this could to a good extent be an illustration of something representative of what was experienced (ie a coupling issue).
View attachment 332470


Interestingly, subject SO is Sean Olive :

https://x.com/seanolive/status/1550173465545560067?s=20

Now you could very well suggest to Sean that he should perform plastic surgery on his head so that the Stealth fit him better, or perhaps less ambitiously wear an elastic strap around his head to press the ear cups more firmly against it, but I think that it's rather easier to suggest that these headphones have a shortcoming in that regard, whether because their acoustic design is too sensitive to coupling issues to begin with, or because their physical design can't compensate for it and ensure as good a seal as it requires on enough individuals. In other words, that the capacity of a pair of headphones to deliver the target it aims at is an intrinsic characteristic, which, luckily, can be quantified - albeit quite impractically I'll admit, particularly in ASR's context.

I'm using that data simply because it includes two headphones from Dan Clark, one of which performs quite a good deal better than the other closed backs, but I'll reiterate that this is nothing new and has been known for years, and there are several articles on that subject published on AES or other journals, including from Harman. Which is why I had a good chuckle when I read that self-aggrandising nonsense :



Maybe instead of literally making things up and misunderstanding the graphs that you quote, you should just... actually read it ?

Now what's quite interesting is that Dan Clark says that attention was given to how sensitive to leakage the E3 is. And given that as we can expect these days from the company, it otherwise measures really, really well, it would be quite brillant if it performed more predictably and consistently than the Stealth when worn by people.

And considering that it merely is a "fitment issue" is doing a disservice to Dan's efforts in that regard if they've proven successful, or to all of the engineers who are actively trying to make their headphones perform better for real humans and not consider that their job ends when their headphones measure nicely on a particular fixture.

Yeah I've found the expanse the most sensitive to fit and positioning issues of my headphones.

If I press them against my head there's a very dramatic change compared to other headphones where the change is much more modest.

 

[srkbear]

E3 was code for "third generation ETHER" because it occupies the spot in our lineup. I was considering other names, and the idea hit me that the symmetry of E3 as a logotype seemed very appealing so we dropped the ETHER and went E3.



The metal looked awkward trying to partially recess the jack so we went with this. Just flowed better with the design...

PLEASE consider offering these and your other wares on Amazon, including cable options—I’ve been tempted so many times to pick up a Stealth or Expanse but the small number of storefronts on Amazon that carry your headphones only offer unbalanced cables! At this point these new E3s are a no-brainer for me with a four pin XLR termination—I simply prefer Amazon Prime over other dealers like headphones.com due to shipping conveniences, the number of uncensored customer reviews and the ease of returns if necessary (not that I expect any returns of your products to be necessary)

 

[Somafunk]

Dan would take a 15% cut on price if he offered them through Amazon as that is their percentage share.

 

[dougq]

Tempting, but they seem very hard to drive, I suspect that my RME ADI-2 is not powerful enough, then I wonder if the problems with the pads they had on previous models still remain.

ADI-2 was good enough for the Stealth and this requires even less power. It is plenty powerful.

 

[Mac Arthur]

I confirm my ADI 2 smoothly drives my Stealth too
No worries

 

[dougq]

Does the e3 come with the 3D printed insert like the stealth and expanse?

 

[staticV3]

Does the e3 come with the 3D printed insert like the stealth and expanse?

Try the product page: https://danclarkaudio.com/e3.html

 

[dougq]

Try the product page: https://danclarkaudio.com/e3.html

I’m asking because I don’t see any indication on the site specifying that information.

 

[oceansize]

I’m asking because I don’t see any indication on the site specifying that information.

E3 is our midrange closed-back headphone. E3 delivers a stunning level of performance and value, incorporating all the technological and ergonomic achievements of our flagship headphones, but at a more accessible price point.


Incorporating an all-new 5th generation driver and our pioneering Acoustic Metamaterial Tuning System, or AMTS, E3 delivers a true flagship performance, without the flagship cost.

It was in the second sentence of the product description, so I can understand it being difficult to find.

 

[dougq]

It was in the second sentence of the product description, so I can understand it being difficult to find.

You pointed out an answer to a question I did not have. I did not ask if it had AMTS. I was asking if it had the protective 3d printed insert. That prevented pad wear.

You guys have a choice when answering questions. you can choose to do it rudely, or at least just answer the question friendly. You must be the life at a party.
AMTS is the core of the product, why would I be asking if it has AMTS? That doesn't even make sense.

 

[staticV3]

@dougq
I'm sorry. When reading "3D printed insert like the Stealth and Expanse", I think most of use would assume you're referring to AMTS.

Perhaps @Dan Clark can answer whether the E3 has that ear pads insert?

 

[Dan Clark]

@dougq
I'm sorry. When reading "3D printed insert like the Stealth and Expanse", I think most of use would assume you're referring to AMTS.

Perhaps @Dan Clark can answer whether the E3 has that ear pads insert?

Yes, it's simply a spacer we provide to keep pad compression low when in the case or stored.

 

[quattro98]

Is the Grace m900 sufficient to drive the E3?