• 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!

Dan Clark E3 Headphone Review

Rate this headphone:

  • 1. Poor (headless panther)

    Votes: 5 1.9%
  • 2. Not terrible (postman panther)

    Votes: 11 4.2%
  • 3. Fine (happy panther)

    Votes: 38 14.7%
  • 4. Great (golfing panther)

    Votes: 205 79.2%

  • Total voters
    259
Are there specific areas in the frequency response we should be looking at for open vs closed sounding?
First, good post as I too find that the effect on open backs is variable and some are good and some are great (Sennheiser HD800s falls in the latter). When correcting deficiencies in 1 to 4 or 5 kHz, I find that boosting that area significantly improves spatial effects -- in both closed and open back headphones. Conversely, they will sound closed if they are recessed in that area.

Personally I find E3 to be a bit more open sounding than Stealth which I suspect is due to having more energy in the above band. I don't find it to sound "open" per se but good enough. The really open sounding headphones have better spatial effects.
 
First, good post as I too find that the effect on open backs is variable and some are good and some are great (Sennheiser HD800s falls in the latter). When correcting deficiencies in 1 to 4 or 5 kHz, I find that boosting that area significantly improves spatial effects -- in both closed and open back headphones. Conversely, they will sound closed if they are recessed in that area.

Personally I find E3 to be a bit more open sounding than Stealth which I suspect is due to having more energy in the above band. I don't find it to sound "open" per se but good enough. The really open sounding headphones have better spatial effects.

I think this varies from person to person. For me it's the opposite with spatial effect. Also quite fatiguing when listening at higher levels.
Maybe I'm just sensitve to ear gain.
 
Probably you are both right, decreasing 1k to 3k like hifiman does does create the psicoacoustic effect of making you believe you are in a non treated room that dampens some of the high mids, whereas having a slope in 1k to 3k that matches proper pina gain, like harman target curve does, helps hear more clearly the spatial information in the recording, if there is any.
 
Probably you are both right, decreasing 1k to 3k like hifiman does does create the psicoacoustic effect of making you believe you are in a non treated room that dampens some of the high mids, whereas having a slope in 1k to 3k that matches proper pina gain, like harman target curve does, helps hear more clearly the spatial information in the recording, if there is any.
maybe I just need m-scalar:) joking aside, looking forward to test the spatial on E3.
 
I can only comment on my subjective experience, but I have found there to be a significant difference between various open sounding closed backs. I also own the Focal Stellia, which is praised as an open sounding closed back. Side by side, the Stealth sounds more open. Same for the the DCA AEON Flow closed.

Perhaps more importantly, I have found some open backs to sound more open than others. For example, the STAX SR-009 gives the impression of openness far more than the DCA Expanse. To me, that backs the idea that internal reflections are the cause. Is that not the case? I know many open backs sound closed if I cup my hands over the back of the ear cups.

Are you saying that this is not due to internal reflections or that the reflections are accounted for by the smooth flat response?

Are there specific areas in the frequency response we should be looking at for open vs closed sounding?
There are several forms of 'open-ness' in the sound.

The ability of hearing things around you
A little bass-roll-off instead of a boost and 'flat' bass response
The absence of 'cuppy' or 'hollow' tone (a peak or dip in the lower mids), this could be caused by cup resonances or other acoustic issues,.
A small bump in the 1-2kHz range gives the appearance of 'open' sound (Stax, K701etc)

When you have a flat response (so little dips and peaks) there usually is no 'typical closed' sound.
Cheaper headphones that don't have a typical closed sound are OLLO S4R(1.2) and AA Hi-X60 for instance.

A headphone could measure well with a perfect seal but could sound cuppy with the seal broken. When this is not measured you might be surprised when it sounds like that. Could well be an open headphone that sounds 'closed' as well.
 
The really open sounding headphones have better spatial effects.
Interesting, when comparing, I did not notice any "worse" spatial effects on either Stealth nor Expanse compared to HD800s. HD800s was just sounding more "open" and more pleasant to me.
 
1701677336876.png


Found this online on head-fi. It is taken from DMS's review of this headphone I believe - post says it is measured with "DMS's 4128 Rig"
(Not sure what is the policy on cross-posting of measurements from other forums, happy to oblige whatever it is)

These actually seem to present a different sound profile compared to the original measurements of this thread, does it not? If you looked at these you'd think you will need to dial in a notch filter at 300Hz, another one at 4000Hz and a sharp one at 10K. But then again, I suppose you can normalize the graph at ca 400Hz and everything will be within bounds again. Am I confused or is it the headphones.com team?
 
View attachment 331611

Found this online on head-fi. It is taken from DMS's review of this headphone I believe - post says it is measured with "DMS's 4128 Rig"
(Not sure what is the policy on cross-posting of measurements from other forums, happy to oblige whatever it is)

These actually seem to present a different sound profile compared to the original measurements of this thread, does it not? If you looked at these you'd think you will need to dial in a notch filter at 300Hz, another one at 4000Hz and a sharp one at 10K. But then again, I suppose you can normalize the graph at ca 400Hz and everything will be within bounds again. Am I confused or is it the headphones.com team?
Yeah, if they shifted the measurements all down by around 3dB then they would all fall within their shaded preference curve area apart from some areas above 10kHz - so I don't know why they're choosing to align it like that, it kind of invalidates their Preference Bounds when aligned like that. Do you mean 5128 rather than 4128? You can't really compare 5128 measurements to GRAS measurements as they're different rigs so they will show peaks and troughs in different areas.
 
Not really. I've already shown that graph before, but I don't think that it's sunk in enough on ASR what this :
View attachment 331081
DCA Stealth, blocked ear canal entrance measurements (https://danishsoundcluster.dk/wp-content/uploads/2022/11/Olive_DSD_2022.pdf)

Means for this :
View attachment 331082

And, unless Sean's unit had issues (but my little stint with DCA's Stealth, Noire and Expanse makes me doubt that), in extenso a review like this :

Basically, to put it admittedly excessively bluntly, the above review is quite misleading, and the fixture's measurements mostly irrelevant to most if not all users. The one thing the first graph above tells me, and this time with rather high confidence, is that most people won't experience in situ what the fixture tells them, and that they'll experience something quite different between them.

This is a blind spot for most review outlets at the moment, alongside sample variation or HRTF matching (but for the Stealth I think that the latter two issues are quite likely of a much lesser importance).

Interestingly I notice that in this video Dan Clark might have tried to work on that problem with the E3 :

The Noire performed quite well in terms of inter-individual variation in Harman's tests above (possibly less well in terms of average transfer though), hopefully the E3 does as well... but that needs to be tested before any consideration of target adherence comes into play.

I can understand that it's highly impractical to perform in situ measurements for each review, like Rtings does to some degree, but some proxy measurements could already provide some clues that would increase or decrease one's confidence in what the fixture's measurements say.



Applying EQ based on measurements performed on a fixture on a pair of headphones which in situ response is unpredictable and variable is like shooting in the dark.
That's one of the reasons why I love my A16 Realizer. It has in ear binaural mics which you insert, then place the headphone on, and only then does it run the measurement curve and create the eq correction curve which is custom for each individual user's head and pinnae, as well as the individual headphone. Absent that, some sort of individual manual equal loudness eq generation software for a range of frequencies seems like the only reasonable substitute. And that type of software is also needed to correct for individual variation in ear canal resonance.
 
Last edited:
That's one of the reasons why I love my A16 Realizer. It has in ear binaural mics which you insert, then place the headphone on, and only then does it run the measurement curve and create the eq correction curve which is custom for each individual user's head and pinnae, as well as the individual headphone. Absent that, some sort of individual manual equal loudness eq generation software for a range of frequencies seems like the only reasonable substitute. And that type of software is also needed to correct for individual variation in ear canal resonance.
Gave you a like for the A16 Realizer comment, I really do think that would work extremely well based on how I understand the process and also what people have said re their experiences of it. I mean it's totally creating your own target curve to your own anatomy that mimics any speakers that have been measured properly in that room during your setup procedure. To me it seems like it's only limited by how good the speakers & room are within which the A16 has been setup - I think it's fantastic. (I haven't experienced it, but I think it's fantastic!)
 
I've had a chance to spend some time with Utopia and Stealth while a friend was on vacation, through his Mjolnir 3 amplifier and ampsandsound Kenzie .... I'm on board with those saying sell everything, simplify your headphones to DCA e3. Amir loves Stealth so if he says the e3 nearly as good, I'd grab one.

Utopia is great in it's way, but not as clean a sound. e3 sounds like the perfect solution.
 
If it was half the price, I'd be ordering one now. As is, still a bit too expensive for me, but definitely a step in the right direction(and definitely something I'll be scouring the used market for in a couple years). Impressive that there really don't seem to be any compromises.
Get the Aeon 2 then. I doubt you'll notice much of a difference.
 
When the newsletter came and it said extraordinary value, I was expecting a sub-$1000 headphone. How disappointing, despite its clearly amazing performance.. Also, this basement gamer aesthetic with the contrast stitching throughout and on top of the headphones need to stop. It looks cheap and tacky. IMO.
While it's not quite my style, I think the design works exceptionally well with the Stealth and is actually quite understated compared to most of it's competitors. Focal and Meze for example.
 
View attachment 331611

Found this online on head-fi. It is taken from DMS's review of this headphone I believe - post says it is measured with "DMS's 4128 Rig"
(Not sure what is the policy on cross-posting of measurements from other forums, happy to oblige whatever it is)
I found the review DMS did for headphones.com. First, the testing is done on B&K 4128 (NOT 5128) HATS which is not compliant with Harman research. He says they have made a correction for it. Such a correction doesn't fully work for 5128 so I am doubtful that it works for 4128. But let's say it does. You need to match the measured response of the headphone to the target at low enough frequency where there are least variations. This is usually around 500 Hz (I use 425 Hz). In what you have above, the matching is done at an odd frequency of 1.1 to 1.2 kHz. That is too high. If you pull down the measured response and match it at 500 Hz, then it actually falls in their gray area.

Better yet, we can grab the measurement from DMS' review video where this is done correctly:

1701750388270.png


See how it is aligned much better with their target at around 410 Hz or so? His commentary confirms the same as far as tonality being excellent and compliant. Just like what I found (objectively and subjectively).
 
Interesting, when comparing, I did not notice any "worse" spatial effects on either Stealth nor Expanse compared to HD800s. HD800s was just sounding more "open" and more pleasant to me.
"More open" falls in the same category as "spatial effects." Anything that either separates instruments or pulls that away from your head is within spatial effects.
 
See how it is aligned much better with their target at around 410 Hz or so? His commentary confirms the same as far as tonality being excellent and compliant. Just like what I found (objectively and subjectively).

I suppose you can normalize the graph at ca 400Hz and everything will be within bounds again. Am I confused or is it the headphones.com team?

Yes, as I thought so as well. Thank you.

It is so easy to misrepresent data and cause people to reach different conclusions by simply playing around with where to normalize the graph. I'm not suggesting that is what DMS or headphones.com did, just a general observation. squik.link is great in that it allows you to play around with the normalization frequency and level. I know there are conventions around this but a bit surprised there isn't a methodology for where to normalize, because I don't think fixed frequency is working well all the time. Maybe someone can create a set of samples and train an "AI" to do it.
 
Excellent value. I would feel tempted to sell my Expanses, but I can't find anything tempting to put my money into.
 
Yes, as I thought so as well. Thank you.

It is so easy to misrepresent data and cause people to reach different conclusions by simply playing around with where to normalize the graph. I'm not suggesting that is what DMS or headphones.com did, just a general observation. squik.link is great in that it allows you to play around with the normalization frequency and level. I know there are conventions around this but a bit surprised there isn't a methodology for where to normalize, because I don't think fixed frequency is working well all the time. Maybe someone can create a set of samples and train an "AI" to do it.
I'd probably align it by eye to get most important frequency range (say 40-8000Hz) area "average aligned" with the target.
 
If I ever replace my Audeze LCD-3 it will be with something from Dan Clark.
I recommend listening at length to it, before you commit.
Audeze and DCA are more different than alike. I have the LCD-XC (closed back), and I have heard the Expanse and the Stealth, side by side with them.
That DCA signature upper-bass hump, gets up my nose! also I find the DCAs a little laid back compared to the XC, so going by Amir's observation that the E3 is a little brighter, that should please me.
At the same time, all three headphones were clear sounding beyond expectation.
DCA headphones are damn comfortable though!
Audeze not so much.
 
His commentary confirms the same as far as tonality being excellent and compliant.

But I find it courious how DMS, Resolve and Golden in their video rate the Expanse so poorly (Stealth a little better), while DMS loving E3.
Atleast they do explain why - the whole in situ perfomance vs measurements, shapes and sizes of human heads...

Still is courious that the headphone doesn't perform on 3 out of 3 people.

 
Back
Top Bottom