Ultimate closed back (portable-ish) when using eq

[candaon]

Basically title but with some more detail: I'm looking for the best closed back in terms of detail/speed/etc. Essentially, everything besides FR (well perhaps low and high extension might count here as I found that you can't eq the lowest lows and the highest highs if the headphones are too rolled off). Ideally these cans will also be somewhat portable (not super light necessarily but at least better than 600g Audeze at least) and money is almost no object (but if a 500$ or 1000$ pair can actually compete with the top dogs then that's great).

I understand most people say FR is sort of the be all/ end all of how a headphone sounds so if you just eq everything to harman (or close enough) it should sound the same. But this has simply not been true in my experience, with some clear differences in sound between different headphones. These differences are most stark between driver types, but admittedly I don't own any high end dynamics (I have LCD2 Closed Classic, Verum One, Gold Planar GL2000, some dyi headphones with supposedly beryllium drivers from aliexpress + a bunch of IEMs). Where I found these differences the largest was when I got my first Stax a few months ago (some old lambda pro classic) and with/ without eq they just sound absolutely unbelievable compared to any of my other headphones (fyi whoever said stax don't have bass hasn't heard a good lambda eq'd to the harman target, the bass on these is far away better than anything I've heard in any planar or dynamic). What I mean by better is kind of hard to quantify precisely, perhaps I just mean distortion, as it's sort of the opposite of loose, crappy bass you would get from something like a cheap headphone (think skullcandy/beats). Maybe this is more apparent in the bass because of the level of bass I prefer (a few db north of harman) which will actually start to distort in all headphones (but not on estats). Interesting side note, I think the elevated bass is preferable for me because I'm quite sensitive to highs (I can also still hear up to 19.5k so that might play into it vs a lot of older ears who prefer a more "neutral" signature). But that also means that extension at the highest end (beyond 14k) actually makes a difference for me and I suspect this is also part of the stax magic (this extensions most likely affects the harmonics for mids/highs but I have no idea of the precise mechanism).

Anyway, this was a bit of a rant, but I'm looking to replicate as much of this sonical performance (of the eq'd stax) in a closed back, somewhat portable set of headphones. If they can be driven directly from something like a qudelix 5k, that would be even better but I'm open to trickier setups as I'm not actually going to use these running or anything that active. Think more moving them around and using them with people around potentially, but would also be nice on walks/public transport if possible.

Some current contenders I was looking at:
- DCA Aeon X closed/2 Noir, Stealth;
- Focal Celestee, Stellia;
- ZMF Verite Closed;
- Final Audio Sonorous x/viii/vi

I would say out of those the Stellia seem like a good mix of everything but maybe beat by the Stealth in outright technical ability (by how much, especially both being eq'd and specifically for the lowest sub-bass and 14k+ extension is the question for anyone who has tried both).

Would be great to hear what people think, especially if you've had a similar goal and have tried various closed back with that in mind.

 

[markanini]

The most portable circumaural closed back I've used are Creative Aurvana Live!, but IMO their sound couldn't be imporved enough with EQ. Maybe consider JBL Tune 710BT, stock response is good and they have all connectivity options for portable use.

Otherwise, if the portability requirement is slightly relaxed, probably DCA Noire. I can get the concept of selecting sets for a better post EQ performance, but I think it's more valid for open backs. With closed backs just evaluate based on the stock sound IMO.

 

[candaon]

The qudelix 5k should have enough power for a fairly large set of cans and can do PEQ. If really wanting to push the boundaries something like a topping g5 can be added for more power surely? But thank you for the contribution.

 

[charleski]

if you just eq everything to harman (or close enough) it should sound the same. But this has simply not been true in my experience,

Unfortunately, if you are EQing based on measurements of other headphones, then you aren't really matching them to the target. Measurements from oratory, for instance, will get you reasonably close to the target, but there's sufficient variation between units, pads, fitting, etc that you're always going to be unsure of where you're actually going.

Sonarworks

If you really want to match the FR of your Stax on another headphone, there's really no substitute for getting a test fixture and measuring yourself (then measuring again with the EQ applied). Since you're just making comparisons for yourself, you should be able to get away with a cheap fixture like the miniDSP Ears or a DIY flat-plate (though seal is something you have to watch carefully). You will at least be measuring the headphones you're actually listening to.

 

[paudio]

Austrian Audio Hi-X60 fold down pretty small and the cables aren't microphonic. Noire and Stealth fold as well but not as small and cables and connectors are bulkier

 

[GaryH]

If you really want to match the FR of your Stax on another headphone, there's really no substitute for getting a test fixture and measuring yourself (then measuring again with the EQ applied). Since you're just making comparisons for yourself, you should be able to get away with a cheap fixture like the miniDSP Ears or a DIY flat-plate (though seal is something you have to watch carefully). You will at least be measuring the headphones you're actually listening to.

Absolutely not. The miniDSP EARS or a even worse a flat plate are a terrible simulation of an actual ear, and this method will produce totally inaccurate results. See here:
https://www.reddit.com/r/headphones/comments/7szpqm/_/dt9pm7d

 

[charleski]

Absolutely not. The miniDSP EARS or a even worse a flat plate are a terrible simulation of an actual ear, and this method will produce totally inaccurate results. See here:
https://www.reddit.com/r/headphones/comments/7szpqm/_/dt9pm7d

No. This is why it’s essential to measure with the EQ applied and adjust the EQ so that the difference between measurements is very small. The differences in acoustic impedance (of both headphone and fixture) make it hard to develop a transfer function that can translate across variations in impedance. The solution is to avoid the need for such a function. Two headphones that measure 3dB different at a particularly frequency may produce a result 6dB different on another fixture. But if the discrepancy is 0.5dB on the first fixture, the result on the second will scale accordingly (and this scale factor will vary, which is the problem oratory ran into). If the two headphones produced identical measurements on the first fixture, they would produce identical measurements on the second (within the limits of precision). (Note that this assumes that the response variations are all minimum-phase, which will be true of the majority of cases, but possibly not all.)

Oratory had a specific goal: to develop a way to compare measurements on the miniDSP Ears to those made on more accurate rigs. But the goal in this situation is different and this allows us to bypass the problem he ran into.

 

[GaryH]

If the two headphones produced identical measurements on the first fixture, they would produce identical measurements on the second (within the limits of precision).

No, they will absolutely not, if the the two fixtures have different acoustic input impedance, which is this case here. This is because different headphones have differing acoustic output impedances that will react differently with the differing acoustic input impedances of the fixtures. This is why you must use a fixture with the same or very similar acoustic input impedance as the ear, so the interaction with the headphone's acoustic output impedance (which will be different for each one due to their differing impedance) will be an accurate simulation. EQing makes no difference here. You cannot bypass this issue.

 

[charleski]

No, they will absolutely not, if the the two fixtures have different acoustic input impedance. This is because different headphones have differing acoustic output impedances that will react differently with differing acoustic input impedances. This is why you must use a fixture with the same or very similar acoustic input impedance as the ear, so the interaction with the headphone's acoustic output impedance (which will be different for each one due to their differing impedance) will be an accurate simulation. EQing makes no difference here. You cannot bypass this issue.

The only area you'll really run into problems is with pipe resonances, which will generally produce a narrow major error (either peak or dip) in the 5-10kHz region. You shouldn't be attempting to EQ these anyway, as they're non-minimum-phase and you need to learn to identify them. If your headphones differ significantly in the positioning of driver relative to ear then you may end up with resonant errors at substantially different frequencies, which will limit the degree of matching you can achieve. But if there's one thing that's notable about oratory's findings, it's that from 20Hz-5kHz the different fixtures had a 'deviation' (not standard deviation as it's asymmetrical, so I'm not entirely sure how he calculated this) that doesn't go beyond 2dB, and for the vast majority of the range it's 1dB or less. The larger errors he got between 5-10kHz are affected by differing resonant frequencies and EQ over 10kHz is generally just left to a high-shelf filter anyway.



If you're building up a database of measurements then you need to use strictly standardised tools. But that's not the aim here (and frankly none of the databases address the issue of unit-to-unit variation in a systematised fashion, which is unsatisfactory). You won't match pipe resonances, and you shouldn't be trying to, but as long as you're clear about the limits of the approach and willing to put in the work needed to maximise the precision of your measurements you can get a very acceptable result.

 

[GaryH]

But if there's one thing that's notable about oratory's findings, it's that from 20Hz-5kHz the different fixtures had a 'deviation' (not standard deviation as it's asymmetrical, so I'm not entirely sure how he calculated this) that doesn't go beyond 2dB, and for the vast majority of the range it's 1dB or less.

Huh? There's a huge 8 dB difference between 2 and 3 kHz (where our ears are most sensitive no less as per the equal loudness contours). And these differences are an average including just two headphones, both with a similar fully open low acoustic impedance design. Other headphones could have drastically different acoustic impedance and so this error curve would look completely different. It's a lottery, which renders the method invalid and prone to gross error.

 

[markanini]

The only area you'll really run into problems is with pipe resonances, which will generally produce a narrow major error (either peak or dip) in the 5-10kHz region. You shouldn't be attempting to EQ these anyway, as they're non-minimum-phase and you need to learn to identify them. If your headphones differ significantly in the positioning of driver relative to ear then you may end up with resonant errors at substantially different frequencies, which will limit the degree of matching you can achieve. But if there's one thing that's notable about oratory's findings, it's that from 20Hz-5kHz the different fixtures had a 'deviation' (not standard deviation as it's asymmetrical, so I'm not entirely sure how he calculated this) that doesn't go beyond 2dB, and for the vast majority of the range it's 1dB or less. The larger errors he got between 5-10kHz are affected by differing resonant frequencies and EQ over 10kHz is generally just left to a high-shelf filter anyway.
View attachment 300210


If you're building up a database of measurements then you need to use strictly standardised tools. But that's not the aim here (and frankly none of the databases address the issue of unit-to-unit variation in a systematised fashion, which is unsatisfactory). You won't match pipe resonances, and you shouldn't be trying to, but as long as you're clear about the limits of the approach and willing to put in the work needed to maximise the precision of your measurements you can get a very acceptable result.

Have you though about getting a clone coupler + pinna? https://www.audiosciencereview.com/...ent-rig-on-aliexpress-anyone-bought-it.43197/

 

[Luke]

I would consider making an in-ear mic, ideally open earcanal, but blocked earcanal mics are more DIY friendly and are likely still more accurate than a 43AG or 5128 for the listener wearing the mic. You won't have a target of course and that is the biggest drawback, but if you already have a headphone you like and use as a reference, well, you now have a target (I use my Harman EQ’d HD800 as my reference).

It’s just a shame that the commercially available in-ear mics are not great. It’s quite important that they don’t protrude too far out from the ear canal entrance.

 

[JeffJ]

I'm on a similar journey, looking for the closed back with the best timbre/tonality, fine w/ EQ, for office usage. My progress so far:

Watch list:
Atrium closed
Australis
Aeon X closed
Aeon Noire

Tried and considering purchase:
Verite closed


Tried but don't like:
Stealth: strange tonality specially in the mid. A bit dull (lack dynamic?)
HD820: poor noise isolation (only tried for 10 min)
Radiance: metallic(?) timbre
Celestee: metallic(?) timbre, but better than Radiance
Meze Liric: sibilant (only tried for 10 min)
Audeze LCD-XC: too heavy
Ether CX: lack excitement and liveness
Denon 7200/9200: too thin
ATH-AP2000Ti: too harsh / piercing
Sundara closed: poor noise isolation (only tried for 10 min)
Maxwell: poor resolution. very small soundstage, closed in

(I also just placed order on a set of estat headphone haha)

 

[markanini]

Closed backs I liked:
Shure SRH 440
AKG K371

I liked less:
Shure SRH 840
Beyedynamic DT250-250
Beyerdynamic DT250-80
Sony MDR7506

Want to try:
JBL Tune 710
JBL N700NC2
Beyer MMX300 or DT770-32with velour pads
DCA Aeon closed X/RT
DCA Aeon closed Noire/2 with perforated pads