I don't get high electrostatic/planar headphones?

[BrokenEnglishGuy]

Those measurements were taken with the miniDSP EARS and are not reliable. The one I posted was taken with an industry standard GRAS set-up by Jude of Head-Fi, just as the Ether Flow ones you posted were, so that allows for a direct comparison.

That's right, measurements from headphones are a bit complicated. In my opinion measurements in headphones aren't as complete and objetive as speakers.
I listened multiple version of that headphone ( 400i, 4xx ) and for example they are clearly below the hifiman he1000/arya/andanda, is clearly these headphones doesn't sound clear as he1000/ananda/arya
The he1000 can sound like a top world speakers in term of resolution, it's very hard to a speakers match these speed and crystal clear sound.

 

[Zensō]

Headphones are like putting your ears in prison.

Headphones are like blocking out the world.

 

[xykreinov]

Who cares about <0.2% THD when it costs $1200 and needs heavy EQ sound fine?. The Stax L300ltd & others all have -16db roll off starting at 70Hz, The LCDX needs EQ just to sound great since it sound wrong stock. But it always weird when they praise the bass but then get offended when balanced armatures sound the same.

Heck the HD650 & HD800S hold up fine despite having 5% THD at 35Hz?.

Not very large an issue in my eyes, as I consider parametric EQ correction mandatory in general. If you're going off of objective measurements and using software like AutoEQ to do the heavy lifting, it can only help.
Nonetheless, the bass roll off you speak of is just an issue with being highly open, not electrostatic. Stax's only closeback, the SR-4070, was actually rather bass-heavy. Plus, beyond the roll off, modern Stax tonality is actually pretty good. The L300 gets a score of 88 on the Harman preference score:

Compared to other similarly open headphones, they require rather minimal EQ to get to HOE 2018.

 

[bobbooo]

That's right, measurements from headphones are a bit complicated. In my opinion measurements in headphones aren't as complete and objetive as speakers.
I listened multiple version of that headphone ( 400i, 4xx ) and for example they are clearly below the hifiman he1000/arya/andanda, is clearly these headphones doesn't sound clear as he1000/ananda/arya
The he1000 can sound like a top world speakers in term of resolution, it's very hard to a speakers match these speed and crystal clear sound.

The 4XX's response is actually closer on average to the Ananda and HE1000 than the 400i models, particularly in the bass, and lower to mid-treble where our ears are most sensitive:

Your subjective judgement that the 4** models are 'clearly below' the Andanda and HE1000 is therefore likely heavily skewed by subconscious pricing bias, as it's simply not logically consistent with the objective data.

 

[egellings]

Maybe bass suffers with headphones because there is no tactile experience of the LF energy hitting your entire body.

 

[BrokenEnglishGuy]

The 4XX's response is actually closer on average to the Ananda and HE1000 than the 400i models, particularly in the bass, and lower to mid-treble where our ears are most sensitive:

View attachment 104179

Your subjective judgement that the 4** models are 'clearly below' the Andanda and HE1000 is therefore likely heavily skewed by subconscious pricing bias, as it's simply not logically consistent with the objective data.

Remember the measurements from headphones clearly needs a lot of job, you can't judge only by measurements in headphones, thd and fr are just insuficient, for example you can't quantify the dynamics, soundstage or the speed of the driver in m/s, etc.
In headphones you need to use your ears, in speakers the measurements have a lot more sense.
HE1000 is clearly superior headphone, is more like a e-stat headphone but planar magnetic, the 4** series just cannot match that headphone, for example if Hifiman reduce the weight of the driver the measurements will not show that.

If a headphone can have really good dynamic and at the same time be very resolving headphone, that headphone is pure gold. The 4** series just lack of dynamics to my ears, no matter what EQ you put in there..


Another example, here are the CSR FROM 400i and HE1K
he400i

HE1000

For my ears is clearly the he1000 it's much clean and resolving, but to my eyes clearly the he1000 is not better

Maybe bass suffers with headphones because there is no tactile experience of the LF energy hitting your entire body.

Headphones currently can show 20hz, especialy the planar magnetic at such low distortion that speakers can dream... a very good headphone will be very quick and clean in bass, there is some headphones that also can show a tacticle experience but also is that reason why in headphones a neutral FR doesn't sound right as speakers, headphones are more resolving and a lot les tactile, a little coloration is better than neutral in headphones, specially if the bass is clean.
I mean there is no '' room interaction '' just clean sound to your ear.

 

[Degru]

Who cares about <0.2% THD when it costs $1200 and needs heavy EQ sound fine?. The Stax L300ltd & others all have -16db roll off starting at 70Hz, The LCDX needs EQ just to sound great since it sound wrong stock. But it always weird when they praise the bass but then get offended when balanced armatures sound the same.

Heck the HD650 & HD800S hold up fine despite having 5% THD at 35Hz?.

You sound like someone who hasn't tried all these headphones.

Modern Stax roll off because they are ported. The prior generations don't roll off. Similarly, not all planars have wonky tonality, that's just an Audeze thing. Don't generalize whole product categories based on cherry-picked mistakes of a few specific models.

Balanced armatures definitely do *not* sound the same as planars or stats, and don't have nearly as much headroom for bass EQ, if they even have any.

If you try to EQ 650 to not have rolled off bass the distortion will go much higher. They definitely don't hold up to a good planar in bass, EQ or not, and it's very obvious if you actually listen to the two side by side.

 

[bobbooo]

Remember the measurements from headphones clearly needs a lot of job, you can't judge only by measurements in headphones, thd and fr are just insuficient, for example you can't quantify the dynamics, soundstage or the speed of the driver in m/s, etc.

Read this, and every other comment by Oratory1990 in that thread (he's a professional acoustic engineer who measures and helps design headphones for a living): https://www.reddit.com/r/oratory1990/comments/gbdi7v/_/fp5bpi6

 

[Ilkless]

Read this, and every other comment by Oratory1990 in that thread (he's a professional acoustic engineer who measures and helps design headphones for a living): https://www.reddit.com/r/oratory1990/comments/gbdi7v/_/fp5bpi6

I love the Dunning-Kruger I always see in replies to him. The arrogance, that their anecdotal experience and intuition is remotely equally valid to his expertise, is shocking.

 

[Degru]

I love the Dunning-Kruger I always see in replies to him. The arrogance, that their anecdotal experience and intuition is remotely equally valid to his expertise, is shocking.

On the contrary I feel like he himself is at the peak of the dunning-kruger curve, and doing his best to bring lots of people up there with him. The "FR is everything" crowd is deluding themselves into not hearing extremely obvious differences and characteristics just because some supposed authority told them they don't exist, when said authority derived those conclusions from their own biased interpretation of the specific research and information they chose to read.

The reason we don't have "measurements" for things like detail yet is because they cannot be quantified into one graph or number, and many of those perceived effects are caused by multiple different specific measured behaviors, as well as being connected to each other in various ways. You can't just put that into one graph and rank different headphones by some "detail rating" in a table; it's something that needs to be explored on a case by case basis to determine how exactly the driver is behaving when given different types of signals and what audible effects that is causing. It's definitely not easy, but better than insisting that a single graph dictates everything that we hear and dismissing everything else as placebo.

 

[Degru]

Read this, and every other comment by Oratory1990 in that thread (he's a professional acoustic engineer who measures and helps design headphones for a living): https://www.reddit.com/r/oratory1990/comments/gbdi7v/_/fp5bpi6

He claims that IR and FR show the exact same thing, when this is pretty easy to disprove. I can take the same measurement of a headphone with two different ADCs and get very different IRs even while the FR remains totally identical, as do things like decay. Plus, REW even lets you generate a minimum phase IR (basically converting FR back into IR) and compare to the actual measured IR from the headphone. I don't know enough about IR to interpret it directly or explain why this is happening, but it seems to directly contradict what oratory said there. Really starts to break down his "FR is everything" narrative IMO.

 

[Feelas]

On the contrary I feel like he himself is at the peak of the dunning-kruger curve, and doing his best to bring lots of people up there with him. The "FR is everything" crowd is deluding themselves into not hearing extremely obvious differences and characteristics just because some supposed authority told them they don't exist, when said authority derived those conclusions from their own biased interpretation of the specific research and information they chose to read.

The reason we don't have "measurements" for things like detail yet is because they cannot be quantified into one graph or number, and many of those perceived effects are caused by multiple different specific measured behaviors, as well as being connected to each other in various ways. You can't just put that into one graph and rank different headphones by some "detail rating" in a table; it's something that needs to be explored on a case by case basis to determine how exactly the driver is behaving when given different types of signals and what audible effects that is causing. It's definitely not easy, but better than insisting that a single graph dictates everything that we hear and dismissing everything else as placebo.

But we do have e.g. THD measurements and even back to Tyll Herstens & Innerfidelity times these were already measured, and somehow they actually correlated with perceived clarity & things esoterically called "black-background" & shizz. And even when Tyll posted measurements, he DID NOT try to make anyone believe that only FR matters - simply because he had contact with the actual pro-audio world, he knew that's not enough. As many shortcomings his methods used, he at least tried to assess as much as possible, and I'm heavily hurting that o1990 doesn't even try to measure anything other than Harman conformance (neither does Crinacle) - I mean, we could use more people cataloguing as much information as possible, if they have the rigs.

Obviously, anything happening on Reddit sooner or later is subject to massive hivemind, which is always disconcerting. I'd ask us all to disagree more often... Anyways, I'd rather have people down in D&K curve, at least having theories, since they might as well in the future try to get more knowledge about topics and verify what they think and whether it happens to agree with whatever science says is happening, or contributing if there's something missing in the research.

He claims that IR and FR show the exact same thing, when this is pretty easy to disprove. I can take the same measurement of a headphone with two different ADCs and get very different IRs even while the FR remains totally identical, as do things like decay. Plus, REW even lets you generate a minimum phase IR (basically converting FR back into IR) and compare to the actual measured IR from the headphone. I don't know enough about IR to interpret it directly or explain why this is happening, but it seems to directly contradict what oratory said there. Really starts to break down his "FR is everything" narrative IMO.

IR and FR show the same information for time-invariant systems, which headphones are not, which we can show, proving that non-linear distortion happens. Which can be assessed by THD+N graphs, for instance.

 

[Ilkless]

On the contrary I feel like he himself is at the peak of the dunning-kruger curve, and doing his best to bring lots of people up there with him. The "FR is everything" crowd is deluding themselves into not hearing extremely obvious differences and characteristics just because some supposed authority told them they don't exist, when said authority derived those conclusions from their own biased interpretation of the specific research and information they chose to read.

The reason we don't have "measurements" for things like detail yet is because they cannot be quantified into one graph or number, and many of those perceived effects are caused by multiple different specific measured behaviors, as well as being connected to each other in various ways. You can't just put that into one graph and rank different headphones by some "detail rating" in a table; it's something that needs to be explored on a case by case basis to determine how exactly the driver is behaving when given different types of signals and what audible effects that is causing. It's definitely not easy, but better than insisting that a single graph dictates everything that we hear and dismissing everything else as placebo.

He has the weight of peer-reviewed evidence on his side. The burden of proof is on the people seeking to falsify said evidence. And no, intuition, anecdotal evidence and conjecture is not even close to what's needed to falsify said evidence. It would take a whole lot of anti-intellectualism and mental gymnastics to think so.

, but better than insisting that a single graph dictates everything that we hear and dismissing everything else as placebo.

An entirely appropriate response considering the counter-claims made are inevitably made in conditions where it is impossible to rule out placebo, as very comprehensively set out in this peer-reviewed published literature review, under Section 3 in particular.

 

[Feelas]

He has the weight of peer-reviewed evidence on his side. The burden of proof is on the people seeking to falsify said evidence. And no, intuition, anecdotal evidence and conjecture is not even close to what's needed to falsify said evidence. It would take a whole lot of anti-intellectualism and mental gymnastics to think so.

Yet he hasn't got the entire evidence on his side, unless you mean sources different than Harman's publications on target curves, since they outright mention that the non-linear behaviour in general (I specifically checked 2013 paper on virtual headphones testing to see it) wasn't part of testing and should be investigated in the future, thus they acknowledge the problem, albeit severely downplaying it; merely because it's about getting the FR broadly right to sell lots of lots of lots & not upping the hi-end headphone market, thus THD is out of what they need, unless it'd severely kill the headphones - and it probably wouldn't. But it is indeed possible to hear & might sabotage the EQ-ing potential, as recent amir's testing would highlight. All in all - omitting anything else than FR is a massive understatement to a large field of work regarding audio gear.

Harman acknowledges the non-linear problems, Oratory just doesn't care.

 

[BrokenEnglishGuy]

Read this, and every other comment by Oratory1990 in that thread (he's a professional acoustic engineer who measures and helps design headphones for a living): https://www.reddit.com/r/oratory1990/comments/gbdi7v/_/fp5bpi6

Faaaake.
Hifiman he1000 is Currently faster because of the lighter membrane, hifiman 4** can have a little bit better damping but still can't even dream with the speed of he1000.
For me this is just fake and wrong. Also sony mdr z1r can have 20hz - 120khz FR, no one in the World think is the most fast headphone. Currently is a slow one.
Csd he1000

Csr he400i

 

[Blujackaal]

Faaaake.
Hifiman he1000 is Currently faster because of the lighter membrane, hifiman 4** can have a little bit better damping but still can't even dream with the speed of he1000.
For me this is just fake and wrong. Also sony mdr z1r can have 20hz - 100khz FR, no one in the World think is the most speed headphone Currently is a slow one.
Csd he1000
View attachment 104567
Csr he400i

View attachment 104568

Common dynamic drivers can't do heavy bass at loud volumes without deforming the centre area from not resting fast enough. BA, IEM dynamic drivers(5 ~ 9mm), planar, electrostatic don't suffer from this. It why the HD650 reaches 5 ~ 10% at 85 ~ 100db or with +8db bass boost.

 

[BrokenEnglishGuy]

Common dynamic drivers can't do heavy bass at loud volumes without deforming the centre area from not resting fast enough. BA, IEM dynamic drivers(5 ~ 9mm), planar, electrostatic don't suffer from this. It why the HD650 reaches 5 ~ 10% at 85 ~ 100db or with +8db bass boost.

there is some very insteresting in ears, for example the Empire ears zeus is currently a DD + BA + EE, the audio64 U18T use 18 BA, 8 in the bass the iem have headroom for bass but isn't tactile as EE odin for example, both are clean.
Hi end in ears can reproduce very detail over all sounding but some of them aren't too tactile.

this is odin
https://cdn.shopify.com/s/files/1/2110/6297/files/Odin.png?v=1595349649

 

[Blujackaal]

BA drivers can have great impact it, BA woofers use less stiff drivers but robust to deforming or in the Etymotic ER4S case just do a +8db low shelf at 80Hz to make the driver flex more.

 

[Degru]

BA drivers can have great impact it, BA woofers use less stiff drivers but robust to deforming or in the Etymotic ER4S case just do a +8db low shelf at 80Hz to make the driver flex more.

BAs can distort a ton if you bass boost.

 

[Degru]

Common dynamic drivers can't do heavy bass at loud volumes without deforming the centre area from not resting fast enough. BA, IEM dynamic drivers(5 ~ 9mm), planar, electrostatic don't suffer from this. It why the HD650 reaches 5 ~ 10% at 85 ~ 100db or with +8db bass boost.

They can, if designed properly. You need to allow the driver to excurse linearly as far as possible and give it a beefy enough magnet structure to control it throughout that motion. However this usually also makes said driver very bassy without significant further tuning. A decent modern example of this is the TH900. There were cans back in the 70s like koss Pro4aaa and pioneer monitor 10-ii that could also do very loud subbass at very low distortion.