Audeze LCD-X Over Ear Open Back Headphone Review

[overkilly]

Pretty sure overkill just is not comprehending that Harman curve is to headphones as predicted in room response is to speakers, except because of ears interacting with headphones, you need to survey human beings to estimate the correct curve. Instead he is thinking it is like Beats coloration that a bunch of stupid people happen to prefer because of their subjectivity, and now he is emotionally invested in his illogic.

Friend, you have no CLUE what you're talking about. The only people emotionally invested in "illogic" are the people who refuse to read and understand what I have posted, and who have not provided any evidence that Harman 1) IS accurate 2) ever CLAIMED to be accurate. As for you, "you need to survey human beings to estimate the correct curve" - that is the height of parroting, and you refer to "stupid people [that] happen to prefer [Beats coloration] because of their subjectivity"? All the while ignoring the subjectivity of the Harman test participators? Emotionally invested in your illogic indeed.

 

[amirm]

All the while ignoring the subjectivity of the Harman test participators?

As if we have any other goals than subjectively please listeners....

 

[overkilly]

What you get out of flatness, I don't know.

Really, you have very confused notions here and you explain them in even more confusing terms. Tons of research has been performed here with respect to listener preference and none starts with your premise of "flat" and that "flat is different for everyone."

Why don't you TRY IT and find out?

And I don't know what's so confusing about recognizing that everyone has different hearing. Must be beyond so many people's abilities to acknowledge this simple fact. If I run a frequency sweep through a pair of headphones and all of the frequencies appear the same volume to ME, then YOU put them on and run a frequency sweep, the likelihood is that YOU will hear some dips and/or peaks because OUR hearing is DIFFERENT. Is that STILL too confusing?

 

[overkilly]

As if we have any other goals than subjectively please listeners....

But it only counts if it's not Beats, because they're stupid. Got it.

 

[amirm]

The only people emotionally invested in "illogic" are the people who refuse to read and understand what I have posted, and who have not provided any evidence that Harman 1) IS accurate 2) ever CLAIMED to be accurate.

I hazard a guess that you have not read any Harman research if you objected to what he wrote. From Sean Olive:

Suggest not bluffing about what you know in these circles.

But you are right that we have no idea what you are saying. As I said, it is not grounded in any reality, nor do you use any terminology that makes sense.

 

[amirm]

But it only counts if it's not Beats, because they're stupid. Got it.

Once again you talk in a way that I can't understand. What on earth do you mean???

Running with it, someone at Beats thought they were doing the right thing just like you think they have done the right thing by your ears. It is through research that we can determine if there is merit to generalizations of such approaches. And that research says what you and them did is not valid because in controlled blind tests, these approaches don't hold up.

 

[overkilly]

I hazard a guess that you have not read any Harman research if you objected to what he wrote. From Sean Olive:
Suggest not bluffing about what you know in these circles.

But you are right that we have no idea what you are saying. As I said, it is not grounded in any reality, nor do you use any terminology that makes sense.

Right, plain English seems to be beyond reality in here, apparently. As for your post above, you already know from my previous posts the definition I gave for "accurate," so you KNOW what I was talking about when I mentioned "accurate." The slides have nothing to do with what I defined, posting this stuff is just posturing on your part. I trust I don't need to go into why things like "accurately predict" has NOTHING to do with producing the same sound that would have been heard at the time of the recording? Or maybe I do, but I won't.

By the way, you DID understand that the logic followed in the slides you posted, starts with a HYPOTHESIS, right? Not proven scientific fact? And that that hypothesis only even tangentially touches on "accurate" as I defined it?

 

[DualTriode]

Once again you talk in a way that I can't understand. What on earth do you mean???

Running with it, someone at Beats thought they were doing the right thing just like you think they have done the right thing by your ears. It is through research that we can determine if there is merit to generalizations of such approaches. And that research says what you and them did is not valid because in controlled blind tests, these approaches don't hold up.

This conversation of logic:

This is a perfect example of “Chick Logic”, a world where shifting personal definitions and feelings trump rigorous technology and measurement.

This is lose lose.

Thanks DT

 

[solderdude]

If I run a frequency sweep through a pair of headphones and all of the frequencies appear the same volume to ME, then YOU put them on and run a frequency sweep, the likelihood is that YOU will hear some dips and/or peaks because OUR hearing is DIFFERENT. Is that STILL too confusing?

There are different things going on here.
1: HpTF is individual for each person but isn't so much about tonal balance being different but treble peaks differing.
2: Individual HpTF is not the same as specific HATS minus correction curve.
3: Personal preference exists but should not be in a standard.
4: Standard conditions are not ones personal conditions which creates personal 'standards'.
5: Flat can mean different things. It can mean 'smooth' (as in no wiggles), it can mean 'perceived loudness at a specific average level' it can mean equal SPL at specific measurement conditions.

The correction curve for a bunch of headphones will differ because of HpTF of the HATS when measured.
This means the Harman correction curve for their specific HATS will have to be an average.
It's why the correction curve is 'smooth' (because averaged) which creates measurement errors for all headphones.
The higher the frequency and the closer the frequencies are in the area where the HATS pinna and earcanal have an influence the bigger the error will be.
When measuring it is best to have an accepted 'standard' which to compare to.
There weren't any that were really applicable to headphones.

Flat, as in SPL over the audible range and smooth (least amount of wiggles) measuring speakers in an anechoic room used in the 'Harman listening room' which is supposed to react similar-ish to (again averaged) listening rooms (rooms where the stereo is usually setup).
The 'flat' sound is modified in the room. This is complex because SPL measurements (not gated) at the 'listening position' differs from what we hear per frequency band. Ears are not microphones.

Harman research showed that most people preferred the sound of speakers that had a smooth response (the least amount of dips and peaks) in their 'standard room' with no modifications of the sound signature (personal preference). Those speakers measured SPL-flat at 1m in anechoic conditions and had good dispersion properties. When such speakers are used in a room (with reflections at listening position) then the recordings sounded realistic. I assume at 'studio listening levels'.

When you measure a headphone's SPL and that would be flat, as in no wiggles and SPL the same for all frequencies the plot will be flat like the perfect speaker would measure at 1m in anechoic conditions on axis.
The thing is no one listens that way to speakers.
Yes, some folks do listen to speakers at 1m (nearfield desktop setups) but not anechoic and bass boost (from the room) will also be 'added' and the brain won't be able to discern low frequencies. Most will listen to speakers in a room at several meters away in furnished rooms.

This means that flat+smooth SPL measured in headphones is not tonally correct like 'flat + smooth' speakers in a room
These 2 have a different sound.

When the goal is to get the correct tonal balance recordings are designed for (flat+smooth speaker in a room) you will have to 'mimic' room effects. Here is where things go sour.
Everybodies room is different. Just like HpTF differs from person to person.

To measure anything you need to have a reference. For most things in life measurements can be fairly exact and when a standard is used we will get similar results.
Not so in everyones room and not so with HpTF, leaving preference aside.
So both HpTF and 'speaker in room' will have to be an average. There is no way around that.

Assuming at Harman they knew what they were doing and established standards (using human hearing + objective measurements) they arrived at their standard which (mostly) differs from SPL measurements in the lows and treble. Bass because of the room, treble because of speaker distance/room and averaging. It is turned into this standard when measured on a specific HATS with a specific configuration and averaging.

So now we have a standard that works in the Harman room which is supposed to mimic 'average' listening rooms with better than average speakers.
That doesn't mean that each individual speaker in each individual room at each specific listening position/distance will yield the correct 'tonal balance'.
Alas, for most people, that sound of their selected speaker in their room with their recordings and music preference will be 'their' reference.
That may wel NOT be the intended sound (optimal speaker in 'optimal' room) but your brain get's used to it.
Just like when listening to 1 headphone only we get 'used' to sound signature (and generally think it is burn-in of the headphone).
The closer the speakers in that room get to the sound sig of the headphone the easier we accept it as 'real' and thus how we perceive bass.

Now here's the thing. The serious listeners all EQ their room/speakers and thus, in general, have to EQ the lows. Mostly downwards in level.
These are usually EQ'ed to a 'gentle sloping' curve and thus deviate from 'flat+smooth speaker in room with no treatment'.
You get used to less 'room made bass boost'.
Chances thus are that audiophiles with treated rooms/room EQ will get used to and prefer less 'bass boost' than the vast majority of 'standard listeners (with room bass boost).

So... in short your (and mine) amount of bass boost differs and amount of treble may also.

Let's all live with our own 'truths' but lets measure according to a standard. Harman is such a standard. And while not perfect for everyone it is a scientifically determined standard and 'better' than FF or DF so why not use it ?

This shouldn't be in an LCD-X thread though.
Maybe these posts should go in another thread where the discussion can easily be found again for those looking for Harman things.

 

[DualTriode]

There are different things going on here.
1: HpTF is individual for each person but isn't so much about tonal balance being different but treble peaks differing.
2: Individual HpTF is not the same as specific HATS minus correction curve.
3: Personal preference exists but should not be in a standard.
4: Standard conditions are not ones personal conditions which creates personal 'standards'.
5: Flat can mean different things. It can mean 'smooth' (as in no wiggles), it can mean 'perceived loudness at a specific average level' it can mean equal SPL at specific measurement conditions.

The correction curve for a bunch of headphones will differ because of HpTF of the HATS when measured.
This means the Harman correction curve for their specific HATS will have to be an average.
It's why the correction curve is 'smooth' (because averaged) which creates measurement errors for all headphones.
The higher the frequency and the closer the frequencies are in the area where the HATS pinna and earcanal have an influence the bigger the error will be.
When measuring it is best to have an accepted 'standard' which to compare to.
There weren't any that were really applicable to headphones.

Flat, as in SPL over the audible range and smooth (least amount of wiggles) measuring speakers in an anechoic room used in the 'Harman listening room' which is supposed to react similar-ish to (again averaged) listening rooms (rooms where the stereo is usually setup).
The 'flat' sound is modified in the room. This is complex because SPL measurements (not gated) at the 'listening position' differs from what we hear per frequency band. Ears are not microphones.

Harman research showed that most people preferred the sound of speakers that had a smooth response (the least amount of dips and peaks) in their 'standard room' with no modifications of the sound signature (personal preference). Those speakers measured SPL-flat at 1m in anechoic conditions and had good dispersion properties. When such speakers are used in a room (with reflections at listening position) then the recordings sounded realistic. I assume at 'studio listening levels'.

When you measure a headphone's SPL and that would be flat, as in no wiggles and SPL the same for all frequencies the plot will be flat like the perfect speaker would measure at 1m in anechoic conditions on axis.
The thing is no one listens that way to speakers.
Yes, some folks do listen to speakers at 1m (nearfield desktop setups) but not anechoic and bass boost (from the room) will also be 'added' and the brain won't be able to discern low frequencies. Most will listen to speakers in a room at several meters away in furnished rooms.

This means that flat+smooth SPL measured in headphones is not tonally correct like 'flat + smooth' speakers in a room
These 2 have a different sound.

When the goal is to get the correct tonal balance recordings are designed for (flat+smooth speaker in a room) you will have to 'mimic' room effects. Here is where things go sour.
Everybodies room is different. Just like HpTF differs from person to person.

To measure anything you need to have a reference. For most things in life measurements can be fairly exact and when a standard is used we will get similar results.
Not so in everyones room and not so with HpTF, leaving preference aside.
So both HpTF and 'speaker in room' will have to be an average. There is no way around that.

Assuming at Harman they knew what they were doing and established standards (using human hearing + objective measurements) they arrived at their standard which (mostly) differs from SPL measurements in the lows and treble. Bass because of the room, treble because of speaker distance/room and averaging. It is turned into this standard when measured on a specific HATS with a specific configuration and averaging.

So now we have a standard that works in the Harman room which is supposed to mimic 'average' listening rooms with better than average speakers.
That doesn't mean that each individual speaker in each individual room at each specific listening position/distance will yield the correct 'tonal balance'.
Alas, for most people, that sound of their selected speaker in their room with their recordings and music preference will be 'their' reference.
That may wel NOT be the intended sound (optimal speaker in 'optimal' room) but your brain get's used to it.
Just like when listening to 1 headphone only we get 'used' to sound signature (and generally think it is burn-in of the headphone).
The closer the speakers in that room get to the sound sig of the headphone the easier we accept it as 'real' and thus how we perceive bass.

Now here's the thing. The serious listeners all EQ their room/speakers and thus, in general, have to EQ the lows. Mostly downwards in level.
These are usually EQ'ed to a 'gentle sloping' curve and thus deviate from 'flat+smooth speaker in room with no treatment'.
You get used to less 'room made bass boost'.
Chances thus are that audiophiles with treated rooms/room EQ will get used to and prefer less 'bass boost' than the vast majority of 'standard listeners (with room bass boost).

So... in short your (and mine) amount of bass boost differs and amount of treble may also.

Let's all live with our own 'truths' but lets measure according to a standard. Harman is such a standard. And while not perfect for everyone it is a scientifically determined standard and 'better' than FF or DF so why not use it ?

This shouldn't be in an LCD-X thread though.
Maybe these posts should go in another thread where the discussion can easily be found again for those looking for Harman things.

It is not about averaging; someone will think that it is about an individual average headphone user. It is not.

Harman tells us that that 1/3 octave measurement points are not enough sampling points for constructing a curve. That is 31 sampling points for 10 octaves are too few.

Harman argues that 1/12 octave sampling points or 121 samples are about right to fabricate a more useful curve.

In the world of statistics we see Multiple Regression, line and curve fitting algorithms in Excel. We see MATLAB Spline fitting routines in commercial software like APx500. We do not average headphone users.

https://www.mathworks.com/help/curvefit/smoothing-splines.html

SD,

Yes a headphone tools and methods wiki would be nice.

Thanks DT

one point at time with a link

 

[tential]

I want to see more reviews of headphones now. Less of amps. Amps should have a really high bar for review. Portable setups, speakers, and headphones are the things that are huge unknowns.

 

[amirm]

The slides have nothing to do with what I defined, posting this stuff is just posturing on your part.

Not at all. I was responding to you saying this: "The only people emotionally invested in "illogic" are the people who refuse to read and understand what I have posted, and who have not provided any evidence that Harman 1) IS accurate 2) ever CLAIMED to be accurate. "

I showed you that very "evidence" as clear as it can be. They target accuracy. So clearly you are wrong with respect to Harman research and what it means.

 

[amirm]

By the way, you DID understand that the logic followed in the slides you posted, starts with a HYPOTHESIS, right? Not proven scientific fact?

It was no hypothesis. They made a logical inference that if people like a certain fidelity in a room with speakers, that the same experience with headphones would garner the same satisfaction. They then proceeded to demonstrate that through countless experiments and research papers across many years of work. None of that follows your proposed scheme of messing with an EQ for every person individually.

And that that hypothesis only even tangentially touches on "accurate" as I defined it?

That can't be because your opening statement was that you don't agree that the reference for headphone frequency response is Harman's. So in no way or shape should you be using Harman as a reference for what you proposing (which is still clear as mud).

You have said that what you consider accurate is not what I would consider accurate which means no one can ever design a headphone for anyone. The research from Harman proves this wrong as it did for speakers. People used to think everyone likes something different in speakers so anything goes. Research from Dr. Toole while at NRC changed all of that. It demonstrate that there are objective measurements that track listener preference. Similar results ensued for headphones although due to challenges of measuring headphones, the accuracy is not as good as it is in speakers.

Bottom line in this argument is that you have sat there and messed with EQ until you liked the sound. Nothing about that is prescriptive for us or anyone else. Who knows if that is even right for you since you have not done anything proper to justify what you have. We are not here to grab any idea thrown out there as valid. We use what we can prove and demonstrate to be right.

 

[overkilly]

Bottom line in this argument is that you have sat there and messed with EQ until you liked the sound.

Okay, I'm just going to come out and say it: the level of intellectual dishonesty is ASTOUNDING here. You either have made absolutely NO effort to read or understand the method I followed and mentioned, or are so invested in Harman you can't pull your head out of whatever Harman paper it's stuck in. The simplicity is such that I can't believe you would be ignorant enough NOT to understand it: there IS a methodology as I have mentioned and referenced before, and the point of it is to make sure all frequencies sound the same volume to the LISTENER, therefore eliminating to the greatest extent possible, colorations which are the product of everything between the recording, to the headphones FR, to the listener's own hearing. That is, using EQ to compensate for the output through the headphones and the listener's ear and brain. Preference plays no part, because removing FR dips and peaks by using EQ while running frequency sweeps has nothing to do with musical enjoyment or preference.

By the way, I have seen other members allude to and provide graphs of such EQ compensation in this very forum, so attempts to say that what I state in plain English is strange or incomprehensible looks like plain gaslighting to me. See https://www.audiosciencereview.com/forum/index.php?threads/what-headphone-s-do-you-own.4285/page-49 for example.

I invited you and others to try it for yourselves, but it seems here what I inspired is some pro-Harman crusade from the posters here, with no honest attempt to explore things they obviously do not know. It makes no difference to me, except for severely diminished amount of respect I can garner for people with such contemptuous attitudes. Perhaps it's time to change the site name to audiodogmaticreview.com, then we can all know where we stand outright.

 

[amirm]

I am going to take you back to your first post:

Sorry but I just don't agree with setting the Harman (or any other) curve as the "correct" sound. Every individual's hearing and preferences are different. As far as an ideal headphone, I think the best one could hope for would be a flat FR curve. Then one could more easily adjust with EQ to one's own preference.

You say every individual's hearing and preferences are different. This is the opposite of what research shows. Indeed we are far more alike than different. What is the ratio of people who like chocolate to those who don't? It is a myth that we all have different preference. Go to audio shows and you hear the same demo tracks. Why do audiophiles like the same recordings this way?

I can quote you ton of research on this that goes back decades. Without this we would be lost in the woods with no idea how to build any sound reproduction system. It would all be a crapshoot but fortunately it is not. This is why I am doing speaker and headphone testing. We know the basics of what makes good sound.

Second you say a headphone has to have a "flat FR curve." A curve is a measurement. You have show no such curve that I can use post measurements of a headphone to relate the data to. This has been the problem with your argument from start. Using your ears to mess with EQ of the headphone is not a curve. A curve is a published response that on can compare a response to. This is why I say you are using improper explanation for what you are doing.

And yes, EQ can and should be used to generate a target curve to taste. Decide how much bass and treble you need relative to each other. That is very different animal than seeing flaws in narrower spectrum and correcting them with EQ. Such correction based on measurements routinely provides excellent results for most listeners.

Instead of telling us to try your scheme, my suggestion is that you try our scheme. And that is, you are in a science forum so learn the science behind what we do and why. If you don't have access to the research, don't pretend that you do and ask us for help. We are happy to show you the relevant pieces.

Finally, we are not married to Harman research. But it is a compass that guides us. Deviations are fine as long as they are justified by more than waving one's hands with no training, qualifications or study of the field. You would be just as dismissive if you were in our shoes. Heck, you dismiss science with ease yet demand that we listen to ad-hoc suggestions. There is nothing logical about this.

 

[RickSanchez]

3: Personal preference exists but should not be in a standard.

^^ This.

 

[BARDUKE]

I looked for the solution to my issue, but all the threads were from years ago.

I have the LCD-X and really enjoy them, however I am tempted to send them back. I bought them a few weeks ago and recently I have been experiencing pretty bad discomfort, so much that I am considering sending them back. Apparently my ears are rubbing against the fazors and I feel like there is sandpaper rubbing against my ears. I contacted Audeze support and they were quick to answer, asking me if I had the old pads or new pads. I have the new style pads that do not accept the spacers they used to send out. I emailed them back and asked what their suggestions would be, haven't received an answer yet.

I want to ask what other people have done to resolve this issue. I have seen some putting vinyl/rubber and putting them under the pads from the inside of the cups. A lot have said good things about Dekoni pads, but I don't want to have the same issue or for the pads to compress over time and I have the same issue again.

Any suggestions or replies would be appreciated. Thanks!

 

[SeaNNyT]

How noticeable is the complete loss of nearly -15dB in the 4khz? How does it change the sound relative to other headphones? if you were blind testing two headphones would it stick out like a sore thumb?

 

[Helicopter]

How noticeable is the complete loss of nearly -15dB in the 4khz? How does it change the sound relative to other headphones? if you were blind testing two headphones would it stick out like a sore thumb?

Compared to Focal Clears, yes, the 4kHz dip in LCDX sticks out like a sore thumb. The one good thing about LCDX without EQ, to me, is the lack of distortion, but against something with low distortion and better FR (Clear), the LCDX sounds totally inferior to me. Music is less enjoyable and realistic.

I did not know about the problems with FR when I formed my opinion either, and I really wanted to prefer the American headphone with novel planar technology, so LCDX probably had the edge for bias going in. At the time I formed my opinion about the sound, I incorrectly attributed the differences to driver type rather than FR.

 

[jcadduono]

thanks to reading this thread while using my LCD-X i decided to give mathaudio headphone eq a go and turned on the sweep and tried to listen for these dips. huh. there is a 6 dB dip in my left channel, and a 2 dB dip in my right channel at 4k. a LOT of volume mismatch going on between 3.8k and 10k.
i had the right driver replaced about a couple months ago after it failed, and they said they replaced both with matching drivers.
doesn't seem to be very matching.
i feel like it would take like 14 different filters to try and volume match the drivers. grrrrrrrrrr. from 4.5k to 8k i swear they are doing some kind of double helix dance
just listening to the right driver sweep there's not a whole lot going on in volume variances, but listening to the left driver sweep it's all over the damn place once you get past 3.7k but returns to normal past 8k.
i must now go try other planars i have laying around!