Audeze LCD-X Over Ear Open Back Headphone Review

[Jimbob54]

https://www.superbestaudiofriends.org/index.php?threads/harman-curve-poll-public.10130/

Maybe not universally liked by goldeneared folks.

The Harmon name is meant funny I believe Harm-on



It's not V-shaped curve. Opposite flat there is an (un-natural but preferred) bass boost and the highs are rolled off somewhat.
Similar as to what a flat measuring speaker would sound like in a room.
You will get some low frequency bass boost and treble will be rolled off a bit giving a warm full bodied sound.

Now here's the thing. When you are used to listening to speakers that may be preferred.
When you are used to listening to 'flat' monitors in near-field you may not prefer the Harman sound.

I personally find the bass boost a tiny bit too much and too steep and due to my age don't like the treble rolling off. That's preference though.

Dang it. Apologies @DualTriode

 

[MRC01]

... there is an (un-natural but preferred) bass boost ... I personally find the bass boost a tiny bit too much

I agree. I like deep extended clean bass, but I don't like it exaggerated.

... and the highs are rolled off somewhat. ...

Depends on what you mean by "highs". The Harman curve boosts the lower treble (2k - 5k) and attenuates the upper treble (above 9k). One could call it "V" shaped due to the boost in the bass and lower treble.
I find that lower treble boost to be a bit "hot" personally. Just a bit, that hump should be there but something like 2-4 dB less sounds natural and "right" to me. And, like you, I prefer not to attenuate anything above 10k.

At least, these are my preferences for music that is well recorded and sounds "natural".

 

[overkilly]

You should really read the HarmAn AES papers.

Also take a close look at some of the ASR speaker reviews, zero in on the in room power response. You will notice that the in room power response is a downward sloping line. If you want a straight line this is it.

Now if you would put on a pair of HarmAn Curve equalized headphones. This set of equalized headphones has the response closest to the in room speakers we just discussed. There is no bass or high frequency accentuation. If you want accurate this is it. If you want personal preference that may be something else.

A better question is, did *you* read and understand both the papers and my previous posts?

1) The Harman target curve comes from feedback of adjustments to bass and treble AFTER the headphones were equalized to match the flat in-room response from loudspeakers. Repeating: Bass and Treble adjustments AFTER the equalization to match the flat in-room response from loudspeakers. Bass and high frequency accentuation is EXACTLY what is happening.

2) All this talk about loudspeakers and flat response in rooms is IRRELEVANT. I am talking about headphones. I stated that the a Harman curve-tuned headphone will not be accurate and I stand by that statement. Let me define "accurate" here: producing the same sound that would have been heard at the time of the recording. If I plucked a guitar string, and the playback through the headphone was exactly the same sound, that headphone is accurate. All headphones, of course, are inaccurate in that respect due to FR characteristics; this is why I stated that a perfectly flat FR would be most desired in a headphone. Harman tuning will NOT correct a FR deficiency, just introduce an euphonic one.

3) Even if a headphone's output were perfectly flat, there is the deficiency of one's hearing - the most common being treble hearing loss as one gets older. Using PEQ to modify my headphone output to my perception of flat FR, means I am also correcting MY OWN hearing deficiency in the process, allowing me to hear different frequencies at the levels they were recorded at, therefore getting me as close as possible to actually accurate* listening. Harman does nothing for this as well. *see definition above

 

[DualTriode]

Yes I have read the HarmAn AES papers and I am ready for the Quiz.

I am sorry, you are mistaken. If the reference speakers we are talking about are equalized to a level flat Frequency Response with a gated impulse or in an anechoic chamber the in room power response is a downward sloping straight line not a level flat Frequency Response.

You are partially correct when you say that the headphone test participants adjust the bass and treble to their preference. The bass and treble is adjustable either up or down. When the test data is collected the test participants collectively prefer a headphone Frequency Response that matches the speaker in room power response. That preferred headphone frequency response is the HarmAn Curve. There is no bass or treble boost or cut, just a faithful accurate curve that is the Transfer Function.

This may help
https://www.mdpi.com/2076-3417/8/6/...ime domain, the,the headphone and the eardrum.

 

[MRC01]

So can the Harman response be differentiated from HRTF, or are they intertwined? Put differently:
HRTF is about psychoacoustics: how an individual perceives sound, which has been changed by his head, face, mouth & nasal cavities, etc. HRTF doesn't involve subjective preference. You can measure it objectively.
The Harman response relies on preference: people adjusting the sound to make it match an in-room sound. HRTF is a factor, since they are listening on headphones. But it also introduces a subjective preference factor, as they adjust the sound.

 

[DualTriode]

So can the Harman response be differentiated from HRTF, or are they intertwined? Put differently:
HRTF is about psychoacoustics: how an individual perceives sound, which has been changed by his head, face, mouth & nasal cavities, etc. HRTF doesn't involve subjective preference. You can measure it objectively.
The Harman response relies on preference: people adjusting the sound to make it match an in-room sound. HRTF is a factor, since they are listening on headphones. But it also introduces a subjective preference factor, as they adjust the sound.

Headphone Transfer Function and Head Related Transfer Function are two different concepts, they are to a degree overlapping, but largely they are distinct from each other.

More later, I am going to the beach.

 

[overkilly]

Yes I have read the HarmAn AES papers and I am ready for the Quiz.

I am sorry, you are mistaken. If the reference speakers we are talking about are equalized to a level flat Frequency Response with a gated impulse or in an anechoic chamber the in room power response is a downward sloping straight line not a level flat Frequency Response.

You are partially correct when you say that the headphone test participants adjust the bass and treble to their preference. The bass and treble is adjustable either up or down. When the test data is collected the test participants collectively prefer a headphone Frequency Response that matches the speaker in room power response. That preferred headphone frequency response is the HarmAn Curve. There is no bass or treble boost or cut, just a faithful accurate curve that is the Transfer Function.

This may help
https://www.mdpi.com/2076-3417/8/6/918/pdf#:~:text=The headphone-to-ear transfer,at the eardrum [1].&text=In the time domain, the,the headphone and the eardrum.

The only mistake is if you infer my comment is to the "latest" Harman curve. See this, since you like references: http://www.juloaudio.sk/Umiestnenie_reprosustav/History of Harman Target Curve.pdf second to last slide. And if we take the curve to be the latest curve it makes no difference to my previous comments.

I also find it funny that the abstract to your latest reference concludes by stating: "The estimation results show that HpTFs vary considerably with headphones and ear canals,which suggests that individualized compensations for HpTFs are necessary for headphones to reproduce desired sounds for different listeners. "

Why would that be the case if the Harman curve "is a curve with "no bass or treble boost or cut, just a faithful accurate curve that is the Transfer Function," I wonder?

But the funniest part is, nowhere in the papers do I find a statement saying the Harman curve produces "accurate" sound. I don't even know where people got this from. It appears a lot of parroting has been happening and not a lot of independent thought.

I stand by my previous statements, collected below and still not disproven:

1) Defining "accurate" as: producing the same sound that would have been heard at the time of the recording, a Harman curve-tuned headphone will not be accurate.

2) The Harman curve is a best fit through listener preferences, and in no way an accurate representation of sound.

I suggest those that REALLY want accurate sound, look into performing PEQ to flatten their headphone FR with respect to THEIR OWN hearing, which will result in much more accurate perception for themselves than adjusting headphones to this Harman dreck. Although, to get back on topic, as previously posted it appears the LCD-X is not the headphone to do so!

 

[solderdude]

Those that want to buy Harman Targetted headphones simply have very few to choose from.
Harman curve afficionado's can EQ almost any headphone to that target and enjoy that signature.
Those that like to depart from that signature can buy any headphone they like and EQ or not the crap out of it any which way they like.

A lot of people don't really care for the Harman curve. It is their right. It is just another target curve based on research. To some it is THE standard and everyone thinking it is not is wrong and mistaken. So be it.

You like it or you don't.
You either believe it is the correct curve or you don't.

 

[DualTriode]

Headphones amps are all analog. To do what you say they have to digitize the input, perform the convolution and then convert back to analog. Best to do that in the host computer with no cost hit, and ability to store unlimited number of convolvers.

Hello,

I thought that I replied to this post, oops.

Sometimes you want to cut the cord and go sit on the park bench down by the river. You may want put your custom equalization curve on your next gen iPod.

Thanks DT

 

[DualTriode]

The only mistake is if you infer my comment is to the "latest" Harman curve. See this, since you like references: http://www.juloaudio.sk/Umiestnenie_reprosustav/History of Harman Target Curve.pdf second to last slide. And if we take the curve to be the latest curve it makes no difference to my previous comments.

I also find it funny that the abstract to your latest reference concludes by stating: "The estimation results show that HpTFs vary considerably with headphones and ear canals,which suggests that individualized compensations for HpTFs are necessary for headphones to reproduce desired sounds for different listeners. "

Why would that be the case if the Harman curve "is a curve with "no bass or treble boost or cut, just a faithful accurate curve that is the Transfer Function," I wonder?

But the funniest part is, nowhere in the papers do I find a statement saying the Harman curve produces "accurate" sound. I don't even know where people got this from. It appears a lot of parroting has been happening and not a lot of independent thought.

I stand by my previous statements, collected below and still not disproven:

1) Defining "accurate" as: producing the same sound that would have been heard at the time of the recording, a Harman curve-tuned headphone will not be accurate.

2) The Harman curve is a best fit through listener preferences, and in no way an accurate representation of sound.

I suggest those that REALLY want accurate sound, look into performing PEQ to flatten their headphone FR with respect to THEIR OWN hearing, which will result in much more accurate perception for themselves than adjusting headphones to this Harman dreck. Although, to get back on topic, as previously posted it appears the LCD-X is not the headphone to do so!

Okay, I get it. You want to cling to your very own equalization curve. You may want to hire someone with the tools to fit you with your very own custom equalization curve. I have earlier in this thread proposed the idea of marketing custom equalization curves.

What grinds me is that you keep going back to flat headphone Frequency Response curves as being most accurate. Search your posts in this thread for the word flat.

Here is my challenge; sort your best most personally preferred headphone equalization curve. Now stick calibrated probe microphones in your ears next to your ear drums. Put on your headphones set up your analyzer and plot a Frequency Response.

That Frequency response has a 68% chance of being plus or minus 1 Standard Deviation of the HartmAn Curve. There is no chance in never of that Frequency Response being flat.

Thanks DT

 

[overkilly]

Okay, I get it. You want to cling to your very own equalization curve. You may want to hire someone with the tools to fit you with your very own custom equalization curve. I have earlier in this thread proposed the idea of marketing custom equalization curves.

What grinds me is that you keep going back to flat headphone Frequency Response curves as being most accurate. Search your posts in this thread for the word flat.

Here is my challenge; sort your best most personally preferred headphone equalization curve. Now stick calibrated probe microphones in your ears next to your ear drums. Put on your headphones set up your analyzer and plot a Frequency Response.

That Frequency response has a 68% chance of being plus or minus 1 Standard Deviation of the HartmAn Curve. There is no chance in never of that Frequency Response being flat.

Thanks DT

Actually, I'm just going to copy-paste my posts below, and YOU tell everyone: WHERE did I state flat HP FR curves are most accurate? I have not edited any of my posts, so people can see for themselves in the original posts.

I DID state:

"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."

and

"For all headphones, there will be 2 components to the sound: what the headphones output, and what my hearing reduces/amplifies. Therefore if the headphone output is perfectly flat, I need only compensate for my own hearing. I don't want extra bass and treble because the average person likes it, I want accurate sound."

and

"Let me define "accurate" here: producing the same sound that would have been heard at the time of the recording. If I plucked a guitar string, and the playback through the headphone was exactly the same sound, that headphone is accurate. All headphones, of course, are inaccurate in that respect due to FR characteristics; this is why I stated that a perfectly flat FR would be most desired in a headphone. Harman tuning will NOT correct a FR deficiency, just introduce an euphonic one.

3) Even if a headphone's output were perfectly flat, there is the deficiency of one's hearing - the most common being treble hearing loss as one gets older. Using PEQ to modify my headphone output to my perception of flat FR, means I am also correcting MY OWN hearing deficiency in the process, allowing me to hear different frequencies at the levels they were recorded at, therefore getting me as close as possible to actually accurate* listening. Harman does nothing for this as well. *see definition above"

Anyone with a shred of intellectual honesty can easily garner what I said above.

Also, that bunch of nonsense about a personally preferred headphone equalization curve is just that, nonsense. The whole point I've been making is that for ONE person using ONE headphone, they can use PEQ to perceive the FR as flat - that is, to make all frequencies appear to sound the same volume to that one listener. Obviously if I change the HP or listener, any curve previously drawn will not have the same effect because everyone's hearing is different. But, to extend your example, if you can show me ONE, just ONE case, where a person has a headphone PEQd to neutral (defined as: all frequencies appear to him as the same volume, no dips or peaks during frequency sweep) for themselves, and they follow your method and end up with a plot of the Harman curve, I'll be happy to recommend the Harman target as blindly as you have.

Now as I said, my previous posts, where I supposedly stated "flat headphone Frequency Response curves as being most accurate."

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.


Well, what I'm referring to is that Harman is just a target, made out of an average of listener preferences. I don't think this is an objective "good." For example, I PEQ my setups to neutral (as close to flat based on MY hearing) using frequency sweeps and equalizing loudness for all frequencies as I perceive it. For all headphones, there will be 2 components to the sound: what the headphones output, and what my hearing reduces/amplifies. Therefore if the headphone output is perfectly flat, I need only compensate for my own hearing. I don't want extra bass and treble because the average person likes it, I want accurate sound.


Insults aside, let me restate what I said before: I DON'T WANT extra bass and treble just because that's what "normal" people find pleasing. I don't see anything wondrous about the Harman curve, it's been known forever that V-shape is a common sound preference. I submit that it is not *accurate* and therefore not MY preference.


A better question is, did *you* read and understand both the papers and my previous posts?

1) The Harman target curve comes from feedback of adjustments to bass and treble AFTER the headphones were equalized to match the flat in-room response from loudspeakers. Repeating: Bass and Treble adjustments AFTER the equalization to match the flat in-room response from loudspeakers. Bass and high frequency accentuation is EXACTLY what is happening.

2) All this talk about loudspeakers and flat response in rooms is IRRELEVANT. I am talking about headphones. I stated that the a Harman curve-tuned headphone will not be accurate and I stand by that statement. Let me define "accurate" here: producing the same sound that would have been heard at the time of the recording. If I plucked a guitar string, and the playback through the headphone was exactly the same sound, that headphone is accurate. All headphones, of course, are inaccurate in that respect due to FR characteristics; this is why I stated that a perfectly flat FR would be most desired in a headphone. Harman tuning will NOT correct a FR deficiency, just introduce an euphonic one.

3) Even if a headphone's output were perfectly flat, there is the deficiency of one's hearing - the most common being treble hearing loss as one gets older. Using PEQ to modify my headphone output to my perception of flat FR, means I am also correcting MY OWN hearing deficiency in the process, allowing me to hear different frequencies at the levels they were recorded at, therefore getting me as close as possible to actually accurate* listening. Harman does nothing for this as well. *see definition above

The only mistake is if you infer my comment is to the "latest" Harman curve. See this, since you like references: http://www.juloaudio.sk/Umiestnenie_reprosustav/History of Harman Target Curve.pdf second to last slide. And if we take the curve to be the latest curve it makes no difference to my previous comments.

I also find it funny that the abstract to your latest reference concludes by stating: "The estimation results show that HpTFs vary considerably with headphones and ear canals,which suggests that individualized compensations for HpTFs are necessary for headphones to reproduce desired sounds for different listeners. "

Why would that be the case if the Harman curve "is a curve with "no bass or treble boost or cut, just a faithful accurate curve that is the Transfer Function," I wonder?

But the funniest part is, nowhere in the papers do I find a statement saying the Harman curve produces "accurate" sound. I don't even know where people got this from. It appears a lot of parroting has been happening and not a lot of independent thought.

I stand by my previous statements, collected below and still not disproven:

1) Defining "accurate" as: producing the same sound that would have been heard at the time of the recording, a Harman curve-tuned headphone will not be accurate.

2) The Harman curve is a best fit through listener preferences, and in no way an accurate representation of sound.

I suggest those that REALLY want accurate sound, look into performing PEQ to flatten their headphone FR with respect to THEIR OWN hearing, which will result in much more accurate perception for themselves than adjusting headphones to this Harman dreck. Although, to get back on topic, as previously posted it appears the LCD-X is not the headphone to do so!

 

[amirm]

2) All this talk about loudspeakers and flat response in rooms is IRRELEVANT. I am talking about headphones. I stated that the a Harman curve-tuned headphone will not be accurate and I stand by that statement. Let me define "accurate" here: producing the same sound that would have been heard at the time of the recording. If I plucked a guitar string, and the playback through the headphone was exactly the same sound, that headphone is accurate. All headphones, of course, are inaccurate in that respect due to FR characteristics; this is why I stated that a perfectly flat FR would be most desired in a headphone. Harman tuning will NOT correct a FR deficiency, just introduce an euphonic one.

There are two severe misunderstandings here:

1. You never know how something sounded when being recorded. Recordings go through many transformation and no way can you know the exact tonality of anything. Furthermore, the talent approved that sound using whatever speakers were used in recording, not the headphone you are using.

2. Any target curve is specific to the measurement gear. A different measurement gear produces a different measurement for the same sound produced from a single headphone. So your notion of "I want it flat"does not at all apply to test fixture I use. Harman has solved this problem by taking in situ measurements in a room and then computing a graph for the fixture that I use. As such, there is no notion of you wanting "flat" response on whatever measurement you see.

 

[overkilly]

There are two severe misunderstandings here:

1. You never know how something sounded when being recorded. Recordings go through many transformation and no way can you know the exact tonality of anything. Furthermore, the talent approved that sound using whatever speakers were used in recording, not the headphone you are using.

2. Any target curve is specific to the measurement gear. A different measurement gear produces a different measurement for the same sound produced from a single headphone. So your notion of "I want it flat"does not at all apply to test fixture I use. Harman has solved this problem by taking in situ measurements in a room and then computing a graph for the fixture that I use. As such, there is no notion of you wanting "flat" response on whatever measurement you see.

Hello Amirm,

I will first let you know I really do appreciate the measurements and the work you do here! I have read the forum for a long time, and finally have some time to participate.

1) This is just nitpicking, not "severe misunderstanding." If you just *need* me to state "as close to whatever the sound picked up by the microphones, within their limited capacity, as modified by the recording media and system and the engineer's taste, within the actual ambient temperature and pressure, electrical and magnetic fields, solar storms at the time of recording, and add whatever etc else you want to as a factor here," then here it is. The comparison, which is just for purposes of a definition, mind you, is the same: a sound existed and was recorded to a certain degree of fidelity - the closer a headphone can reproduce this, the more accurate it is. And this is leaving aside the effects of personal hearing (or assuming a perfect listener), as this leads to more cans of worms that distract from the simple truth of the definition.

2) This does not seem to bear any relation to anything I have written about, as I have not mentioned any of your, or anyone else's, measurement gear or measurements. Nor have I any notion of "I want it flat" as you call it. Yes, I do understand that different measurement gear used to measure/tune the output of a headphone would give different results - so? That is true for ANY target, not just flat, and one can only get so close and so far. I confess my AMAZEMENT at this level of criticism at a preference to get a headphone output tuned to be perceived flat, all the while, the HARMAN curve, based on preferences and subjective measurements, oh THAT was the answer all along! It was a BEST FIT through those preferences! Nevermind the input in this same thread, with people stating they find either bass or treble not to their liking with a Harman tuning! Guess they weren't part of the, what, less than 200 people that took part of their studies? Less than 200 people that represent the 6 billion people on Earth, let's not go into that.

As far as Harman being a "solution" to a "problem," the only way I can see it as such is that they have solved the "problem" of producing a target curve based on listener preferences for headphones so that they will not sound like crap out of the box for a majority of people and therefore present huge losses for the company producing them. This has no relation to accuracy, as I have previously defined. And I notice, several posts afterward, that no one has provided evidence that Harman 1) IS accurate 2) ever CLAIMED to be accurate.

 

[amirm]

1) This is just nitpicking, not "severe misunderstanding." If you just *need* me to state "as close to whatever the sound picked up by the microphones, within their limited capacity, as modified by the recording media and system and the engineer's taste, within the actual ambient temperature and pressure, electrical and magnetic fields, solar storms at the time of recording, and add whatever etc else you want to as a factor here," then here it is.

It is not a nitpick at all. Or any of the factors you mention. Recordings are not photographs. They are paintings. They are art in themselves, different than any live presentation. Any reference then to "I want a flute to sound like I know what flute sounds like" is wrong.

All we can aspire to is be truthful to the recording. Not anything before that. We must agree to this or there is no sense in going further.

Once there, we then have to take into account that there are no standards for recording. Any speaker can be used to mix and master in. And presented to the talent and label in yet another room different than ours.

At high level then, this is an unsolvable problem. Accuracy that is.

We could throw our hands up in the air or do something about this. And that something is to experiment. Play different sounds to people with different speakers/headphones and see what the listeners say is "correct." However they define it in their head.

When done like above, a speaker winds up having a certain set of measurements to approximate preference. And headphones do as well.

Headphone variability however is higher than speakers due to fixture and testing accuracy. For this reason, I don't advocate any strict adherence to any curve. Judgement needs to be approximate and variable based on which frequency range we are talking about.

Importantly, the target curve or how much bass we need relative to treble should be to taste. I provide a recommendation with my EQ in my reviews which you can use as starting point. Or not. Others have produced EQ that hug the Harman curve. I have not.

Where it becomes important to more or less follow Harman is upper bass, midrange and lower treble. I have tested some 20 headphones now and my preference for accuracy is that. Again, not 100% but close.

If you want to advocate anything else, you need to show the research or personal experimentation like above. Otherwise it gets tiring to just hear "harman curve is wrong." It is directionally quite correct. And the ideal does not exist.

 

[overkilly]

Any reference then to "I want a flute to sound like I know what flute sounds like" is wrong.

Well, no. I have a flute. I play it daily. I make a recording of my flute. I make a FLAC file of this recording. When I play it through my headphones, I want it to sound like I know what my flute sounds like.

If this is wrong to you, then let's really not go any further.

To reproduce the sound of my flute, I use my EQ to make it so I perceive all frequencies at the same volume when listening through my headphones. When I play the FLAC file, this is the closest I get to the actual sound information contained in the FLAC file (to get away from the whole "issue" of what a recording is). How do I know this? Because any deviation from a purely flat FR will exaggerate or diminish volume of certain frequencies relative to the others - therefore introducing coloration. If I play the file through, say, my HD800, I will hear the flute through the HD800 FR: diminishing bass, treble peak that we all know. Now let's say I add EQ to my HD800 after the Harman curve. Have I achieved the actual sound in the file? Of course not, maybe the bass has been corrected a little, but now the additional treble will punish my ears. And if I use a headphone that has a FR profile that follows Harman? Well, I will hear accentuated bass and treble - again, away from the actual sound information contained in the file.

 

[amirm]

Because any deviation from a purely flat FR will exaggerate or diminish volume of certain frequencies relative to the others - therefore introducing coloration.

So you are saying your HD800 plays that flute just like your ears hear it live?

 

[Robbo99999]

"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."

About your "Flat FR Curve" idea. It depends on how you measure it. Are you talking about measuring it at the eardrum on a dummy head, or are you talking about measuring it in other ways (because for sure we don't want a measured flat frequency response if measured at the actual eardrum - would sound terrible - because your ear & head anatomy attenuate different frequencies by different amounts before it gets to your eardrum)? If I give you the most benefit of doubt, are you talking about trying to measure the frequency response of a headphone in "free air" not attached to head anatomy? If the latter is true to what you're suggesting, then is your thinking that if a manufacturer were to build a headphone with this flat frequency response, then you're saying that this would allow your own ear anatomy to hear that as flat or close to neutral/natural when you put it on your head. The problem with this is that as soon as you place the headphone on your head it's not behaving like a speaker in a room anymore for a number of reasons if I think it through:

• speakers are in a room in front of you and frequency response is already influenced by the shape of your head (and the room) before it even gets to the outer parts of your ear. (some of the elements of HRTF)
• the angle at which the sound is coming from is different, speakers in front, headphones literally from each side of your head so this is influencing how sound propagates around & down your ear, affecting the frequency response.
• I think there are some interactions/implications related to bass and "direct pressure waves"/"coupling" (potentially wrong & vague terminology recalled by my memory??) by having the headphone drivers so close to your ears in a virtually sealed system along with your eardrum which has further effects on frequency response.

So I think your "Flat FR Curve" idea, even beyond the question of how you mean you're measuring it, is flawed. I think the best we can hope for at the moment in relation to neutral sound for the largest percentage of the population is the research done by Harman, which is the Harman Headphone Curve, which is relying on measurements from a dummy head in a room with speakers along with "most preferred" tone control preference added onto that curve in the target creation as a later step. This is gonna be the most effective starting point for end users to do any further EQ of their own, because statistically that starting point is likely to be closest to natural sound for the most people. (Harman Headphone Curve tries to address all my bulleted points in its creation)

 

[DualTriode]

About your "Flat FR Curve" idea. It depends on how you measure it. Are you talking about measuring it at the eardrum on a dummy head, or are you talking about measuring it in other ways (because for sure we don't want a measured flat frequency response if measured at the actual eardrum - would sound terrible - because your ear & head anatomy attenuate different frequencies by different amounts before it gets to your eardrum)? If I give you the most benefit of doubt, are you talking about trying to measure the frequency response of a headphone in "free air" not attached to head anatomy? If the latter is true to what you're suggesting, then is your thinking that if a manufacturer were to build a headphone with this flat frequency response, then you're saying that this would allow your own ear anatomy to hear that as flat or close to neutral/natural when you put it on your head. The problem with this is that as soon as you place the headphone on your head it's not behaving like a speaker in a room anymore for a number of reasons if I think it through:

• speakers are in a room in front of you and frequency response is already influenced by the shape of your head (and the room) before it even gets to the outer parts of your ear. (some of the elements of HRTF)
• the angle at which the sound is coming from is different, speakers in front, headphones literally from each side of your head so this is influencing how sound propagates around & down your ear, affecting the frequency response.
• I think there are some interactions/implications related to bass and "direct pressure waves"/"coupling" (potentially wrong & vague terminology recalled by my memory??) by having the headphone drivers so close to your ears in a virtually sealed system along with your eardrum which has further effects on frequency response.

So I think your "Flat FR Curve" idea, even beyond the question of how you mean you're measuring it, is flawed. I think the best we can hope for at the moment in relation to neutral sound for the largest percentage of the population is the research done by Harman, which is the Harman Headphone Curve, which is relying on measurements from a dummy head in a room with speakers along with "most preferred" tone control preference added onto that curve in the target creation as a later step. This is gonna be the most effective starting point for end users to do any further EQ of their own, because statistically that starting point is likely to be closest to natural sound for the most people. (Harman Headphone Curve tries to address all my bulleted points in its creation)

Hello,

This is very much on target. Human hearing is not flat.

The Harman Curve is not only based on preference. Using multi-variate analysis the preference data has a very strong correlation with real world measurement of human hearing.

Thanks DT

 

[overkilly]

So you are saying your HD800 plays that flute just like your ears hear it live?

No, I stated : "If I play the file through, say, my HD800, I will hear the flute through the HD800 FR: diminishing bass, treble peak that we all know. "

 

[overkilly]

About your "Flat FR Curve" idea. It depends on how you measure it. Are you talking about measuring it at the eardrum on a dummy head, or are you talking about measuring it in other ways (because for sure we don't want a measured flat frequency response if measured at the actual eardrum - would sound terrible - because your ear & head anatomy attenuate different frequencies by different amounts before it gets to your eardrum)?

You are ASSUMING I introduce some gear or equipment to measure the sound and create the curve, this is not so. The "curve" is entirely subjective to ME: that is, I calibrate to flat AS I HEAR IT. I do this by using frequency sweeps to identify peaks/dips, and using PEQ to correct them. I also compare random frequencies (say, 1500 Hz and 3000 Hz) to ensure they both sound equally loud TO ME in different regions). There IS no measurement gear involved, this is about the ultimate perception of the actual listener.

Here is an example - an old correction curve picture I have. This is using the Neutron PEQ. Let's say I was using my HD598 (this is old and I can't recall which HP I developed this for).

If I listen with no PEQ activated, I will hear the music as colored by the headphone/chain. Probably will sound very close to the available FR plots measured for those headphones. But if I, (meaning only me, because this curve is the result of me compensating for MY hearing and the HP/system) activate this curve, then my perception is of a flat FR - a frequency response sweep is even in volume throughout, and coloration disappears. In this case, it is obvious I had to compensate for diminishing bass, and some treble dip and peak (countered by the graph's peak and large dip). And this is a combination of the system/HP coloration, and my OWN hearing flaws (it could be MY hearing needs that extra treble peak, and not the HP, but I am the listener and it is my perception that is corrected as well).