Harman preference curve for headphones - am I the only one that doesn't like this curve?

[Jimbob54]

Yeah I don't know how Oratory got such a high rating except maybe he got a better sample or something else is different. We use the same GRAS test equipment.

It would be cool if multiple reviewers got the same objective score based on measurements. Of course, that would NEVER happen in a million years based on subjective listening

Two years ago we compared correlation between our predicted headphone sound quality ratings based on deviation from Harman Target Curve and ratings given by review organizations & found CR had highest correlation (0.67-0.75). CNET and PC.mag were the lowest.
Rtings was surprising mid-way. It would be interesting to revisit this now that they have fully adopted the Harman Target Curve albeit using a different measurement rig (Heads Acoustics).

The lesson learned is that take what you read from reviews with a grain of salt, especially if they are based only on uncontrolled listening with a sample of 1. Measurements are my go-to lie detectors

Question from the k371 review thread please. To the best of your knowledge, does the k371 BT have the same sonics as the wired version when the electronics are off and it's wired up? Thanks

 

[Music1969]

One thing I noticed is you are comparing near-field monitors @ 1m with the headphone. Our target curve began by measuring neutral speakers in our reference listening room at 10 feet which is much further away. That could explain why the Target does not match your monitors. It was not based on near-field listening

Thanks @Sean Olive !

One other thing - my experience trying the Harman Curve is full dependant on how accurate orgatory1990's measurements are for my particular headphones...

I am EQ'ing based on his measurement and his recommended adjustments. Any errors in his measurement will affect the EQ to be applied and thus my experience with the Harman Curve.

In other words, I may not actually be hearing the Harman Curve when I think I am !

 

[Robbo99999]

Thanks @Sean Olive !

One other thing - my experience trying the Harman Curve is full dependant on how accurate orgatory1990's measurements are for my particular headphones...

I am EQ'ing based on his measurement and his recommended adjustments. Any errors in his measurement will affect the EQ to be applied and thus my experience with the Harman Curve.

In other words, I may not actually be hearing the Harman Curve when I think I am !

True, but I think his measurements are likely to be the best on net in terms of using the curve for EQ purposes, by virtue of the intelligent averaging he does of a number of measurements (ie disregarding outliers). To truly know if you like the Harman Curve or not it would probably help if you'd EQ'd more than one headphone to the Harman Curve, as they don't all sound the same when EQ'd to the Harman Curve....so by comparing a number of headphones that have all been EQ'd to the Harman Curve you can get a real picture of how much you like the curve and how much variance to expect between headphones. For instance my AKG K702 / NAD HP50 / and Senn HD600 all sound quite similar and spot on when EQ'd to the Harman Curve, whereas the Hifiman HE4XX EQ'd to the same is a bit dull in it's tone in contrast to the others, so it's almost like the HE4XX needs to be EQ'd to a different curve for me......but I've tested enough headphones to know that I like the Headphone Harman Curve, (and the EQ on the HE4XX was an improvement over stock).

 

[watchnerd]

It was tested using classical, pop and rock. And we use jazz sometimes as well. We didn't test it on synthetic pop and techno which you say the Target Curve shines.

I don't believe in the common notion that audio equipment excels in one music genre but not the other. If we assume that most music regardless of genre is recorded and mastered on flattish speakers (anechoic flat but not in-room flat), which is a pretty assumption, then consumers should choose speakers and headphones with a similar response

That is an absolute prerequisite for solving Audio's Circle of Confusion.

My preference in the bass by genre may be heavily influenced by the fact that I'm a bass player, both acoustic double bass (classical, jazz) and electric (jazz, fusion).

I also do recordings, of both music I perform and that of others, so am accustomed to monitoring / mixing my own bass playing or groups I play in. I often use these as reference tracks, including playback in the same room in which the recording was made.

If I hear an instrument that I play in a genre, I'm more attentive to how the bass sounds than in a genre where it doesn't exist.

Tremolo and legato, in particular, can get easily messed up.

Demographically, I'm also in the "less bass is better" cohort, although just barely.

 

[Sean Olive]

@Sean Olive by calculating hearing loss at lows based on the research data you pointed out. And only referd towards what Harman curve really is (in its current state).
https://asa.scitation.org/doi/10.1121/1.4894719

I still do not see the connection between hearing loss in the ASA reference and the Harman Target Curve. We never designed it to address hearing loss for a particular age of listeners, and our listeners included a wide distribution of age ranges, many of whom had normal audiometric hearing. There was wide acceptance across age and audio experience.


I only brought up the ASA reference about hearing in regard to a prior question about hearing above 12 kHz, and female vs male hearing.

 

[Sean Olive]

True, but I think his measurements are likely to be the best on net in terms of using the curve for EQ purposes, by virtue of the intelligent averaging he does of a number of measurements (ie disregarding outliers). To truly know if you like the Harman Curve or not it would probably help if you'd EQ'd more than one headphone to the Harman Curve, as they don't all sound the same when EQ'd to the Harman Curve....so by comparing a number of headphones that have all been EQ'd to the Harman Curve you can get a real picture of how much you like the curve and how much variance to expect between headphones. For instance my AKG K702 / NAD HP50 / and Senn HD600 all sound quite similar and spot on when EQ'd to the Harman Curve, whereas the Hifiman HE4XX EQ'd to the same is a bit dull in it's tone in contrast to the others, so it's almost like the HE4XX needs to be EQ'd to a different curve for me......but I've tested enough headphones to know that I like the Headphone Harman Curve, (and the EQ on the HE4XX was an improvement over stock).

One of "huge" assumptions here that Oratory's (or any EQ correction) sample of the headphone is similar in response to yours, so that his EQ will work for your headphone. How many samples did he measure? One-two? How many samples is Sonar Works EQ based on?

My experience with measuring multiple samples of any given model of headphone is that the tolerances in manufacturing are simply too wide to make that assumption. You may get better quality control for the premium models above $600- $1k but not necessarily in sub $300 headphones.

So unless you have access to good measurements and can use them to EQ your own headphone, all bets are off in regards to what you are listening to.

 

[Sean Olive]

Question from the k371 review thread please. To the best of your knowledge, does the k371 BT have the same sonics as the wired version when the electronics are off and it's wired up? Thanks

I have a sample at home that I've yet to measure so I don't know the answer yet. But my guess is that the K371 is closer to the Target.

 

[Sean Olive]

This is the actual in-room measurement of the Revel speaker before an after it was equalized to the target curve, We also show the predicted in-room curve from anechoic measurements. It is about a 10 dB slope across the bandwidth.

It's from the referenced paper.


https://www.aes.org/e-lib/browse.cfm?elib=17042

 

[sejarzo]

...I've had many conversations over decades with the founder of Etymotic Dr. Mead Killion about the tuning of their headphones, which are closer to DF, particularly the bass which is flat. I think their headphones are excellent except they are thin in the bass.

That's my take, too. Their frequency response is so smooth that it's relatively simple to add a low shelf boost using PEQ to compensate for the lack of bass, and they respond to that very well. I also use a relatively broad peak filter at 3.2 kHz Q 1.0 to cut that region by 3 dB or so for most rock/pop material as most of it seems overly boosted in that range. When it comes to well recorded classical/acoustic music, that's not always necessary.

 

[Robbo99999]

One of "huge" assumptions here that Oratory's (or any EQ correction) sample of the headphone is similar in response to yours, so that his EQ will work for your headphone. How many samples did he measure? One-two? How many samples is Sonar Works EQ based on?

My experience with measuring multiple samples of any given model of headphone is that the tolerances in manufacturing are simply too wide to make that assumption. You may get better quality control for the premium models above $600- $1k but not necessarily in sub $300 headphones.

So unless you have access to good measurements and can use them to EQ your own headphone, all bets are off in regards to what you are listening to.

For the HE4XX, which is the headphone that sounds most different from my other headphones when EQ'd to the Harman Curve, that headphone was measured between 3 seperate units if I recall correctly. @bobbooo sent in his own headphone, and Oratory said there was really close matching between the different headphone samples for that particular headphone, so from that you'd think that would mean that the HE4XX would have some of the least variability between units.....and given that this one sounds the most "off" to me when EQ'd to the Harman Curve, I would wager there's something else at play here, perhaps specific re how that headphone interacts with my specific physiology which makes the GRAS measurement less relevant to me than my other headphones. I would postulate that for some reason the "headphone transfer function" for the HE4XX for some reason varies for me significantly more from the GRAS simulation than that of my other headphones. Unless I had bad luck and my particular HE4XX varies significantly from the 3 that measured very close together when Oratory measured them. However, my other 3 headphones (Senn HD600, NAD HP50, and AKG K702) all sound really very similar when EQ'd to the Harman Curve......so I find it hard to accept that there is large significant widespread variation between headphones of the same model (I could imagine a few outliers, but a high percentage of the headphones of any given model showing wide variation?), as it's not shown itself in my experience (apart from potentially the HE4XX).

Didn't your research show that price of a headphone wasn't indicative of it's preference? If quality control is "all over the shop" in sub $300 headphones then how come this was the conclusion you came to, because surely this lack of quality control would decrease preference if the sub $300 headphones you tested had QC issues that affected the frequency response? Or did you buy say 10 units of each headphone & then measure them all and throw out the outlying units from the preference tests?
EDIT: I remember now, you EQ'd a K712 to mimic the other headphones.....but you still had to measure & choose which frequency response to use from the sub $300 headphones though....so I think there is still some relevance to this last paragraph of mine.

 

[ZolaIII]

I still do not see the connection between hearing loss in the ASA reference and the Harman Target Curve. We never designed it to address hearing loss for a particular age of listeners, and our listeners included a wide distribution of age ranges, many of whom had normal audiometric hearing. There was wide acceptance across age and audio experience.


I only brought up the ASA reference about hearing in regard to a prior question about hearing above 12 kHz, and female vs male hearing.

I didn't know what you designed it for but I told you to what it only corresponds to (50~60 years old).
Did you test hearing abilities of participants before it? We all think that we hear good which usually isn't true, actually most people don't. Hearing lose above 12.5 KHz is of little interest when it comes to audio reproduction, not a lot informations there to start with (not that I think it should be disregarded), however lose in lower end of the spectrum does make a difference.
I would really like that we make a brake trogh regarding bass level variance paradigm, at least I think I had.

Any how thanks for your attention and best regards.

 

[andreasmaaan]

Most people loudspeakers that measure anechoically flat (around the design axis) but when we put them in semi-reflective rooms and measure the steady-state response they are no longer flat.

The measurement includes a combination of the direct sound (flat) but many reflected sounds that originate from off-axis (not flat) and are further attenutated/diffraced by the room surfaces. Add to that influence of LF room modes and boundary gain you no longer have a flat in-room measurement but something that is smooth but tilted down about 6-10 dB from 50 Hz- 15kHz. We've done studies where we equalized the loudspeaker in the room to be flat and people did not like it. It sounded too bright and thin.

If you measure the same speaker at the ear drum of a listener it is not flat either, but rather approximates the Harman Target Curve. We've tested it against the Diffuse Sound Field Target (which is flat in the bass) and listeners said it sounded flat and thin also.

Hi @Sean Olive, great to see you on the forum, and thanks for elaborating on my earlier post While we're in conversation, I hope you don't mind me asking you a couple of questions about the heaphone target you nd Welti derived?

Firstly, I wonder why the starting point used in your 2015 study with Welti was that of a pair of loudspeakers equalised to have a flat steady-state response, and not a flat diffuse-field response? Also, did you conduct research indicating to what extent the in-ear response differed for a pair of loudspeakers equalised flat in-room vs. a flat diffuse-field response? If so, what were the differences? (I.e. perhaps the two responses when measured at the ears are essentially the same, which actually seems plausible for measurements performed beyond the room's critical distance).

Secondly, I wonder if you have a theory as to whether (and if so, how/why) specific design characteristics of a given headphone may affect a given listener's preferred target curve (and by corollary, the average listener's preferred curve).

This second question is informed by two factors. Firstly, your research suggests that the preferred curve is different for IEMs than it is for headphones. And secondly, my (obviously limited) personal experience is that larger open-backed phones tend to sound more bass-heavy than both IEMs and smaller closed-backed phones when EQ'd to the same target response. I speculate that this may be due to the differing accoustic impedances in the volume of air between driver and eardrum, and/or differeing driver-to-eardrum distances (assuming, of course, that my own headphones do not measure significantly differently to those on which the correction curves I've used were based).

 

[Sean Olive]

For the HE4XX, which is the headphone that sounds most different from my other headphones when EQ'd to the Harman Curve, that headphone was measured between 3 seperate units if I recall correctly. @bobbooo sent in his own headphone, and Oratory said there was really close matching between the different headphone samples for that particular headphone, so from that you'd think that would mean that the HE4XX would have some of the least variability between units.....and given that this one sounds the most "off" to me when EQ'd to the Harman Curve, I would wager there's something else at play here, perhaps specific re how that headphone interacts with my specific physiology which makes the GRAS measurement less relevant to me than my other headphones. I would postulate that for some reason the "headphone transfer function" for the HE4XX for some reason varies for me significantly more from the GRAS simulation than that of my other headphones. Unless I had bad luck and my particular HE4XX varies significantly from the 3 that measured very close together when Oratory measured them. However, my other 3 headphones (Senn HD600, NAD HP50, and AKG K702) all sound really very similar when EQ'd to the Harman Curve......so I find it hard to accept that there is large significant widespread variation between headphones of the same model (I could imagine a few outliers, but a high percentage of the headphones of any given model showing wide variation?), as it's not shown itself in my experience (apart from potentially the HE4XX).

Didn't your research show that price of a headphone wasn't indicative of it's preference? If quality control is "all over the shop" in sub $300 headphones then how come this was the conclusion you came to, because surely this lack of quality control would decrease preference if the sub $300 headphones you tested had QC issues that affected the frequency response? Or did you buy say 10 units of each headphone & then measure them all and throw out the outlying units from the preference tests?
EDIT: I remember now, you EQ'd a K712 to mimic the other headphones.....but you still had to measure & choose which frequency response to use from the sub $300 headphones though....so I think there is still some relevance to this last paragraph of mine.

My statement is based on a a project a couple of years ago where we purchased and measured 10+ units of the most popular headphones and that was the conclusion: that QC and tolerances were quite wide. When you go to Chinese manufacturers, the prices goes up the tighter the tolerances and lower the volumes. The exception here are the Apple Earbuds which have remarkable consistency for something they used to include "free" with their devices. They have the benefit of volume and they send engineers to the factories to teach and monitor what is being made.

Yes, there is poor correlation between price and sound quality but that doesn't imply that QC doesn't improve with price.

It just means that companies don't understand how to optimize the tuning for best sound quality, and I'd like to think it is partially due to the lack of scientific knowledge about the relationship between subjective and objective measurements, which we have addressed.

If you look at the distribution of headphones vs. price, the range in sound quality does get narrower with increased price meaning you get fewer really bad ones. However, the scores do not increase monotonically with price.




What is the HE4XX headphone?

 

[Schackmannen]

What is the HE4XX headphone?

A collaboration between Drop and HIFIMAN.

 

[Jimbob54]

My preference in the bass by genre may be heavily influenced by the fact that I'm a bass player, both acoustic double bass (classical, jazz) and electric (jazz, fusion).

I also do recordings, of both music I perform and that of others, so am accustomed to monitoring / mixing my own bass playing or groups I play in. I often use these as reference tracks, including playback in the same room in which the recording was made.

If I hear an instrument that I play in a genre, I'm more attentive to how the bass sounds than in a genre where it doesn't exist.

Tremolo and legato, in particular, can get easily messed up.

Demographically, I'm also in the "less bass is better" cohort, although just barely.

Moar bass better

 

[bobbooo]

I've just posted a YouTube video of a presentation we gave at the AES 146th Dublin Convention, "Segmentation of Listeners Based on Their Preferred Headphone Sound Quality Profiles" It explains the research behind the Harman AE/OE Headphone Target .
https://youtu.be/1nRYyd23Czk

Great presentation, thanks. I have a few questions on it and the original paper it's based on. What's the red curve in the graph on slide 6 (entitled 'New Harman AE-OE Target Curve')?

On slides 8-10 the headphones tested in the study are described and listed. Were the Bluetooth and ANC capable headphones tested with these functions, or wired and with ANC off (if possible)?

I see the AKG N90Q was part of the study ('HP2' in the study comparing with Keith Howard's measurements?), which has AKG's 'TruNote' auto correction feature that auto-calibrates the frequency response to the listener's personalised ear acoustics via playing and recording tone sweeps when worn. Was this feature run on the GRAS artificial head/ear before taking the measurements? The N90Q also has simultaneous bass/treble EQ tone control (which I presume was kept at its neutral mid-point), as well as a 'Stage Control' feature for spatial emulation of speakers ('Standard', '2.1 Studio' and '5.1 Surround Sound'). Which of these three spatial modes was the headphone set to for your measurements (I presume one of the first two)? I would have thought that the '2.1 Studio' would be the best choice as the aim of headphones should be to emulate the sound heard through speakers in the mastering studio (I tried to infer by comparing your HP2 measurement with Innerfidelity's measurements of the three modes but it's difficult due to the non-standard rig Tyll used).

 

[Sean Olive]

Hi @Sean Olive, great to see you on the forum, and thanks for elaborating on my earlier post While we're in conversation, I hope you don't mind me asking you a couple of questions about the heaphone target you nd Welti derived?

Firstly, I wonder why the starting point used in your 2015 study with Welti was that of a pair of loudspeakers equalised to have a flat steady-state response, and not a flat diffuse-field response? Also, did you conduct research indicating to what extent the in-ear response differed for a pair of loudspeakers equalised flat in-room vs. a flat diffuse-field response? If so, what were the differences? (I.e. perhaps the two responses when measured at the ears are essentially the same, which actually seems plausible for measurements performed beyond the room's critical distance).

Secondly, I wonder if you have a theory as to whether (and if so, how/why) specific design characteristics of a given headphone may affect a given listener's preferred target curve (and by corollary, the average listener's preferred curve).

This second question is informed by two factors. Firstly, your research suggests that the preferred curve is different for IEMs than it is for headphones. And secondly, my (obviously limited) personal experience is that larger open-backed phones tend to sound more bass-heavy than both IEMs and smaller closed-backed phones when EQ'd to the same target response. I speculate that this may be due to the differing accoustic impedances in the volume of air between driver and eardrum, and/or differeing driver-to-eardrum distances (assuming, of course, that my own headphones do not measure significantly differently to those on which the correction curves I've used were based).

I had to re-read that study as it was more than 5 years ago, and some of the details I forgot.

1. We measured the steady-state response in the room of the Revel F208 and equalized the bass to the Harman target curve produce the curve below. You can see that it's relatively smooth and approximates the predicted in room curve above 100 Hz. There is no EQ applied to the treble at this point, just the raw speaker.

We then applied then applied a broadband shelving filter in the bass and treble to adjust the relative bass and treble levels of the speaker (as well as the headphone). These filters could be adjusted to get back to the original in-room target curve that you see in the graph. The point of this exercise was to see if people preferred more bass or treble than the Harman Target Curve or preferred to keep in the same as in a previous study. The same method of adjustment was done to a headphone by moving bass and treble shelving filters up and down.

People on average preferred to leave Harman Target in-room curve alone in the treble (no shelving boost or cut to the treble) and a bass shelf within 1-2 dB of what the original Harman Target in-room curve was. The headphone target was close to the in-room target within 2 dB of the bass and 1 dB in the treble.

I'm not sure what you mean by adjusting the speaker to flat steady-state vs flat-diffuse field? The flat steady-state accounts for both direct and reflected sounds from the speaker. A flat diffuse field is impossible to measure in typical listening room because it is nowhere close to a diffuse field as defined in the standards. It would have to be a reverberation chamber. We included headphones in our study equalized to the DF target and found them to be thin and bright.

2. I don't have a theory at least proven and tested about why preferred IE targets have 4 dB more bass than AE/OE targets. With in-ear earphones (closed ear canal), yes the acoustic impedance interaction are greater, the insertion depth has an effect on the ear volume, & at high frequencies 1/2 wave resonance. These are mostly issues you don't have to deal with with open back or closed headphones. I think you can see those effects when measuring IE headphones and you vary the insertion depth.

What we know is that when those variables are removed from the listening test as we did in our studies people are remarkably in agreement about what the IE target response should be (see attached graph). In this study we controlled leakage and the same replicator headphone was used to simulate all the responses. So if there were any acoustic impedance interactions between the replicator earphone and the individual listener's ear canal they did not significantly impact the listening test results. Otherwise you would see a lot of disagreement.

 

[amirm]

I don't think it got a rating from ASR.

It was posted here: https://www.audiosciencereview.com/...eview-closed-back-headphone.19657/post-649070

Score no EQ: 63.6

I don't know though if Maiky's process for computing it is the same as Sean's.

 

[amirm]

BTW, when people say this and that EQ doesn't sound good, it doesn't mean anything to me. You need to post the music samples where that is evident. That way we can check to see if we hear the same thing and/or it is an issue with the recording.

 

[andreasmaaan]

I'm not sure what you mean by adjusting the speaker to flat steady-state vs flat-diffuse field? The flat steady-state accounts for both direct and reflected sounds from the speaker. A flat diffuse field is impossible to measure in typical listening room because it is nowhere close to a diffuse field as defined in the standards. It would have to be a reverberation chamber. We included headphones in our study equalized to the DF target and found them to be thin and bright.

Many thanks for the reply @Sean Olive. I realise that measuring an in-ear DF response would require a reverberation chamber. My question referred to these comments from your 2015 paper (p. 3, where you outline the conditions under which you conducted the listening experiments (I hope you don't mind my reproucing your words here):

The headphone was equalized at the DRP to match the flattened in-room steady-state response of a reference loudspeaker (Revel Performa F208) calibrated in the Harman reference listening room. Although the loudspeaker had a flat anechoic on-axis response it was equalized in-situ to be flat as a baseline condition.

I took this to mean that essentially the following process had been followed:

1. Steady-state response of Revel F208 in Harman listening room adjusted to be flat at the listening position (which presumably meant that the anechoic response of the speaker was not flat, but rather recessed in the bass/boosted in the treble.
2. In-ear measurement taken using HATS system.
3. In-ear measurement of test headphones taken on same HATS system.
4. EQ used to adjust the headphones' response so that it matched that of Step 1 when measured on the same GRAS system, above.

And that this headphone response had then provided the starting condition from which subjects were asked to adjust bass/treble shelving filters to taste.

Did I understand that correctly? If so, my question was simply why this particular starting condition was chosen. Was this supposed to be an approximationg of a flat DF response? Or was there a particular reason why a flat steady-state response of a loudspeaker in a listening room was chosen? And, more broadly speaking, why was a flat steady-state response (as opposed to any other particular starting condition) chosen?

I realise the study was a few years ago, so no expectation here for a detailed answer It's just a question that I'd wondered about since reading the study.

2. I don't have a theory at least proven and tested about why preferred IE targets have 4 dB more bass than AE/OE targets. With in-ear earphones (closed ear canal), yes the acoustic impedance interaction are greater, the insertion depth has an effect on the ear volume, & at high frequencies 1/2 wave resonance. These are mostly issues you don't have to deal with with open back or closed headphones. I think you can see those effects when measuring IE headphones and you vary the insertion depth.

What we know is that when those variables are removed from the listening test as we did in our studies people are remarkably in agreement about what the IE target response should be (see attached graph). In this study we controlled leakage and the same replicator headphone was used to simulate all the responses. So if there were any acoustic impedance interactions between the replicator earphone and the individual listener's ear canal they did not significantly impact the listening test results. Otherwise you would see a lot of disagreement.

That's very interesting, thanks. It seems the aparent discrepancy between IEM and over-ear headphone bass preferences remains a bit of a riddle.