Master Complaint Thread About Headphone Measurements

[Robbo99999]

Thank you very much for the reply, Robbo99999.

I appreciate the tips on exporting the raw data.

I currently use the Equalizer APO's Configuration Editor as my EQ, and also for some general graphing and plotting jobs. And have already made a high resolution plot of Oratory's DF curve in there. So if I can figure out some way to import the raw ASR headphone data, then I could probably use the Editor's stacking function to compute some equivalent DF curves from that.

I'd be curious to see the DF curve that was provided with the ASR GRAS setup as well, for comparison to the one that Oratory is currently using. Because Oratory recently made some changes to his DF compensation curve. And I'm not really sure why that was done.

It is getting rather late here. So I'll have to get back to you on some of the other stuff. Maybe there are some other knowledgeable folks here who can dig back in their memories a bit, and try to explain how DF compensation was supposed to be used to make the data between different rigs more easily comparable and interchangeable though. Because I haven't had much luck explaining this concept to folks on some of the other forums. (I'll give it a try though, if nobody else feels like it.)

I'm afraid the hype about the Harman curve has led to alot of confusion and misunderstanding though about some of the original uses and purposes for both diffuse and free-field compensation. Because these still can have some very beneficial uses for data comparison purposes.

Re comparing DF compensations of headphone measurements between different measurement rigs.....they're different heads & ears so you'll get different raw results as well as compensation results. Even though you'd think the DF Target Curve creation procedure would be the same for the different heads, but there's still different transfer functions for any given headphone on any given dummy head, so the headphones will react differently on each dummy head and wouldn't give the same deviations from the DF Targets even when you're talking about the compensated graph (ie difference from target). That's my thinking on it based on using what I know about the topic. I mean you might find somekind of rough correlations in comparing DF compensated graphs of any given headphone between different measurement rigs, but even if you got correlations (no matter how accurate or innacurate), then what do you do with the information, why is it important?

 

[ADU]

Re comparing DF compensations of headphone measurements between different measurement rigs.....they're different heads & ears so you'll get different raw results as well as compensation results. Even though you'd think the DF Target Curve creation procedure would be the same for the different heads, but there's still different transfer functions for any given headphone on any given dummy head, so the headphones will react differently on each dummy head and wouldn't give the same deviations from the DF Targets even when you're talking about the compensated graph (ie difference from target). That's my thinking on it based on using what I know about the topic. I mean you might find somekind of rough correlations in comparing DF compensated graphs of any given headphone between different measurement rigs, but even if you got correlations (no matter how accurate or innacurate), then what do you do with the information, why is it important?

This is a hard concept for alot of people to grasp, Robbo99999.

The main reason for the existence of compensation in the first place though was for the purpose of normalizing or standardizing the measurements between different rigs of different designs. And calibrating them to a common reference point. The idea was that you compared the response of each rig to a common, easily reproducible reference or stimulus, such as a diffuse sound field. And then use the differences between their responses to that common stimulus to offset or calibrate the other measurements made on those rigs, so that they were a better match to one another.

So instead of standardizing everything to the same raw frequency response curve (like Harman wants people to do), which only works effectively with a single apparatus, you were standardizing the measurements to a common stimulus or sound field that's independent of the response of any particular rig.

I'm not suggesting this was a perfect solution to the problem of standardization. But it was certainly better than none at all.

Nobody really knew or understood what a "neutral response" looked like when these approaches to compensation were developed. So the main purpose of DF and FF compensation was simply to apply some objective standard or reference point to the measurements between different rigs, so they would all yield approximately similar results when measuring the same stimulus.

For this process to work, the DF curves must be different for each manufacturer's rigs, because they will each have a somewhat different response to the reference stimulus (a diffuse sound field, in this case). And it is that difference between the DF curves that allows the other measurements on the different rigs to be compensated or calibrated in such a way that they are a closer match to one another, despite the minor differences in some of their physical and acoustic characteristics.

I can't really think of a much better way to explain it, so I hope some of this is making some sense.

 

[Robbo99999]

This is a hard concept for alot of people to grasp, Robbo99999.

The main reason for the existence of compensation in the first place though was for the purpose of normalizing or standardizing the measurements between different rigs of different designs. And calibrating them to a common reference point. The idea was that you compared the response of each rig to a common, easily reproducible reference or stimulus, such as a diffuse field. And then use the differences between their responses to that common stimulus to offset or calibrate the other measurements made on those rigs, so that they were a better match to one another.

So instead of standardizing everything to the same raw frequency response curve (like Harman wants people to do), which only works effectively with a single apparatus, you were standardizing the measurements to a common stimulus or sound field.

I'm not suggesting this was a perfect solution to the problem of standardization. But it was certainly better than none at all.

Nobody really knew or understood what a "neutral response" looked like when these approaches to compensation were developed. So the main purpose of DF and FF compensation was simply to apply some objective standard or reference point to the measurements between different rigs, so they would all yield an approximately similar results when measuring the same stimulus.

For this process to work, the DF curves must be different for each manufacturer's rigs, because they will each have a somewhat different response to the reference stimulus (a diffuse sound field, in this case). And it is that difference between the DF curves that allows the other measurements on the different rigs to be compensated or calibrated in such a way that they are a closer match to one another, despite the minor differences in some of their physical and acoustic characteristics.

I can't really think of a much better way to explain it, so I hope some of this is making some sense.

Yeah, that makes sense, that's already how I understood the mechanism of your proposed Diffuse Field Compensation Curves, but the point I was making was that creating a Target Curve is one thing but expecting a specific model of headphone to have the "same" predictable reaction to each of the rigs is not reasonable, and I'm talking variance beyond the compensation. It's for the same reasons that two different models of headphones EQ'd to the same target curve on one apparatus will not sound identical when you wear them on your own head - there is an element of unpredictability about how a model of headphone will interact with your ear structure, and that comes down to not just your ear structures but also the design of the headphone, it works both ways (this is different to & beyond the compensation curve and how speakers react to the rig - creation of the target curve). So your diffuse field compensation process can't take that variable into account. But anyway, even if you ignore everything that I wrote earlier in this paragraph (or if it's debateable what I wrote) what's the point of comparing headphones measured on different rigs via diffuse field compensations.....I don't really see the value in it? The point with the Harman Curve is that it's some proper research that links "speakers in a room" along with "preference studies" to arrive at a final curve that seems to bring a lot of enjoyment to people...(it just means you have to measure your headphones on a GRAS unit as a "limitation"), I don't really see Diffuse Field & Diffuse Field Compensations to compare headphones on different rigs....I don't see how this really comes into it & is of benefit?

 

[ADU]

Thank you for your responses on this stuff, Robbo99999.

I'll take the last question above first, since it's one of the easier ones to answer (I think?)... There are a variety of reasons why you might want to compare the measurements between different rigs.

A. To verify and validate that the results are accurate. Reviewers and graphers routinely do this to try to figure out where there might be some noticeable errors or discrepancies in their setups and measurements. Even Amir does this with his spinorama plots, to try to compensate them and improve their accuracy. The same theory also applies to headphones.

I will also often use the measurements of several different sites to EQ my own headphones. And in order to do that, I have to figure some way of adapting the data from each of them, so the results are similar. Which is no small task. The diffuse field graphs, primarily in Oratory's graphing tool are my current graphs of choice though, because they're the easiest for me to relate to something fairly concrete. (Namely, a speaker's sound power response.) And because many of the other sites have simply dropped DF compensation or put it behind a paywall, sometimes in favor Harman compensation or another target of their choice that's similar but maybe also flatter in the bass.

B. Because standardization is a good thing. For starters, without some form of standardization, it's not really possible for users of disparate measurement systems to accurately compare and contrast their results. Standardization also makes it easier and faster though (time is money they say) for a wider audience to use and interpret a measurement system's results. If the measurements are more standardized, then users have to spend less time learning how to intepret a particular reviewer or grapher's plots. And can spend more of their time just comparing and analyzing the brightness, darkness, warmness, bassiness, sibilance, tonal balance, etc. of some different headphones. So it's alot more efficient and useful.

I don't know if you and others here have noticed this, Robbo99999. But it seems like there's been a somewhat growing movement in certain audio circles, and perhaps also among some ordinary Joe-audio types against the use of measurements. And I think one possible reason for this could be the lack of standardization in measurements. Because the harder they are to read and interpret, the more frustrated people get. And the more likely they probably are to just throw up their hands, and say this is all meaningless!

Laugh if want, but some of these folks apparently have a pretty big following among some audio users and enthusiasts.

Reading graphs isn't particular easy either. There's probably only a tiny fraction of the headphone users out there who can actually make sense of most frequency response graphs. And really understand what they mean. And even many of the so-called experts on the subject are still in the dark about alot of stuff. (I think even Resolve might possibly concur with me on that.)

Standardizing the measurements in such a way that the data is more readable, usable and comprehensible to a variety of users, and more easily interpretable across the different measurement platforms would, I think, not only help alot more people with their buying decisions. But it might also help to put the brakes on some of the anti-measurement sentiment.

I've spent a good part of the last decade or so looking into this subject, for example, and I still have some difficulty accurately pointing out to others what a neutral response looks like on both raw and also compensated plots! The Harman curve made some good steps towards answering those questions. But it really didn't go far enough in my opinion. And based on some recent comments from one or two of the key researchers at Harman, I think they also believe that there's probably still some more work to be done on this subject as well. (Which was actually somewhat of a relief to me.) This brings me to point number three...

C. If someone like Harman could ever come up with a truly neutral response, then the standardization of measurements between different rigs (via something like DF compensation) would make it alot easier to accurately translate or adapt that target response across a much wider variety of measurement platforms. Instead of having to use it on just one closed system like the GRAS 45C, or something else. And it would also save folks like Jaakko the trouble of having to figure out how to adapt the target to the measurements of several other different systems.

What if someone using a different measurement system, like, say the new HBK 5128, or something by Head Acoustics, came up with a better and more accurate target than the Harman curve. And then decided they didn't want to share it with the users of other measurement systems? I wonder how that would make all of the GRAS users feel? Because that wouldn't be very cool. And I suspect that many folks using systems other than GRAS probably feel similarly about the Harman curve.

Tyll Hertsens of Inner Fidelity was one of the key figures who promoted, publicized and supported Harman's work on a new target headphone response curve, through his articles and videos. He was even invited to Harman's reference room to do some in-ear measurements on his rig. The Head Acoustics system that Tyll was using was different than the system Harman was using though. So in the end, there was really no easy way for him to deploy the new target response curve (that he wrote so many loving articles about) for his own measurements.

I watched Tyll struggle with this problem for months and years. To find a way to make the target work correctly with his measurements. And I tried to also help out a little where I could. And also suggested using DF compensastion as a possible method of translation. But that idea was rebuffed by some of the "smarter" kids in the room. It was like watching someone trying to put a square peg into a round hole though. And it wasn't that enjoyable for him, I suspect, or for the rest of his readers.

I'm sure Tyll had other more important reasons for closing up shop, and leaving the profession. But I suspect that experience left a somewhat bitter taste in his mouth, after all the support he'd given to the folks at Harman... Water under the bridge as the say.

So standardizing the measurements would make this kind of thing more open and accessible to a wider variety of graphers. And consequently also to a wider variety of headphone manufacturers, and also consumers. I think that was actually Harman's original goal in the beginning. And why they made most of the research on the subject of their target public. But then somewhere along the way, the Harman target ended up becoming more of a "perk" for all the users who happened to own systems similar to Harman. And a bit more of a hurdle to get over for just about everyone else. Which was a bit of a shame.

I'm sure I'll think of some other good reasons to compare measurements as well. But since this post is already gettin gargantuan in size, I'll stop with those. And maybe come back to this again another time.

 

[amirm]

The diffuse field graphs, primarily in Oratory's graphing tool are my current graphs of choice though, because they're the easiest for me to relate to something fairly concrete.

Unfortunately research shows that they are the wrong target. Our rooms are not fully diffused so it reasons that such a target can't be the one we use. This has been extensively researched and results have been quite conclusive. Here is Dr. Olive in one of his presentations:

Notice that generation of diffused field is not an exact thing either. Here are the results of the listening tests:

You might say to provide both targets but that makes an already difficult to read graphs, more difficult.

 

[thewas]

If I understood @ADU correctly he doesn't recommend the DF as a target but as measurement reference baseline to make the understanding easier like for example by a better comparability with loudspeaker measurements (where the sound power has a similar shape and background) but I am sure he will reply himself on it.

 

[ADU]

thewas is exactly correct.

And thank you also, Amir for diving in on this on as well. I really do appreciate all that you folks have been doing here to educate people on sound measurements.

I am familiar with the Harman tests you're referencing in the images above. And what they did was tune the response of the headphones to match sevaral different DF compensation curves. Which is not what I'm talking about here.

I'm not even necessarily suggesting that anyone else use DF compensation in precisely the way that I'm using it, which is simply to compare the diffuse field response of headphones with the sound power responses of loudspeakers. If the Harman curve is your preference, so be it. It at least makes a pretty good starting point for adjusting a headphone's raw response.

I see a very high correlation between the sound power response of good loudspeakers and the DF response of good headphones though. Which I think has the potential to also be a good model for a neutral headphone response. The Harman curve can actually be used sort of as a tool to prove this point.

Here are two plots of the 2018 Harman Over-Ear Target. Both of which have been compensated with Oratory's Diffuse Field compensation. The first version uses Oratory's pre-2021 DF compensation curve, and the second one uses his current 2021 DF compensation curve.

2018 HARMAN OVER-EAR TARGET WITH DF COMPENSATION

PRE-2021:

CURRENT 2021:

There are some minor differences between the two curves in the treble, due to Oratory's update. But for the most part they are pretty similar.

What's interesting about these curves and also the diffuse field plots of some of the headphones that I posted above and below (which include the AKG K371, that is based on the above Harman target), is their rather striking similarity to the sound power responses of the better loudspeakers. Especially the ones with built-in subs, and good extension in the bass.

I am in the process right now of computing some new higher resolution plots that will show the average sound power responses of some of the better loudspeakers in Pierre A.'s spinorama database. But here is an early test image of the combined response of three of the best-extended speakers with a flat direct response, for comparison to the above. These are the sound power curves of the Dutch Dutch 8C, Infinity Prelude MTS and Mesanovic RTM10 combined.

SOUND POWER RESPONSE OF 3 WELL-EXTENDED LOUDSPEAKERS:

The units on the graphs aren't the same, because all I did was stack the 3 sound power curves together in Equalizer APO's Configuration Editor, and then take a snapshot of their combined response. All three speakers have a very similar slope though to the diffuse field response of the Harman curve. And also the slope of the diffuse field response of headphones which are close to the Harman target, shown below and also in my previous post above. This is still pretty rough, but hopefully you can see some of the rather striking similarities in responses of all of these curves.

DIFFUSE FIELD RESPONSE OF 4 DIFFERENT HEADPHONES:

 

[amirm]

What if someone using a different measurement system, like, say the new HBK 5128, or something by Head Acoustics, came up with a better, and more accurate target than the Harman curve.

The only entity that can do this is Harman. No one else has the capability and motivation to run such extensive listening tests to develop a preference curve. Despite being one of the first customers for 5128, Harman never developed a preference curve for it.

FYI I got an evaluation unit of the 5128 but eventually had to return it since Harman does not have a preference curve for it: https://www.audiosciencereview.com/forum/index.php?threads/headphone-measurements-using-brüel-kjær-5128-hats.15352/

We tried to develop one using a modified diffused field but it just did not work. Nothing measured on it correlated much with any other measurements either. As such, this unit is only useful for internal R&D research, not for review work.

 

[amirm]

I see a very high correlation between the sound power response of good loudspeakers and the DF response of good headphones though. Which I think has the potential to also be a good model for a neutral headphone response. The Harman curve can actually be used sort of as a tool to prove this point.

After testing some 60 headphones now and getting them to conform to Harman target, I am realizing that when done, these headphones produce a unique and in many cases superior experience to speakers. Even Harman has realized that post their headphone research, that the in-room target for speakers should be different than they classically thought.

The deep sub-bass for example is not something we have a handle on with regular speakers since they rarely have flat response to 10 Hz like headphones do.

 

[Deleted member 16543]

The diffuse field response approach would make sense if the following could be verifiable:
Take several headphones and measure them on rig A. Calculate the correction EQ for each of them to a DF response.
Do the same with rig B.
Then, it the DF worked as a 'unifying' common denominator for headphone measurements, the difference between each correction EQ, for each pair of headphones, would be exactly the same.
Unfortunately I very much doubt that would be the case, because a diffuse field interacts with a rig in a very different way than a localized, near source such as a headphone. And all headphones interact with different rigs in different ways, due to the geometry boundary change.
It could be that, for whatever reason I can't honestly think of, DF comparison is a little more robust.. But I wouldn't bet on it.

Aside that.. how are these measurements taken? Especially for speakers, the gating affects the resulting response so much that the correction EQ for a pair of headphone that one is trying to match to a good pair of speakers is not univocally determined.
That's why it's very important to do measurements with gating settings that make sense in a psycho acoustic kind of way.
Incidentally, this also explains a lot of skepticism from some people when it comes to judging sound quality based on measurements. How are these measurements taken? It is one thing to measure a DAC. Pretty straight forward. It is another thing to boil down an acoustic field, made up of a direct and many delayed sound waves (all with their frequency shape and direction, therefore weighted differently through the HRTFs), into a FR plot that 'makes sense' from the human sound perception point of view.

 

[TimF]

Excuse me for a question that is off topic. Does ASR have an estimate of the distribution of common listening level volumes of its members. Every now and then I pull out a sound level meter and check what the volume levels I am playing music. It is rare that I hit 80 dB although I do get into the 70 - 80 dB range when I am playing louder than normal. Sometimes I like to turn it UP and that is when I bump the 80 dB level. I don't like it louder than that. A lot of commercial recordings especially from 20 or more years ago just aren't very dynamic.

 

[ADU]

The only entity that can do this is Harman. No one else has the capability and motivation to run such extensive listening tests to develop a preference curve. Despite being one of the first customers for 5128, Harman never developed a preference curve for it.

FYI I got an evaluation unit of the 5128 but eventually had to return it since Harman does not have a preference curve for it: https://www.audiosciencereview.com/forum/index.php?threads/headphone-measurements-using-brüel-kjær-5128-hats.15352/

We tried to develop one using a modified diffused field but it just did not work. Nothing measured on it correlated much with any other measurements either. As such, this unit is only useful for internal R&D research, not for review work.

Thank you again for hearing my out on alot of this, Amir. I noticed you did some tests with the HBK 5128 rig. But hadn't found may way to the above thread yet on that topic. So I appreciate the link.

I recently found the DF and FF compensation curves for the 5128 as well here. So I may also try doing some DF plots for this rig.

https://www.bksv.com/-/media/literature/Product-Data/bp2573.ashx

In theory, one or the other diffuse field versions of the Harman target that I posted above could probably be used, at least as a beginning point for a target response curve for the 5128. I think sound power might work almost as well though (if not even better?). There aren't very many headphone measurements available for the 5128 rig though, so I'm probably already getting a bit ahead of myself on that.

After testing some 60 headphones now and getting them to conform to Harman target, I am realizing that when done, these headphones produce a unique and in many cases superior experience to speakers. Even Harman has realized that post their headphone research, that the in-room target for speakers should be different than they classically thought.

The deep sub-bass for example is not something we have a handle on with regular speakers since they rarely have flat response to 10 Hz like headphones do.

All good points. Even if you could accurately match the frequency response of a loudspeaker in a room within a pair of headphones (which is somewhat debatable, and would likely require some fairly sophisticated DSP), they probably would still sound and feel different, because there are other properties in their sound and feel that still wouldn't match.

All I'm personally interested in though is a rough approximation of the steady-state response or tonal balance in a room. Or something pretty close to that. Some of the approaches above seem to be getting me closer. But there are always some strange measurement anomalies in the rigs that are used for these kinds of plots that also need to be taken into account in same way. So what I'll mostly be using for my target response curves are probably the actual measured responses of other headphones that have a rep for their neutral sound, and are close to some of the target curves that I've posted above. And also similar to a -1.0 to -1.5 dB per octave slope, which is the approximate average slope of the Harman curve (with DF comp.), and also the sound power response curves of the better loudspeakers.

Nothing is engraved in stone though yet. Because I'm still learning more about all of this stuff. Tyll H and his articles and videos were certainly a big help to me on headphone measurements. And Floyd Toole's articles and videos have so far been my best guides to some of the thornier aspects of speaker measurements. But there's still alot more to learn on these subjects. (And I'm sure most of the speaker buffs here are already well-acquainted with Dr. Toole's work.)

 

[amirm]

Excuse me for a question that is off topic. Does ASR have an estimate of the distribution of common listening level volumes of its members. Every now and then I pull out a sound level meter and check what the volume levels I am playing music. It is rare that I hit 80 dB although I do get into the 70 - 80 dB range when I am playing louder than normal. Sometimes I like to turn it UP and that is when I bump the 80 dB level. I don't like it louder than that. A lot of commercial recordings especially from 20 or more years ago just aren't very dynamic.

Single value SPL numbers are notoriously unreliable. And further, few if any member is in a position to correctly measure SPL of any headphone. This is why some reviews are provided as relative SPL and not even absolute values! I had to buy a specialized calibrator my my Gras measurement gear to get proper SPL numbers.

I have done a video on this: https://www.audiosciencereview.com/forum/index.php?threads/music-how-loud-is-loud-video.22434/

 

[Robbo99999]

@ADU , you've done a lot of detailed posts there, one of which was in response to me. I've got to go to work now, it's morning here in the UK, so I'll respond to you after work, about 9hrs or so, ha! To put you in the picture preliminarily one of the questions is about your "2018 HARMAN OVER-EAR TARGET WITH DF COMPENSATION" graphs you showed, the Y-axis is all negative, but if you're comparing Harman 2018 Target with Diffuse Field Target then you should have a zero line with values above & below that.....for instance Harman has more bass than Diffuse Field so it should be positive in that area in your graphs (or vice versa depending on which you're subtracting from the other). I'm sure there's some other points I want to bring up but not enough time before work. Oh, and what's your background, anything audio related or are you an enthusiast?

EDIT: @ADU , I think the stuff in your post here that you're doing (https://www.audiosciencereview.com/...bout-headphone-measurements.18451/post-862618 ) doesn't quite add up logically (I think you might have overlooked a step of logic in your own creative process), but before I jump the gun I'll wait to hear from you first re the question above first as it might aid in my understanding of what you're doing there.

 

[Robbo99999]

Thank you for your responses on this stuff, Robbo99999.

I'll take the last question above first, since it's one of the easier ones to answer (I think?)... There are a variety of reasons why you might want to compare the measurements between different rigs.

A. To verify and validate that the results are accurate. Reviewers and graphers routinely do this to try to figure out where there might be some noticeable errors or discrepancies in their setups and measurements. Even Amir does this with his spinorama plots, to try to compensate them and improve their accuracy. The same theory also applies to headphones.

I will also often use the measurements of several different sites to EQ my own headphones. And in order to do that, I have to figure some way of adapting the data from each of them, so the results are similar. Which is no small task. The diffuse field graphs, primarily in Oratory's graphing tool are my current graphs of choice though, because they're the easiest for me to relate to something fairly concrete. (Namely, a speaker's sound power response.) And because many of the other sites have simply dropped DF compensation or put it behind a paywall, sometimes in favor Harman compensation or another target of their choice that's similar but maybe also flatter in the bass.

B. Because standardization is a good thing. For starters, without some form of standardization, it's not really possible for users of disparate measurement systems to accurately compare and contrast their results. Standardization also makes it easier and faster though (time is money they say) for a wider audience to use and interpret a measurement system's results. If the measurements are more standardized, then users have to spend less time learning how to intepret a particular reviewer or grapher's plots. And can spend more of their time just comparing and analyzing the brightness, darkness, warmness, bassiness, sibilance, tonal balance, etc. of some different headphones. So it's alot more efficient and useful.

I don't know if you and others here have noticed this, Robbo99999. But it seems like there's been a somewhat growing movement in certain audio circles, and perhaps also among some ordinary Joe-audio types against the use of measurements. And I think one possible reason for this could be the lack of standardization in measurements. Because the harder they are to read and interpret, the more frustrated people get. And the more likely they probably are to just throw up their hands, and say this is all meaningless!

Laugh if want, but some of these folks apparently have a pretty big following among some audio users and enthusiasts.

Reading graphs isn't particular easy either. There's probably only a tiny fraction of the headphone users out there who can actually make sense of most frequency response graphs. And really understand what they mean. And even many of the so-called experts on the subject are still in the dark about alot of stuff. (I think even Resolve might possibly concur with me on that.)

Standardizing the measurements in such a way that the data is more readable, usable and comprehensible to a variety of users, and more easily interpretable across the different measurement platforms would, I think, not only help alot more people with their buying decisions. But it might also help to put the brakes on some of the anti-measurement sentiment.

I've spent a good part of the last decade or so looking into this subject, for example, and I still have some difficulty accurately pointing out to others what a neutral response looks like on both raw and also compensated plots! The Harman curve made some good steps towards answering those questions. But it really didn't go far enough in my opinion. And based on some recent comments from one or two of the key researchers at Harman, I think they also believe that there's probably still some more work to be done on this subject as well. (Which was actually somewhat of a relief to me.) This brings me to point number three...

C. If someone like Harman could ever come up with a truly neutral response, then the standardization of measurements between different rigs (via something like DF compensation) would make it alot easier to accurately translate or adapt that target response across a much wider variety of measurement platforms. Instead of having to use it on just one closed system like the GRAS 45C, or something else. And it would also save folks like Jaakko the trouble of having to figure out how to adapt the target to the measurements of several other different systems.

What if someone using a different measurement system, like, say the new HBK 5128, or something by Head Acoustics, came up with a better and more accurate target than the Harman curve. And then decided they didn't want to share it with the users of other measurement systems? I wonder how that would make all of the GRAS users feel? Because that wouldn't be very cool. And I suspect that many folks using systems other than GRAS probably feel similarly about the Harman curve.

Tyll Hertsens of Inner Fidelity was one of the key figures who promoted, publicized and supported Harman's work on a new target headphone response curve, through his articles and videos. He was even invited to Harman's reference room to do some in-ear measurements on his rig. The Head Acoustics system that Tyll was using was different than the system Harman was using though. So in the end, there was really no easy way for him to deploy the new target response curve (that he wrote so many loving articles about) for his own measurements.

I watched Tyll struggle with this problem for months and years. To find a way to make the target work correctly with his measurements. And I tried to also help out a little where I could. And also suggested using DF compensastion as a possible method of translation. But that idea was rebuffed by some of the "smarter" kids in the room. It was like watching someone trying to put a square peg into a round hole though. And it wasn't that enjoyable for him, I suspect, or for the rest of his readers.

I'm sure Tyll had other more important reasons for closing up shop, and leaving the profession. But I suspect that experience left a somewhat bitter taste in his mouth, after all the support he'd given to the folks at Harman... Water under the bridge as the say.

So standardizing the measurements would make this kind of thing more open and accessible to a wider variety of graphers. And consequently also to a wider variety of headphone manufacturers, and also consumers. I think that was actually Harman's original goal in the beginning. And why they made most of the research on the subject of their target public. But then somewhere along the way, the Harman target ended up becoming more of a "perk" for all the users who happened to own systems similar to Harman. And a bit more of a hurdle to get over for just about everyone else. Which was a bit of a shame.

I'm sure I'll think of some other good reasons to compare measurements as well. But since this post is already gettin gargantuan in size, I'll stop with those. And maybe come back to this again another time.

Ah, ok, so the reason you want to have Diffuse Field Compensated Graphs on here is so that you can compare measurements on this site vs that of others. Well that's a reasonable request, but even if you just consider Harman, then there are quite a few people out there who are doing measurements on industry standard GRAS units (Oratory / Amir / Crinacle / Headphones.com), so there's lots of other resources where you can directly compare measurements based on the Harman Target on those GRAS units, so I would think the relatively small number of people that go into the detail of comparing measurements between sites could have quite a bit of joy already by using those sites I mentioned......so praps going to the trouble of Diffuse Field Compensations to compare headphones between sites is all a bit unnecessarily complicated as well as adding another variable of inaccuracy . I mean I recently bought an HD560s, and did exactly this, I compared measurements from Oratory / Crinacle & Headphones.com and came up with the following to create my own EQ (interestingly the average curve of all 3 is pretty much Oratory's curve exactly):

And following is my EQ of the HD560s based on the average curve above (just if people are curious or they want to try it):


I'm not a massive fan of compensation graphs, I'd much rather see the raw frequency response with the Target overlaid (which is what Amir currently does), to me it's more elucidating than a compensated graph and makes people have a greater understanding of what's really happening as well as an appreciation of different targets or what the target really is. The more people that are exposed to the raw frequency response graphs that Amir shows then the more people will start to understand the difference between headphones & speakers, so I don't really see compensated graphs as a necessity to "dumb it down" for the readership, people don't want to be dumbed down, and people like to join in the discussion so that means learning about the differences between headphones & speakers. (I also prefer to see the raw measurement and target when I'm doing my headphone EQ's in REW, it feels more intuitive to me than converting it to a compensated graph....and it helps you align the target on the measurement more effectively before you start EQ'ing.)

I do understand the mechanism of your idea re using Diffuse Field Compensated Graphs of headphone measurements to compare between different measurement rigs. How accurate can this be though, it's only accurate if the exact same procedure & test environment is used in the Diffuse Field Target creation for the different rigs.....and I don't know if & how tightly this Diffuse Field Target creation process is standardised between all the measurement rig manufacturers. The other variable of inaccuracy is one that I've mentioned twice already in my previous posts - ie that any specific model of headphone can & would react slightly unpredictably with each of the different measurement rigs vs just the Diffuse Field Target that was created by the manufacturer. What I mean here (which I've tried to explain in my other 2 posts), is the phenomenon that 2 different models of headphones EQ'd to the same curve on the same measurement rig will not sound exactly the same when you then place them on your own head - by extension this means that the same headphone (that you so wish to compare) when measured on 2 different rigs will also have the same unpredictable reactions, and I'm talking beyond the diffuse field compensation that you mention. How significant the differences could be related to this last variable I mention I am not sure, but I expect it would be similar in scope to the difference in sound between 2 different models of headphones that have been EQ'd to the same curve on the same measurement rig.

All-in-all, I'm not personally converted to the idea of Diffuse Field Compensations as being a particularly useful addition. I could say, "well maybe Amir could just include them if he wants to, what's the harm", but I guess we don't want to add stuff that is unnecessary as it would confuse people as to why it was there & it's importance (or lack thereof), so it's probably best to include stuff in the reviews that is strongly in-line with what is deemed fundamental.....there's enough material to wade through without including "excess extraneous".

 

[ADU]

The diffuse field response approach would make sense if the following could be verifiable:
Take several headphones and measure them on rig A. Calculate the correction EQ for each of them to a DF response.
Do the same with rig B.
Then, it the DF worked as a 'unifying' common denominator for headphone measurements, the difference between each correction EQ, for each pair of headphones, would be exactly the same.
Unfortunately I very much doubt that would be the case, because a diffuse field interacts with a rig in a very different way than a localized, near source such as a headphone. And all headphones interact with different rigs in different ways, due to the geometry boundary change.
It could be that, for whatever reason I can't honestly think of, DF comparison is a little more robust.. But I wouldn't bet on it.

Aside that.. how are these measurements taken? Especially for speakers, the gating affects the resulting response so much that the correction EQ for a pair of headphone that one is trying to match to a good pair of speakers is not univocally determined.
That's why it's very important to do measurements with gating settings that make sense in a psycho acoustic kind of way.
Incidentally, this also explains a lot of skepticism from some people when it comes to judging sound quality based on measurements. How are these measurements taken? It is one thing to measure a DAC. Pretty straight forward. It is another thing to boil down an acoustic field, made up of a direct and many delayed sound waves (all with their frequency shape and direction, therefore weighted differently through the HRTFs), into a FR plot that 'makes sense' from the human sound perception point of view.

Howdy, sax512. And thanks for posting the above.

I agree with your analogy re the measurement of a DAC vs. something like a speaker in the more complex sound field in a room. I'm flaking a bit at the moment on "gating" though. (Although I've heard the term used in these types of discussions before.) And will probably have to do a bit more research before offering comment on that. If there are some other good discussions or info you could point me to on that subject, that would be cool. And might help get me back up to speed on the subject.

I don't have any degrees or certificates, and am mostly just self-taught on the subject of audio, speakers, headphones, and acoustics. (Albeit with the help of some good articles, videos, white papers, and so forth from some fairly bright bulbs in the field that I've been fortunate to find or be directed to on the web.) So there are still some significant gaps in my knowledge in this area, especially re speaker and transducer design. I'm sure that someone like Steve Guttenberg could probably run circles around me all day on the design, manufacture and installation of speakers in different kinds of spaces, for example. Even though I don't necessarily agree with some of his opinions on measurements. I try to pick up a few new pieces of info though wherever I can, including from folks like this that I don't necessarily agree with.

 

[ADU]

@ADU , you've done a lot of detailed posts there, one of which was in response to me. I've got to go to work now, it's morning here in the UK, so I'll respond to you after work, about 9hrs or so, ha! To put you in the picture preliminarily one of the questions is about your "2018 HARMAN OVER-EAR TARGET WITH DF COMPENSATION" graphs you showed, the Y-axis is all negative, but if you're comparing Harman 2018 Target with Diffuse Field Target then you should have a zero line with values above & below that.....for instance Harman has more bass than Diffuse Field so it should be positive in that area in your graphs (or vice versa depending on which you're subtracting from the other). I'm sure there's some other points I want to bring up but not enough time before work. Oh, and what's your background, anything audio related or are you an enthusiast?

EDIT: @ADU , I think the stuff in your post here that you're doing (https://www.audiosciencereview.com/...bout-headphone-measurements.18451/post-862618 ) doesn't quite add up logically (I think you might have overlooked a step of logic in your own creative process), but before I jump the gun I'll wait to hear from you first re the question above first as it might aid in my understanding of what you're doing there.

Howdy again, Robbo99999... I completely understand the work thingy. Some of the posts you refer to above actually took quite a bit of time & energy to put together. So I also needed a break to reset. I appreciate all of your and other folks' opinions and feedback on things though. Because it helps me drill down more, and see where some potential flaws in my reasoning and assumptions may be. So that is all good. Sometimes it may take awhile for me to respond though, because I also have a few other responsibilities to contend with,... now and then.

As mentioned above, my background in this subject is mostly self-taught. I've always been involved with music in some form though since I was little. So that's where alot of my interest comes from. I wouldn't really characterize myself as an "audiophile" though. Because the most expensive headphones I've ever owned cost about $150. And my current amp, and DAC together ran me about a whopping $120! Whooohoo! Yeah!

I try to put components together that I think will do a good job within the very limited income I have to spend on this hobby though. I also have quite a bit of experience in the visual arts, including computer graphics. So analyzing graphical info may come a bit more easily and naturally to me than for some other folks with less experience in this area.

I put together a couple new pictures to hopefully give a little better idea how I arrived at some of my previous plots and images.

For starters, here's a slightly better version of my plot of the combined/average sound power response of the Dutch Dutch 8C, Infinity Prelude MTS, and Mesanovic RTM10. Which also includes the amplitude levels in dBs listed on the side. The dBs shown on this graph represents the average levels of the three speakers combined, which is different than what will appear on the next graph that follows...

AVERAGE SOUND POWER RESPONSE OF 3 SPEAKERS WITH GOOD BASS EXTENSION:

It's a bit hard to read but the lowest dB value in the lower left is -15 dB. So the change in amplitude in the entire curve from the highest point in the bass at around 45 Hz, to the lowest point in treble at 20 kHz is about -13.5 dB overall.

Here is how the above plot was created in Equalizer APO's Configuration Editor...

The sound power response curves of the three different speakers were each loaded into the stack as variable graphic EQ filters. A variable graphic EQ in this program is simply one which can have any number or configuration of points or "bands" in its curve. If you look closely, you can actually see that points or frequency bands on each curve are numbered, and the first plot of the Dutch Dutch 8C has 74 points/bands. The Infinity Prelude MTS plot has 51 points/bands. And the Mesanovic RTM10 plot has 54 points/bands in all. So these are fairly detailed and accurate plots of each speaker.

The Analysis Panel at the bottom shows the total combined amplitude of the 3 curves together. So the dB levels levels are basically 3x the values shown on my previous plot of the average response above. And all I really did was change the values on that graph to give the levels for the average response.

The combined amplitude of all three speakers slightly exceeded the baseline level of 0 dB in the program. And to fix that, and normalize the peak level in the bass closer to 0 dB, I also added a Preamp with a -3.10 dB setting. If I had not done that, then the area of the curve that exceeded 0 dB on the graph would've become red. Because it would have been over the level of unity gain (0 dBFS) for the program.

The above plot represents a very small sampling of sound power response curves, btw. So I wouldn't attempt to read too much into it yet. At least not until the sampling gets a bit larger. The 3 speakers used for this sampling are not untypical of the responses of some of the better loudspeakers with good extension in the bass though. So they weren't such a bad place to start imo. My main criteria for inclusion in these sound power plots though is the flatness and linearity of the speaker's direct response. The amount of extension it has in the lower frequencies. And to a somewhat lesser degree, the smoothness/linearity of the speaker's sound power, and directivity. I make all my assessments on these things simply by eye. Though the speakers that meet my criteria on the above will also tend to rank highly in the same areas in Pierre Aubert's rankings.

Small program-related note: The "Instant Mode" auto-save feature is enabled by default in Equalizer APO's Configuration Editor. So any other Equalizer APO project you currently have saved to the config.sys file in the EqualizerAPO/config folder will automatically be over-written by the program when it's first opened. Unless you happen to save the config.sys under a different file name first.

 

[Deleted member 16543]

Howdy, sax12. And thanks for posting the above.

I agree with your analogy re the measurement of a DAC vs. something like a speaker in the more complex sound field in a room. I'm flaking a bit at the moment on "gating" though. (Although I've heard the term used in these types of discussions before.) And will probably have to do a bit more research before offering comment on that. If there are some other good discussions or info you could point me to on that subject, that would be cool. And might help get me back up to speed on the subject.

I don't have any degrees or certificates, and am mostly just self-taught on the subject of audio, speakers, headphones, and acoustics. (Albeit with the help of some good articles, videos, white papers, and so forth from some fairly bright bulbs in the field that I've been fortunate to find or be directed to on the web.) So there are still some significant gaps in my knowledge in this area, especially re speaker and transducer design. I'm sure that someone like Steve Guttenberg could probably run circles around me all day on the design, manufacture and installation of speakers in different kinds of spaces, for example. Even though I don't necessarily agree with some of his opinions on measurements. I try to pick up a few new pieces of info though wherever I can, including from folks like this that I don't necessarily agree with.

Gating means (in this context) using time interval definitions (specific to a frequency range, usually) outside of which the presence of a certain frequency in the measured response is weighted much less in the calculation of the total power at that frequency.
This time frame is larger at lower frequencies and decreased as frequency goes up. All this is based on how humans perceive sounds, psychoacoustically.
Presence of a certain frequency outside of its most relevant time window is associated more to a sense of ambience than an increase in power perception at that frequency.
This is a very coarse explanation, of course. If you like to research this stuff I suggest this book. Very informative and easy to understand.
https://www.amazon.com/Accurate-Sou...ords=accurate+dsp+mitch&qid=1627688702&sr=8-2

 

[Robbo99999]

Howdy again, Robbo99999... I completely understand the work thingy. Some of the posts you refer to above actually took quite a bit of time & energy to put together. So I also needed a break to reset. I appreciate all of your and other folks' opinions and feedback on things though. Because it helps me drill down more, and see where some potential flaws in my reasoning and assumptions may be. So that is all good. Sometimes it may take awhile for me to respond though, because I also have a few other responsibilities to contend with,... now and then.

As mentioned above, my background in this subject is mostly self-taught. I've always been involved with music in some form though since I was little. So that's where alot of my interest comes from. I wouldn't really characterize myself as an "audiophile" though. Because the most expensive headphones I've ever owned cost about $150. And my current amp, and DAC together ran me about a whopping $120! Whooohoo! Yeah!

I try to put components together that I think will do a good job within the very limited income I have to spend on this hobby though. I also have quite a bit of experience in the visual arts, including computer graphics. So analyzing graphical info may come a bit more easily and naturally to me than for some other folks with less experience in this area.

I put together a couple new pictures to hopefully give a little better idea how I arrived at some of my previous plots and images.

For starters, here's a slightly better version of my plot of the combined/average sound power response of the Dutch Dutch 8C, Infinity Prelude MTS, and Mesanovic RTM10. Which also includes the amplitude levels in dBs listed on the side. The dBs shown on this graph represents the average levels of the three speakers combined, which is different than what will appear on the next graph that follows...

AVERAGE SOUND POWER RESPONSE OF 3 SPEAKERS WITH GOOD BASS EXTENSION:

View attachment 144565

It's a bit hard to read but the lowest dB value in the lower left is -15 dB. So the change in amplitude in the entire curve from the highest point in the bass at around 45 Hz, to the lowest point in treble at 20 kHz is about -13.5 dB overall.

Here is how the above plot was created in Equalizer APO's Configuration Editor...

View attachment 144566

The sound power response curves of the three different speakers were each loaded into the stack as variable graphic EQ filters. A variable graphic EQ in this program is simply one which can have any number or configuration of points or "bands" in its curve. If you look closely, you can actually see that points or frequency bands on each curve are numbered, and the first plot of the Dutch Dutch 8C has 74 points/bands. The Infinity Prelude MTS plot has 51 points/bands. And the Mesanovic RTM10 plot has 54 points/bands in all. So these are fairly detailed and accurate plots of each speaker.

The Analysis Panel at the bottom shows the total combined amplitude of the 3 curves together. So the dB levels levels are basically 3x the values shown on my previous plot of the average response above. And all I really did was change the values on that graph to give the levels for the average response.

The combined amplitude of all three speakers slightly exceeded the baseline level of 0 dB in the program. And to fix that, and normalize the peak level in the bass closer to 0 dB, I also added a Preamp with a -3.10 dB setting. If I had not done that, then the area of the curve that exceeded 0 dB on the graph would've become red. Because it would have been over the level of unity gain (0 dBFS) for the program.

The above plot represents a very small sampling of sound power response curves, btw. So I wouldn't attempt to read too much into it yet. At least not until the sampling gets a bit larger. The 3 speakers used for this sampling are not untypical of the responses of some of the better loudspeakers with good extension in the bass though. So they weren't such a bad place to start imo. My main criteria for inclusion in these sound power plots though is the flatness and linearity of the speaker's direct response. The amount of extension it has in the lower frequencies. And to a somewhat lesser degree, the smoothness/linearity of the speaker's sound power, and directivity. I make all my assessments on these things simply by eye. Though the speakers that meet my criteria on the above will also tend to rank highly in the same areas in Pierre Aubert's rankings.

Small program-related note: The "Instant Mode" auto-save feature is enabled by default in Equalizer APO's Configuration Editor. So any other Equalizer APO project you currently have saved to the config.sys file in the EqualizerAPO/config folder will automatically be over-written by the program when it's first opened. Unless you happen to save the config.sys under a different file name first.

Hi, well that's good, I'm self taught too in a similar way to you, but I'm not a musician, instead I have a scientific background. Ah, the plots I was referring to that I wanted clarification on were the following ones, not the speaker curves:

And this is what I was asking you based on your graphs above:
"one of the questions is about your "2018 HARMAN OVER-EAR TARGET WITH DF COMPENSATION" graphs you showed, the Y-axis is all negative, but if you're comparing Harman 2018 Target with Diffuse Field Target then you should have a zero line with values above & below that.....for instance Harman has more bass than Diffuse Field so it should be positive in that area in your graphs (or vice versa depending on which you're subtracting from the other)."


(You didn't reply to my other post https://www.audiosciencereview.com/...bout-headphone-measurements.18451/post-863259 but you probably ran out of time)

 

[ADU]

Hi, well that's good, I'm self taught too in a similar way to you, but I'm not a musician, instead I have a scientific background. Ah, the plots I was referring to that I wanted clarification on were the following ones, not the speaker curves:

And this is what I was asking you based on your graphs above:
"one of the questions is about your "2018 HARMAN OVER-EAR TARGET WITH DF COMPENSATION" graphs you showed, the Y-axis is all negative, but if you're comparing Harman 2018 Target with Diffuse Field Target then you should have a zero line with values above & below that.....for instance Harman has more bass than Diffuse Field so it should be positive in that area in your graphs (or vice versa depending on which you're subtracting from the other)."

Sorry about that. I was coming to this as well, Robbo. But got sidetracked by some other stuff.

The diffuse field plots of the 2018 Harman over-ear curve were produced essentially the same way as the sound power plot above in Equalizer APO's Configuration Editor.

You are absolutely correct that the bass on the Harman curve should rise above the baseline. This is true also on a diffuse field compensated plot, because the DF curve is essentially flat below about the 160-200 Hz range in the bass. Without the negative Preamp enabled, you can see above that the compensated plot in the Analysis Panel does indeed rise above the baseline of 0 dBFS.

Just to clarify what's going on above, the raw plot of the Harman curve is the first filter or item in the stack, plotted using a variable graphic EQ curve. The second curve below that is Oratory's new 2021 diffuse field curve, turned upside-down (inverted to the geeks). Both curves are plotted in high precision, using a large number of points or bands The third filter is a Preamp control, which is currently disabled above. (You can tell it's off because its "power button" on the left isn't highlighted in light blue like the other filters.)

The bass level on Oratory's new 2021 DF curve is slightly above 0 dB. So when it's flipped upside down, it adds a small amount of negative gain on the order of -0.3 to -0.4 dB. Which brings the peak bass level on the Harman curve's compensated plot down by just a small amount. Otherwise the peak level in the Harman curve's bass response would be closer to 6.1 dB above the baseline. You can see the difference though after the two curves were combined in the lower right of the Analysis Panel, where it says "Peak gain: 5.7 dB". The area that excurses above the 0 dBFS baseline is shown in red, btw, to warn you that audio in that area could be clipped, because it exceeds the unity gain level.

With the -8.00 dB Preamp filter enabled though, the whole curve falls somewhat below the 0 dBFS clip point, So the bass is no longer red, as shown below.

Since these curves and plots are mostly for my own informational/analytical purposes, and not for actual listening use (at least not in this particular form). And my primary concerns are the overall shape, tonal balance, and slope of the curve, what I'll generally do is drop the curve low enough so it's below the 0 dBFS clip point. And in the case of a diffuse field or sound power plot, I'll usually try to align or normalize the resulting response curve to one of the horizontal gridlines at around 1 kHz or thereabouts, using the Preamp control. (If I'm in a hurry though, sometimes I may forget to do this.)

The choice of the normalization frequency is somewhat arbitrary though. And sometimes I may try to normalize other parts of the curve, such as the bass instead. It just sort of depends on what I'm focusing on at the moment.

Normalizing to a point somewhere around 1 kHz in the midrange though helps me to better assess how much of a dip there is in the curve at the crossover point of the treble and midrange at around 2 to 2.5 kHz. The level of that dip is particularly important to me because my ears seem overly sensitive in that range. Probably because alot of the pop recordings I listen to are mixed too hot there, because the in-room monitors used for mixing and mastering the content likely have a similar depression there in their dispersion and off-axis response.

It also helps me to gauge some of the differences in the levels between the midrange and bass. And also the midrange and the treble, particularly around 6 to 8 kHz in the low treble. Imo, the treble in that 6 to 8k range on Oratory's DF graphs probably shouldn't exceed the level in the mids at about 500 Hz to 1kHz by too much, or the headphones may begin to sound a bit too bright. (This is just a general rule of thumb though, which may not apply to all headphones or listeners.)

The Harman curve does not represent a perfectly neutral response to my ears btw. But I can go into some more detail on that another time. The main area that I have some issues with though is the lack of air in the upper treble. There is some unevenness in the upper mids as well imo, which leads me to believe that it's possibly also a bit too forward in some spots there. Perhaps around 1.5 to 2 kHz, and maybe also in some other area around 4k or so. This is still something I'm experimenting with though.

The dip in the Harman curve's DF response in the upper mids is positioned a little higher in frequency than a typical loudspeaker's midrange-tweeter crossover. Which tends to be more around the 2 to 2.5k band range (though it can technically occur almost anywhere from 1 to 3 kHz, depending on the speaker's design). The artificial pinna that Oratory uses complicates the response in the upper mids and treble though. So it's not quite an ideal match to the average sound power response of the good loudspeakers. So that's why I use various references, including the measured responses of headphones, to try to get a better handle on what should be going on in that area. I think Ora's pinna can produce both bumps and dips/notches around that whole area though, which may make his plots look slightly off, or uneven, or overly bright or depressed in some parts of that range.

I'll have some more thoughts on diffuse field, and on your other post above with the HD560S plots another day.

That is another headphone I've been looking at pretty closely though lately. Although I haven't tried them yet myself, the new polymer drivers and good bass extension supposedly give those HPs a some wider sounding soundstage than some of Sennheiser's other classic 6-series HPs. Some reviewers also claim that they have a pretty neutral sound. And it looks that way from the graphs I've seen so far. They look like they might be just a bit uneven in some spots in the upper mids and treble though. And maybe just a tad bright around 4.5k or so. Which should be easily correctable with an EQ.

Although the Senn 5 and 6 series headphones tend to have pretty good driver symmetry overall, I've noticed that some of the models in those lines will sometimes appear a little imbalanced in some spots in the upper mids or treble. And it looks like the HD560S may possibly fall into this category as well, due to a little dip in the right channel at around 2k. Based on this compensated Rtings graph...

https://www.rtings.com/headphones/1-5/graph#18492/7903

A feature like this is slightly harder (but certainly not impossible) to correct with an EQ, because it requires a somewhat different adjustment for the right and left channels.

I don't currently have any open back HPs (or Senns) though. And this looks like it could be one of the better options in the Sennheiser lineup in my general price range. So the HD560S is still very much on my radar. There's a local store where I can easily get one, which also helps.

The audio gear I'm currently using is listed in the link in my signature btw.