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

[bobbooo]

Their choice of preferred EQ is likely dependent on volume so if you don't control that variable you are comparing apples vs oranges. I would only allow them to adjust volume if that is treated as a measured variable. So maybe they adjust the EQ to preference at some standardized low/medium/high volumes. It would be interesting to know what their preferred volume is although that data exists somewhere.

Related to controlling for volume, I discussed in this exchange with @Floyd Toole the difficulties of loudness normalization when doing blind tone control / EQ preference tests. What specific loudness normalization procedure would you say would be best if blind testing say two EQ profiles on the same headphone e.g. one with Harman-level bass and another with zero bass 'boost'? And is this normalization really necessary, considering your findings in this paper which (if I understood correctly) seem to suggest loudness normalization during MOA tone control tests did not have a statistically significant effect on preferred bass shelf level and starting frequency (at least for IEMs)?

 

[Sean Olive]

It is good to have a considered reference for comparison.

The Harman curve is the result of much research but is no by no means compulsory. Preference is allowed. One can see where the preference differs from the Harman curve and match that different response to other products.

Exactly.

It's a good starting point, and our data tells us it satisfies most of the listeners (64%) we tested. Some (21%) will want less bass, and even less (15%) will want even more bass. Which segment you fall into may depend on your age, hearing loss, and listening experience. And of course, your taste.

 

[raistlin65]

Their choice of preferred EQ is likely dependent on volume so if you don't control that variable you are comparing apples vs oranges. I would only allow them to adjust volume if that is treated as a measured variable. So maybe they adjust the EQ to preference at some standardized low/medium/high volumes. It would be interesting to know what their preferred volume is although that data exists somewhere.

That is exactly the point. The listener wants the frequency response to fit their preferred volume.

That's sort of the principle behind Audyssey Dynamic Volume (but let's not debate how effective it actually is lol).

While as a researcher, you might want to know what volume people are choosing, it's not necessary. You just need the user to calibrate with the source material to their preferred listening volume before they take the test.

What reference volume level does HTR assume everyone listens at? Might be interesting to see the results when people are choosing their own volume.

 

[Robbo99999]

This is an open-back headphone so you would think that it's not related to acoustic impedance. What else could explain it? But getting accurate measurements would rule out accuracy of the EQ.

Yesterday, it prompted me to investigate further. I decided to "exactly" match the frequency response of my HE4XX to my EQ'd K702 by using REW to match, as previously I was trying to EQ them to the Harman Curve optimally rather than in the same way:
K702:

HE4XX(so the target curve in this one is actually the K702 EQ'd predicted curve above, shown in light blue in the following pic):


I found that this brings the tonality closer between the two, with the notable adjustments being on the HE4XX that I rolled off the bass at the same place as the K702 and I completely pumped up the relatively sharp dip at around 5kHz all the way up to the Harman Curve (as previously I only partially boosted that). This prompted me to try the completely boosted 5kHz dip with the extended bass potential of the HE4XX:

I found that this still maintained the clarity and general tone of the previous, so I think boosting the 5kHz dip all the way to the Harman Curve was instrumental in the difference I was experiencing. I'd need to do some more testing to confirm, but it's possible that with this change that the HE4XX is not as far off as I thought. It's hard to compare headphones because there's the time lag between replacing the different headphones, as well as volume matching issues, even though I do listen to the differences on my select tracks I use for EQ experimentation (so I know those very well). Further testing required by me. I'm quite certain that it's not the same sonic experience as the K702 even when EQ'd to the same curve (spacial and clarity differences I think), but it's possible general tonality is comparable, but more testing required.

 

[Sean Olive]

Related to controlling for volume, I discussed in this exchange with @Floyd Toole the difficulties of loudness normalization when doing blind tone control / EQ preference tests. What specific loudness normalization procedure would you say would be best if blind testing say two EQ profiles on the same headphone e.g. one with Harman-level bass and another with zero bass 'boost'? And is this normalization really necessary, considering your findings in this paper which (if I understood correctly) seem to suggest loudness normalization during MOA tone control tests did not have a statistically significant effect on preferred bass shelf level and starting frequency (at least for IEMs)?

Yes, loudness matching is tricky especially when you have speakers or headphones with widely different frequency responses. It can work for some programs but not well for others. The more homogeneous the program spectra, and the more constant the loudness of the music, the better the loudness normalization will work.


For loudspeaker comparison, we historically used B-weighting (Slow) playing pink noise through the speaker and measuring at the listening seat. In recent years, we adopted ITU-R 1770 loudness. It's not perfect either and there is a question how well it works at different playback levels.

I'm glad you brought up that study where we asked listeners to adjust the level and frequency of the bass shelf filter of the headphone calibrated to the Harman Target. We repeated the measurement where the leakage and the loudness effects from the bass level adjustment were controlled and uncontrolled.

It turned out that the loudness normalization had little effect on the preferred bass level. When leakage was controlled they reduced the level slightly, as you might expect.

There was a small effect from loudness normalization on the bass filter frequency selected.Most people preferred the LF shelving filter to be set below 200 Hz.

But the loudness differences in these experiments are are small in comparison to giving someone a volume control and adjusting the broadband level of music over a range of 60 dB and setting the bass levels. I would expect to see much, much larger effects on where they set the bass level than the effects observed in these experiments.


https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1770-3-201208-S!!PDF-E.pdf

 

[Sean Olive]

Hi @Sean Olive

What are your thoughts of crossfeed methods for headphones like "Bauer stereophonic-to-binaural DSP":

http://bs2b.sourceforge.net

Have you used it and do you like it?

No, I have not tried it..but I understand the rationale behind it.

Let's be honest. Listening to stereo music through headphones has nothing to do with reality in a spatial sense. It is a spatially-compromised experience. Yet, many people get pleasure from it, and that is fine.

Personally, I am amused when audiophiles start talking about the better imaging of this headphone versus this headphone. For me, it's mostly inside my head or slightly outside, and rarely if ever out in front of me.. Maybe I'm drinking the wrong proof of scotch or smoking the wrong stuff ?


In the realm of spatial experiences, all headphone listening sucks compared to a 7.4 or 22.2 loudspeaker experience. I wonder if these people have ever had such an experience??


Binaural recordings with head-tracking (and personalized HRFTS) at least gets images outside your head, and done well you can localize in 3D space and get externalized auditory images in front of you. That is the future of headphone listening IMO. This is what we are working on now. It's a project where we are measuring individualized HRTF's different ways, using machine learning to predict HRTFs from sparse data sets, and applying them to binaural playback over headphones and loudspeakers ---so you don't get a spatially compromised experience.

 

[amirm]

Personally, I am amused about audiophiles talking about the better imaging of this headphone versus this headphone. For me, it's mostly inside my head or slightly outside, and rarely if ever out in front of me.. Maybe I'm drinking the wrong proof of scotch or smoking the wrong stuff ?

You and me both.

 

[Sean Olive]

No, I have not tried it..but I understand the rationale behind it.

Let's be honest. Listening to stereo music through headphones has nothing to do with reality in a spatial sense. It is a spatially-compromised experience. Yet, many people get pleasure from it, and that is fine.

Personally, I am amused when audiophiles start talking about the better imaging of this headphone versus this headphone. For me, it's mostly inside my head or slightly outside, and rarely if ever out in front of me.. Maybe I'm drinking the wrong proof of scotch or smoking the wrong stuff ?


In the realm of spatial experiences, all headphone listening sucks compared to a 7.4 or 22.2 loudspeaker experience. I wonder if these people have ever had such an experience??


Binaural recordings with head-tracking (and personalized HRFTS) at least get images outside your head, and done well you can localize sounds in 3D space and get externalized auditory images in front of you. That is the future of headphone listening IMO. And that is what we are working on now.

 

[Sean Olive]

Except most people would likely prefer the local soul/bbq over the McRib under blind conditions, which would make the Harman curve more similar to the local soul/bbq.

I get what you mean, but I see it as different. The Harman curve is what most people will prefer under blind conditions. Having been to both, I have a hard time believing that most would prefer Olive Garden's Lasagna over Trattoria Lisina's .

That said, I do love Olive Garden, and McDonalds can certainly hit the spot at times, and you're making me hungry now .

I would not eat at any restaurant with my last name. There is too much negative expectation bias and cognitive dissonance to accept that someone named Olive could cook an edible meal

 

[buz]

You and me both.

And here I was thinking something is wrong with my ears, maybe I don't need that hd800s after all...

 

[Sean Olive]

And here I was thinking something is wrong with my ears, maybe I don't need that hd800s after all...

Save the money, and buy some decent scotch or red wine.. After 3 glasses everything will sound better.

 

[buz]

Save the money, and buy some decent scotch or red wine.. After 3 glasses everything will sound better.

I have a few bottles of exceptional scotch around that I don't really drink anymore because of the hangovers it gives me. Maybe the smoking part would work, though

 

[bluefuzz]

would you care to speculate as to why the preferred headphone curve equates to the Harman speaker room curve?

Not Sean (obviously!), but I would have thought this was fairly uncomplicated: speakers in a room sound like everyhing else in a room, i.e. natural to our evolved sense of 'what stuff sounds like' in a room. We thus want our headphones to sound similar. As far as I can tell, this is exactly what the Harman headphone curve achieves - a gently downward sloping even response from 20Hz - 20 kHz ...

 

[Music1969]

Binaural recordings with head-tracking (and personalized HRFTS) at least gets images outside your head, and done well you can localize in 3D space and get externalized auditory images in front of you. That is the future of headphone listening IMO. This is what we are working on now.

Isn't it the present, not just future?

Apple's Airpods Pro and Max have already hit mass market in a huge way, although the numbers using the Spatial Audio feature probably aren't huge yet.

 

[Sean Olive]

Isn't it the present, not just future?

Apple's Airpods Pro and Max have already hit mass market in a huge way, although the numbers using the Spatial Audio feature probably aren't huge yet.

Yes, it is coming slowly.

Those products just came to market and do allow you to experience DOLBY ATMOS and 7.1 movies but the rendering is not based on personalized HRTF but generic ones. How good is the experience compared to personalized renderings?

The other roadblock is content, especially immersive music. ATMOS Music titles are becoming available on Tidal and Amazon Music HD, as is Sony Reality 360 available on Tidal, Amazon and Deezer?.

But unlike previous attempts to sell multichannel music on physical discs (remember SACD, DVD-A, DTS on CD,etc) the content can be more easily distributed and enjoyed anywhere since it's streamed from the cloud to portable and mobile devices. So there is a better chance it may succeed because the distribution costs are lower, don't require a 7.1.4 loudspeaker setup and are included in your music streaming service. Given that people are driving around in cars with 14-32 speakers in them most of the hardware is already in place.

 

[Sean Olive]

I have a few bottles of exceptional scotch around that I don't really drink anymore because of the hangovers it gives me. Maybe the smoking part would work, though

Given that your name here is "Buz" I think you already know the answer to that question

 

[Sean Olive]

Not Sean (obviously!), but I would have thought this was fairly uncomplicated: speakers in a room sound like everyhing else in a room, i.e. natural to our evolved sense of 'what stuff sounds like' in a room. We thus want our headphones to sound similar. As far as I can tell, this is exactly what the Harman headphone curve achieves - a gently downward sloping even response from 20Hz - 20 kHz ...

Yes this is essentially it. The hypothesis is that recordings are made to sound good through loudspeakers in rooms so try to emulate at least the spectral balance of the experience through headphones. We started by acoustically measuring a loudspeaker in a room through an ear simulator/head and from there slightly tweaked the bass.

Why tweak the bass? Because there is not a 1:1 relationship between bass heard in a room through loudspeakers and via a headphone. We discovered this years ago when doing binaural reproduction of car audio systems through headphones to do controlled double blind A/B tests of different car audio systems. When you compare the headphone experience with the car speakers there was some missing bass. We measured the wholebody vibration in the car and found that the missing tactile experience in the headphones was equivalent to 2-3 db bass

http://www.aes.org/e-lib/browse.cfm?elib=15150

 

[Sean Olive]

That is exactly the point. The listener wants the frequency response to fit their preferred volume.

That's sort of the principle behind Audyssey Dynamic Volume (but let's not debate how effective it actually is lol).

While as a researcher, you might want to know what volume people are choosing, it's not necessary. You just need the user to calibrate with the source material to their preferred listening volume before they take the test.

What reference volume level does HTR assume everyone listens at? Might be interesting to see the results when people are choosing their own volume.

If I tried to publish a paper on preferred headphone/loudspeaker targets and didn't know the playback levels at which the adjustments were made, the paper would be rejected for scientific reasons.

There has been some research on the topic of preferred listening levels. For dialog on television dialog it's 58 dBA and home theater ~ 64 dBA. The preferred level depends on background noise and people tend to turn it up with increased background noise. Dolby et al need to know this for setting dialog normalization

http://www.aes.org/e-lib/browse.cfm?elib=15150

And now people are studying preferred levels in headphones because they are trying to limit the exposure to noise to limit hearing loss.

 

[Frank Dernie]

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.

Precisely what I would expect.
I suppose difference between L and R in a test may give an indication of the magnitude of production variation, but even that would be luck.
To know one would need to test multiple units from multiple production batches - so basically never going to happen.
I would only choose anything, speaker or headphone, which needed minimal or zero EQ personally and have generally bought the actual item I auditioned, not a shop model of the same "model".
Very hard to do nowadays, so hifi is real "pot-luck"

 

[Feelas]

Precisely what I would expect.
I suppose difference between L and R in a test may give an indication of the magnitude of production variation, but even that would be luck.
To know one would need to test multiple units from multiple production batches - so basically never going to happen.
I would only choose anything, speaker or headphone, which needed minimal or zero EQ personally and have generally bought the actual item I auditioned, not a shop model of the same "model".
Very hard to do nowadays, so hifi is real "pot-luck"

IMO we should mostly check which models have the best L-R matching, since with in-ear mics (one-time effort) one can easily EQ a well-matched pair to preference, and that'll be hard with mismatched L&R.