Headphones and the Harman target curve

[Senior NEET Engineer]

Upper mids is being referred to as "too much" pretty often as I can see - how is it felt, exactly? Is it too forward, harsh? What gear did you try for it to appear so? Would be interesting to know. Maybe combining Harman with an already-bright presentation over 10kHz lends itself to be absolutely over-the-top, in theory it'd make sense and I'm asking people about that.

Too loud... feeling of unpleasantness. I tried Galaxy Buds+.

 

[solderdude]

I found the K371 to be fairly neutral. It only needed a little lowering of the upper mids/lower treble (hinting to shoutiness).
Fixable.

Bass was not too much. Subbass was a bit elevated but don't mind.

Horizontal trace is audible neutral to me, a small downward slope from 1kHz to 20kHz (a few dB tops) is not objectionable as long as the treble response is fairly 'even' and not 'spiked'.

There were other reasons why I do not own one but it was not tonal balance issues.

 

[Dreyfus]

Upper mids is being referred to as "too much" pretty often as I can see - how is it felt, exactly? Is it too forward, harsh?

Exactly that.

To me the emphasis on the presence region feels like they are trying to force the frontal localization. IMO this sounds rather fake because it limits the preceived width of the soundstage and makes the sound very hard and tinny. It adds fatigue.

Not all but most of the headphones I own have this issue when EQ'd to Harman (following Oratory's settings). It's not perfect. Still a good starting point as already mentioned.

For me, the bass should be somewhere between the flat in-room response and the average preference curve. Then a little bit of attenuation in the presence area. Works quite well for me.

 

[Feelas]

Exactly that.

To me the emphasis on the presence region feels like they are trying to force the frontal localization. IMO this sounds rather fake because it limits the preceived width of the soundstage and makes the sound very hard and tinny. It adds fatigue.

Can't disagree about the fatigue happening. VERY recording dependent, though.

Let's think it through. If we:
- assume the curve is was taken from speakers angled some degrees (I don't know exactly how much it was) from the listener,
- the resultant ear behaviour in itself is variable due to angle,
then it'd be natural to assume, that the curve derived that way would remind us of such a speaker setup. Whether it's good or not - different topic.

I think it's possible that we're being hit here by the "pan law", which makes the center sound a different volume on speakers and on headphones. I'd try to explain what it more, but I don't get the thing that well to explain it properly, yet I feel that's the proper way to think through. It might result in phantom center being "smaller" on some recordings. It might not just be Harman sounding wrong; it might be that speaker HRTF w/o cross-ear interaction on headphones just doesn't sound "true" and will never be truly universal, working on many cases but not on all. Spatial perception isn't the matter of the Harman "preference experiment", anyways.

Also, if you think how binaural recordings are produced, it makes some sense to think that way: the crossfeed is taken "into the recording". Harman's tonal character could actually be neutral, yet the presentation as a whole wouldn't be and would probably ask for driving too loud, since stereo doesn't take the crossfeed. In the end, you'd end up with either harsh/tinny recordings (because of distant centre due to panning "for speakers" with a different pan law) and the ones suffering in a variable amount from the combination of these, plus the recordings done "wrong" for speakers, but sounding fine on Harmans (due to not sounding "too distant" when on proper loudness). Maaaaaaaaan!

I also wonder whether Harman works better on open vs closed cans. If the problem was crossfeed-related (not just crossfeed, the whole room interaction), I believe open-backed designs produce some amount of crossfeed and that could benefit the tinny/narrow character. See - Sundaras are open-backed cans with nearly-Harman (sparing the bass) FR and they're pretty popular & not mentioned as harsh, not so much, at least. I've seen numerous people bothered with K371s being problematic.

Not all but most of the headphones I own have this issue when EQ'd to Harman (following Oratory's settings). It's not perfect. Still a good starting point as already mentioned.

Well, oratory mentions that he listens for distortion happening - but due to production variance you might get a pair with imbalanced channels, where one driver starts resonating and the one isn't yet. Going further...

... the whole "let's EQ everything to Harman" affair shows up the inherent problems of statistical stuff - how averaging things mocks up the details, sometimes absolutely ruining the sense. Some AutoEQ measurements show absolute nonsense due to averaged measurements, when you dig into the details. Crinacle provides three original measurements + L/R separated graph (obviously not accounting for the variance), o1990 doesn't give out neither original measurements nor the L/R separated ones. Hm.

For me, the bass should be somewhere between the flat in-room response and the average preference curve. Then a little bit of attenuation in the presence area. Works quite well for me.

Wonder whether the preference score is tied to how long you have to listen to stuff. Maybe endurance-testing of curves would show us more reliable information.

It'd be necessary to provide example material for people against Harman to show them that it works sometimes and some stuff to show that oftentimes it just doesn't work for the rest. We'd probably all gain a lot of perspective from that.

 

[KeithPhantom]

the whole "let's EQ everything to Harman" affair shows up the inherent problems of statistical stuff - how averaging things mocks up the details, sometimes absolutely ruining the sense. Some AutoEQ measurements show absolute nonsense due to averaged measurements, when you dig into the details. Crinacle provides three original measurements + L/R separated graph (obviously not accounting for the variance), o1990 doesn't give out neither original measurements nor the L/R separated ones. Hm.

This is one issue I have with Harman. It used the perception of trained and untrained testers and then did some statistical averaging to find a perception curve, but the premise in question was wrong in my opinion for a scientific-derived curve. Different people experience the same signal differently, and holding everything else equal, there is no way their experiences are directly correlated and other control factors the same such as the way the head is shaped (in a few words, what are the correlation statistic values of these estimations, the presence of outliers, and how the curve was averaged) Not only that, people both have different experiences and records of "what is neutral". Is it the intention of the mastering/mixing engineer neutrality, or do they even how that sounds. Can people even remember how "neutrality" sounds if their audio memory is extremely short? Also, isn't music somewhat flawed as constant? Should we use broadband noise for testing neutrality? How do you know a song or a composition to be "neutral"? These are questions that are in my head right now and I would like someone with more knowledge to answer.

 

[Pretorious]

Not only that, people both have different experiences and records of "what is neutral". Is it the intention of the mastering/mixing engineer neutrality, or do they even how that sounds. Can people even remember how "neutrality" sounds if their audio memory is extremely short? Also, isn't music somewhat flawed as constant? Should we use broadband noise for testing neutrality? How do you know a song or a composition to be "neutral"? These are questions that are in my head right now and I would like someone with more knowledge to answer.

These are some of the very same questions I have asked myself. Keep in mind that I am not as well-versed in the technical aspect as many others here, but I do try and apply scientific thinking to the whole equation.

With that said, I keep going back to the origin. That is, defining what the source of "neutrality" is. To me that would be fidelity to the original source, not the engineer's idea of neutrality. If the engineer has manipulated the recording away from the natural sound of the source, then neutral cannot exist.

This, however, leads to a circular argument, and once again we come back to "what is neutral"? Take a recording of orchestral music, for example. Already, any and all recordings will sound nothing like sitting in the audience of a music hall. Instead, the sound is all from on the stage. Is that fidelity to our "neutral"? Probably not. Is it fidelity to the neutrality of the instruments? More so, I say. What if then we change venues? Recording an orchestra in one hall as opposed to another will lead to different timbre and ambiance, within a certain threshold. We can all identify what a violin or clarinet sounds like no matter the venue; but the reflections and positioning will change our perception of so much of what we hear. When recording, do we take the venue into consideration, and is that then a "neutral" to draw a baseline from?

These are not easy answers even for myself who proposed them. I come back to this consistently. It's interesting from a scientific aspect; but sometimes I just want to enjoy the music. However I can do that has merit, in my opinion. But I want the sound as uncolored as can possibly be. In that case, it is faith to the recording that becomes foremost.

 

[Feelas]

These are some of the very same questions I have asked myself. Keep in mind that I am not as well-versed in the technical aspect as many others here, but I do try and apply scientific thinking to the whole equation.

With that said, I keep going back to the origin. That is, defining what the source of "neutrality" is. To me that would be fidelity to the original source, not the engineer's idea of neutrality. If the engineer has manipulated the recording away from the natural sound of the source, then neutral cannot exist.

This, however, leads to a circular argument, and once again we come back to "what is neutral"? Take a recording of orchestral music, for example. Already, any and all recordings will sound nothing like sitting in the audience of a music hall. Instead, the sound is all from on the stage. Is that fidelity to our "neutral"? Probably not. Is it fidelity to the neutrality of the instruments? More so, I say. What if then we change venues? Recording an orchestra in one hall as opposed to another will lead to different timbre and ambiance, within a certain threshold. We can all identify what a violin or clarinet sounds like no matter the venue; but the reflections and positioning will change our perception of so much of what we hear. When recording, do we take the venue into consideration, and is that then a "neutral" to draw a baseline from?

These are not easy answers even for myself who proposed them. I come back to this consistently. It's interesting from a scientific aspect; but sometimes I just want to enjoy the music. However I can do that has merit, in my opinion. But I want the sound as uncolored as can possibly be. In that case, it is faith to the recording that becomes foremost.

Broadly the "circle of confusion", as Floyd Toole called it. The best we can strive for is reproducing faihtfully and exactly what the engineer put on the disk.

 

[kkeretic]

After moving from kinda strange (mid-bass boosted & hyped on top), very clear and "right-there" KNS-8400 onto K371s I start to feel that the veil kinda should be there. The perception of harmonics (and thus proper instrument character, where the ratio of fundamentals to harmonics is pretty much well-defined) depends heavily on whether the HF is correctly reproduced or not. I think the clarity might be taken as a must, where many recordings just aren't meant to sound "clear", you're perhaps trading off the feeling of hi-end high-resolution for depth perception and placement.

After reading your post I got the urge to re-examine the situation around the Harman target curve and the ATH-M50x. I processed the frequency response curve that Oratory has for those headphones through the AutoEq and got the curve similar to that from the AutoEq repository, only more precise with 17 parametric equalizers. Went to sleep and the next day used those parameters with familiar tracks. And I am really pleased with the result. The key point was obviously going to rest. . Thanks.

 

[Pretorious]

Broadly the "circle of confusion", as Floyd Toole called it. The best we can strive for is reproducing faihtfully and exactly what the engineer put on the disk.

Indeed that would seem to be the only constant we can use as our so-called neutral point, and is what I strive for in my playback equipment. As I'm sure we all do.

 

[Feelas]

Indeed that would seem to be the only constant we can use as our so-called neutral point, and is what I strive for in my playback equipment. As I'm sure we all do.

I think it's the best position if you want to stay sane I mean, unless something amazing happens. Much like Fourier transform moved us into an entirely new realm, if we were to "decompose" the recording, somehow, into an "instrument domain" and then process it somehow as parts - everything would be fine. Or just get raw instrument tracks & simulate the environment using metadata supplied with tracks. MQA nonsense happened, and yet virtualized rooms still didn't and we're stuck with surrounds, instead of just putting the instruments where we'd want. Smyth Realizer is a step towards the best, yet imagine if you could compose the soundstage (or just "slip" into one the engineer intended).

Even if we were to know what was in engineer's head that day and knew what he eff'd up (and he's keeping him up at night as a big life fault), it's impossible to reverse anything that's bigger than a simple "it went to the disk with wrong EQ set up just before recording". Anything past the mastering stage and you're going to deal with something missing.

And that one night when you thought that musicians are pricks, when they've canceled the gig because of bad hall acoustics? Guess they were probably feeling as much helpless, just a thousand times more AND towards to something they were trying to perform. Makes you think...

 

[Dreyfus]

it might be that speaker HRTF w/o cross-ear interaction on headphones just doesn't sound "true"

Doesn't the Harman target include an approximation of the interaural level differences by measuring both the left and right speaker channel for each ear?

Also, if you think how binaural recordings are produced, it makes some sense to think that way: the crossfeed is taken "into the recording".

That is an important aspect. Binaural recordings can be technically even more demanding, though. Studio mixes are mostly tied to the 30° speaker placement. Binaural recordings however do highly depend the calibration of the dummy head which may expect a "flat" response at 0°, 30°, 90° or anything between, DF or FF. Maybe it isn't even calibrated at all, as we can see with all the DIY rigs and cheaper 3Dio-ish startup products.
That's where headphone calibration (which has to be matched to the dummy head calibration at the reference point, of course) can get very tricky. The standards are not very well defined - and most importantly respected - compared to studio mixes.

Some AutoEQ measurements show absolute nonsense due to averaged measurements, when you dig into the details

That's simply the limit of generic headphone measurements. Without knowing the exact response on YOUR particular head and ears you HAVE to smooth and average. Otherwise you will make things much worse by overcompensating resonances that may not appear (at the same level or frequency) on your own head with your particular headphone model.

It is a field of uncertainty. Headphone measurements can only be objective as long as the headphone remains on the fixture.

 

[Feelas]

Doesn't the Harman target include an approximation of the interaural level differences by measuring both the left and right speaker channel for each ear?

I'd think it does, yet it just bothers me that I don't know exactly. I've read a bunch of their papers, yet not encountered the part where they describe the GRAS speaker measurement process. Gonna have to dig some on the weekend.

Still, even if each ear gets the "proper" timbre, it'd end up short by not getting the delayed sound that the head expects -> the phantom centre is dead. The recording doesn't have the necessary cross-ear info. Reminds me of incorrect toe-in on very directive speakers. It's very confusing when that happens, yet the timbre isn't broken.

Remember that all the steps for deriving the curve were done on open-backed cans - that'd enforce the idea that cross-ear interaction could be missing on closed-backs. I guess you did EQ some opens onto Harman - do they sound the same harsh way w/ a narrow centre, or does it open up the nicely-behaving cans?

That is an important aspect. Binaural recordings can be technically even more demanding, though. Studio mixes are mostly tied to the 30° speaker placement. Binaural recordings however do highly depend the calibration of the dummy head which may expect a "flat" response at 0°, 30°, 90° or anything between, DF or FF. Maybe it isn't even calibrated at all, as we can see with all the DIY rigs and cheaper 3Dio-ish startup products. That's where headphone calibration (which has to be matched to the dummy head calibration, of course) can get very tricky. The standards are not very well defined - and most importantly respected - compared to studio mixes.

Well, a sad state of affairs. Impressive summary of the binaural problem! Standarization is hard to push through, anywhere, especially with niches such as binaurals.

 

[Dreyfus]

Still, even if each ear gets the "proper" timbre, it'd end up short by not getting the delayed sound that the head expects -> the phantom centre is dead. The recording doesn't have the necessary cross-ear info.

A center position on the median plane requires the same time delay on both ears. You get that by playing both drivers simultaneously.
The issue is rather that there isn't anything else other than continuous phase matching. That is one of the explinations for the in-head-localization.

Remember that all the steps for deriving the curve were done on open-backed cans - that'd enforce the idea that cross-ear interaction could be missing on closed-backs. I guess you did EQ some opens onto Harman - do they sound the same harsh way w/ a narrow centre, or does it open up the nicely-behaving cans?

I mostly use open back cans. The effect (harsh w/ narrow center) is present on most of the samples I have tried so far. Overall I would say that the Harman target does not improve the soundstage, at least not for me. If at all, it narrows the stage to the front ... without sounding realistic, for my taste.

 

[blse59]

Apparently, the Harman target curve tries to mimic the sound that the artist/engineer had in mind when creating the music. So I'm curious, do I just have a taste and hearing that differs from the majority, do people here generally like this kind of sound?

I've never recorded so is that really the case? It seems to me when mixing and engineering a record, it would make more sense to work with a pallet of frequencies with equal volume i.e. flat. The Harman target has a big bass boost.

The Etymotic ER4 has what I would call accurate, truthful sound since it has a flat frequency response. If you want to hear what a sound recording actually sounds like, what you hear through the ER4 would be it. The Harman would have what I would call a subjectively good, but not accurate, sound. People get caught up in the fact that it was scientifically derived by Ph.d's. Well, the taste of Doritos potato chips, Coca-Cola soft drinks, and McDonalds hamburgers were scientifically derived by food scientists as well. I think we can all agree they taste good.

 

[terasankka]

I've never recorded so is that really the case? It seems to me when mixing and engineering a record, it would make more sense to work with a pallet of frequencies with equal volume i.e. flat. The Harman target has a big bass boost.

I mix and master records (not professionally though). You are not really looking for EQ that much. You just mix using your ears and apply correction to the EQ using your experience. Like "too muddy.... substract 1-2db around 300hz...." "need more air and shimmer... try a high shelf at around 8K... " etc. Then you take the finished mix and listen to it with a mobile phone, in you car, bluetooth speaker, monitors, Hifi systems. etc... and then make corrections.

There is a certain EQ curve though that is kind of a "starting place" for mixing that it pretty similar to the Harman target. But you try to reach that already in recording and balancing the instruments to fill the whole audio spectrum.

 

[Koeitje]

I have Sony WH-1000XM3's that are extremely bass heavy with recessed treble. The moment I put it on I noticed it sounded nothing like my loudspeakers. With Wavelet on Android I got them equalized to the Harman Curve (Wavelet uses AutoEQ curves) and it sounded much better. But I still boosted the bass a bit afterwards, because I enjoy that . For reference, the bass is about 6dB too heavy compared to the Harman Curve. I opted for 3dB too heavy in the end. The equalisation did fix the midrange and treble.

 

[terasankka]

Can someone point me to a resource where I can find Harman Curve 2013 and 2018 EQ settings for Audeze LCD-2 Classic?

 

[dominikz]

Can someone point me to a resource where I can find Harman Curve 2013 and 2018 EQ settings for Audeze LCD-2 Classic?

This is from Oratory1990: https://www.dropbox.com/s/4341dbn3nrcou4k/Audeze LCD-2 Classic.pdf?dl=0

Full index of presets: https://www.reddit.com/r/oratory1990/wiki/index/list_of_presets

 

[terasankka]

This is from Oratory1990: https://www.dropbox.com/s/4341dbn3nrcou4k/Audeze LCD-2 Classic.pdf?dl=0

Full index of presets: https://www.reddit.com/r/oratory1990/wiki/index/list_of_presets

What is Oratory1990?

 

[dominikz]

What is Oratory1990?

A reddit user / acoustic engineer who does headphone measurements and corresponding EQ presets aiming at Harman 2018 curve.