Headphones and the Harman target curve

[isostasy]

Ah, ok, I was assuming it was a Senn from their 'Made in Ireland' period, which I'm a big fan of.
I recently bought and refurbed a pair of HD555, just so could I could complete the nomenclature from HD525 to HD565 (I know, drrrp)
I did the well-documented mod of removing the piece of adhesive rubber from inside the baffles and it transformed the sound from something mediocre to something quite impressive!
Assuming the HD560 comes later/higher in the more recent range?

Not a huge fan of the later mid-range models tbh and you can clearly see where costs have been trimmed and cheaper materials and components are used.
The older 5*5 series were far superior, and comfortable!
Here's a fun fact- did you know the driver capsules from the 5*5 range and the totl 6** range drivers were all interchangeable?
I've had a lot of fun swapping drivers around in different chassis. HD650 drivers in a closed-back HD265 chassis for sh!ts n giggles for example. The business end of the FR literally made my brain wobble!


View attachment 242980
View attachment 242981

Hey I recently got a HD265 and I've been through what I assume are your posts on head-fi, unless there's someone else with such an impressive HD5x5 collection!

Wondering what pads you're using on them now that Sennheiser no longer make them. Mine came with some pleather pads but I have no idea if they're original; the material is more like stretchy bin liner to be honest. I'm thinking my best bet is to just put HD600 velour pads on and remove the backs to essentially convert them into a bargain HD565.

I have measurements of the HD265 in various forms (pleather vs velour pads, open vs closed) against my HD6XX which I can post but don't want to take up too much space here on an off-topic.

Regarding your initial post much earlier on, it is interesting to note that all those Sennheiser headphones you've got there, bar the HD265, conform well to the Harman target, with the only notable divergence being in sub-bass response. I think you can lay your worries over conformity to rest: headphones are for reproducing recordings, you can choose what to listen to. If anything having a known standard for something like headphone frequency response is a win in the battle against conformity as it allows everyone to more transparently listen to the recorded material, i.e. it allows you to express opinions on the material as it is rather than as it sounds coloured by your reproduction device. This is getting a bit too philosophical... no replies expected

I think the debate is set to roll on and that as with cables, there will be 2 camps; the believers and the non-believers.

I would say rather, in this case, there are those who have read and understood the Harman research, and those who haven't.

 

[someguyontheinternet]

In its current form: unfortunately I can´t.

Layman voters should either be removed from skewing the results to unneutral BS-direction, or have two separate curves:



Harman Amateur - any moron can vote (current Harman curve). Harman curve for general mass. Main purpose is to please, neutrality = file(s) not found.


Harman Professional - only oath made audiophiles / AES-card holders / dual-decker hardcore ASR-forum members vote. True Harman curve for serious audio people. The primary target is neutrality, pleasing comes secondary.


If current trend continues, my prediction of year 2040 Harman curve, is bass emphasis tickling +10 dB.

How did you come to the conclusion that neutrality and preference don't map to the same target? Flat speakers are preferred by most listeners so why should the reverse not hold for headphones?
Also Harman research found no difference in preference rankings between experience groups:

 

[fpitas]

As easy as it is to EQ headphones these days, this seems like a lot of furor over nothing. Harman wants to sell headphones, and wanted a design curve. It's no surprise somebody won't like it.

 

[Sombreuil]

I don't think that's how science works, especially when both the Sennheiser HD 600/650 match the Harman curve pretty well.

 

[Robbo99999]

In its current form: unfortunately I can´t.

Layman voters should either be removed from skewing the results to unneutral BS-direction, or have two separate curves:



Harman Amateur - any moron can vote (current Harman curve). Harman curve for general mass. Main purpose is to please, neutrality = file(s) not found.


Harman Professional - only oath made audiophiles / AES-card holders / dual-decker hardcore ASR-forum members vote. True Harman curve for serious audio people. The primary target is neutrality, pleasing comes secondary.


If current trend continues, my prediction of year 2040 Harman curve, is bass emphasis tickling +10 dB.

Try the Harman 2013 Headphone Curve (blue line above), it's theoretically more exactly true to flat anechoic speakers in a room. It's basically equal to the tilt of the Harman Speaker Curve applied to the dummy in ear measurement of in-room flat EQ'd speakers - so essentially it's a direct representation of the Harman Speaker Curve without any other added bass & treble. If you already have a headphone EQ'd to the 2018 Curve then you can you use the attached filters to convert your 2018 EQ into a 2013 EQ (EqualiserAPO config file attached to post):

Filter: ON LSC Fc 1000 Hz Gain -0.5 dB Q 0.707
Filter: ON LSC Fc 160 Hz Gain -1.6 dB Q 0.707
Filter: ON PK Fc 30 Hz Gain -0.7 dB Q 2
Filter: ON HSC Fc 3350 Hz Gain -1.6 dB Q 0.707
Filter: ON PK Fc 2800 Hz Gain 1.6 dB Q 2

So you'd run these in addition to your 2018 EQ.

 

[someguyontheinternet]

By "scientific" method: looking at the graph. Harman bass is about 5 dB lifted i.e. hotter than neutral.
Wonder what kind of flute music and/or silencer in the a** -SPL-levels they might been using in those listening tests. That much bass overdose should get super fatiguing on already bassy program material and/or if volume is non-backround level. At least for most sane persons.

Those "experienced" Harman curve -test group members might have been some half-deaf DJ and university´s janitor? The rest being just senior students of university, about half probably in hangover?

By the way, experience alone won´t do expert or judgmentally capable technical person. Otherwise my late grandma would have been ultimate expert in telephone and television matters: she had used both of them on regular basis since 60´s.



AES -member card would be good practical criteria for that suggested Harman Professional -curve test crew. Person with such could be trusted in high statistical probability to not be typical bass-dimwit regarding audio.

"I don't understand where it comes from" is no valid argument for criticism. The HRTFs simply do not map to a flat line which results in peaks and valleys.
There is some variance in the bass preference (and to a lesser degree, treble), but this is also explicitly disclosed in Harman research. That is simply no refutation of Harman research as you seem to claim.

Also you don't seem to understand the point of experience group comparisons. Between all experience levels the rankings were very stable. This is not a factor that seems to impact preference rating. Although there is a difference in the scores by groups (less experience -> smaller score ranges, more experience -> larger score ranges) there is currently no indication whatsoever that listener experience has any sort of significant influence on preference scores.

If you make the claim that experience or a professional occupation has such a huge influence, I would like to ask you to provide evidence that goes beyond "trust me I know better than you".

 

[isostasy]

By "scientific" method: looking at the graph. Harman bass is about 5 dB lifted i.e. hotter than neutral.
Wonder what kind of flute music and/or silencer in the a** -SPL-levels they might been using in those listening tests. That much bass overdose should get super fatiguing on already bassy program material and/or if volume is non-backround level. At least for most sane persons.

Those "experienced" Harman curve -test group members might have been some half-deaf DJ and university´s janitor? The rest being just senior students of university, about half probably in hangover?

By the way, experience alone won´t do expert or judgmentally capable technical person. Otherwise my late grandma would have been ultimate expert in telephone and television matters: she had used both of them on regular basis since 60´s.



AES -member card would be good practical criteria for that suggested Harman Professional -curve test crew. Person with such could be trusted in high statistical probability to not be typical bass-dimwit regarding audio.

The mere act of looking at graphs is not 'scientific method'. The scientific method is a process of experimentation with controlled variables which is upheld or disputed through attempts at replication. 'Looking at the graph' is an act of interpreting the data. According to your line of thinking, journalists employ the 'scientific method' every time they read and report on a piece of research.

You make unsubstantiated assertions about the sample selection used in the research, rooted in prejudices you hold regarding people who put on live performances for a living, people responsible for the upkeep of university buildings, and students.

'By the way, experience alone won't do expert or judgmentally capable technical person'. This is nonsensical, but I think you're trying to make a comment on the supposed expertise on the sample, which, again, you need to substantiate based on the selection criteria in the original research, otherwise you are simply making stuff up.

Your last point demonstrates some fundamental misunderstandings around sample selection. There is large statistical variance amongst small sample sizes, but this is reduced the larger sample you use. If Sean Olive had decided his 'trained listener' group would solely consist of AES members, as you suggest, all you would end up with is data on the average preferences of AES members

 

[fpitas]

Your last point demonstrates some fundamental misunderstandings around sample selection. There is large statistical variance amongst small sample sizes, but this is reduced the larger sample you use. If Sean Olive had decided his 'trained listener' group would solely consist of AES members, as you suggest, all you would end up with is data on the average preferences of AES members

And since their target market is not AES members, that would be silly. They're in business to sell headphones, people, not conform to what someone thinks is "right".

 

[Astoneroad]

They're in business to sell... not conform to what someone thinks is "right".

This is a universal statement that transcends the audio market and is underlying to every headline for the last 2000 years.

 

[fpitas]

This is a universal statement that transcends the audio market and is underlying to every headline for the last 2000 years.

Indeed

 

[Sean Olive]

By "scientific" method: looking at the graph. Harman bass is about 5 dB lifted i.e. hotter than neutral.
Wonder what kind of flute music and/or silencer in the a** -SPL-levels they might been using in those listening tests. That much bass overdose should get super fatiguing on already bassy program material and/or if volume is non-backround level. At least for most sane persons.

Those "experienced" Harman curve -test group members might have been some half-deaf DJ and university´s janitor? The rest being just senior students of university, about half probably in hangover?

By the way, experience alone won´t do expert or judgmentally capable technical person. Otherwise my late grandma would have been ultimate expert in telephone and television matters: she had used both of them on regular basis since 60´s.



AES -member card would be good practical criteria for that suggested Harman Professional -curve test crew. Person with such could be trusted in high statistical probability to not be typical bass-dimwit regarding audio.

AES- member card as a criteria? Have you not seen the audiometry data collected by the House Ear Institute at AES convention over-the-past few years? Attendees have more hearing loss than the general population Hearing loss is an occupational hazard in the audio industry and on average AES members wouldn't qualify as trained listeners in the Harman panels b/c they would fail our audiometey screening process. We don't make hearing aids yet.

 

[Sean Olive]

Question is mostly about bass. Neither of them have that Harman bass boost.

Yes, HD 600 is my reference headphone too. Kinda wheel of headphone world, which re-inventions have failed.

Studio monitors are flat on -axis in the free-field but not when you put them in a semi-reflective room and measure the steady-state response. The response is a combination of the direct and reflected sound and will have a downward sloped curve. That is what the Harman Target is based on.

A headphone with flat bass will sound thin and not correct. That is what our subjective data tells us and there is a logical explanation for it.

It also sounds thin to me and I am a trained musician, Tonmeister, AES fellow and past AES President who still has some hearing left

 

[Keith_W]

Sean do you have a copy of the paper that you could link to. I have a lot of experience reading scientific papers and I might be able to help shed some light on this.

 

[Mulder]

I don´t think just one headphone is ideal for any situation. Or for any person. Harman Curve or not, I think it is impossible to find a one size fit all for any reference curve. When I am listening to headphones outdoors, in trafic etc, I prefer an elevated bass compared to listening at home. Jut as an example.

 

[Cars-N-Cans]

Studio monitors are flat on -axis in the free-field but not when you put them in a semi-reflective room and measure the steady-state response. The response is a combination of the direct and reflected sound and will have a downward sloped curve. That is what the Harman Target is based on.

With respect to speakers, this seems to be one of the most misunderstood aspects of the in-room target. Many seem to think its sloped for "more bass" and I think this is at least in part where some confusion comes from. As you state its what a neutral response of a typical speaker looks like in most listening environments. Its definitely something that could benefit its own discussion of how the room and the radiation pattern/directivity of the speaker combine to produce the in-room measurement. Not being suggestive in saying that person would be you (although your generosity would be greatly appreciated! ) But, it would be a good topic in its own right in one form or another for completeness so people understand why things look the way they do.

 

[solderdude]

You clearly fall within the 25% group that likes less bass than the 'average' group of people that enjoy music.
Nothing wrong with that.
Dismissing research because you like less bass is another thing though.
Believe it or not but most people like headphones to sound like their speakers in their living room.
(I am in the 25% group as well though, the group that prefer less bass than Harman, somewhere between no boost and Harman boost which is a subset of that 25% group)

 

[Robbo99999]

Ok... so Harman curve is actually based on severely colored "loudspeaker-in-room" -sound. Bad goal for audiophile/Hi-Fi purposes...

I'm not gonna pick up this argument with you, but I will pick out this one point in your post. You've got it wrong, Sean is not saying that, he's saying it's based on an anechoic flat speaker in a room, so the opposite of what you're saying.

 

[AdamG247]

I'm not gonna pick up this argument with you, but I will pick out this one point in your post. You've got it wrong, Sean is not saying that, he's saying it's based on an anechoic flat speaker in a room, so the opposite of what you're saying.

Don’t hold your breath waiting on a reply

 

[Cars-N-Cans]

I'm not gonna pick up this argument with you, but I will pick out this one point in your post. You've got it wrong, Sean is not saying that, he's saying it's based on an anechoic flat speaker in a room, so the opposite of what you're saying.

To add to Sean’s comments, I think another thing is the lack of tactile bass. On a speaker system that can be A/B’d with headphones simultaneously, the lack of tactile feedback makes it seem like the headphones have boosted bass. It’s not until each is level matched to a level low enough to take the bass out of the equation that I realize they are in fact the same. And one has to remember that the speaker is shooting for tonally neutral sound reproduction to start with. If I record sound using a decent mic array and listen to it back thru the headphones, it’s clear they are reproducing the sound tonally. This may not be so for all users, but it at least shows for most it will provide a perceptually flat response.

 

[isostasy]

To add to Sean’s comments, I think another thing is the lack of tactile bass. On a speaker system that can be A/B’d with headphones simultaneously, the lack of tactile feedback makes it seem like the headphones have boosted bass. It’s not until each is level matched to a level low enough to take the bass out of the equation that I realize they are in fact the same. And one has to remember that the speaker is shooting for tonally neutral sound reproduction to start with. If I record sound using a decent mic array and listen to it back thru the headphones, it’s clear they are reproducing the sound tonally. This may not be so for all users, but it at least shows for most it will provide a perceptually flat response.

You mean the other way round, surely? If the headphones produce the same bass response at the ear, it will seem as though they have less bass due to lack of tactile feedback, not 'boosted'?