• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

HBK Headphone Measurement Talks from Head-Fi and Sean Olive

MayaTlab

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
924
Likes
1,512
For highs I intend the frequencies above which directivity plays a role in the measured FR and a lumped EQ is not applicable anymore. Broadly speaking, I think it should be around 2 kHz, if I remember some of the graphs I've seen in the past correctly.
1 kHz is probably still in the direction independent area, which I think (but don't quote me on this) can be corrected universally by EQ, at least under perfect seal conditions (which is like saying that practically speaking the distributed acoustic impedance behaviour could apply to most of the audio range, especially with HPs with poor sealing).

By directivity, what exactly do you mean ? Changes during spatial averaging when measuring headphones on a test rig ? HRTF ?

I'm trying to understand what you have in mind in the context, for example, of HRTF measurements being mostly done at the ear canal entrance, which seems to be sufficient to gather most direction dependent information up to at least around 7 kHz according to a few articles I've read on the subject, or in the context of headphones still operating under pressure chamber conditions up to a few kHz at least, unless I'm mistaken.
 
D

Deleted member 16543

Guest
By directivity, what exactly do you mean ? Changes during spatial averaging when measuring headphones on a test rig ? HRTF ?

I'm trying to understand what you have in mind in the context, for example, of HRTF measurements being mostly done at the ear canal entrance, which seems to be sufficient to gather most direction dependent information up to at least around 7 kHz according to a few articles I've read on the subject, or in the context of headphones still operating under pressure chamber conditions up to a few kHz at least, unless I'm mistaken.

I mean that the sound wave inside the ear canal is not a flat front type of wave until it gets way down into the canal, right up to the eardrum, and it's distribution in space still depends on where the sound is coming from.
I'm not talking about intensity change with frequency like you see with HRTFs.
I'm talking about the movement of air particles not having assumed a flat front wave kind of behavior inside the canal at a certain depth, yet.

This is evident at 'properly high' frequencies (10 kHz and up) based on some study @Mad_Economist was kind enough to point me to a while ago, but comparing speaker measurements at different azimuth they seem to start as low as at least 3 kHz, which wasn't too clear in that study if I remember correctly.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,408
Location
Sweden
I mean that the sound wave inside the ear canal is not a flat front type of wave until it gets way down into the canal, right up to the eardrum, and it's distribution in space still depends on where the sound is coming from.
I'm not talking about intensity change with frequency like you see with HRTFs.
I'm talking about the movement of air particles not having assumed a flat front wave kind of behavior inside the canal at a certain depth, yet.

This is evident at 'properly high' frequencies (10 kHz and up) based on some study @Mad_Economist was kind enough to point me to a while ago, but comparing speaker measurements at different azimuth they seem to start as low as at least 3 kHz, which wasn't too clear in that study if I remember correctly.
A bit OT, but do you have a link to such studies? Curious what the difference in wavefront is at the entry vs. eardrum with respect to azimuth.
 

CedarX

Senior Member
Forum Donor
Joined
Jul 1, 2021
Messages
411
Likes
625
Location
USA
I'm late to the party. I'll catch up on the previous posts sooner or later, but I just read this and I wanted to provide a possible explanation for the lack of bass and different high frequency shape.
I apologize if somebody already did.

The 5128 employs a mic capsule that supposedly matches the eardrum's own acoustic impedance. The eardrum has less sensitivity at low frequencies, therefore the lower curve in that area.
For the highs, instead, the difference is partly due to said impedance and partly due to the fact that the ear canal in the 5128 is completely anthropomorphic, while GRAS's is only accurate up to about 1 cm, with the rest of the canal continuing as a cylinder.

The eardrum impedance matching by the 5128's capsule is, by its own nature, a lumped direction independent parameter, so that part could easily be accounted for by a correction EQ, for ALL headphones.
But the canal shape is a distributed direction dependant parameter, so that can't be corrected with a one size fits all EQ.

The fact that you 'discovered' that target translation can't happen among rigs is therefore not surprising at all (I've been saying that for a long time myself, but whatever...).

The assumption that this proves that the GRAS rig is more 'true' because of the preference scale drop is only a superficial analysis.
Once a target curve is provided for the 5128, it will be most likely 'truer', and definitely not any less accurate than GRAS's one.
But once again, GRAS's rig IS the most readily usable rig at the moment, given the 5128's absence of said target curve.

There is value in the measurements you do. I even feel confident to say it is probably the highest value possible at the moment, speaking in practical terms.

But it is definitely not the highest value possible in absolute. All it takes is for B&K to produce a target curve.
I have no idea what's taking those guys so long.
If I worked there I would get that done in a couple weeks, tops.. And I'm fairly lazy.
Weird.
Isn’t there a more fundamental issue than the correlation question between GRAS and BK rigs? If I am not mistaken, the GRAS target curve was developed by Harman to maximize the preference scores across a representative population of individuals. If the FR correlation between GRAS and BK is non-linear, isn’t it wrong to assume that the BK target you would get using this FR-based transfer function remains the one that maximizes the preference scores?
 
D

Deleted member 16543

Guest
Isn’t there a more fundamental issue than the correlation question between GRAS and BK rigs? If I am not mistaken, the GRAS target curve was developed by Harman to maximize the preference scores across a representative population of individuals. If the FR correlation between GRAS and BK is non-linear, isn’t it wrong to assume that the BK target you would get using this FR-based transfer function remains the one that maximizes the preference scores?

Not unless the majority of people don't like balanced speakers, or they expect a different balance when listening to HPs instead of speakers.
This second possibility is maybe not that farfetched. Didn't Harman determine that was the case? That's why I added the bass preference compensation EQ to the equation, in my proposed method to find a target for the 5128.
 
Last edited by a moderator:
D

Deleted member 16543

Guest
A bit OT, but do you have a link to such studies? Curious what the difference in wavefront is at the entry vs. eardrum with respect to azimuth.

I'll have to dig it out of some past posts. I'm pretty sure it was in a conversation on diyaudio with a fellow that was doing something similar to my ear models, but 3d printed the rest of the head as well.
Pretty smart and determined guy, as a few I have had the privilege to virtually meet here and there in the DIY community, thanks to the weird audio interests I share with some of them.

To be fair, the study was using a closed entrance canal to measure what you're looking for, while to do things properly you would need an infinitesimally small probe microphone placed at the canal entrance (there's that distributed impedance thing again.. Geometric boundary have to be exact, for measurements to be exact).

Nonetheless, it was indicative of the fact that at the canal entrance the directionally dependent part of the outer ear (as in pinna+canal) is still effective down to at least 3 kHz.
Funny enough, the few dBs of variation in response depending on the direction were summarily brushed off by the people who did the study, saying that they were not that meaningful.
That part of the study I beg to differ with. Just ask any recording engineer if a 2-3 dB bump or dip at 3 kHz makes no difference.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,408
Location
Sweden
I'll have to dig it out of some past posts. I'm pretty sure it was in a conversation on diyaudio with a fellow that was doing something similar to my ear models, but 3d printed the rest of the head as well.
Pretty smart and determined guy, as a few I have had the privilege to virtually meet here and there in the DIY community, thanks to the weird audio interests I share with some of them.

To be fair, the study was using a closed entrance canal to measure what you're looking for, while to do things properly you would need an infinitesimally small probe microphone placed at the canal entrance (there's that distributed impedance thing again.. Geometric boundary have to be exact, for measurements to be exact).

Nonetheless, it was indicative of the fact that at the canal entrance the directionally dependent part of the outer ear (as in pinna+canal) is still effective down to at least 3 kHz.
Funny enough, the few dBs of variation in response depending on the direction were summarily brushed off by the people who did the study, saying that they were not that meaningful.
That part of the study I beg to differ with. Just ask any recording engineer if a 2-3 dB bump or dip at 3 kHz makes no difference.
I'll have to dig it out of some past posts. I'm pretty sure it was in a conversation on diyaudio with a fellow that was doing something similar to my ear models, but 3d printed the rest of the head as well.
Pretty smart and determined guy, as a few I have had the privilege to virtually meet here and there in the DIY community, thanks to the weird audio interests I share with some of them.

To be fair, the study was using a closed entrance canal to measure what you're looking for, while to do things properly you would need an infinitesimally small probe microphone placed at the canal entrance (there's that distributed impedance thing again.. Geometric boundary have to be exact, for measurements to be exact).

Nonetheless, it was indicative of the fact that at the canal entrance the directionally dependent part of the outer ear (as in pinna+canal) is still effective down to at least 3 kHz.
Funny enough, the few dBs of variation in response depending on the direction were summarily brushed off by the people who did the study, saying that they were not that meaningful.
That part of the study I beg to differ with. Just ask any recording engineer if a 2-3 dB bump or dip at 3 kHz makes no difference.

Thanks, would be much appreciated if you can find them. I can look at DIY audio as well.

My primary interest is to have reasonable target curves for headphones mainly over ears HPs, for binaural recordings. Most related to timbre correction. Direction would need head tracking as well, but that is more than I need.
 
D

Deleted member 16543

Guest
Isn’t there a more fundamental issue than the correlation question between GRAS and BK rigs? If I am not mistaken, the GRAS target curve was developed by Harman to maximize the preference scores across a representative population of individuals. If the FR correlation between GRAS and BK is non-linear, isn’t it wrong to assume that the BK target you would get using this FR-based transfer function remains the one that maximizes the preference scores?

Sorry, I think I read your question hastily, earlier..
Are you asking if trying to find a transfer function between the two rigs, to apply the GRAS target curve to the 5128, is appropriate?
I think Sean answered himself. It's not accurate. It couldn't possibly be. But it's probably a good ballpark starting point, without doing some 5128 based research.
 

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
I mean that the sound wave inside the ear canal is not a flat front type of wave until it gets way down into the canal, right up to the eardrum, and it's distribution in space still depends on where the sound is coming from.
I'm not talking about intensity change with frequency like you see with HRTFs.
I'm talking about the movement of air particles not having assumed a flat front wave kind of behavior inside the canal at a certain depth, yet.

This is evident at 'properly high' frequencies (10 kHz and up) based on some study @Mad_Economist was kind enough to point me to a while ago, but comparing speaker measurements at different azimuth they seem to start as low as at least 3 kHz, which wasn't too clear in that study if I remember correctly.
Tsunoda, Hara, & Nageno 2017 seems to disagree with this:
1633717331957.png

1633717348817.png

From our past DM dialogues on the subject, I would also point out this figure from Middlebrooks et al 1989

index.php


And this one from Hammershøi & Møller 1996:
index.php

Broadly, directional effects of the canal structure are, if they exist, present only 1, at very high frequencies, and 2, in magnitudes which are similar to the measurement variation found at said frequencies...
 

adude995

Member
Joined
May 17, 2019
Messages
15
Likes
5
Location
aut + fin
Would make sense, since the wavelength of a 20kHz tone is roughly 1,7cm.
So the needed half of a wavelength for wave propagation is about 0,85cm which is more than the diameter of the ear canal in total.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,537
Location
Seattle Area
I'm late to the party. I'll catch up on the previous posts sooner or later, but I just read this and I wanted to provide a possible explanation for the lack of bass and different high frequency shape.
I apologize if somebody already did.
We don't need "possible explanation." We need real explanation from BK if one exists.

Regardless, if you read the rest of the thread and understood what I wrote as summary in OP, you realize that it doesn't matter. The only research into what we think is a good sounding headphone is from Harman and that research and correlation to measurements do not work for BK 5128. There is an attempt to make some correction but otherwise what I have been saying all along is true: you can't use BK 5128 or any other fixture to draw conclusions with respect to preference for a headphone sound. This is why I bought the GRAS fixture and not 5128.

As to your tired argument of 5128 being more accurate, that is a hypothesis in need of proof. Just because BK has done this and that research doesn't mean any goals have been met. They, the company, needs to do that research and provide the verification. What has gone into design of something is not the same as achieving said results.

Also the whole notion of an average of many people's ears being a model that is most accurate is non sequitur to me. You can't average a bunch of stuff, i.e. filter, and wind up with a specific thing that is accurate. Average temps today may be 60 degrees but that says nothing about the low of 40 and high of 80.

The goal then should be to see how well a measurement predicts listener preference and not how "accurate" it is. We need actionable data. Not random technical goals being met.

Now, it may turn out that a brand new, multi year research with BK 5128 leads to better correlation with its measurements. Until that is done though, the arguments about accuracy are moot. We don't know that it is more accurate, nor better predictor of listener preference.

As far as I am concerned, post measurement Equalization and listening test is key to finalizing the measurement data.
 

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
As to your tired argument of 5128 being more accurate, that is a hypothesis in need of proof. Just because BK has done this and that research doesn't mean any goals have been met. They, the company, needs to do that research and provide the verification. What has gone into design of something is not the same as achieving said results.
Objectively, the 5128 is more accurate in terms of ear impedance than the 45CA - that's proven simply by comparing the impedance of the two fixtures to one of a number of different measures of human ear Z, including those undertaken by B&K in the 5128 research, and Gunnar Rasmussen and Per V. Brüel's classic work which lead to the IEC711. Equally, this isn't likely to matter much for circumaural headphones...at all.

I would honestly be pretty surprised if the 5128 produced meaningfully better correlation with subjective sound quality than the 45CA for over/on-ear headphones. For in-ear designs, it may end up being better (or worse - Sam Vafaei at RTings viewed it as worse, I'm inclined towards better), but for circumaural stuff the differences that aren't just "different pinna" stuff are confined to the frequency extremes, which matter less.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,408
Location
Sweden
Tsunoda, Hara, & Nageno 2017 seems to disagree with this:
View attachment 157928
View attachment 157930
From our past DM dialogues on the subject, I would also point out this figure from Middlebrooks et al 1989

index.php


And this one from Hammershøi & Møller 1996:
index.php

Broadly, directional effects of the canal structure are, if they exist, present only 1, at very high frequencies, and 2, in magnitudes which are similar to the measurement variation found at said frequencies...
Interesting. I need to get hold of the JASA article somehow.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,537
Location
Seattle Area

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
Again, that is design criteria, not verification.
The design criterion there is...human ear impedance as a function of frequency, though. It's definitely true that the behavior of the 5128 (below 100hz and above 8khz) is more accurate to the impedance of a human ear than an IEC711/60318-4 sim. That also means it will be more accurate, all else equal, in terms of headphone sound pressure level at the eardrum, but all else is far from equal at those frequencies, so it's really not worth much concern outside of, as said, in-ear monitors perhaps.
Interesting. I need to get hold of the JASA article somehow.
It's on researchgate
 

MayaTlab

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
924
Likes
1,512
but for circumaural stuff the differences that aren't just "different pinna" stuff are confined to the frequency extremes, which matter less.

If you have seen the slides with the individual measurements from Sean Olive's presentation, what would be according to you likely hypotheses to explain the differences observed for, for example, the K550 between 300 and 5kHz, or the Bose QC35II between 1 and 5kHz or so ?
 

Mad_Economist

Addicted to Fun and Learning
Audio Company
Joined
Nov 29, 2017
Messages
543
Likes
1,618
If you have seen the slides with the individual measurements from Sean Olive's presentation, what would be according to you likely hypotheses to explain the differences observed for, for example, the K550 between 300 and 5kHz, or the Bose QC35II between 1 and 5kHz or so ?
I have, and in fact scraped the data and put it into Excel (pardon the smoothing - scraping a plot from a pixeled video screenshot is jagged).
1633726048588.png

It's pretty typical for different HATS to have variations on this order with a given headphone, particularly those which compress the pinnae somewhat, and thus are particularly sensitive to geometry of the ear both on terms of the air volume presented to the driver, and how the ear tends to fold or bend. In comparisons of a 4128 and KEMAR (both using the same B&K IEC60318-4 ear simulators), I saw comparable or larger variations in that same band on specific headphones.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,408
Location
Sweden
The design criterion there is...human ear impedance as a function of frequency, though. It's definitely true that the behavior of the 5128 (below 100hz and above 8khz) is more accurate to the impedance of a human ear than an IEC711/60318-4 sim. That also means it will be more accurate, all else equal, in terms of headphone sound pressure level at the eardrum, but all else is far from equal at those frequencies, so it's really not worth much concern outside of, as said, in-ear monitors perhaps.

It's on researchgate
Thanks!
 

DualTriode

Addicted to Fun and Learning
Joined
Oct 24, 2019
Messages
893
Likes
593
snip...

Regardless, if you read the rest of the thread and understood what I wrote as summary in OP, you realize that it doesn't matter. The only research into what we think is a good sounding headphone is from Harman and that research and correlation to measurements do not work for BK 5128. There is an attempt to make some correction but otherwise what I have been saying all along is true: you can't use BK 5128 or any other fixture to draw conclusions with respect to preference for a headphone sound. This is why I bought the GRAS fixture and not 5128.

snip...

We agree that the 5128 is not the "gold standard" or standard yardstick.

The Harman target curve was developed with measurement data from the JBL Modified GRAS 45 CA. The JBL Modified GRAS 45 CA used different couplers and pinna.

The same is true for the GRAS 45 CA-10. It uses different couplers and pinna than the JBL Modified GRAS 45CA did . There is not a direct conversion between the two different GRAS measurement instruments.

Same as the 5128 the GRAS 45 CA-10 is also not the "gold standard" or standard yardstick we are looking for.

Thanks DT
 
  • Like
Reactions: ADU
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,537
Location
Seattle Area
The design criterion there is...human ear impedance as a function of frequency, though.
As I said, design criteria is not proof of accomplishment. Independent testing and research needs to be performed to ascertain that.
 
Top Bottom