Headphone Measurements Using Brüel & Kjær 5128 HATS

[JIW]

A note about the distortion levels of the silverscreen, these are taken in a headphone store where the owner was moving around an working.
The black screen is in a bedroom in the house.
For this reason you may see some peaks and elevation that isn't there in reality.

Also note that the tonal differences between the black and silver driver are primarily caused by the different compression of the pads.
Below the black driver with new pads (taken at home) vs the 2017 (silverscreen) taken in the mentioned store.
View attachment 78494

The figures I posted were all described as having new pads apart from relative distortion for the black driver. However, after looking at the figure you posted, the last figure in my previous post seems to compare the black driver for new vs. old pads rather than black vs. silver drivers.

 

[amirm]

Someone asked about EARS comparison. Here is that for HD-650 in a couple of positions:

Amazing how flat it looks compared to 5128:

Deviation starts at around 500 Hz and gets much worse due to lack of simulation for ear resonances. Calibration can't fix that because one doesn't know how much energy there is in that region for each headphone to resonate.

 

[LeftCoastTim]

@amirm Have you considered that the Harman target applies only to the particular measurement rig used by Harman?

 

[amirm]

Comparing to @solderdude's measurements of various HD650, it seems you have what is generally called the black (screen) driver version.

I just looked and it looks like the silver version! I bought it maybe 2 to 3 years ago. When did they switch from one type to the other?

 

[amirm]

@amirm Have you considered that the Harman target applies only to the particular measurement rig used by Harman?

Of course. The target is a computed one based on Free-field and diffused-field HRTFs from BK and then using research that uses those two measurements to create a mixed target for a real room based on its characteristics. This response is then modified based on how Harman has modified a standard target to suite listener preference. @Mad_Economist explains that in the intro and spreadsheet.

It is the best thing we have until Harman creates one for 5128 (which they own).

 

[Jimbob54]

View attachment 78499

Can I say that I love that there’s a handle on the right side of this thing? Would love to see someone carrying it down the street like a briefcase.

See Thomas's post about taking it to the camper.

 

[zermak]

Great to see some number and our experts debating about it and how to move forward from here. I like this community sharing of ideas and so on

@amirm If I may ask, how sensitive are the mics in it? Could it record a sweep (at like 85dBA/C) from a stereo loudspeakers setup (already measured and close to the Harman curve; and if I recall you own a pair of Salon?) and then compare the raw data (and the compensated ones by @Mad_Economist ) to have an idea of how off/close are the compensation curves?

If my understanding was correct this HATS should be a step forward from the competition to measure frequencies over 10kHz but I see it is very inconsistent and very headphone and position dependant; am I missing something?

 

[amirm]

Oh and no sneaking off with it to cuddle in the camper van ! They definitely won't want it back after that.

What kind of comment is that for heaven's sake Thomas? The thing is happily married. Its spouse is waiting for it at home....

 

[Mad_Economist]

If my understanding was correct this HATS should be a step forward from the competition to measure frequencies over 10kHz but I see it is very inconsistent and very headphone and position dependant; am I missing something?

The 5128 is a step forward in terms of the accuracy of its emulation of the human ear above 10khz - the trouble is, an accurate emulation of the human ear will still display substantial variability with headphone positioning above 10khz because...human ears do that.

 

[zermak]

The 5128 is a step forward in terms of the accuracy of its emulation of the human ear above 10khz - the trouble is, an accurate emulation of the human ear will still display substantial variability with headphone positioning above 10khz because...human ears do that.

Yes I am aware of that but does it mean Amir will always have to take like 5 (probably more?) measurements (from a perfect position and seal to just around the hears positions) to show the average response of a headphone or will he aim to get the perfect position and seal to show the best response the headphones can have (like the brand of it intended to)?

 

[JIW]

I just looked and it looks like the silver version! I bought it maybe 2 to 3 years ago. When did they switch from one type to the other?

I found this article from October 2009 describing the change. Maybe the differences are due to unit variation or the condition of the pads.

 

[Mad_Economist]

Yes, excellent post here. Seems to me it is an unsolvable problem as it is being approached. All this kind of effort reminds me of epicycles in the Ptolemaic system. Though a flawed idea, various little refinements actually made it predict the planetary movements fairly well eventually. So much so that when the more correct Copernican explanation came along its biggest problem was that it lacked the precision of the old patched together Ptolemaic explanation.

This headphone problem is going to need something new and overlooked as an approach for it to actually work. Like Kepler's use of the ellipse in the Copernican system finally giving high accuracy in the correct explanation.

Wished I were the headphone Kepler, but I'm not. It is apparent the current approach isn't getting things too much further along according to my opinion.

I think the idea our brain continually recalibrates itself is supported by quite a bit of known information about how the brain works. One I've mentioned is within 2-3 weeks artificial pinna implants that totally ruin directional hearing are fully recalibrated to original accuracy. And when those are removed directional accuracy returns to a prior calibration even more quickly.

So what kind of testing can recalibrate on the fly in a way that it predicts headphones with the clearest most accurate response or do any of them? IEMs would seem to simplify matters because they bypass the pinna altogether. Still not an easy to solve problem even from there.

The bolded section is pretty tautological to any approach to target headphone response that relies on some sort of HRTF as a target - and, for that matter, to any consideration of speaker listening that accommodates the idea that the same speaker can sound the same to different people at the same time, or to the same person in different instances. Perceived response is in relation to perceived acoustic source, which in turn our brains infer in part from data from our eyes, but also from the continuously-adjusted set of "HRTF filters" which we acquire simply by living and having ears.

What I'm entirely confused by is that you seem to regard this as presenting an issue for a given compensation or pinna construction - I'd suggest that you haven't thought the implications of the relationship between our HRTF features and the frequency response elements our brain expects in subjectively "flat" sound through entirely...

 

[Mad_Economist]

Yes I am aware of that but does it mean Amir will always have to take like 5 (probably more?) measurements (from a perfect position and seal to just around the hears positions) to show the average response of a headphone or will he aim to get the perfect position and seal to show the best response the headphones can have (like the brand of it intended to)?

That's a good question! @amirm, any thoughts?

Personally, I regard "average response", "edge case behavior", and "best-case behavior" as all interesting aspects of a headphone. From a designer's standpoint, I absolutely consider robustness to ex. variation in coupling to be a very significant figure of merit (because end users aren't going to spend 15 minutes positioning these things), and I'd love to see it included in more test sets (as Keith Howard presently does at HeadphoneTestLab).

 

[Chocomel]

Yeah positional variance, how the headphone reacts to breaking of the seal is all part of the performance of a headphone and ideally a set of Measurements would show this behavior.

 

[Mad_Economist]

That's the speaker curve, not the headphone curve, they are quite different. IEMs are different again.

I'm not sure I'd say "quite different"

FWIW, Sean has a paper regarding subgroups of headphone preference.

 

[zermak]

Personally, I regard "average response", "edge case behavior", and "best-case behavior" as all interesting aspects of a headphone. From a designer's standpoint, I absolutely consider robustness to ex. variation in coupling to be a very significant figure of merit (because end users aren't going to spend 15 minutes positioning these things), and I'd love to see it included in more test sets (as Keith Howard presently does at HeadphoneTestLab).

I agree with you mainly because we all have different ears and so someone can figure it out just looking at the best-to-worst case scenario to consider an headphone or find something else and I am pretty sure the variations will be even higher with the on-ears. So far the in-ears looks like the easiest to measure when you have the average simulated ears canals (and maybe this is were to start to see improvements over other HATS, removing the pinna variable).

 

[Mad_Economist]

I agree with you mainly because we all have different ears and so someone can figure it out just looking at the best-to-worst case scenario to consider an headphone or find something else and I am pretty sure the variations will be even higher with the on-ears. So far the in-ears looks like the easiest to measure when you have the average simulated ears canals (and maybe this is were to start to see improvements over other HATS, removing the pinna variable).

FWIW, on-ears may vary less than you'd think with good fit - the small pads limit the range of positioning to some extent, and anthropomorphic HATS' pinnae (e.g. modern pinnae for KEMAR, 4128, 5128, etc) are designed to deform in ways that mirror human ears.

Vis-a-vis in-ears, that's actually something of a point of controversy/interest regarding the 5128, it consistently shows different low-frequency behavior to conventional IEC711/60318-4 couplers - Sam Vafaei (formerly of RTings) is of the view that this constitutes a problem, whereas Jude Mansilla of Head-Fi regards it as a feature. I remain undecided for lack of thorough analysis.

 

[Chocomel]

I agree with you mainly because we all have different ears and so someone can figure it out just looking at the best-to-worst case scenario to consider an headphone or find something else and I am pretty sure the variations will be even higher with the on-ears. So far the in-ears looks like the easiest to measure when you have the average simulated ears canals (and maybe this is were to start to see improvements over other HATS, removing the pinna variable).

It goes even further than your ears but also the shape of your head where some people might not be able to get a good seal and in turn get wildly different responses. For example the HD820 .

 

[solderdude]

I found this article from October 2009 describing the change. Maybe the differences are due to unit variation or the condition of the pads.

Yes it was about 10 years ago. My (black) driver was bought in 2004.
As far as I know the only change was the material used around the driver. The pads seem to have a different internal shaped foam and slightly different velour as well. No idea id that happened around the same time.
In any case the differences are small for the driver, bigger for the pads. Of course the foam did also deteriorate over the last 16 years that may certainly have a lot to do with it.

I prefer the old pads with EQ specifically for that combo.

 

[amirm]

One issue I am finding with the pinnae on 5128 is its material characteristics. It is soft and flexible which is its claim to fame. But it also has high stiction with similar material such as the silicone tips of IEMs. They are like magnets and instantly stick to each other, making it harder to find the ear canal and push it in there. Hard pinnae don't have this issue. Real ear has a bit of this and the reason I like silicon tips but not as much as what I am seeing on 5128.

I guess the solution is to use foam tips.

But I also worry whether the pinnae is grabbing onto the pads of circumaural headphones. Hard to know how bad this issue is as the pinnae is occluded by the headphone pads.