AKG K371 Review (closed back headphone)

[watchnerd]

The K371 has been known for a while to have poor frequency response consistency and seal variation.

Subjectively, I find this to be true for my ears.

They move around a lot, it's hard to get a good seal, and when I do, relatively small shifts in body position or head seem to wiggle them out of place easily, which changes the sound pretty dramatically.

 

[markanini]

Subjectively, I find this to be true for my ears.

They move around a lot, it's hard to get a good seal, and when I do, relatively small shifts in body position or head seem to wiggle them out of place easily, which changes the sound pretty dramatically.

I was able to get a more consistent seal after fixing the headband with zip ties. It does something to the angling of the drivers. IMO the failure of AKG is using a janky headband for such a seal dependent headphone.

 

[Jimbob54]

Talking of distortion, I see there's no absolute distortion vs frequency plot for this headphone. Bit of an odd omission considering all the previous vitriol and furore over other reviewers only posting 'misleading' percentage distortion plots...despite them being commonly used in academic papers (log scaled no less) by leading acoustic scientists in the field of headphone research such as Dr. Sean Olive.

Ain't got no time for a full round of ASR Cleudo. Spit it out.

 

[watchnerd]

I was able to get a more consistent seal after fixing the headband with zip ties.

Fixing meaning making it so it wouldn't move?

 

[DualTriode]

Hello All,

The cognitive gymnastics and conversation about seal, Frequency response and distortion that circle these closed back headphones is interesting.

What do we know about these headphones that is different than others? The first thing that comes to my mind is that there is a sealed air chamber on the backside of the diaphragm where all the ideal gas laws apply, as in PV=NRT.

If you recall the conversations about the distortion caused by Compression Drivers and Horns the same thing happens with a closed back headphone. (Google the JBL Technical Note I think #108). Air is not a very linear spring especially in an enclosed space such as the space in the closed back of a headphone.

Equally interesting is that when we measure a headphone of for that matter when we put a headphone on our head and listen there are a enclosed spaces to some extent sealed to our head around our ears. If the seal is tight to our head or test fixture the sealed air chamber on the backside of the diaphragm is roughly equivalent to the sealed air chamber on the front side of the diaphragm. The linearity or lack of linearity between the front side and backside of the headphone will balance out or cancel.



This leads me to think a couple of things.

First,

The headphone seal to the test fixture or your head is always important, perhaps more important with closed back headphones. Leaks are critical.

Second,

The sealed space between the headphone and the test fixture or your head is a pressure sensitive environment. Carefully calibrated pressure microphones are required for accurate measurements.



Thanks DT

 

[Feelas]

First,

The headphone seal to the test fixture or your head is always important, perhaps more important with closed back headphones. Leaks are critical.

This is yet another reason for Harman using the standardized test-rig w/ simulated response curves. Firstly - ability to commit blind-testing of frequency responses, where one won't be affected by the price factor (you never know the FR by heart, thus you won't be influenced by LCD-2s costing a lot more than MDR-7506). The second one they identified, are the leaks, which are incredibly important and can ruin the sense of FR measurements on real subjects, if you have to account for variable leak.

 

[Robbo99999]

Hello All,

The cognitive gymnastics and conversation about seal, Frequency response and distortion that circle these closed back headphones is interesting.

What do we know about these headphones that is different than others/? The first thing that comes to my mind is that there is a sealed air chamber on the backside of the diaphragm where all the ideal gas laws apply, as in PV=NRT.

If you recall the conversations about the distortion caused by Compression Drivers and Horns the same thing happens with a closed back headphone. (Google the JBL Technical Note I think #108). Air is not a very linear spring especially in an enclosed space such as the space in the closed back of a headphone.

Equally interesting is that when we measure a headphone of for that matter when we put a headphone on our head and listen there are a enclosed spaces to some extent sealed to our head around our ears. If the seal is tight to our head or test fixture the sealed air chamber on the backside of the diaphragm is roughly equivalent to the sealed air chamber on the front side of the diaphragm. The linearity or lack of linearity between the front side and backside of the headphone will balance out or cancel.



This leads me to think a couple of things.

First,

The headphone seal to the test fixture or your head is always important, perhaps more important with closed back headphones. Leaks are critical.

Second,

The sealed space between the headphone and the test fixture or your head is a pressure sensitive environment. Carefully calibrated pressure microphones are required for accurate measurements.



Thanks DT

That's complicated. In reality though all you can do is choose a headphone that will seal effectively on your own head, and when you place your headphone on your head you just do your best to get it sitting right, and if you're gonna go the extra mile perhaps listen briefly to a small section of a bass heavy track you know to verify it's sealed & outputting the correct amount of bass. There's not much else you can do in practice right.

If your theory is correct though, then during headphone reviews when measuring distortion (& frequency response) you'd want to ensure a good seal to the measuring equipment (within reason - you couldn't use unnatural clamping to get there, as it wouldn't be reflective of real world usage).

 

[markanini]

Fixing meaning making it so it wouldn't move?

Yes.

 

[Lbstyling]

Question: The k371-BT (Bluetooth version) has analogue in for cables as well as Bluetooth connectivity.

Would anyone know if the analogue (xlr) in bypasses the DAC for the Bluetooth, or if it does ADC to DAC conversion of the incoming xlr?

Thanks.

 

[Pmelb]

As a long time reader of ASR I can't help notice that the reviews are now somewhat quite subjective "Amir's trust your ears reviews".

Some other non-science sites really dig deeper into the science and perform other tests/measurements on headphones to validate their findings. Many of us know that psychological bias is a strong beast to tame.

I know these headphone reviews are the early ones, and hopefully Amir will beef up the tests so a better understanding of the performance of the headphones can be assessed objectively in future.

 

[jhaider]

Question: The k371-BT (Bluetooth version) has analogue in for cables as well as Bluetooth connectivity.

Would anyone know if the analogue (xlr) in bypasses the DAC for the Bluetooth, or if it does ADC to DAC conversion of the incoming xlr?

Thanks.

You can use them powered off, so I expect the electronics are not in use.

 

[amirm]

As a long time reader of ASR I can't help notice that the reviews are now somewhat quite subjective "Amir's trust your ears reviews".

The gold standard in headphone and speaker research is subjective listening. Measurements are then correlated with them. So let's make sure subjectivism in that regard is not looked down upon.

In my case, the measurements are the "true north" when it comes to performance of a headphone. What my listening tests do is probe the findings of the measurements to assess their real audible effect and degree to which they need attention/correction.

There is no future in headphone measurements that give you reliable results that you can take to the bank. It cannot and will not exist due to variations in the way headphones are placed on fixture, how people wear them, etc. So no new development is in the cards there.

I hope you are not confusing what I am doing with pure subjectivist reviews which completely go by random listening tests. If they are not using measurements and strict listening test protocols, you can indeed ignore them as useless. I read and watch a lot of these and chuckle when they say X and my measurements show the opposite of X. As do my listening tests.

Your concern about psychological bias was addressed by me in a number of threads. In a nutshell, that effect is very low in controlled listening tests I perform due to large response variations we are talking about. You may have noticed that I do not use any small amplitude corrections where bias factor may be stronger than the tonal effect. When I boost the bass by 8 dB, it would be impossible to make a case that I am hearing placebo effect than real bass boost.

As I said in my recent video interview, there is no guarantee here. Ultimately there is some variability. Question is whether I can get you 80% of the way there. And I firmly believe I am.

Even with speakers where our measurements are much more accurate I still believe in value of subjective testing. Recently I found a resonance that was clearly audible but while measurements provided a clue, they did not at all ring the alarm bells like my ears did.

Anyway, I plan to write an article on this. For now, please post follow ups to the complaint thread and not these review ones.

 

[Doodski]

@Pmelb the complaint thread is here.

 

[Newman]

The gold standard in headphone and speaker research is subjective listening.....

.... as long as the listening conditions are controlled to eliminate all non-sonic variables.

Without such controls, the standard is certainly not gold, and arguably not silver or bronze either.

...For now, please post follow ups to the complaint thread and not these review ones

Apart from correcting your misleadingly-incomplete sentence above, I will certainly comply.

 

[amirm]

.... as long as the listening conditions are controlled to eliminate all non-sonic variables.

Without such controls, the standard is certainly not gold, and arguably not silver or bronze either.

"All" non-sonic variables are impossible to remove in headphone testing. Subject can tell what headphone he is listening to based on feel of the cups even in blind listening tests. For this reason, Harman resorted to emulating headphones using equalization of a virtual headphone. Another compromise.

So no, you don't need to remove "all" variables. You just need to remove enough to get high confidence results. That is what the research did, and that is what I am doing.

You want to wait for pink elephants to arrive in the form of perfect measurement and and testing? Do it on someone else's time, not mine.

Apart from correcting your misleadingly-incomplete sentence above, I will certainly comply.

I suggest you learn the topic better before getting cute with me this way.

 

[Maiky76]

Hi,

Here are some thoughts about the EQ.

Notes about the EQ design:

• The average L/R is used to calculate the score.
• The resolution is 12 points per octave interpolated from the raw data (provided by @amirm)
• A Genetic Algorithm is used to optimize the EQ.
• The EQ Score is designed to MAXIMIZE the Score WHILE fitting the Harman target curve with a fixed complexity.
  This will avoid weird results if one only optimizes for the Score.
  It will probably flatten the Error regression doing so, the tonal balance should be more neutral.
• The EQs are starting point and may require tuning (certainly at LF).
• The range above 10kHz is usually not EQed unless smooth enough to do so.
• I am using PEQ (PK) as from my experience the definition is more consistent across different DSP/platform implementations than shelves.
• With some HP/amp combo the boosts and preamp gain need to be carefully considered to avoid issues

I probably went overboard with the EQ but the results look quite interesting.

Average L/R match.
I have generated one EQ, the APO config files are attached.
Score no EQ: 63.6
Score Armirm: 70.7
Score with EQ: 101.3

AKG K371 APO EQ Score 96000Hz
January252021-120118

Preamp: -4.4 dB

Filter 1: ON PK Fc 27 Hz Gain -5.6 dB Q 1
Filter 2: ON PK Fc 66 Hz Gain 2.5 dB Q 3
Filter 3: ON PK Fc 112.5 Hz Gain 1.5 dB Q 5.81
Filter 4: ON PK Fc 130 Hz Gain -5.37 dB Q 0.31
Filter 5: ON PK Fc 1204 Hz Gain -2.69 dB Q 0.95
Filter 6: ON PK Fc 3945 Hz Gain 5 dB Q 5.93
Filter 7: ON PK Fc 6804 Hz Gain -2.63 dB Q 3.7
Filter 8: ON PK Fc 8496 Hz Gain -2 dB Q 6
Filter 9: ON PK Fc 9995 Hz Gain 3 dB Q 4.5
Filter 10: ON PK Fc 13313 Hz Gain -3 dB Q 4

 

[thyristor]

Are the pads memory foam like K712?

 

[Robbo99999]

The gold standard in headphone and speaker research is subjective listening. Measurements are then correlated with them. So let's make sure subjectivism in that regard is not looked down upon.

In my case, the measurements are the "true north" when it comes to performance of a headphone. What my listening tests do is probe the findings of the measurements to assess their real audible effect and degree to which they need attention/correction.

There is no future in headphone measurements that give you reliable results that you can take to the bank. It cannot and will not exist due to variations in the way headphones are placed on fixture, how people wear them, etc. So no new development is in the cards there.

I hope you are not confusing what I am doing with pure subjectivist reviews which completely go by random listening tests. If they are not using measurements and strict listening test protocols, you can indeed ignore them as useless. I read and watch a lot of these and chuckle when they say X and my measurements show the opposite of X. As do my listening tests.

Your concern about psychological bias was addressed by me in a number of threads. In a nutshell, that effect is very low in controlled listening tests I perform due to large response variations we are talking about. You may have noticed that I do not use any small amplitude corrections where bias factor may be stronger than the tonal effect. When I boost the bass by 8 dB, it would be impossible to make a case that I am hearing placebo effect than real bass boost.

As I said in my recent video interview, there is no guarantee here. Ultimately there is some variability. Question is whether I can get you 80% of the way there. And I firmly believe I am.

Even with speakers where our measurements are much more accurate I still believe in value of subjective testing. Recently I found a resonance that was clearly audible but while measurements provided a clue, they did not at all ring the alarm bells like my ears did.

Anyway, I plan to write an article on this. For now, please post follow ups to the complaint thread and not these review ones.

I agree with all you've said there, although when it comes to EQ I personally think & take the approach of accurate EQ based on measurements up 1kHz in terms of matching the Harman Curve as closely as possible (less than 1dB and smooth) - because measurements up to 1kHz are very reliable as long as you disregard bad bass seal outlier measurements. I also EQ more accurately to the curve in the treble than you do, but I don't take it to the extreme that I do in the up to 1kHz zone. The further up the frequency range I go the less I'm bothered about doing everything I can to match the Harman Curve spot on. I have the intuition that EQ'ing a smooth and accurate curve up to 1kHz can really help create that lovely accurate instrumental/vocal quality that the Sennheiser HD600 has "out of the box", which is a headphone that is exceedingly smooth in the mids in terms of measured frequency response.

 

[Frank Dernie]

You can use them powered off, so I expect the electronics are not in use.

I hadn't noticed that!
I will give it a try!

 

[daftcombo]

I borrowed @renaudrenaud 's pair of K371 last week (after his pair of K361 two weeks ago) to listen to music at the office.

I can only apply EQ in Foobar because I can't install EqualizerAPO, since I would need administrator rights in Windows.


Right now listening in Foobar with Amir's EQ settings. It is excellent.
Without EQ, I prefer the K361 (a little less bright and more neutral bass).