• 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!

Resolution, speed, do these things really exist?

Chyżwar

Active Member
Joined
Jun 11, 2021
Messages
152
Likes
177
What does it mean that some headphones are more resolving or faster than others? I have read many times that more expensive headphones are more resolving or faster than cheaper ones because they have better drivers. But these have always been subjective impressions, can you measure these things in any way?

(Please don't attack me if I said something wrong. I'm a musician, not an audiophile or a technical expert. I'm just trying to understand certain things. When I asked this question to audiophiles, they said I had to hear these differences. I also read Amir reviews but that didn't help me either. My subjective impressions are completely inconsistent. For example, I thought the Beyerdynamic T1 (which Amir doesn't like) are really technically a bit better than the DT880 or HD600 (more transparency, more precision), but other times I think the cheap Sony MH1C (after some equalization) are so good as Campfire Andromeda or any expensive headphones I've heard, and detail, resolution is mainly dependent on equalization...)
 

o7_brother

Active Member
Joined
Oct 8, 2019
Messages
106
Likes
326
Location
Portugal
I just so happened to write a Reddit post about the topic of "resolution" and why it's not really a thing, and frequency response is the most important factor by far.

First of all, any good headphones produce distortion that is low enough not to be audible. This is helpful because it simplifies our reasoning. Now that we've put distortion aside for a second, your question can be divided into two parts:

1. Headphones/IEMs being minimum-phase systems means FR fully describes their output.

Minimum-phase is a mathematical concept that applies not just to loudspeakers. In this case, it means that at any time, the loudspeaker's movement (and the resulting sound pressure at the ear drum) is within one phase-cycle of the input signal, and does not lag behind. This means there is no such thing as "attack", "decay", "driver speed", etc because the driver is tracking the original signal perfectly in the time domain. Now, you can totally get a subjective sense of "this headphone has bad decay compared to this other headphone", but what you're describing is a difference between their FRs.

"Wait, you're saying the headphone tracks the input signal, how come different headphones sound different?" Well, how a headphone changes the input signal is called frequency response, but the point is the "not lagging behind" thing (the output is stable and time-invariant).

2. Since FR is what matters, this means FR is what determines the subjective sense of resolution and soundstage. How can that be?

So what does this mean for headphones like Stax, where the driver is super thin and lightweight? Surely the lighter membrane can accelerate faster, stop moving faster, and can therefore track every little nook and cranny of the analog waveform of the recording, thus extracting more details, right?

Well yes, but just about any driver can move fast enough to do that. They are minimum-phase, remember? The mass of the driver is important when designing a headphone because it affects FR, but that's something the headphone designers need to worry about, not us.

CD quality cuts off at 22.05 kHz, not to mention the hearing of most adults cuts off below that. This means any driver that can do that many wiggles per second (at a usable amplitude or volume) is fast enough to extract every "detail" in the recording. Even an entry-level headphone like a Koss or a cheap Grado can do this. You want your driver to go faster? Sure, there are headphones that can reach up to 40kHz or whatever, but it doesn't really matter because you can't hear details that would require more than 20kHz to describe.

"Ok, sure. If two drivers are playing a 20kHz tone at the same volume, their drivers are moving at the same speed. However, that's just a static sine wave, not a complex musical signal with multiple frequencies of varying amplitudes!"

How the driver reacts to a complex signal like music, which has multiple frequencies of varying amplitudes, that's called "frequency response" :)

When reviewers say X headphone is more detailed than Y, they are describing differences in FR, sometimes with price bias or other cognitive biases thrown in the mix. The classic example of this is Stax, not because the drivers are these super light membranes with extremely low distortion, but because they generally lack sub-bass and have exaggerated upper midrange and treble, things that get associated with the sense of "detail".

"But wait, we can EQ two headphones to have the same FR and they won't sound the same! You can't EQ Stax levels of detail into an HD650, therefore there must be something other than FR at work here!"

That's because we're not actually EQ'ing to the same FR. If two headphones have the exact same FR at the ear drum (not just on a measurement rig, but on your actual ear drums), they would sound the same. This is impossible to do in practice because a) the measurement rig's ears aren't shaped the same as your individual human ears, which affects FR of the treble, b) simply taking a headphone off your head and putting it back on will change the FR in the treble due to imprecise seating c) the bass response will be affected by how tight of a seal you can get on your head vs on the measurement rig. These are all frequency response differences, mind you.

Oratory1990 has mentioned a few things that a headphone needs in order to respond well to EQ:

- perform reliably, with repeatable seal across multiple users
- easily obtain the amount of seal that it was designed for (rip glasses users or people with large beards)
- have good quality control = little unit variation and no channel imbalance
- have a relatively smooth FR free from high-q artifacts (sharp peaks and dips)
- deform the pinna as little as possible
- have little reflections inside the earcup, especially those that lead to destructive interference. You can't fix a notch in the FR with EQ (non-flat excess group delay).
- have suitably low distortion

Most headphones do not meet all of these conditions which affect FR, so their FR will be a pain to EQ accurately. What I'm trying to explain is that there will always be a FR difference when comparing two headphones, even with EQ. Therefore, there doesn't "need" to be some other variable at play, and indeed if you do a blind test, FR tracks very closely with listener preference, but no other metric does.

With IEMs, everything I said applies, except it's even simpler. They don't interact with the outer ear, only with the ear canal. Depending on the shape of your individual ear canals, the treble response will be affected, which crinacle has covered in his article about interpretation of FR graphs. You can have quite significant variations in the treble response just by inserting IEMs differently. This may affect one's subjective notion of "detail" and "resolution". Are your ear canals identical to mine? Are you sure the two of us are not just hearing a different FR?

Now for the soundstage thing. When it comes to speakers in a room, soundstage size is determined by the directivity characteristics of the speaker, which in turn affect how the sound reflects around the room and back to the listener's ears.

In headphones, it's basically frequency response. the "room" is the interior of each ear cup, so does that mean the ear cup reflections are responsible for soundstage? Well... ear cup reflections affect group delay, which in turn affects FR, like oratory's quote explained above. There have been attempts at measuring soundstage, to questionable degrees of accuracy (see RTings' attempts). We know it is affected by the FR of the headphone and how your ears affect the FR that reaches your ear drum (called HRTF/PRTF), although there would also seem to be a "trick" where large earcups that do not touch your ears contribute to the perception of this effect (think HD800 vs HD650).
 

Zensō

Major Contributor
Joined
Mar 11, 2020
Messages
2,740
Likes
6,740
Location
California
The main contributors to how a headphone sounds are frequency response and distortion. A headphone with a FR that generally matches a target that mimics a flat speaker in a good room, without too many deviations, and with low distortion, is going to sound good to most people. Descriptive terms like fast, dark, bloated, detailed, resolving, etc, can usually be correlated to FR and distortion.

Soundstage is related to FR, but also to the physical positioning of the drivers, the HD800S being an example of this.
 
OP
Chyżwar

Chyżwar

Active Member
Joined
Jun 11, 2021
Messages
152
Likes
177
I just so happened to write a Reddit post about the topic of "resolution" and why it's not really a thing, and frequency response is the most important factor by far.

Thanks! A very interesting article!

I think I will sell all my headphones except the Etymotic and HD560S because I like it without EQ.
 

markanini

Major Contributor
Joined
Feb 15, 2019
Messages
1,718
Likes
1,768
Location
Scania
Good question. The only certain thing is that it exists at the very least as an idea among audio enthusiasts, and as a marketing term. There exists no agreed upon definition, of any model that determines the magnitude of resolution in such a way that you can measure and predict it.

The implications of that would make audiophiles mad if considered realisticlally. It means that when one person says the hears a certain resolution in a headphone, and another person listening to the same headphone says the same thing, there is a non-zero chance that they might be reacting to totally different aspects of the headphones performance.
 
Last edited:

Feelas

Senior Member
Joined
Nov 20, 2020
Messages
390
Likes
315
Good question. The only certain thing is that it exists at the very least as an idea among audio enthusiasts, and as a marketing term. There exists no agreed upon definition, of any model that determines the magnitude of resolution in such a way that you can measure and predict it.

The implications of that would make audiophiles mad if considered realisticlally. It means that when one person says the hears a certain resolution in a headphone, and another person listening to the same headphone says the same thing, there is a non-zero chance that they might be reacting to totally different aspects of the headphones performance.
Having a low THD will contribute to a better overall performance at all times, never will a worse THD pair be more accurate than the better. THD + FR near the actual hearing curves + lack of major dips and peaks and you're pretty much set.
 
OP
Chyżwar

Chyżwar

Active Member
Joined
Jun 11, 2021
Messages
152
Likes
177
Having a low THD will contribute to a better overall performance at all times, never will a worse THD pair be more accurate than the better. THD + FR near the actual hearing curves + lack of major dips and peaks and you're pretty much set.

Yes, but most good headphones have low THD.

"While total harmonic distortion (THD) of a few tenths of 1 % can be detected in a single-
frequency tone under carefully contrived experimental conditions, the just- detectable
distortion for speech and music is somewhat higher. Near threshold, the harmonics pro-
duced by distortion may be simply inaudible. At high levels, the apparent distortion in
the ear masks the harmonics. A 2 % THD is generally inaudible for speech and music
between 50 and 90 dB SPL. Below 50 dB SPL the inaudible distortion rises to 10 % at
30 dB SPL. Above 90 dB SPL, the inaudible distortion rises linearly to 10 % at 110 dB
SPL. The derivation of the limits can be found in Killion."
 

Feelas

Senior Member
Joined
Nov 20, 2020
Messages
390
Likes
315
Yes, but most good headphones have low THD.

"While total harmonic distortion (THD) of a few tenths of 1 % can be detected in a single-
frequency tone under carefully contrived experimental conditions, the just- detectable
distortion for speech and music is somewhat higher. Near threshold, the harmonics pro-
duced by distortion may be simply inaudible. At high levels, the apparent distortion in
the ear masks the harmonics. A 2 % THD is generally inaudible for speech and music
between 50 and 90 dB SPL. Below 50 dB SPL the inaudible distortion rises to 10 % at
30 dB SPL. Above 90 dB SPL, the inaudible distortion rises linearly to 10 % at 110 dB
SPL. The derivation of the limits can be found in Killion."

And while that is all correct, remember that perceptibility versus what is heard (the sound w/ distortion) in the result might be two different aspects. Just wondering whether imperceptible THD might mask potentially hearable details due to simply being a bit louder. As in where we couldn't say for sure that the THD is there or not, but the details themselves are lost.
 

Yorkshire Mouth

Major Contributor
Joined
Apr 22, 2020
Messages
1,254
Likes
1,154
Location
God's County - Yorkshire
Yes, but most good headphones have low THD.

"While total harmonic distortion (THD) of a few tenths of 1 % can be detected in a single-
frequency tone under carefully contrived experimental conditions, the just- detectable
distortion for speech and music is somewhat higher. Near threshold, the harmonics pro-
duced by distortion may be simply inaudible. At high levels, the apparent distortion in
the ear masks the harmonics. A 2 % THD is generally inaudible for speech and music
between 50 and 90 dB SPL. Below 50 dB SPL the inaudible distortion rises to 10 % at
30 dB SPL. Above 90 dB SPL, the inaudible distortion rises linearly to 10 % at 110 dB
SPL. The derivation of the limits can be found in Killion."

In addition, I understand that distortion is less noticeable (at the same %) in lower frequencies.

https://www.axiomaudio.com/blog/distortion
 

DeepFried

Active Member
Joined
Jun 19, 2020
Messages
106
Likes
116
Location
UK England
I always assumed 'speed' and 'resolution' both referred to decay. i.e. fast decay = high resolution. Though i'm not saying thats any sort of accepted definition, its just what I thought people meant.

I've no idea actually how audible decay is, instinctively I'd expect a slow decay to blur some of the detail, but perhaps its such a slight effect it makes no difference. no idea!
 
Last edited:

Dentin

Member
Joined
Jun 18, 2018
Messages
5
Likes
2
Lower and upper harmonics are clearly things that exist, and headphones/IEMs said to be fast like Focal's open-backs, everything by Stax, Hifiman's HEK series, and the Empire Ears Odin have boosted upper midrange or lower treble that emphasize the former and de-emphasize the latter. Having the upper treble trail-off reduced (without being cut off too much and inaudible altogether) will naturally lead to less treble transient bleed and greater perceived resolution.

"Headphones' sound is defined by FR, therefore technicalities do not exist" is a hot take posted by people who don't actually understand how to read and interpret the measurements of the headphones that they're complaining about beyond "Harman good."
 
Last edited:

Jose Hidalgo

Addicted to Fun and Learning
Joined
Aug 15, 2020
Messages
961
Likes
1,032
Location
France
Thanks @o7_brother . Great post! It's consistent with my own experience, only much better explained than I could.

AFAIK, the main variables that we hear are : frequency response, distortion, and internal resonances. Other things are related to such variables.
And the main reasons why two people hear differently are : ear shape, and ear canal shape/diameter. And aging of course. Because ultimately these things affect FR (the FR that we actually hear, the FR that reaches our brains through our ears, not the FR that the headphones produce).

I'm a big fan of EQ, and I think everybody should use it. It's the best/easiest way to get the most out of any headphone. But of course two headphones EQed to the same target curve (e.g. Harman AE-OE 2018) won't sound exactly the same. That can be because of distortion, or good/bad seal, or even EQ preset imperfections, but I find that the thing I notice the most is internal reflections.

For example, that's why my EQed Audioquest Nighthawk can't sound exactly like my EQed HD 600 or Sundara. After EQ, both sound signatures become very similar, and yet something remains different. It's not distortion (it remains below audible levels in my case, even @ 20-40 Hz, because I don't listen at very high loudness levels), and I guess it's not seal either. I think it has more to do with the Nighthawk's semi-closed nature and internal reflections. That's what I think I hear.

Subjectively, I find that it doesn't translate into the HD 600 or the Sundara being 'better' than the Nighthawk. In my opinion, it translates into 'the HD 600 or the Sundara are definitely closer to the naked truth, while the Nighthawk is sultry and sexier'. Like two beautiful women : one naked and the other in a beautiful dress. So in the end it's apples and oranges, even after EQ. Some songs sound 'better' on one headphone while others sound 'better' on the other one.

Anyway, thanks again for explaining things very clearly and debunking some very old myths. :)
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,747
Likes
16,181
But of course two headphones EQed to the same target curve (e.g. Harman AE-OE 2018) won't sound exactly the same. That can be because of distortion, or good/bad seal, or even EQ preset imperfections, but I find that the thing I notice the most is internal reflections.
Also EQing them to having the same response on a measurement rig doesn't mean they will have the same response on your own ears due to the different geometries, that's why also the frequency response difference of two headphones measured at different rigs or ears is usually different too.
 

sfdoddsy

Active Member
Joined
Jun 6, 2019
Messages
293
Likes
438
100% agree with the above.

There's an old saw that the closer you get to perfection the more minimal differences become.

For me, the biggest problem with attempting an objective approach to headphones and IEMs is that no-one can even agree what perfection is..

With speakers, there is general agreement that a flat anechoic measurement with even dispersion results in good in-room performance.

But well-respected headphone reviewer Crinacle has 11 different targets on his graph site. Some of his also respected brethren have even more.

Harman themselves continuously change their target.

When it comes to transducers, speed and attack are not seperate from frequency response. It is either accurate or not.

Speakers with the same anechoic response can have different in room responses due to dispersion differences.

Headphones have fewer variables and so should be easier to make great.

And they are.

I recently bought my son a $5 earbud because it had the right connection for his new phone. It's performance is far closer to my fancy headphones than a $5 speaker would be to my fancy speakers.
 

ADU

Major Contributor
Joined
Jul 29, 2021
Messages
1,584
Likes
1,086
I apparently don't know enough about how transducers work to comment intelligently on this. If FR is the only thing that matters though, then does this mean that the oft claimed sound quality differences between planar-magnetics and dynamic headphones simply boil down to just FR, and have nothing to do with the other differences in their physical design and drivers?

I think o7_brother also menthioned e-stats. And suggested that the differences in sound quality between an e-stat and a good dynamic (or planar?) headphone is also related solely or primarily to differences in their FR. There are a few e-stats that have FRs which are close to a typical dynamic or planar headphone though. So maybe those would make a better test case for something like this than the Stax, which has a more unusual response?
 

ADU

Major Contributor
Joined
Jul 29, 2021
Messages
1,584
Likes
1,086
If o7_brother's comments are correct, then wouldn't this also have to mean that you should be able to get any halfway decent pair of $100 headphones off the shelf at Guitar Center (like say an AT M40x?), and make it sound as good as the best-reviewed, best-sounding, high-end headphone currently on the market, simply through the application of some EQ?
 

Tim Link

Addicted to Fun and Learning
Forum Donor
Joined
Apr 10, 2020
Messages
722
Likes
638
Location
Eugene, OR
I apparently don't know enough about how transducers work to comment intelligently on this. If FR is the only thing that matters though, then does this mean that the oft claimed sound quality differences between planar-magnetics and dynamic headphones simply boil down to just FR, and have nothing to do with the other differences in their physical design and drivers?

I think o7_brother also menthioned e-stats. And suggested that the differences in sound quality between an e-stat and a good dynamic (or planar?) headphone is also related solely or primarily to differences in their FR. There are a few e-stats that have FRs which are close to a typical dynamic or planar headphone though. So maybe those would make a better test case for something like this than the Stax, which has a more unusual response?
I think it does come down to frequency response at the eardrum. Since our ears are all shaped a little differently and since the sound is going into our rear at a different angle than it typically would when listening to a distant source, and since the headphone driver is close enough to actually acoustically couple with our outer ear and create resonances in that tiny space between the driver and the eardrum, it's complicated to come up with any single headphone design that's going to produce the best response at the eardrum for everyone.
 
Last edited:
  • Like
Reactions: ADU

ADU

Major Contributor
Joined
Jul 29, 2021
Messages
1,584
Likes
1,086
I think it does come down to frequency response at the eardrum. Since our ears are all shaped a little differently and since the sound is going into our rear at a different angle than it typically would when listening to a distant source, and since the headphone driver is close enough to actually acoustically couple with our outer ear and create resonances in that tiny space between the driver and the eardrum, it's complicated to come up with any single headphone design that's going to produce the best response at the eardrum for everyone.

Thank you for the reply and insights on this, Tim Link.

There seems to me to be general, though certainly not universal agreement among audiophiles on what constitutes a superior sounding headphone in terms of sound quality and so-called "technicalities" or technical performance. And I wonder how your idea about the different ear shapes of individuals would fit into that, if at all.

Many audiophiles and reviewers of higher-end audio products also seem to be able to separate a headphone's FR performance from its technical performance. And seem to believe that the latter has more to do with with things like a headphone's distortion, and its behavior in the time domain. And that there can be headphones with both excellent and also poor FR which are capable of delivering good technical performance. (I believe Crin actually separates these two things in his headphone ratings and rankings, for example.) Which seems counter to o7_brother's and some others' opinions on this subject.

I suppose the part of o7_brother's comments that I'm having the most difficulty with is his take on the time component. Maybe I'm misreading or not understanding all of his remarks, but he seems to be suggesting that time isn't (or maybe shouldn't?) be a factor at all in a headphone's sound quality,... because headphones are minimum phase devices. (?) And I'm not really sure that one necessarily follows from the other. Or, for that matter, that headphones always strictly obey the rules of a minimum phase device.

I'm afraid that I just don't know enough about this subject, and exactly what a minimum phase device is (though I have heard the term tossed around quite a bit by Floyd Toole and some other audio experts) to really offer a good opinion on this. But I personally have tended to think that most or all of the aspects of a headphone's sound quality, including its so-called technical performance, can probably be reduced to just its FR, distortion, and also (notably) its performance in the time domain. Which seems fairly reasonable to me... But it seems as if some here are now sayin, or at least suggesting that the time component of a headphone is mostly or completely irrelevant to its sound quality. Which doesn't seem to quite fit with my own experiences of listening to a variety of different headphones over the years (mostly lower-end ones, I should say, in the sub-$300 range... though the price is probably not really the object here).
 
Last edited:

ADU

Major Contributor
Joined
Jul 29, 2021
Messages
1,584
Likes
1,086
I think there probably are also audible differences in distortion between some headphones. Though I'm unable to comment on whether this would be a significant or noteworthy factor in some higher-end products that exhibit very low distortion characteristics.

There are many different kinds of distortion though. And I'm not sure that we really understanding which types and how much of them can really effect our perceptions of a speaker or headphone's sound quality yet. Research still seems to be in the early stages on this, particularly with regard to headphones.

In a recent study on this subject, I believe Sean Olive stated that he was able to hear some differences in the sound characteristics of headphones with different types or amounts of distortion, which had been equalized to approximately the same frequency response. But he was, for the most part, unable to identify which ones he strongly preferred over the other. And I have to wonder whether his ability to differentiate and also rank headphones based on the different amounts and types of distortion could improve, if he spent some more time training his ears to better hear and identify the differences. In much the same way that Harman trains its research subjects to better hear and distinguish the differences in frequency response.
 
Last edited:

Aerith Gainsborough

Addicted to Fun and Learning
Joined
May 4, 2020
Messages
853
Likes
1,280
Just wondering whether imperceptible THD might mask potentially hearable details due to simply being a bit louder. As in where we couldn't say for sure that the THD is there or not, but the details themselves are lost.
If the Harmonic distortion product in question is louder than the detail it masks yet is already too quiet to be perceivable, it stands to reason, that the masked detail is too quiet to be heard in the first place.
In practice, there would be no loss of information.
 
Top Bottom