[bigshot]

Bleh.

 

[Phronesis]

You have a wrong idea of how IEMs work. The key factor to understanding IEMs is that acoustics in a closed ear canal is fundamentally different from acoustics in free air.

First of all, it's no problem whatsoever to make small diaphragms vibrate with 15 cycles per second (i.e. 15Hz). Neither for your iPhone speakers, nor for even smaller IEM drivers.

But your iPhone speakers have two disadvantages compared to IEMs: first, the coupling (transmission of energy from diaphragm to air) is far less efficient in free air than in a closed ear canal. And second, the lower you go in frequency, the less directional the sound waves become in free air. Whereas in the closed canal, they have nowhere else to go. That's why seal is so important with IEMs.

To illustrate the latter, here's an example of IEMs measured with a full seal (blue) vs. a slightly leaky seal (red). You can see that it makes next to no difference for higher (shorter wavelength and more directional) frequencies. But it makes a substantial difference in the bass range.

Here's the measured frequency response of the IEMs I did the 15Hz test with. They have 1% THD@20Hz and no bass roll-off whatsoever.

I played around a bit using pure tones with my Audeze i4 IEM, which is an unusual IEM because it's open back and has a large driver for an IEM, but I have a good seal into the ear. I used my iphone, the synalski online tone generator, and my Chord Mojo to generate the tones. I don't know what the FR limits of that setup are, but assuming the range is at least 20 to 20 kHz, my results are below.

Down to 40 Hz, I hear distinct pitched notes. Between 20 and 40 Hz, I readily perceive the bass, and I can hear pitch differences if I change the frequency, but I don't hear distinct pitched notes; towards the lower end of that range, it becomes unpleasant to listen to the pure tone continuously. Dropping below 20 Hz to about 17 Hz, my perception of the sound fades and it becomes more like 'pressure' than sound. I become unable to detect anything below 17 Hz even by increasing the volume, not sure it that's due to my ears or the gear not producing those frequencies.

At the upper end, my hearing limit seems to be about 15.8 kHz (consistent with other tests I've done), and my perception of the sound drops off pretty quickly around that frequency. Near that limit, the sound is like a very high pitched ringing in the ears, and is unpleasant to listen to continuously despite being subtle. So I think it's safe to say that 18 kHz is ultrasonic for me, and if I perceive it subconsciously at all, it may need to be transient in order for it to not be unpleasant.

At this point, I'm doubtful that such ultrasonics would add much value to my enjoyment of music, so I'm not concerned about missing them in my gear. And of course, if my gear is giving me up to 20 kHz, I should be getting a range of about 16 to 20 kHz which is already ultrasonic for me, and ultrasonics beyond 20 kHz may be swamped in importance by the ultrasonics from 16 to 20 kHz. I'm just not seeing a possible significant benefit of 25 or 30 kHz for me, though I suppose it may matter for a kid who can sonically hear beyond 20 kHz (I wish I knew my limit when I was young, but never tested it).

I'll try my SE535 later, which is fully sealed, to see how that compares.

Addendum:

I tried the SE535. I hear an FR from about 30 to 15.3 kHz, which is a narrower range than the i4. I assume the difference is the IEM rather than my ears.

 

[castleofargh]

Nope, we’re talking science, which aims to develop ever better models of reality, so I still say we need a nuanced definition of a fact.

Is it a fact that space and time are absolute and independent dimensions of reality? Is it a fact that mass is the amount of stuff and is always conserved? For engineering and designing audio gear, we can adamantly say yes, but relativity theory says it’s erroneous to call them facts in that model.

In engineering surveying, we often treat the earth surface as completely flat, and only if the scale gets a lot bigger do we treat it as curved (is the shape of the earth even constant in relativity theory?). It’s usually fine and better to say in ordinary life that the earth isn’t moving (do anyone of us feel it moving?), and that the sun goes around it, rather than to constantly visualize the solar system with the sun at the center, the earth on elliptical orbit around it, and the earth spinning.

Facts are always relative to a model, and relative to models, there can be facts. You just have to accept that you’re working with a model (which is not reality itself), and know which model you’re working with. You don't need to qualify which model you're using when you say you have two hands, because it will likely be two hands in any model we might choose to use, but lots of audio and perception stuff isn't that straightforward, so we do need to know and state our models (and therefore assumptions) when we talk about 'facts'.

Note the qualified and provisional definition of scientific fact here:

https://ncse.com/library-resource/definitions-fact-theory-law-scientific-work

"Fact: In science, an observation that has been repeatedly confirmed and for all practical purposes is accepted as “true.” Truth in science, however, is never final and what is accepted as a fact today may be modified or even discarded tomorrow."
​

And BTW, I didn’t say it was plausible that anyone could hear 250 dB below music, I said that maybe Watts was measuring that number in his gear somewhere, and the resulting gear change made a difference somewhere else that was audible. I thought we already covered this ad nauseum.

I have zero issue with the definition of fact that you quoted. as I said, if later on we get new evidence demanding to reconsider, we'll do that then, when the new evidence will call for it. because we act by relying on data, not on faith that in the future anything will be possible and we're always wrong.
about all your examples of conditional truth, the answer is in the name.

as for

but lots of audio and perception stuff isn't that straightforward, so we do need to know and state our models (and therefore assumptions) when we talk about 'facts'.

that's in part why I said we needed to be clear on what we mean by hearing. and why I brought up my understanding of part of the cochlea. I'm certainly with you on having to define conditions and models.

 

[castleofargh]

I don't disagree with your first point - but I do have a semantic issue with it. When I learned science, I was taught that you don't EVER "just decide that something is a fact". At best you decide to act AS IF it's a fact for the same of convenience and practicality. (Perhaps the proper term would be that "you agree to treat it provisionally as a fact".) However, at some deeper level, you keep it in the back of your mind that it may someday turn out to be untrue - or untrue in some context you haven't tried yet. (For example, for any prediction you might make based on any so-called "law of physics", quantum theory now demonstrates that the result may not be what you expect. And, of course, tomorrow we may find out that quantum theory is incomplete or slightly wrong. Newton's laws of motion USUALLY work. However, it's possible that, due to quantum uncertainty, you could fire a gun away from you, and have the bullet turn around in midair and hit you in the face. It's extremely unlikely that it will happen in the next billion years... in our galaxy... but it isn't impossible.) We make models and, in many cases, we agree to treat those models AS IF they were simply facts, because it serves no practical purpose to keep repeating out loud "of course this is really only an extremely high probability".... We also neglect to acknowledge even more likely outcomes - for example, we don't say: "I'm going to a movie tonight, unless I have a heart attack, or get hit by a meteorite". However, in that case, we are willing to acknowledge that those things really could happen. I guess it was just hammered so thoroughly into me in science class in college that we must never forget this that I am unwilling to pretend to do so.

I also agree that the goal of most musical recordings is NOT an accurate reproduction of an original - and that it's quite possible that no such original exists. And, obviously, we may not actually prefer the original if it does exist. We don't do our best to reproduce the delays between songs, the occasional clinker hit by a musician, nor the air conditioning failure at the concert hall the night we attended, and we usually edit out the coughs and chair scrapes from the audience. What we are generally trying to do is to create an idealized version of the experience... or a wholly artificial but "ideal" experience. (Note, however, that some people find the noise of the musicians turning pages to be a distraction, while others find that it adds an air of realism to the recording, so we don't all agree on all the details.)

It should also be obvious that, once you acknowledge that, the practical details of that goal will be different for different people. I personally find the experience with headphones to be quite different from a live experience, and both to be quite difference than hearing a recording. However, this is as much a choice as a limitation. If you REALLY want a headphone experience that is closer to listening with speakers you can buy a Smyth Realiser". The Realiser simulates speakers - using a processor box and a tracking transmitter you attach to your headphones. It accepts a surround sound input, tracks the exact position of your head as you move it, and plays through your headphones a binaural signal that simulates what you would be hearing if you were listening to a set of speakers, playing that content in a real room. They even offer various plugin modules that allow you to "listen through" specific sets of popular models of high end speakers. People who have heard one insist that it works quite well and produces a rather convincing illusion. The fact that it remains quite expensive, and only one company sells such a product as far as I know, probably suggests that this is NOT the goal most audiophiles pursue. (It actually costs less than many high-end DACs, so the cost is not impractical in audiophile terms.)

To me it's obvious that our perception is far from a simple thing. I have one pair of open air electrostatic headphones that are rated to go down to 8 Hz (no tolerance given - Koss ESP/950's). When I listen to something with really low bass on them, it "almost feels as if I can feel my feet vibrating from the bass". I do not get a similar experience with most other headphones. Obviously there is something going on there with my perception. (It could be that they reproduce bass harmonics especially clearly, or that they press against my head in a certain way, or simply that I'm imagining it. One guess might be that, either because they accurately reproduce a cue others miss, or fail to produce some distraction that other headphones produce, my brain simply considers the illusion they produce to be "more real", and so does a better job of filling in the other details that aren't actually there. There is a whole area of brain science dedicated to how our brains fill in details that aren't actually there, based on memories, previous experience, and other less well defined factors.)

I find all of this fascinating... and, yes, it is SCIENCE.

However, I can see how others may find it irrelevant to their day-to-day experience.

in principle, I spend my days complaining about empty claims and people claiming way too many stuff on the forum. I'd be happy if everybody was to talk the science way and always go for "the experience/the data strongly supports ...." instead of making claims. I really love discussing with people who do that TBH. I sort of grew more suspicious of confident people for some reason. but that's just me, and this is Head-fi. can you confidently say that it's going to be beneficial for fellow members to have all the guys who know what they're doing, talking in degrees of confidence, while all the idiots with too much ego, will keep spamming stuff with absolute certainty? most people with a question want a yes or no answer, so they will consistently go for that type of answer that a rigorous scientist will rarely make.
so that's another issue where I'm divided. there is my idea of science and the care for truth on one hand, and there is fighting the obscurantism and the myths propagated by amateur audiophiles to themselves on the other. if we show doubt while some fool doesn't, sadly, the readers will often take the confident claim for truth. in this section we expect that people will be slightly more curious and slightly more demanding when it comes to substantiating claims, but in the rest of the forum... when you try to help a fellow audiophile with his problem and give an accurate answer sounding a lot like "it depends on....", while some other guy will come and say something like "this is the best, trust me". should we still be satisfied that we expressed ourselves the most accurately? wasn't the main objective to help the guy? in that respect, I think @bigshot is doing a great job fighting fire with fire. and in respect to making accurate and substantiated claims, he drives me mad once or twice a week. ^_^
I don't know who is doing the right thing.

about the Realiser I joined the kickstarter campaign for the A16 10mn after I learned about it. it was the second or third day of the campaign, I believe dinosaurs still roamed the earth back then. and since I've been waiting for them to finally deliver ... something.


as for feeling like we're getting tactile bass from headphones, I get what you mean and I admit that I'm not sure what gives that feeling and why some headphones with a lot of bass don't. Floyd Toole seem to believe that some physical shaking(even at the wrong frequency) can be significant in giving such an impression. so it would be interesting to see how much some headphones really shake and how much is felt by the skin. I somehow got the idea that large drivers did that better and imagined that they were just pushing over a bigger surface of skin, helping the subjective impression. but if the driver is bigger, it seems plausible that the weight and force involved will in effect make the headphone vibrate more(unless it's super heavy with the clamping of death). so my hypothesis may just be what Mr Toole is talking about.
now an even funnier case, and @james444 can maybe relate to that as he has owned a few IEMs with serious bass. I've had some IEMs when lying down in my bed and listening to something with abundance of low end, I started to feel like my lungs were shaking with the bass ^_^. I'd certainly be interested in understanding what exactly can trigger such subjective impression.

 

[gregorio]

[1] And it is here that ANY timing errors - even in sub microsecond range - become audible.
[2] And that means recording techniques have to mimic the human perception of sound best they can to begin with ... Good as in being true to the sound as heard by the listener in audience.

1. This being the sound Science forum, I presume you haven't just completely made that assertion up and have evidence to support it? For everyone else (in this universe): Even the very best musicians are only accurate to within a few milli-seconds and are NEVER consistently so. Do these always present timing errors sound many thousands of times worse than apparently audible sub micro-second timing errors?

2. Make-up your mind, which is it: "Mimic the human perception of sound" OR "be true to the sound as heard by the listener in the audience"? You can't have both, it's one or the other, either we are "true to the sound entering a listener's ear" or we "mimic the human perception of sound". The only way we could have both would be if human perception did not exist and therefore what we would "hear" would always be identical to the sound which enters our ears. Of course though, that is purely hypothetical, we are all humans, we therefore all have human perception and recordings are made by humans for humans (a simple obvious fact, that's somehow often ignored by audiophiles). So we (recording engineers/music producers) have a choice, a choice which first started being explored in the 1950's and which continues to be explored today. As we're dealing with art (music) in the first place, it was unsurprisingly found (and continues to be found) that consumers vastly prefer the former and that's why virtually all acoustic commercial music recordings are made using multi-mic'ing techniques, because only that technique gives us the flexibility to change what would enter our ears and "mimic the human perception of sound". Furthermore, even if we ignore human perception, we cannot "be true to the sound as heard by the listener in the audience" anyway! For example: There are no (and cannot be) any perfect microphones, everyone has at least a somewhat different HRTF and a listener in the audience never has their skull bolted in a fixed position.

There's no need to address/refute the other points in the quoted post, as it should (hopefully) already be obvious that it's incorrect.

G

 

[castleofargh]

I agree that the overal percentage is LESS than the figures cited.

And that's why I started recording - at 44 of age, I had enough of multimiked stuff to the point a person can not show over his or hers head with extended arm anymore.

I am perfectly OK with multimiking in case of electric instruments that without loudspeakers of one sort or another can not be heard - at least not the way as normally heard on amplified concetrs. Have you ever heard an electric guitar played without being plugged into "something electrical " ? If not, I can tell you it's about the most pathetic sound you are likely to hear coming from a musical instrument. Here, I agree recording can be and usually is a form of art all by itself - and it is a VERY creative form of art, allowing the use of methods and gear that are nothing but detrimental in acoustic music recording to a good, often novel approach to sound and music in general.

When it comes to recording anything acoustical in a decent acoustics place, only then does the term " high fidelity ", " absolute sound ", " "true to the sound heard live" or " ANYTHING TO THAT EFFECT " really come into play. And it is here that ANY timing errors - even in sub microsecond range - become audible. And that means recording techniques have to mimic the human perception of sound best they can to begin with - way before recording methods come into play.
To be blunt - a recording on a Sony WMD-6(C) ( a pro walkman cassette recorder ) fed from any simple 2 mic recording technique will trounce any digital recorder fed from mulimiked session mix.

It is much harder to do, as most of the work is BEFORE the actual music recording - because everything, first and foremost the all important " balance among instruments and voices " ( the term musicians themselves are most frequently using - NOT an objective term in scientific sense ), spatiality, reverb, etc - HAS TO BE DONE IN SITU - and not in the post production that multimiking allows for and leaves an enormous window open for the recording>mastering engineer to manipulate - past any reason, unfortunately in not small percentage of all cases. It does take more time to find that "sweet spot" for both the musicians and the mike - and, since time is money, such recordings can well end up costing more than multimiked sessions, where the formula " we can fix everything in the mix " can be - and usually is - applied.

It was human wish to fly like a bird ... for how many thousands of years, before the first human made object, heavier than air, actually took to the air ?
How many times it has been said it can not be done, how many times the attempts have failed, many times requiring the ultimate sacrifice ?

Just like with airplanes driven by a piston engine and propeller the maximum speed is limited to approx 800 km/h ( single propeller ) , up to almost 900 km/h ( turboprop, two contra rotating propellers ) , regardless of the motor of the power that drives these propellers, the speeed can not be icreased above that limit - because, once the propeller blade tips approach or reach the speed of sound , the efficiency of the prepeller drops to zero.

It is similar with any attempt to piece together a sound captured with multimiked session. WHICH microphone should be the reference for time ? Whichever is finally asigned, it is WRONG decision.

That's why the binaural can be so good. Good as in being true to the sound as heard by the listener in audience.
And bad in the extreme for - say - a sahophone player, who would like to have recorded as precise instrument handling as possible - and that means close miking, often with the mic clipped on the sax itself ,preferably eschewing any room influence. The first can be heard ( or very similar ) by anyone in the audience, the second the sax player alone.

Which you can relate better to ( asuming you are NOT a sax player ) ?

It is high time that recordings - or pigs, for that matter - learn to *fly* in correct time frame. Hopefully, it will not be another > two thosand years, before that is first understood, then accepted as normal - similar to the frequency response for human hearing from 20Hz - 20kHz is accepted today.

well I'm not going to agree with a lot of what you said, but I certainly have nothing against people working to reproduce a given event or a given experience of the event as close as it can possibly be done. I see value in that. but I also see value in the various artistic interpretations brought by recording and mastering techniques. when I take a photograph, sometimes I will hope to get what I saw and will work toward that goal even in post processing. but most of the time even when taking the picture I have already some more or less specific vision of what I want as a result that isn't the real scene. and then sometimes I don't really have a clue, I just try stuff and settle on what I like the most. I expect that it's very similar for recording/mastering, and that all the approaches have value, if only for being different, which can really be attractive to an artist.

 

[gregorio]

You're talking about reproducing "the sound of a kick drum"....
I'm talking about reproducing exactly the pressure wave the drum produces...
(Just to be clear here, I'm talking about the actual movements of the air, rather than what might be recorded.)

What I was pointing out was simply that, if you observe the INITIAL PHYSICAL MOTION of the drumhead....
The beater pushes it forward, and it then turns around, and returns to its original position.
(It then overshoots and continues to resonate - while internal energy produces other motions and sound.)

If you actually look at a graph of that initial movement - it's going to look like a hill.
And that hill represents half of a sine wave.

No, you are NOT talking about reproducing exactly the pressure wave the drum produces, you are talking about the fundamental frequency the drum produces but a drum (string or any other physical object in the real world) does NOT produce ONLY a fundamental frequency it also, at the very least, produces a set of harmonics which is part of that "pressure wave". This was discovered about 2500 years ago (by Pythogoras) and not since demonstrated to be false. You can call what you're claiming "Pure Science" or anything else you want but actually you're contradicting some of the oldest and most well established actual/real science known to man!

G

 

[SoundAndMotion]

No, you are NOT talking about reproducing exactly the pressure wave the drum produces, you are talking about the fundamental frequency the drum produces but a drum (string or any other physical object in the real world) does NOT produce ONLY a fundamental frequency it also, at the very least, produces a set of harmonics which is part of that "pressure wave". This was discovered about 2500 years ago (by Pythogoras) and not since demonstrated to be false. You can call what you're claiming "Pure Science" or anything else you want but actually you're contradicting some of the oldest and most well established actual/real science known to man!

G

What did Pythagoras discover? Are you talking about the ringing hammers of the blacksmiths? That tale has been debunked. Maybe it was something else?

 

[Phronesis]

in principle, I spend my days complaining about empty claims and people claiming way too many stuff on the forum. I'd be happy if everybody was to talk the science way and always go for "the experience/the data strongly supports ...." instead of making claims. I really love discussing with people who do that TBH. I sort of grew more suspicious of confident people for some reason. but that's just me, and this is Head-fi. can you confidently say that it's going to be beneficial for fellow members to have all the guys who know what they're doing, talking in degrees of confidence, while all the idiots with too much ego, will keep spamming stuff with absolute certainty? most people with a question want a yes or no answer, so they will consistently go for that type of answer that a rigorous scientist will rarely make.
so that's another issue where I'm divided. there is my idea of science and the care for truth on one hand, and there is fighting the obscurantism and the myths propagated by amateur audiophiles to themselves on the other. if we show doubt while some fool doesn't, sadly, the readers will often take the confident claim for truth. in this section we expect that people will be slightly more curious and slightly more demanding when it comes to substantiating claims, but in the rest of the forum... when you try to help a fellow audiophile with his problem and give an accurate answer sounding a lot like "it depends on....", while some other guy will come and say something like "this is the best, trust me". should we still be satisfied that we expressed ourselves the most accurately? wasn't the main objective to help the guy? in that respect, I think @bigshot is doing a great job fighting fire with fire. and in respect to making accurate and substantiated claims, he drives me mad once or twice a week. ^_^
I don't know who is doing the right thing.

I see two largely separate worlds in head-fi, Sound Science and the rest of head-fi. I spend time in both worlds, but they're separate worlds and don't really mix. In the rest of head-fi, I talk about headphones, and I pretty much ignore whatever people say about DACs, amps, and cables (well, sometimes I challenge it, but carefully).

Sound Science could be focused on undermining subjective claims of big differences, but if those 'subjectivists' generally don't spend much time in Sound Science, isn't it just mostly preaching to the choir? Do we really care about some unknown audience that may be eavesdropping on our conversations? Like almost everyone else, I don't post under my real name, and I don't really care about who may be listening in. And unless someone can be convinced to do controlled testing for themselves, I don't think anyone is going to stop trusting their ears based on arguments presented in Sound Science, especially if they don't have the technical background to even understand those arguments.

I personally am much more interested in what you describe above, as far as taking a scientific approach to exploring questions on which there isn't a consensus on the answers among people who are reasonably well informed on the science. To have worthwhile conversations along those lines, we need to have common ground on some basic assumptions. Let's see if people can agree on these:

- The electronics of gear are well understood.

- There are design choices in gear which result in measurable differences in their outputs.

- Measurements can tell us a lot, but no finite set of measurements can fully characterize a complex musical signal.

- Acoustics is pretty well understood, but it's complex, and there's still 'art' involved in designing transducers.

- Perception is a complex area of science. We know a lot about it, but there's a lot we don't know.

- We're prone to auditory misperception to a degree that we don't and can't consciously realize. This accounts for at least the vast majority of big differences people hear in gear other than transducers.

- Controlled tests need to be carefully designed, conducted, and interpreted to try to answer specific questions.

- A null result on a blind test doesn't "prove" that two things sound indistinguishably "the same", it just indicates that the listener was unable to consistently discern a difference in the test conditions. That could be interpreted as evidence that there's no big difference, but it doesn't rule out the possibility of a small difference which might detected in a different test, or in another trial by the same listener, or by a different listener, or in a larger test, or experienced as a difference in normal listening. And conversely, a positive result on a blind test doesn't "prove" that a difference exists either, since the test could have flaws (e.g., "tells") or the positive result could be due to luck (which is why tests need to have some scale and you need to run stats).

 

[KeithEmo]

The headphones I notice that with to a significant degree are the Koss ESP/950 electrostatics.

Electrically, they claim a frequency response "8 Hz to 35 kHz" - no tolerance given.
I've never seen a graph, but, from the way they sound, they aren't actually extremely flat.
They are an open backed around-ear headphone with a diaphragm that extends far past your ears.

Also, unlike many headphones, they are very light and fit very loosely.
(They feel like they're going to fall off - even though they never do.)

To, me, subjectively, they don't sound at all bass heavy.
You tend to hear a lot of the details in bass notes but the bass doesn't seem especially loud.
To me this contributes to the odd way they seem so sound.
They don't SEEM to make A LOT of bass, and the bass seems rather thin, but it seems to extend incredibly low.
Their sound also doesn't seem to depend much on their fit (moving them around doesn't change it).

I've seen various measurements.
Here's one:
http://audioinvestigations.blogspot.com/2011/10/koss-esp-950-very-nice.html

However, considering that it's difficult to measure headphones, the measurements often don't seem to correlate well with the sound, headphones that sound flat tend not to measure flat and vice-versa, and their large diaphragms and loose fit no doubt complicate matters..... who knows.

...................................


as for feeling like we're getting tactile bass from headphones, I get what you mean and I admit that I'm not sure what gives that feeling and why some headphones with a lot of bass don't. Floyd Toole seem to believe that some physical shaking(even at the wrong frequency) can be significant in giving such an impression. so it would be interesting to see how much some headphones really shake and how much is felt by the skin. I somehow got the idea that large drivers did that better and imagined that they were just pushing over a bigger surface of skin, helping the subjective impression. but if the driver is bigger, it seems plausible that the weight and force involved will in effect make the headphone vibrate more(unless it's super heavy with the clamping of death). so my hypothesis may just be what Mr Toole is talking about.
now an even funnier case, and @james444 can maybe relate to that as he has owned a few IEMs with serious bass. I've had some IEMs when lying down in my bed and listening to something with abundance of low end, I started to feel like my lungs were shaking with the bass ^_^. I'd certainly be interested in understanding what exactly can trigger such subjective impression.

 

[gregorio]

Let's see if people can agree on these:
[1] - The electronics of gear are well understood.
[2] - There are design choices in gear which result in measurable differences in their outputs.
[3] - Measurements can tell us a lot, but no finite set of measurements can fully characterize a complex musical signal.
[4] - Acoustics is pretty well understood, but it's complex, and there's still 'art' involved in designing transducers.
[5] - Perception is a complex area of science. We know a lot about it, but there's a lot we don't know.
[6] - We're prone to auditory misperception to a degree that we don't and can't consciously realize. This accounts for at least the vast majority of big differences people hear in gear other than transducers.
[7] - Controlled tests need to be carefully designed, conducted, and interpreted to try to answer specific questions.

1. Agreed.
2. Agreed.
3. I don't agree with this one. Just one measurement completely characterises any arbitrarily complex signal (musical or not), the measurement of amplitude over time.
4. Agreed.
5. Agreed.
6. Agreed.
7. I somewhat agree. Your statement is true in some/many cases, it depends what we're testing.

G

 

[Phronesis]

1. Agreed.
2. Agreed.
3. I don't agree with this one. Just one measurement completely characterises any arbitrarily complex signal (musical or not), the measurement of amplitude over time.
4. Agreed.
5. Agreed.
6. Agreed.
7. I somewhat agree. Your statement is true in some/many cases, it depends what we're testing.

G

I actually agree on 3 also, but a musical signal has a lot of wiggles and we can't infer much from just looking at all the wiggles. As we discussed, we can do a null test of two signals, but if the difference signal isn't zero, the door is still open for debate about the what the non-zero signal indicates if the amplitude isn't always super low.

Glad we generally agree on all the rest. Progress!

 

[analogsurviver]

The headphones I notice that with to a significant degree are the Koss ESP/950 electrostatics.

Electrically, they claim a frequency response "8 Hz to 35 kHz" - no tolerance given.
I've never seen a graph, but, from the way they sound, they aren't actually extremely flat.
They are an open backed around-ear headphone with a diaphragm that extends far past your ears.

Also, unlike many headphones, they are very light and fit very loosely.
(They feel like they're going to fall off - even though they never do.)

To, me, subjectively, they don't sound at all bass heavy.
You tend to hear a lot of the details in bass notes but the bass doesn't seem especially loud.
To me this contributes to the odd way they seem so sound.
They don't SEEM to make A LOT of bass, and the bass seems rather thin, but it seems to extend incredibly low.
Their sound also doesn't seem to depend much on their fit (moving them around doesn't change it).

I've seen various measurements.
Here's one:
http://audioinvestigations.blogspot.com/2011/10/koss-esp-950-very-nice.html

However, considering that it's difficult to measure headphones, the measurements often don't seem to correlate well with the sound, headphones that sound flat tend not to measure flat and vice-versa, and their large diaphragms and loose fit no doubt complicate matters..... who knows.

The headphones I use for serious listening and evaluation - particularly for bass - are Stax Lambda Pro, driven by its proprietary ( made specially to produce overall correct response with Lambda Pro ) diffusse field equalizer Monitor ED-1 feeding the SRM1MK2 amplifier.

Lambda Pro came into being because of - Mercedes Benz. They wanted to research automotive acoustics and noise levels using binaural recordings - and the best headphone at the time, Stax Lambda, was simply not up to the task; it had limitations in both the dynamic range and low end ( infrasonic ) frequency response. The solution has been to enlarge the stator to diaphragm distance and increase the polarizing voltage - now, it was enough for proper monitoring of the car recordings. The Pro version of Lambda has been embraced by the audiophile community in no time - and, the rest is history.

Lambda Pro is so good in bass extension that ED-1 Monitor has to be modified in order not to loose that edge in bass Lambda Pro does have when driven only by SRM1MK2 ( or any of the still better amps ). IIRC - about an order of magnitude bigger value coupling cap(s) are required for that in the signal path of equalizer ( it has been almost a decade now, would have to check for the exact schematics ) - THEN you can have both the proper EQ and BASS that isvery close to second to none - as far as the accuracy is concerned. Bassheads for dubstep etc would still require louder bass, but for anything acoustical, it is - just - enough. The new L700 headset offers additional 3-4 dB of SPL over Lambda Pro - which, sometimes, would be dearly required for monitoring live mic feed without any limitting/compression.

This setup will reveal ANY component/recording with rolled-off bass - no way it can escape unnoticed.

As good as Stax Lambda Pro is, it can not compete in bass with better IEMs - even not those from its own family, the often called Baby Stax, the SR-001 and its succesor, SR-002.

I have not even seen, let alone heard the Koss ESP-950s ( they are exceedingly rare in Europe, and transport/duties works to their disadvantage pricewise compared to US ) - but here a reasonably good comparison with Stax : https://www.stereophile.com/content/koss-esp950-electrostatic-stereophones-page-3

 

[KeithEmo]

1)

All of the studies I've seen that relate to humans being able to distinguish timing variations down into the microseconds are talking about being able to distinguish variations in arrival times between the two ears. What they're saying is that, if you delay the signal reaching one ear by as little as 10 microseconds, we can notice a perceptible shift in the apparent location of the source. (The generally agreed number is that we can hear differences in inter-aural delay of as little as 10 uS with a 1 kHz signal.)

Here are a few links to studies that mention it:
https://www.pnas.org/content/98/24/14050
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3663869/

(If you Google "inter-aural delay" or "interaural time difference" you will find more.)

2)

Maybe you can't have both, or all three, at the same time... but they are all individually possible.

I could:
a) seek to duplicate the perceptual experience of a listener in the audience
b) seek to exactly duplicate the sound waves, as measured in the air, at the listener's position
c) seek to duplicate the sound waves "at the source"

I believe that, in most consumer applications, what's REALLY going on is b).
We try to design our reproduction system so it "delivers the correct audio information to the listener's ear".

Of course, as you say, in "art", "right" is whatever the artist or the recording engineer says it is.

I would also point out, however, that, while you are technically correct in your final assertion...
You are specifically looking at the matter from a "practical" direction.

You are correct... there cannot ever be "a perfect microphone". HOWEVER, if I am willing to budget enough money, or to accept other limitations, there are quite often scientific methods and equipment that are far better than anything likely to ever be used in a studio. I would also reiterate that, from a scientific point of view, technology does sometimes change very rapidly. When I went to school, a 1 watt LASER was the size of a refrigerator, cost hundreds of thousands of dollars, and required a rather heavy power main to operate. Today, I can buy one for $50, that fits in my pocket, and runs on AA batteries. (So, for example, if we WERE to decide that there was a real need for microphones that could record 30 kHz at a level of 170 dB, in twenty or thirty years you might be able to find one at your favorite pro shop for $100. Quite often "the limits of practical technology" are simply determined by what there is a market for.)

People have been predicting for some time that someday we will bypass the ears entirely - and plug directly into the brain. We haven't been able to do that yet... but modern cochlear implants do manage to bypass a significant portion of the air/ear signal chain. (And, no, they don't work very well yet, but they do continue to improve.)

1. This being the sound Science forum, I presume you haven't just completely made that assertion up and have evidence to support it? For everyone else (in this universe): Even the very best musicians are only accurate to within a few milli-seconds and are NEVER consistently so. Do these always present timing errors sound many thousands of times worse than apparently audible sub micro-second timing errors?

2. Make-up your mind, which is it: "Mimic the human perception of sound" OR "be true to the sound as heard by the listener in the audience"? You can't have both, it's one or the other, either we are "true to the sound entering a listener's ear" or we "mimic the human perception of sound". The only way we could have both would be if human perception did not exist and therefore what we would "hear" would always be identical to the sound which enters our ears. Of course though, that is purely hypothetical, we are all humans, we therefore all have human perception and recordings are made by humans for humans (a simple obvious fact, that's somehow often ignored by audiophiles). So we (recording engineers/music producers) have a choice, a choice which first started being explored in the 1950's and which continues to be explored today. As we're dealing with art (music) in the first place, it was unsurprisingly found (and continues to be found) that consumers vastly prefer the former and that's why virtually all acoustic commercial music recordings are made using multi-mic'ing techniques, because only that technique gives us the flexibility to change what would enter our ears and "mimic the human perception of sound". Furthermore, even if we ignore human perception, we cannot "be true to the sound as heard by the listener in the audience" anyway! For example: There are no (and cannot be) any perfect microphones, everyone has at least a somewhat different HRTF and a listener in the audience never has their skull bolted in a fixed position.

There's no need to address/refute the other points in the quoted post, as it should (hopefully) already be obvious that it's incorrect.

G

 

[KeithEmo]

Yes, and, in order to reproduce the entire signal correctly, you must reproduce all the parts correctly, and in the correct relationship. If some limitation prevents you from producing the fundamental accurately, then you have already failed... because, if any piece isn't right, then the total cannot be right. (So, if there is a "piece", whether it's the fundamental or a sub-harmonic, that exists at 10 Hz, and you're filtering it out, then you cannot possibly be reproducing the entire signal accurately. What you end up with may "sound the same" - but technically it will be incorrect.

No, you are NOT talking about reproducing exactly the pressure wave the drum produces, you are talking about the fundamental frequency the drum produces but a drum (string or any other physical object in the real world) does NOT produce ONLY a fundamental frequency it also, at the very least, produces a set of harmonics which is part of that "pressure wave". This was discovered about 2500 years ago (by Pythogoras) and not since demonstrated to be false. You can call what you're claiming "Pure Science" or anything else you want but actually you're contradicting some of the oldest and most well established actual/real science known to man!

G