Testing audiophile claims and myths

[2leftears]

So you did go to uni at some point!

Yes I did indeed, spent some 8 years in academia in The Netherlands and the UK; MSc in applied physics and PhD in process systems engineering.

Kind of embarrassed to admit that since you reminded me of a few things I should have remembered. But the relevant signal processing theory I studied as part of my MSc in applied physics was almost 33 years ago, and I haven't really used it since. Teaches me for wading into an argument thinking I remembered more about all the relevant detail than I actually do, making myself look a little foolish...

I quoted which properties can be deduced from the frequency response alone: phase shifts, group delays, transient response. These are all related to the linear distortion of a system. All linear (and time-invariant) systems can be described by either its frequency response function, transfer function or its impulse response. They are linked together by certain mathematical operations and one can't change without the other. This should also ring a bell.
SNR (or THD, IMD, jitter and so on) can't be calculated from the frequency response, these are measurements to characterize nonlinear distortion while the frequency response only describes all the linear distortions.

Yes, my bad, I already marked my question as to be ignored just as you responded to it, unfortunate timing. I stupidly looked at the wrong quote when scrolling and mixed up which three properties you were referring to and thought you suggested S/N ratio could be inferred from the frequency response, which obviously you weren't implying at all.

You don't need some fancy lab equipment to gather some basic data about the distortion of your amp! If you have a soundcard (even an integrated one) with a line-in, you can absolutely use that. Your PC is effectively a digital oscilloscope and spectrum analyzer with a sample rate of 96kHz or more, you just need some software for it.
Of course, if the interface has considerably more distortion than your amp then the measurement will show the effects of that, not the performance of the amp, there is no way around that.

Yeah, I would need to buy some extra bits and bobs to make that work; but not sure if there is much point: I also have a Lehmann Linear headphone amp which is well-regarded, and TBH I don't think I could tell the two apart in a blind test.

 

[bigshot]

You’re probably right that you won’t be able to hear a difference between headphone amps as long as the power and impedance works. There’s no reason to expect any difference. You can test that to help you know for yourself, or to figure out testing controls. But there are bigger fish to fry when it comes to achieving optimal sound fidelity.

 

[castleofargh]

Kind of embarrassed to admit that since you reminded me of a few things I should have remembered. But the relevant signal processing theory I studied as part of my MSc in applied physics was almost 33 years ago, and I haven't really used it since. Teaches me for wading into an argument thinking I remembered more about all the relevant detail than I actually do, making myself look a little foolish...

My brain has a rule: use it or lose it.
Over the years, I've often forgotten some math, chemistry or whatever, relearned a bunch online when it would serve me, let time pass and forget again. Strictly for audio I think I've peaked the day I had a question, and Googling it gave me my own post giving the answer in great details. No memory of ever discussing the topic, no memory of ever knowing about it. I taught myself something reading the post.

 

[bigshot]

The sound science posts come up all the time in my google searches, including my own. We bring a lot of people into this forum with the content we create.

 

[2leftears]

I think that’s true. On the other hand, we all know about experts who, guided by the science, have been proven colossally wrong by further research. How much harm did doctors do when they prescribed low fat diets? To make matters worse, the food industry jumped on the bandwagon, and ended up doing things like sticking sugar in low fat food to make it palatable. Hubris is what happens when you equate science with reality. Science is not nature; science is knowledge *about* nature, as well as the method for pursuing it. Qua knowledge, it is always incomplete and subject to human fallibility. This is not to say that nothing is known for certain, only that humility is advisable.

Yes, I understand your point and I agree. I myself was inappropriately non-specific when I mentioned the hubristic nature of science. I should have been more specific. It is not science at fault, it is the fallible nature of some scientists as humans.

As I responded to @VNandor in my previous post, I spent some 8 years in academia. Half of it was at a university that was largely tax-payer funded. The other half was at a research laboratory associated with (another) university, but which was largely industry-funded. The attitude of the academics at both institutions towards research, publication of research, and peer review were, let's say, "different". This was due to both a difference in culture, as well as the economic environment they operated in.

At the industry-funded research lab it was not uncommon for research papers to be submitted for peer review where, on grounds of IP, unavoidably, critical information had been left out. Hence, proper independent verification and peer review of the results were not really possible, yet they often ended up approved for publication. This wasn't intentional poor practice, it was simply the result of financial pressures not providing the appropriate environment for science to be conducted the way it ought to be.

Suspect is in-house research done by companies, which is "published" (i.e. advertised) but has not been peer reviewed in any form, in order to make claims about proprietary novel technologies. That has marketing value, but no scientific value.

 

[2leftears]

My brain has a rule: use it or lose it.
Over the years, I've often forgotten some math, chemistry or whatever, relearned a bunch online when it would serve me, let time pass and forget again. Strictly for audio I think I've peaked the day I had a question, and Googling it gave me my own post giving the answer in great details. No memory of ever discussing the topic, no memory of ever knowing about it. I taught myself something reading the post.

That provides an interesting perspective on the concept of a self-taught man...

"A self-taught man usually has a poor teacher and a worse student" - Henny Youngman

 

[gregorio]

This, to all intents and purposes, has left the consumer with reassuring stats for sure, but not ones that are particularly useful when it comes to choosing between different devices. Some will claim that those stats mean it doesn't matter which device you choose, they will all sound great. Others think they can still hear a difference.

OK, so you’ve now identified and separated published specs from measurements. The former is a marketing tool and the latter an objective quantification of performance and/or specific aspects of performance. Of course though, published specs are supposed to be and are based on objective measurements, although due to being a marketing tool, they are commonly somewhat manipulated or optimised to compare better with similar products. Even so, I would generally disagree that specs are not particularly useful, they are generally particularly useful but that does not, of course, mean they are completely comprehensive or perfectly useful.

Measurements and specs require some understanding, first and foremost, that they are metrics of performance and NOT metrics of perception! It’s strange there seems to be this confusion in the audiophile world. In the car world for example, it is fully understood that measurements/specs are a different thing to perception. For instance, the 0-100kph time does not characterise the entire performance of a car, nor does it characterise any aspect of perception, it is just an objective measurement of an aspect of performance. We obviously don’t measure the 0-100kph time by trying to measure the perception of the driver/s of that car. The difference in the car world is that a majority of, or at least many, consumers not only understand that two vehicles with the same 0-100kph time can be significantly different but also, they are able to somewhat correlate a quoted 0-100kph time with their individual, personal perception. Additionally, many/most car consumers tend to have a better understanding of scale and magnitude. For example, that humans can tell the difference between 0 bhp and 1 bhp, between 300 bhp and 600 bhp but not between 600 bhp and 601 bhp, even though it’s the same objective difference as between 0 and 1 bhp.

Whilst the scientist and audio engineers are very well aware of that discrepancy (they should be!), I have not seen this come through in additional easy-to-understand performance measures made accessible to the average consumer.

There’s two obvious problems there. Firstly, you cannot have “easy to understand performance measures” for the average consumer because the average consumer is not able to correlate their personal perception to performance measures or even necessarily understand that a performance measure is not a measure of perception. Secondly, as you simplify/condense various different measures into single, simple measures then inevitably and deliberately, you will loose the precision of the different measures it contains. For example SINAD is a useful single/simple metric that’s particularly easy to understand when comparing with another similar piece of equipment but the balance between noise and distortion is lost as they’ve been combined and because as it’s a single metric it does not specify where in the spectrum it’s occurring and therefore how audible it may or may not be. It is still a useful comparative tool but requires more understanding about what it’s actually telling us and just as importantly, what it’s not telling us (it’s limitations)! So, what is apparently wanted is a contradiction; a single or a few measurements, which are objective, comprehensive and accurate on one hand but are easy to understand and correlate to human hearing on the other. That’s simply impossible, it’s one or the other, not both. Maybe AI *might* be able to provide a partial solution in the future but it will never be more than a generalisation that’s only applicable to some rather than all.

I have no idea how much research in science has been done re. the ability of humans to perceive e.g. phase differences in sound, differences in perception re. group delay, perceptibility of transient response, etc. I imagine quite a lot of research has been done, and is probably ongoing.

Quite a lot has been done and in some cases it is not ongoing, because it’s rather pointless (and therefore funding cannot be raised) to do further research on something that has already been well established.

Now, separate from this, there is an interesting philosophical discussion to be had about the appropriateness of segregating the domains of audio engineering science from that of the science of audio perception

It definitely is appropriate to segregate/separate them. One is based on established, mathematically proven, well understood and demonstrated laws, while the other (psychoacoustics) is based on anatomy, neuroscience and psychology. Furthermore, there’s an obvious separation between technology that records and reproduces sound from the perception of sound by human beings.

On the other hand, we all know about experts who, guided by the science, have been proven colossally wrong by further research. How much harm did doctors do when they prescribed low fat diets?

That’s an entirely different and non-analogous scenario. Doctors need to try to treat life threatening conditions, almost always with at least somewhat incomplete and sometimes extremely limited medical science upon which to recommend a treatment. That is not the case with audio, which is based on complete classical laws of physics discovered by the 1880’s. Do you have any recent examples of experts guided by that science who “have been proven colossally wrong” in this field?

To make matters worse, the food industry jumped on the bandwagon, and ended up doing things like sticking sugar in low fat food to make it palatable. Hubris is what happens when you equate science with reality. Science is not nature; science is knowledge *about* nature, as well as the method for pursuing it. Qua knowledge, it is always incomplete and subject to human fallibility.

Hubris is also what happens when you equate reality with something other than the proven/demonstrated science! Science is not “knowledge about nature” or the method for pursuing it, or rather, science is not ONLY that. What you are describing is “basic research” but in addition to basic research there is also “applied research”, BOTH of which constitute science! Tape recorders, ADCs, DACs, electronic amps, headphones, etc., are not mined, do not grow on trees or exist in nature, so obviously we are dealing with applied research/science rather than basic research/science. So hubris would be excessive self-confidence based on not understanding that fundamental fact and conflating the two!

Kind of embarrassed to admit that since you reminded me of a few things I should have remembered. But the relevant signal processing theory I studied as part of my MSc in applied physics was almost 33 years ago, and I haven't really used it since. Teaches me for wading into an argument thinking I remembered more about all the relevant detail than I actually do, making myself look a little foolish...

OK, that’s changed my opinion of you. Audiophiles will typically just double down on their rhetoric and/or start quoting audiophile reviews/marketing to defend their beliefs. It’s very uncommon that an audiophile will admit an error that invalidates their assertion/s. So kudos to you!!

G

 

[gregorio]

I spent some 8 years in academia. Half of it was at a university that was largely tax-payer funded. The other half was at a research laboratory associated with (another) university, but which was largely industry-funded.

I spent 10 years in academia, 4 years as a student and then nearly 20 years later, another 6 years on the other side of the fence, as a senior lecturer. If anything, the audio world is even more complex because we also have independent research, industry funded and in-house publications but in addition, we also have a lot of research by international bodies, due to the requirement of standardisation. Initially to allow national telegraph systems to communicate with other national telegraph systems but then also to allow international telephony/telecommunications and entertainment (radio, film and TV).

So we have numerous international bodies employing tens of thousands of scientists and engineers to conduct research, such as the ITU, AES, EBU, SMPTE, IEEE, ATSC and others. Some of which are intergovernmentally funded, while others are funded by a membership comprised of national and multi-national companies. Much of this research is not publicly available but typically it is verified by others (effectively “peer reviewed”, though not the same process). Furthermore, industry funded research in the audio field and even in-house publications is commonly (though not always) reliable, indeed, most of what we have in the audio world comes from in-house publications or other industry funded research, from amplifiers to digital audio, the simplification of Maxwell’s laws (Heaviside), speakers, headphones, electronic recording and more besides. So if you have access to the research, much of which is behind paywalls or only available to members of the international bodies, the majority of it is entirely reliable but it still occasionally warrants a very heavy dose of skepticism, good knowledge of the industry is really the only way of knowing.

None of the above really applies to the audiophile world though, audiophile manufacturers are typically boutique type businesses who are not members of the international bodies, do not participate in scientific research or standardisation procedures and commonly rely on misinforming it’s customers about all of this!

G

 

[2leftears]

OK, that’s changed my opinion of you. Audiophiles will typically just double down on their rhetoric and/or start quoting audiophile reviews/marketing to defend their beliefs. It’s very uncommon that an audiophile will admit an error that invalidates their assertion/s. So kudos to you!!

Happy to hear that. The irony is that I am not an audiophile at all. My inventory of audio equipment is nothing too extravagant. Especially the cable inventory is bog standard consumer grade by audiophile standards. Some of my stuff is old, and rather obsolete from an "audiophile" perspective, but simply good enough for me. Not looking for some missing "magic". I do have a couple of more expensive components, but I consciously bought those not for their sound quality but rather for their non-sound-related features that suited me and the mechanical durability aspect of that specific CD transport, a brand whose reliability I have extensive positive experience with from the past. I am happy to spend a bit more for equipment if I know it suits me well feature-wise and is going to last a lot longer. TBH, planned obsolescence and a growing pile of electronic waste is more of a concern to me than inaudible audiophile differences. No upgradeitis here.

 

[Prog Rock Man]

I have no way of testing this, but I am confident, having listened to music I am very familiar with, that the new headphones I have, which allow listening via Bluetooth to my mobile phone, USB cable to my laptop and analogue cable to my amp, I would fail an ABX source test.

Ironically, the Bluetooth headphones are replacements for wired in-ears that I use when travelling and on holiday, because of, to me, the in-ear's poor SQ. They (Bose) were more expensive than the Bluetooth (Austrian Audio) that replace them.

That just further reinforces my belief that it is the speaker and personal sound preference that ultimately make the real difference, and everything else is ancillary.

 

[gregorio]

The irony is that I am not an audiophile at all. My inventory of audio equipment is nothing too extravagant. Especially the cable inventory is bog standard consumer grade by audiophile standards. Some of my stuff is old, and rather obsolete from an "audiophile" perspective, but simply good enough for me.

None of that indicates you are not an audiophile at all. The literal meaning of “audiophile” is a lover of audio, by implication someone who seeks high fidelity. In the audiophile community the term “audiophile” appears to have a somewhat different meaning, a lover of audio equipment rather than audio and apparently a high level of gullibility and perceptual susceptibility to audiophile marketing BS. Bog standard consumer grade cables do not affect audible fidelity and some older stuff, even going back to some CD players, DACs and amps in the 1990’s likewise did not affect audible fidelity. So you could indeed be an audiophile in the literal sense (as am I), although I do use some very expensive gear (professionally).

G

 

[2leftears]

None of that indicates you are not an audiophile at all. The literal meaning of “audiophile” is a lover of audio, by implication someone who seeks high fidelity. In the audiophile community the term “audiophile” appears to have a somewhat different meaning, a lover of audio equipment rather than audio and apparently a high level of gullibility and perceptual susceptibility to audiophile marketing BS. Bog standard consumer grade cables do not affect audible fidelity and some older stuff, even going back to some CD players, DACs and amps in the 1990’s likewise did not affect audible fidelity. So you could indeed be an audiophile in the literal sense (as am I), although I do use some very expensive gear (professionally).

G

Fair point.

 

[Brahmsian]

None of that indicates you are not an audiophile at all. The literal meaning of “audiophile” is a lover of audio, by implication someone who seeks high fidelity. In the audiophile community the term “audiophile” appears to have a somewhat different meaning, a lover of audio equipment rather than audio and apparently a high level of gullibility and perceptual susceptibility to audiophile marketing BS. Bog standard consumer grade cables do not affect audible fidelity and some older stuff, even going back to some CD players, DACs and amps in the 1990’s likewise did not affect audible fidelity. So you could indeed be an audiophile in the literal sense (as am I), although I do use some very expensive gear (professionally).

G

We talk a lot about audiophiles here, so let’s see if we can flesh out what an audiophile is. My method for evaluating the sonics of a recording is very simple: usually, it either pleases me or it doesn’t. This has little to do with the standard metrics, as there are very old mono recordings that fall gratefully on my ear and modern recordings that do the opposite. In certain cases I might focus on the mix or remastering, but usually I focus on the performance, with the sonics akin to air that I see right through without noticing it. That’s why I don’t consider myself an audiophile despite my marginal interest in audio. I take an audiophile to be someone who places enough importance on the sonics that audio quality becomes something to concern yourself about as distinct from the music. A lot of people don’t place much importance on audio quality even if they love music. Having placed such on importance on audio quality, it’s only natural that it tends to lead to an interest in audio gear.

There are some audiophiles for whom digital audio is a problem. We can talk all we want about the technical merits of digital, but there are at least two legitimate reasons why it might have left a bad impression: the “glassy” sound of some early CDs and the loudness wars. The latter in particular has been the bane of many an audiophile, and with good reason. Most people listen to music in their cars, at the gym, in noisy nightclubs, or from cheap laptop speakers as they type away, do their online shopping, or scroll through social media. Often, the music is not the center of attention. Music is sort of like a soundtrack to the movie that is people’s lives. As Jonathan Sterne writes in Interdisciplinary Science Reviews,

[The] list of environments where heavy audio compression is appropriate quickly outpaces a list of environments where recorded music should be heard with maximum dynamic range. In fact, beyond recording studios and living rooms, there are relatively few environments where people can hear the edges of a recording's definition. Distracted listening in non-ideal circumstances is the norm for music listening and has been for some time. It is very rare for people to sit quietly in a room and just listen to recorded music coming out of speakers.​

Hopefully, many audiophiles are still part of this rare subset of people who “sit quietly in a room and just listen to music coming out of speakers.” I certainly do. In fact, I listen to very little music otherwise. I usually give music my undivided attention. (I find music distracting when I’m doing something else.) Since I listen almost exclusively to classical, dynamic range is not often a problem for me. The loudness wars have not infected this genre. But for most everything else, is it any wonder that so many audiophiles have developed a distaste for digital audio and are constantly on the lookout for ways to improve their listening experience? And since most people don’t sit quietly in a room and completely focus on music, is it any wonder that audiophiles and their concerns are not understood by the vast majority of people?

Don’t get me wrong, I love digital audio. Not only does it sound fantastic to me, but the ease and convenience it offers is such that I don’t think I would ever go back to vinyl. But I also understand how the loudness wars have made some audiophiles sick. I wouldn’t be surprised if we find that it adversely affects them physiologically and/or psychologically. This level of dynamic compression is not something inherent to digital audio, but it has gotten so prevalent that it might as well be. If we can’t run studies that are based on the premise that PCM affects people adversely, perhaps there’s still room to see whether or not dynamic compression and loudness do so. Do you know of any such study?

 

[bigshot]

I’m not an audiophile because my interest predates that word. I’m a “hi-fi nut”.

 

[XTF1]

We talk a lot about audiophiles here, so let’s see if we can flesh out what an audiophile is. My method for evaluating the sonics of a recording is very simple: usually, it either pleases me or it doesn’t. This has little to do with the standard metrics, as there are very old mono recordings that fall gratefully on my ear and modern recordings that do the opposite. In certain cases I might focus on the mix or remastering, but usually I focus on the performance, with the sonics akin to air that I see right through without noticing it. That’s why I don’t consider myself an audiophile despite my marginal interest in audio. I take an audiophile to be someone who places enough importance on the sonics that audio quality becomes something to concern yourself about as distinct from the music. A lot of people don’t place much importance on audio quality even if they love music. Having placed such on importance on audio quality, it’s only natural that it tends to lead to an interest in audio gear.

There are some audiophiles for whom digital audio is a problem. We can talk all we want about the technical merits of digital, but there are at least two legitimate reasons why it might have left a bad impression: the “glassy” sound of some early CDs and the loudness wars. The latter in particular has been the bane of many an audiophile, and with good reason. Most people listen to music in their cars, at the gym, in noisy nightclubs, or from cheap laptop speakers as they type away, do their online shopping, or scroll through social media. Often, the music is not the center of attention. Music is sort of like a soundtrack to the movie that is people’s lives. As Jonathan Sterne writes in Interdisciplinary Science Reviews,

[The] list of environments where heavy audio compression is appropriate quickly outpaces a list of environments where recorded music should be heard with maximum dynamic range. In fact, beyond recording studios and living rooms, there are relatively few environments where people can hear the edges of a recording's definition. Distracted listening in non-ideal circumstances is the norm for music listening and has been for some time. It is very rare for people to sit quietly in a room and just listen to recorded music coming out of speakers.​

Hopefully, many audiophiles are still part of this rare subset of people who “sit quietly in a room and just listen to music coming out of speakers.” I certainly do. In fact, I listen to very little music otherwise. I usually give music my undivided attention. (I find music distracting when I’m doing something else.) Since I listen almost exclusively to classical, dynamic range is not often a problem for me. The loudness wars have not infected this genre. But for most everything else, is it any wonder that so many audiophiles have developed a distaste for digital audio and are constantly on the lookout for ways to improve their listening experience? And since most people don’t sit quietly in a room and completely focus on music, is it any wonder that audiophiles and their concerns are not understood by the vast majority of people?

Don’t get me wrong, I love digital audio. Not only does it sound fantastic to me, but the ease and convenience it offers is such that I don’t think I would ever go back to vinyl. But I also understand how the loudness wars have made some audiophiles sick. I wouldn’t be surprised if we find that it adversely affects them physiologically and/or psychologically. This level of dynamic compression is not something inherent to digital audio, but it has gotten so prevalent that it might as well be. If we can’t run studies that are based on the premise that PCM affects people adversely, perhaps there’s still room to see whether or not dynamic compression and loudness do so. Do you know of any such study?

This is a very nice depiction of what is an audiophiles!

I won’t attempt to define an “audiophile” from a scientific perspective (way too ambitious for me!), but they appear to share two common characteristics: an incontrovertible trust in their ears as the absolute gauge for any audio-related topic and, an incontrovertible trust in their brain as the absolute judge for any audio-related topic. No measurement equipment will ever supersede their ears, no scientific test will ever change their opinions.

This thread may be able to debunk some of the audiophiles claims and myths… but it wont impact self-proclaimed audiophile’s opinions…