[Giru]

It's always amusing to read/listen/watch these "objective" vs "subjective" debates about audio equipment on the "enternet". Apparently it's the place to be if you want to appear smart and flash your recently polished ego with whatever "Noledge" that you've gained by regurgitating facts and ideas that ultimately still have little to no scientific validation or worse: have minimal real-world application.

It is true that we can test and assess the performance of electronic and electric systems to remarkable accuracy these days but time and again we fail to understand that such scientific assays have little value when we talk of their "application/translation" to human "perception".

Sound stimuli (just like light stimuli) have a varied "perception" by individuals that depends on a plethora of factors and hence there is no "objective" method of quantifying the "experience" of such stimulus. A very good example of this is asking yourself "do we all see the same way?" Or "do we all see the same colours?"





The above two videos explain how we all might be following the same "patterns" of "experiences" while responding to a similar visual stimulus but eventually our "personal" experience to a visual stimulus is almost unique. Not only that; our "experience" can also be modified by a variety of factors.

It is a similar story with sound. We all grossly follow the same patterns of "perception" but our individual "experience" is unique and highly dependant on various factors. Infact; given how sound transmits predominately by physical/mechanical pathways (as compared to visual stimuli) we may have even more individual variation in its perception.

A simple analogy: let's say that Amir (no pun intended) buys a mass spectrometer and goes around town trying to find the restaurant that makes the best spaghetti. He would show you how restaurant A's spaghetti has more sodium chloride and a high percentage of aromatic compounds as compared to restaurant B. But does that truly make A's spagehtti better?
Obviously not. It's a personal choice. Ofcourse certain objective measurements would tell you why A's spaghetti might be more palatable but all that goes out the window if I'm suffering from a cold, have a low taste perception, have a shorter tongue and most importantly have never tasted spaghetti! There is no question of this test being able to correctly correlate to my real life "experience".

Certain individuals (and therefore manufacturers) are hell bent on these measurements being absolute indicators of how good an equipment is while it couldn't be far from the truth. There are a lot of factors at play and individual perception might make that -8 to -10 db (or etc) of difference absolutely useless.

This makes "hobbyist" measurements of audio equipment to conclude whether it is "superior" or "inferior" to certain other equipment that don't test well, rather moot and comical. I pray for the souls of people who waste their time on measuring stuff and or gushing over objective numbers than actually enjoying their music (which is afterall the core of audiophilia).

This doesn't mean however that measurements are useless. Measurements and technical analysis of equipment in the domain of R&D and manufacturing is an important tool to ensure that devices perform consistently with no gross deviation from normal that might induce a subjective "perception" away from the common pattern of experience.

TLDR: Measurements are not completely synchronus with individual perception and hence your time is better spent listening and believing your own ears than to blindly go with data that is spit out of a machine.

Thanks for reading! Enjoy your weekend!

Cheers!

PS: Amir did not have any economic interests in restaurant A's business.

 

[bigshot]

It helps to know a little about the thresholds of human perception. Those have been studied and specs have been established, but a lot of audiophiles have never done that research.

 

[Giru]

It helps to know a little about the thresholds of human perception. Those have been studied and specs have been established, but a lot of audiophiles have never done that research.

I have a simple contention: if objectivity is an absolute determinant to performance, why does an equipment sound better than the best measuring one?

This simple question breaks down the very concept of "superiority". What is "better" if objectivity cannot be relied on? How does one decide? If an equipment doesn't sound good even though it tests well, what's the use?

Moreover, people bicker about SINAD, THD, crosstalk, etc but does it even matter to one's perceptibility?

Again, I think the whole debate is preposterous. Given how varied our bodies and environments are, absolute measurements of devices carry very little value to be of use in the real world.

No one can tell you what will suit you. You have to find that out for yourself.

 

[gregorio]

A simple analogy: let's say that Amir (no pun intended) buys a mass spectrometer and goes around town trying to find the restaurant that makes the best spaghetti.

Good analogy. Obviously this Amir would be a complete idiot because a mass spectrometer doesn’t measure how good spaghetti tastes to people, it just measures the elements spaghetti contains. If you want to measure how good the spaghetti tastes you need completely different equipment and a completely different measurement. For example, equipment that measures the pleasure centres of the brain.

There is no question of this test being able to correctly correlate to my real life "experience".

Absolutely there is no question of that! A mass spectrometer only tells you what the spaghetti is made of, it doesn’t tell anything at all about your “real life experience”, it is isn’t even connected to your brain!

Measurements are not completely synchronus with individual perception

They would be if they were measurements of individual perception but audio measurements are obviously measurements of audio, not of your individual brain/perception. How is that not obvious?

and hence your time is better spent listening and believing your own ears than to blindly go with data that is spit out of a machine.

If I want to know what my personal preference is, then of course all I can rely on is my own ears/brain but if I want to know if there is a difference in the audio or sound I’m listening to, what that difference actually is or if I’m just imagining it, then obviously I’m going to ultimately rely on audio or sound measurements. Again, how is that not obvious?

This all seems to be an obvious example of misinterpreting audio measurements, of thinking they’re something other than just measurements of audio performance.

G

 

[gregorio]

I have a simple contention: if objectivity is an absolute determinant to performance, why does an equipment sound better than the best measuring one?

Simple, either they are looking at a limited measurement/s that only measures some aspects of audio performance (but not others), they are testing incorrectly, or because they have a preference for lower fidelity that they then falsely describe as “better” simply because they prefer it.

This simple question breaks down the very concept of "superiority".

No it doesn’t, not if you answer it correctly as above!

What is "better" if objectivity cannot be relied on?

Objectivity can be relied on, that’s the whole point of objectivity! What can’t be relied on is audiophiles testing in a way that removes perceptual errors or audiophiles recognising that their preference does not necessarily correlate with better performance (fidelity).

How does one decide? If an equipment doesn't sound good even though it tests well, what's the use?

If equipment tests well but isn’t perceived/experienced as sounding good, then either it’s not being operated correctly or the person is not actually able to recognise “good” sound and is just describing poor sound as “good” because they prefer lower/poorer fidelity.

G

 

[Giru]

For example, equipment that measures the pleasure centres of the brain.

Ahhh. For a moment there I almost took you seriously.

Have a great weekend buddy. There's no problem if you like tones better than tunes

 

[bigshot]

There is fidelity, and there’s personal taste. You start with fidelity. You do the measurements and listening tests and aim for accuracy. Once you’ve achieved that as a baseline, there is nothing wrong with adjusting things a bit, like adding some salt and pepper to your food. It’s your stereo, you can do what you want with it. That’s what tone controls are for. But if you don’t start with accuracy, you have no anchor to your coloration, and it’s easy to drift into randomness.

 

[gregorio]

For a moment there I almost took you seriously.

There was never a moment when I took you seriously. Again, how is it not obvious that audio measurements are measurements of audio, not of your brain’s perception/experience/preference?

Have a great weekend buddy. There's no problem if you like tones better than tunes

You have a good weekend too but you have no idea what I like or apparently of the difference between audio measurements and human perception, even though you paradoxically posted videos about it!

G

 

[Giru]

There is fidelity, and there’s personal taste. You start with fidelity. You do the measurements and listening tests and aim for accuracy. Once you’ve achieved that as a baseline, there is nothing wrong with adjusting things a bit, like adding some salt and pepper to your food. It’s your stereo, you can do what you want with it. That’s what tone controls are for. But if you don’t start with accuracy, you have no anchor to your coloration, and it’s easy to drift into randomness.

Again, how do you quantify this "accuracy" for application to human use? You run a few tests and do some measurements to prove that the device has certain quantifiable attributes but that won't tell you whether it will sound good to someone or not. You cannot use objectivity to quantify something as subjective as music, especially given the anatomic and biologic factors that I've mentioned above. Moreover, adding that "salt and pepper" basically does away with your measurements and the holy "accuracy" that you were after, so why chase it in the first place?

Also it is important to note that most objective tests can only quantify a few attributes that can be verified by human test subjects (if such a test were to be done). Most major indicators of "fidelity" such as resolution, timbre, dynamics, etc are still beyond the realm of audio measurements. There is a lot of subjectivity to audio that is not quantifiable.

My point is simply this: You cannot use empirical data obtained from objective tests to proclaim a device "is best sounding". Sorry. It might be "best measuring" or "best technically capable" but I won't crown it as"superior sounding" to other equipment.

I know that this sounds as if I'm defending the snake oils that many a manufacturers peddle in the name of audiophilia with the basic idea that objectivity is of no value and all those voodoo that they do makes their products "sound" better. No.

I'm just saying that instead of going by objective tests, subjective reviews and opinions, one must try to trust their own ears. Nothing more. Nothing less.

 

[gregorio]

Again, how do you quantify this "accuracy" for application to human use?

Fidelity is an objective value, relatively easily measured, for any application, human or otherwise. IE. How closely does the output signal match the input signal of a particular device or chain of devices. For human application, we only need a level of fidelity that is audible. IE. Differences between the input and output signal that are below the level of audibility are irrelevant.

You run a few tests and do some measurements to prove that the device has certain quantifiable attributes but that won't tell you whether it will sound good to someone or not.

The above objective measurement obtained with a Null Test, provides the level of fidelity but of course higher fidelity will only sound good to someone who prefers higher fidelity.

You cannot use objectivity to quantify something as subjective as music, especially given the anatomic and biologic factors that I've mentioned above.

But we’re not using objective audio measurements to quantify music we use them to quantify audio performance!! Audio measurements are measurements of audio, nothing else. They are not measurements of your perception or subjective experience of music. How many times?

Most major indicators of "fidelity" such as resolution, timbre, dynamics, etc are still beyond the realm of audio measurements.

No, not a single one of those is beyond the realm of audio measurements. As digital audio is itself a measurement, if what you stated were true, then no digital audio recording would ever have any resolution, timbre, dynamics, etc. Surely you’re not claiming that are you?

There is a lot of subjectivity to audio that is not quantifiable.

No, there’s no subjectivity to audio at all. Audio is an electrical signal, how can you have a subjective preference to an electrical current you can’t perceive? AND, if it were not quantifiable then we would not be able to record or reproduce it and audio recordings would not exist. Again, you’re not seriously claiming that are you?

My point is simply this: You cannot use empirical data obtained from objective tests to proclaim a device "is best sounding".

Of course you can, except of course to those who feel lower fidelity sounds better than high fidelity.

It might be "best measuring" or "best technically capable" but I won't crown it as"superior sounding" to other equipment.

True, it won’t sound better than a piece of equipment with poorer measurements, if those poorer measurements are below the level of audibility, it will sound the same. And again, if the poorer measurements are audible, the better measuring device won’t necessarily sound superior to someone who prefers lower fidelity.

I know that this sounds as if I'm defending the snake oils that many a manufacturers peddle in the name of audiophilia with the basic idea that objectivity is of no value and all those voodoo that they do makes their products "sound" better.

Of course it sounds like you’re defending snake oil salesmen. As you’re making the exact same false assertions that snake oil salesmen make, there’s nothing else it can sound like!

G

 

[someguyontheinternet]

I think human variance should not be ignored, but at the same time humans are not as varied as some might think.
To understand the actual limitations of measurements also means to understand not only how much humans can differ, but also how likely it is for a number of randomly picked humans to differ significantly from each other. The overwhelming majority of humans will be much more similar than different.

 

[geek707]

I think human variance should not be ignored, but at the same time humans are not as varied as some might think.
To understand the actual limitations of measurements also means to understand not only how much humans can differ, but also how likely it is for a number of randomly picked humans to differ significantly from each other. The overwhelming majority of humans will be much more similar than different.

Define "same".
Define how you are determining "same". Define majority.
I think you will find that statement is not able to scale up.
A measurement taken with a device under defined conditions should be repeatable. If the measurement does not repeat you should have data to tell you what was different.

 

[someguyontheinternet]

Define "same".
Define how you are determining "same". Define majority.
I think you will find that statement is not able to scale up.
A measurement taken with a device under defined conditions should be repeatable. If the measurement does not repeat you should have data to tell you what was different.

Human traits and perception are no undiscovered fields. There are different methods and approaches that have been created to quantify differences in humans.

Measurements are repeatable. One of the big advantages of measuring devices is the ability to export and import test setups which only require the press of a button to execute in exactly the same manner on a different instance of the device.

 

[bfreedma]

Ahhh. For a moment there I almost took you seriously.

Have a great weekend buddy. There's no problem if you like tones better than tunes

You’re not addressing the topic at hand. The first step and what is being discussed here is identifying any audible differences. If and only if audible differences are established, one can then start evaluating personal preference.

 

[Giru]

I think human variance should not be ignored, but at the same time humans are not as varied as some might think.
To understand the actual limitations of measurements also means to understand not only how much humans can differ, but also how likely it is for a number of randomly picked humans to differ significantly from each other. The overwhelming majority of humans will be much more similar than different.

It's not about inter-subject variance per say, as much as it is about the translation of most objective measurements to humans.

This is what I'm trying to emphasize here: we know how to measure a device's technical "performance" (quite accurately infact) but do these metrics "always" translate to the real world performance while listening to music? I don't think so. The simple reason is that there's a lot of subjectivity involved that we still can't "objectify" with currently existing methods.

Most tests done by these "hobbyists" are basically benchmarking tests used by manufacturer's for internal R&D of their devices and only reflect certain technical proficiencies/deficiencies of their equipment. These are by no means an indicator of how good or bad the device might be for you or for listening to music in general. There are more descriptors to audio (especially when consumed in the form of music) that cannot be serviced by these tests.