Audio Science: Does it explain everything about how something sounds?

[amirm]

Yes, this is very clear from all of Amir's many posts. My question is why do the people taking the blind listening tests prefer one speaker to another? Are they thinking back to some reference of live music and think this particular speaker sounds more real or do they simply like one speaker more than another for any number of reasons? Or is it because one has the flattest frequency response and that is what they associate with live unamplified music or are they thinking that chocolate is a flavor, just like warm sound is a flavor, and they prefer vanilla?

Just a quick correction that the preference is for a "smooth" frequency response, not flat. A smooth response is free of sudden changes in frequency response but can deviate from flat by having a slope down from low to high frequencies.

Answering your question is hard without one experiencing the test. But let me give an example. If I took your system and inserted a box that boosted the mid frequencies by 5 dB, you would most likely find that change objectionable. It is that aberration that stands out, not its absence. In other words in the test, you notice the faults of certain loudspeakers and the most blameless one stands out as the best. Think of comparing an amplifier that is distorting versus not. It is the distortion that would cause you to mark the one down, not the betterness of the other.

The notion of unamplified sound does not come to mind. The loudspeakers are so different sounding that you judge them based on that differential. Something we can't normally do without the quick switching in their tests.

Do we or they know why they prefer one speaker over the others? Perhaps it does not even matter why they universally select the same speaker in these blind tests. Maybe it is as simple as Harmon gathering the data and then selling the speaker the blind listeners tell them they prefer?

Again, they are not picking the same loudspeaker. They are picking loudspeakers with certain characteristics and that is the smooth frequency response I mentioned. That aside, I can't explain the why. It just seems that when comparing loudspeakers, we seem to prefer the one that doesn't boost or reduce the frequency response. It is a relief that we are good judges of what is technically correct fidelity, sans the sloping down response.

 

[jkeny]

More excellent questions, but I think you're talking about completely different kinds of sighted bias. That early psychoacoustic model develops, for example, the ability to recognize, in all it's variations, the human voice as human. You may see a young woman and expect higher pitch, sweeter tonality. You may see a big, rough-looking middle-aged man and expect something completely different. But you don't look at the quality of his shirt and expect him to sound "better." And that's what happens in sighted bias for audio gear. We expect the bigger speaker, the heavier amplifier, the more respected brand to be "better." And so we hear it as better. Immediately. If the component has a sonic flaw, can it begin to wear on us over time? Sure. I've never doubted that. What I doubt is that accuracy, naturalness is not self-evident at first the blind listening, and only reveals itself over time. I doubt it and the Harman studies confirm it. I think that "revelation" is more about acclimation, pride of ownership and purchase justification. YMMV.

Tim

But, Tim, we also develop a psychoacoustic model for a guitar, drum, etc - shiny ones, expensive ones, etc. - they all come with expectation attached - so how do we distill out this bias & come to our universal agreement on what constitutes realistic sound?

 

[amirm]

You are asking the proper question. In order to debate it we must analyze how the preference was established and measured and in what exact conditions it was carried. IMHO Harman tests were carried in non audiophile conditions - no synergy considerations, no fine tuning, no desire to sound life or reaching a live sound reference.

Haman tests are according to accepted audio science. The conditions you mention are not. Until such time that you can demonstrate those changes to be material to outcome of testing loudspeakers which sound wildly different from each other, you don't have a basis for arguing these points with the rest of us who believe in the soundness of the scientific method. Heck, you have not even presented a lay opinion of that which sticks.

If you deprive stereo sound reproduction from what some of use consider its best but very difficult to reach attributes, including sounding like real instruments, perhaps people will focus on other aspects, probably determined by the conditions of the test.

And you don't think you should have to demonstrate this belief using any kind of scientifically valid methodology? It should be accepted at face value and that is that?

BTW, as far as I know science does not accept paradoxes, it should explain them ...

Let's not use the word science in a post that is devoid of it please.

 

[Tony Lauck]

Your perspective is interesting as Living Presence recordings are generally noted for being well-engineered and excellent sounding. That said, I find it difficult to believe it’s the recording or that your system is truly flat. And it most likely is not because Living Presence intended its recordings to excel with speakers that roll-off at the frequency extremes.

Rather than be so quick to blame the Living Presence or similar labels, I suspect that fingernails on the chalkboard affect is most likely due to 1 of 3 distinct possibilities or combinations thereof:

1. Your system is distorted a bit more than others that in turn induces a bit more ear fatigue.

Though many deny it, it’s a fact that every last system is plagued with very serious distortions that utterly cripple their performance. Thus making some-to-many recordings difficult to tolerate. However, it would not be that unusual if you happened to possess a unique component and/or cable that takes distortions just a bit over the top.

2. In contrast, it could be that you possess a unique component and/or cable that is actually a bit more revealing than most.

Every once in a while we read about somebody reporting their findings about a new cable or component upgrade. They essentially report that their upgraded product was too detailed and as a result lack “musicality” so they removed the product from the system. Of course, there is no such thing as product being too detailed or 101% detail. Even though there are a few mfg’ers out there that like to zip up the highs a bit in the attempt to give the appearance their product is more detailed but that’s fairly rare.

In such cases, the “too detailed” product is really just more revealing and a truly more revealing product is indiscriminate about what it’s revealing. Whether it’s music or distortions. And if per chance this is the case, if more music is revealed it's a given that more distortions would be revealed as well.

3. It could be your hearing is perhaps just a bit more keen than average (not all that unusual) so that when a superior recording i.e. Living Presence is playing on your system, your ears become a bit more aware that something (else) isn’t right.

You’ve not mentioned anything about recent component upgrades, but in a sense are you not performing an upgrade every time you listen to a well-engineered or superior recording?

It’s a bit of a crapshoot, but none of these possibilities are unusual. They seem to be the only reasonable possibilities that could more fully explain what you may be hearing.

The system, out of the box, was flat from 1 kHz up to 20 kHz, measured with a calibrated microphone at my near field listening position. I ran a number of tests and confirmed this. Of course it was nowhere near flat in the bass, thanks to room modes. There wasn't enough room to add bass traps, but I did get a sub and after tweaking the cross-over adjustments got flat response from 30 to 1 kHz with the exception of two peaks of +8 and +9 dB. I took these out with a software parametric equalizer (done by HQPlayer). The highs were still harsh and the Living Presence recordings unlistenable. This is because of the near field listening which doesn't add any high frequency roll off due to room response and speaker polars. The cure for the excessive brightness was very simple: dial down the high frequencies using the tweeter control on the back of the monitors, applying a house curve that gradually declines from 0 dB at 1 kHz to -3.5 dB at 10 kHz and levels off at about -4 dB above that. (I didn't measure this, I adjusted the controls half way down, listened to a few dozen reference recordings and noted that all of them were now neither bright nor dull. Then I measured the results with the calibrated microphone.)

So now, when I play the Ivan Fischer Mahler 9th in DSD I get a good approximation, tonality wise, to what I am used to hearing at a live concert in a good hall sitting in row 20. Also, playing solo piano recordings, non-audiphiles comment that they never have before heard a recording of a piano sound like a live instrument. There is no ear fatigue listening to classical music recordings. Playing jazz at front row club volumes then there will be a problem, because I can get clean sound up to 115 dB at the listening position, reproducing the fatigue (and eventual hearing damage) of live music that's too loud. (However, the blast of a loud jazz trumpet can definitely raise the hairs on one's arms.)

So yes, as set up originally with default speaker cross-over settings and no room EQ something was definitely not right. I analyzed the problem and corrected it. By the way, I am far from the only person who has said that the Mercury recordings were too bright.

 

[jkeny]

But, Tim, we also develop a psychoacoustic model for a guitar, drum, etc - shiny ones, expensive ones, etc. - they all come with expectation attached - so how do we distill out this bias & come to our universal agreement on what constitutes realistic sound?

This was posted from a phone so few words were used. What I'm trying to establish is that expectation bias is a perennial attachment to all our perceptions i.e to everything we hear - it doesn't only apply to audio devices. If this is the case & we all develop a universal psychoacoustic model for what is realistic sounding, then there must be some way of overcoming this bias, otherwise we couldn't all converge on the same psychoacoustic model. I'm suggesting that long term exposure to the sound is the key to overcoming these biasing factors i.e long term listening allows us to judge what remains fixed in the sound (the actual acoustic) from day to day listening & what varies (the biasing factors) - in this way we come to realise the essence of the sound.

Just to throw a spanner in the works - In ways all our listening is controlled by expectation i.e these psychoacoustic models would seem to have an expectation for the attack sustain decay release envelope of a piano note, for instance. If the reproduced piano sound doesn't correlate to this expectation, within certain parameters, then we find the piano less convincing, less realistic - our expectation of the perception of the sound's journey through attack, sustain, decay, release envelope is not met in the way a real piano would meet it.

 

[amirm]

This was posted from a phone so few words were used. What I'm trying to establish is that expectation bias is a perennial attachment to all our perceptions i.e to everything we hear - it doesn't only apply to audio devices. If this is the case & we all develop a universal psychoacoustic model for what is realistic sounding, then there must be some way of overcoming this bias, otherwise we couldn't all converge on the same psychoacoustic model. I'm suggesting that long term exposure to the sound is the key to overcoming these biasing factors i.e long term listening allows us to judge what remains fixed in the sound (the actual acoustic) from day to day listening & what varies (the biasing factors) - in this way we come to realise the essence of the sound.

Hi John. The issue with the long term testing is that the theory has been tested and shown to not be correct. This was documented in Clark's 1991 paper, Ten Years of ABX Testing. Here is the excerpt:

Distortions that were clearly audible in double blind ABX tests, were not in take home, slow switching tests. This very much matches my own experience. If you had me for example repeat all of those high resolution tests with long term listening, I am confident I would fail them all . It was the near instant switching that allowed me to use my short term memory effectively to tell the samples apart.

Do you know of any tests that show otherwise?

 

[jkeny]

Hi John. The issue with the long term testing is that the theory has been tested and shown to not be correct. This was documented in Clark's 1991 paper, Ten Years of ABX Testing.

Distortions that were clearly audible in double blind ABX tests, were not in take home, slow switching tests. This very much matches my own experience. If you had me for example repeat all of those high resolution tests with long term listening, I am confident I would fail them all . It was the near instant switching that allowed me to use my short term memory effectively to tell the samples apart.

Do you know of any tests that show otherwise?

I don't know enough about Clarke's test to comment in detail but, if I understand it correctly, it doesn't say that the people were given two identical boxes to take home (one of which had 2.5% distortion) & to identify which they thought was better. This is what is done in the A/B element of the test & what is typically done in long-term listening - a period of listening to a new device is used to acclimatise to the sound & then the old device put back in to see if there is an appreciable difference.

I've seen another such long-term Vs A/B experiment used to undermine the appropriateness of long-term testing - I think it was by Nousaine - & again it makes the same mistake, except in this test the A/B testers were given training before the A/B test, doh!! & weren't given any training before take-it-home long-term testing, doh!!!!

Long-term listening doesn't preclude A/B listening - it can give the listener the perception that two devices sound different without the listener being able to identify what exactly the difference is. Moving to an ABX style test can often confuse matters as we are then put in a position of trying to identify X which is really another level of differentiation. As you know, if all we have is a gut feeling of the difference & can't identify a specific "tell" then ABX testing can often befuddle & confuse.

Also, Amir, in your high-res ABX tests, did sighted listening not first allow you to decide if there was an audible difference (or a feeling that there was an audible difference) & then to zone in on & begin to try to isolate what this difference is? Surely, this comes from your built in model of what these particular distortions sound like in particular audio circumstances?

I see a recent posting on HA (yes the home of ABX ) where Mzil is testing some ITDs & the tell changes as he goes down to nearer his JND for ITDs (about 10uS) - so it ain't even just a simple, consistent "tell" that's being identified in ABX testing.

From two posts of his

My understanding is that when we reach the JND points it is not at all uncommon for the listener to lose sight of what exactly it is that they are keying on. Is it level? Tone? Direction? Other?

....
Well in some instances my perception is clearly slight alterations in lateral positioning, just as we would expect, however in other instances I suspect my brain hears the two impulses and reinterprets them as two crests of a wave form, two cycles long in duration, and hears them as slightly different frequencies. [Sort of a singular "thunk" vs "thump/thud" sound.] This second detection method is not what we are testing for, ITD, but I can't stop my brain from using it as a "cheating method" now can I?

Although I can see how using impulses in isolation would be useful to increase sensitivity and discrimination even down to the 10 microsecond level, or so, I fear there may be other things at play such as frequency detection that I'm subconsciously keying on. Listening instead to music or perhaps even better, continuous correlated pink noise, would be a possible solution. [I can't hear a "frequency" to the thump noise when there is no thump!]

I never would have thought it, but I know I can hear ITD even down to the small single digits of milliseconds when listening to correlated pink noise. I'm not sure if I can get down to the microsecond levels but I would be more than glad to try if you post it. To the best of my knowledge using this would cure the issue I'm worrying about.

Just shows what little he knows about psychoacoustics - he thinks he's "cheating" because he hears an audible difference which is not to his expectation of how ITD SHOULD be heard, "slight alterations in lateral positioning" doh!!

Oh, Amir, if you read that thread, we are being lumped together & called "the organic twins" (in a derogatory sense) as in "Other than the usual vulnerability of the test tone section (which people like the organic twins will surely exploit without admitting to it)"

 

[PeterA]

No reference. It is quite a paradox in how we can tell the good from the bad without one.

Amir, your response is to my post: "Thinking of these questions led me to consider whether or not the listening panel at Harmon considers any reference when they are asked for their preferences about the speakers under evaluation. Do they respond with the idea that one sounds more "real" than another, and if so, based on what, or do they simply prefer one speaker to another because they like the way it sounds? Perhaps Amir has some insight about this."

You clearly state: "No reference." I take that to mean literally, "No reference to anything at all" because I did not qualify "any reference" in my statement.

That's interesting. How do you know the blindfolded subjects where not considering a reference or thinking about what they remember a guitar or voice (for example) sounding like? Did they state this in some questionaire or form or where they instructed not to consider the sound of actual instruments and voices when judging the quality of the loudspeakers? What can a preference for anything possibly be based on if not some personal experience or reference about the subject being tested?

If they were only asked to compare one speaker to another, what were the criteria for passing judgment or forming an opinion, if the subjects could not refer to anything at all, including the other speakers in the test?

How is it even possible to compare one speaker to another if you can not consider the other speaker as a reference during the comparison?

Are you saying that we can distinguish between good and bad without any reference to anything, including our own experiences? How is that possible? I had always thought "good" and "bad" were relative rather than absolute terms and therefore that they require some context to be understood. Devoid of any context or reference, what does, "It is good" even mean? Could you explain this in some more detail? Thanks.

 

[microstrip]

Haman tests are according to accepted audio science. The conditions you mention are not. Until such time that you can demonstrate those changes to be material to outcome of testing loudspeakers which sound wildly different from each other, you don't have a basis for arguing these points with the rest of us who believe in the soundness of the scientific method. Heck, you have not even presented a lay opinion of that which sticks.

And you don't think you should have to demonstrate this belief using any kind of scientifically valid methodology? It should be accepted at face value and that is that?


Let's not use the word science in a post that is devoid of it please.

Weak arguments disguised with the soundness of the scientific method. I am referring to facts that are accepted by the larger part audiophile community of WBF, I do not have to demonstrate anything. And yes, the post had no science and did not pretend to. My point has always been that most (not all, surely, of the stereo high-end sound) can not be proved scientifically by current scientific knowledge. You want to live in the part that can be proved and ignore all else. Please go on doing it in your crusade.

 

[amirm]

Also, Amir, in your high-res ABX tests, did sighted listening not first allow you to decide if there was an audible difference (or a feeling that there was an audible difference) & then to zone in on & begin to try to isolate what this difference is? Surely, this comes from your built in model of what these particular distortions sound like in particular audio circumstances?

No and no . While I usually do that, in these tests I just jumped directly into blind testing with Foobar since that was the most convenient and what the task eventually was. You can see my trials to find the differences in this sample results:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/31 15:18:41

File A: C:\Users\Amir\Music\AIX AVS Test files\On_The_Street_Where_You_Live_A2.mp3
File B: C:\Users\Amir\Music\AIX AVS Test files\On_The_Street_Where_You_Live_A2.wav

15:18:41 : Test started.
15:19:18 : 01/01 50.0%
15:19:30 : 01/02 75.0%
15:19:44 : 01/03 87.5%
15:20:35 : 02/04 68.8%
15:20:46 : 02/05 81.3%
15:21:39 : 03/06 65.6% <--- Difference found
15:21:47 : 04/07 50.0%
15:21:54 : 04/08 63.7% <--- Dog barked!
15:22:06 : 05/09 50.0%
15:22:19 : 06/10 37.7%
15:22:31 : 07/11 27.4%
15:22:44 : 08/12 19.4%
15:22:51 : 09/13 13.3%
15:22:58 : 10/14 9.0%
15:23:06 : 11/15 5.9%
15:23:14 : 12/16 3.8%
15:23:23 : 13/17 2.5%
15:23:33 : 14/18 1.5%
15:23:42 : 15/19 1.0%
15:23:54 : 16/20 0.6%
15:24:06 : 17/21 0.4%
15:24:15 : 18/22 0.2%
15:24:23 : 19/23 0.1%
15:24:34 : 20/24 0.1%
15:24:43 : 21/25 0.0%
15:24:52 : 22/26 0.0%
15:24:57 : Test finished.

----------
Total: 22/26 (0.0%)
You can see my failures to find the differences at the beginning where I kept getting the answer wrong.

And my success did not depend on any model of what was right (other than MP3 tests). But rather, finding a difference and stopping there.

 

[jkeny]

Another, BTW, about that HA thread - I see when I go back to it from yesterday some post shave been added.
His comment above about the "organic twins" is in relation to his positive ABX result for a new Arny ABX test (I know, I know - another one - it defines his world & him).
Anyway, it appears that Mzil aced this ABX test by hearing a faint difference in the background noise during the IMD test tones that Arny has kept on his new test files.
"The audible difference was during the test tone sequence. The faint background noise changes. I guess you could argue it is IM on either my end or your end but boy is it faint." Hence his suggestion that positive ABX test results for this test in the past were "cheats"

I wonder if this "faint background noise changes" is actually differences in noise modulation between the RB & High-res test tones?

 

[amirm]

Weak arguments disguised with the soundness of the scientific method. I am referring to facts that are accepted by the larger part audiophile community of WBF, I do not have to demonstrate anything.

You do if are choosing to argue with the rest of the WBF membership which doesn't ignore audio science.

And yes, the post had no science and did not pretend to. My point has always been that most (not all, surely, of the stereo high-end sound) can not be proved scientifically by current scientific knowledge. You want to live in the part that can be proved and ignore all else. Please go on doing it in your crusade.

For something to be proven, the hypothesis needs to be right. As I mentioned, your argument doesn't even pass lay logic let alone scientific. You are telling me that if I conduct a test of orange soda against coke, that the type of cup used would reverse the outcome of what people preferred. Don't you think you need some shred of evidence to make it plausible that changing the rest of the system would change the outcome of preference tests for loudspeakers?

 

[jkeny]

No and no . While I usually do that, in these tests I just jumped directly into blind testing with Foobar since that was the most convenient and what the task eventually was. You can see my trials to find the differences in this sample results:

---------- [/COLOR]
Total: 22/26 (0.0%)
You can see my failures to find the differences at the beginning where I kept getting the answer wrong.

And my success did not depend on any model of what was right (other than MP3 tests). But rather, finding a difference and stopping there.

Ok, but that was the weakest, the most speculative point, in all of my post - refuting it does not answer anything else in my post.

 

[amirm]

That's interesting. How do you know the blindfolded subjects where not considering a reference or thinking about what they remember a guitar or voice (for example) sounding like? Did they state this in some questionaire or form or where they instructed not to consider the sound of actual instruments and voices when judging the quality of the loudspeakers? What can a preference for anything possibly be based on if not some personal experience or reference about the subject being tested?

??? I have taken the test twice so what I shared was my experience which was identical to what the formal listening tests consisted of. You are not given any instructions on how to evaluate one loudspeaker against the other. It is up to you to do that however you choose. When you sit there, you are immediately presented with loudspeakers which tonally sound hugely different. It is that difference that you wind up judging.

If they were only asked to compare one speaker to another, what were the criteria for passing judgment or forming an opinion, if the subjects could not refer to anything at all, including the other speakers in the test?

You hear all the samples being tested, say 3 loudspeakers, and you give them score 1 to 10. No different than if you walked into a showroom and were choosing among them. There is no fourth loudspeaker that is the reference.

How is it even possible to compare one speaker to another if you can not consider the other speaker as a reference during the comparison?

How is it possible to do it in our everyday sighted testing? We compare loudspeaker A to B and decide one sounds better than the other. This test is the same except that the conditions of listening to each loudspeaker is identical and you don't know the identity of the loudspeakers. And switching happens in seconds.

Are you saying that we can distinguish between good and bad without any reference to anything, including our own experiences?

I am saying that I don't know how we do it . We just do. We can tell a loudspeaker that has a dip between 2 and 3 Khz as not sounding as good as one that doesn't. We can tell that one loudspeaker has overboosted low frequencies not being as good as one that doesn't. In that kind of identification, we are drawing on some kind of internal reference I guess. What that is, I can't tell you. Other than just like my experience, we know that it is absence of problems that we are detecting, not realism of the alternative.

How is that possible? I had always thought "good" and "bad" were relative rather than absolute terms and therefore that they require some context to be understood. Devoid of any context or reference, what does, "It is good" even mean? Could you explain this in some more detail? Thanks.

Let me quote Dr. Toole in his book:

"Descriptors like pleasantness and preference must therefore be considered
as ranking in importance with accuracy and fidelity. This may seem like a dangerous
path to take, risking the corruption of all that is revered in the purity of
an original live performance. Fortunately, it turns out that when given the
opportunity to judge without bias, human listeners are excellent detectors of
artifacts and distortions; they are remarkably trustworthy guardians of what is
good. Having only a vague concept of what might be correct, listeners recognize
what is wrong. An absence of problems becomes a measure of excellence. By
the end of this book, we will see that technical excellence turns out to be a high
correlate of both perceived accuracy and emotional gratification, and most of us
can recognize it when we hear it."

What I have bolded is exactly what went on in my mind when I took the test.

Perhaps if we had a true reference, our preferences in such tests would be different. But given the impossibility of that, most of seem to fall inline with the above observations of "why."

 

[PeterA]

That's what I've been saying all along - we have internalised a model of how things sound in the world & this is our reference point - some reproductions sound more realistic than others because they better match the criteria that this model uses for evaluation.

Two immediate things spring to mind:
- we don't know these criteria & we don't know the model fully yet. Harmon's results are one aspect which seem to fit but what is the psychoacoustic rule here - that the frequency should not have any major deviations along it's spectrum (what's the deviation limit that is noticeable?) & that the off-axis frequency spectrum should match closely to the on-axis (how close a match is needed?)
- when we were building this psychoacoustic model(early in life) how come doing so sighted didn't cause us all to have a different set of criteria - how come we all came to roughly the same psychoacoustic model? I say it's because over time we average out any sighted biases through repetition - we hear through our biases - one day we might be biased towards one aspect, another day another aspect - what remains fixed is the underlying sound which we tease out over time. This is also the basis on which long-term sighted listening works!!

jkeny, Amir wrote that the Harmon blindfolded subjects used "no reference." I assume he meant that literally, ie, no reference at all. I don't understand how that is possible. I agree with you that they must have used some internal reference based on their own personal experience of what something sounds like, whatever that might be. In other words, they referred to something while they were judging the speakers and voting their preferences.

 

[microstrip]

You do if are choosing to argue with the rest of the WBF membership which doesn't ignore audio science.


For something to be proven, the hypothesis needs to be right. As I mentioned, your argument doesn't even pass lay logic let alone scientific. You are telling me that if I conduct a test of orange soda against coke, that the type of cup used would reverse the outcome of what people preferred. Don't you think you need some shred of evidence to make it plausible that changing the rest of the system would change the outcome of preference tests for loudspeakers?

Again, the question is not about ignoring science. The point is wether current SOTA stereo reproduction can or can't be explained TOTALY by current science.If not science alone can not be used to debate it.

And yes, it is my believe that different speakers can need different systems and rooms, otherwise I would have another hobby and I would not read WBF. YMMV, as they say.

 

[Tony Lauck]

You do if are choosing to argue with the rest of the WBF membership which doesn't ignore audio science.


For something to be proven, the hypothesis needs to be right. As I mentioned, your argument doesn't even pass lay logic let alone scientific. You are telling me that if I conduct a test of orange soda against coke, that the type of cup used would reverse the outcome of what people preferred. Don't you think you need some shred of evidence to make it plausible that changing the rest of the system would change the outcome of preference tests for loudspeakers?

Count me out. I'm one of those people who ignores audio science when it pleases me. I make decisions on my own using my own mind and senses, not following any rituals created by authority figures. In particular, I ignore "audio science" when invoked in arguments that set off my BS detector.

Now I see that we have polylogism, to use a phrase coined by Ludwig von Mises. One logic for laymen and another for scientists. I think not.

Are you a wine connoisseur? If you are, you would know that different wine glasses are commonly used for different wines and that this is related to perception and enjoyment. Why would this be any different when it comes to soft drinks? Or any other field of activity involving human sense perception, something that is by its very nature subjective and can be objectified only by turning human subjects into pigeons in the laboratory of some Harvard scientist.

 

[amirm]

Again, the question is not about ignoring science. The point is wether current SOTA stereo reproduction can or can't be explained TOTALY by current science.

How can that be the only explanation? Can I say that I want to take megadoses of vitamins to treat my cancer and the fact that science can't explain why that works, is the fault of current science? Surely there is another possibility that what you believe, is wrong.

If not science alone can not be used to debate it.

Again, if you are going to argue with people who believe in science, then what you just said is non-sequitur. Don't engage us if all you have to offer is opinions that you can't back other than "other people believe it too." Many people believe in treatments for cancer that science doesn't recognize. That doesn't make their opinion right.

And yes, it is my believe that different speakers can need different systems and rooms, otherwise I would have another hobby and I would not read WBF. YMMV, as they say.

You are claiming that if something other than loudspeakers is changed, listener preference for a loudspeaker changes. And hence the outcome of the tests is not valid. That has nothing to do with what the loudspeaker "needs." You need to show that if you did the test with a different amplifier, the outcome would change. Do you have any such data or we are just dealing with speculation with no foundation?

 

[PeterA]

??? I have taken the test twice so what I shared was my experience which was identical to what the formal listening tests consisted of. You are not given any instructions on how to evaluate one loudspeaker against the other. It is up to you to do that however you choose. When you sit there, you are immediately presented with loudspeakers which tonally sound hugely different. It is that difference that you wind up judging.


You hear all the samples being tested, say 3 loudspeakers, and you give them score 1 to 10. No different than if you walked into a showroom and were choosing among them. There is no fourth loudspeaker that is the reference.


How is it possible to do it in our everyday sighted testing? We compare loudspeaker A to B and decide one sounds better than the other. This test is the same except that the conditions of listening to each loudspeaker is identical and you don't know the identity of the loudspeakers. And switching happens in seconds.


I am saying that I don't know how we do it . We just do. We can tell a loudspeaker that has a dip between 2 and 3 Khz as not sounding as good as one that doesn't. We can tell that one loudspeaker has overboosted low frequencies not being as good as one that doesn't. In that kind of identification, we are drawing on some kind of internal reference I guess. What that is, I can't tell you. Other than just like my experience, we know that it is absence of problems that we are detecting, not realism of the alternative.


Let me quote Dr. Toole in his book:

"Descriptors like pleasantness and preference must therefore be considered
as ranking in importance with accuracy and fidelity. This may seem like a dangerous
path to take, risking the corruption of all that is revered in the purity of
an original live performance. Fortunately, it turns out that when given the
opportunity to judge without bias, human listeners are excellent detectors of
artifacts and distortions; they are remarkably trustworthy guardians of what is
good. Having only a vague concept of what might be correct, listeners recognize
what is wrong. An absence of problems becomes a measure of excellence. By
the end of this book, we will see that technical excellence turns out to be a high
correlate of both perceived accuracy and emotional gratification, and most of us
can recognize it when we hear it."

What I have bolded is exactly what went on in my mind when I took the test.

Perhaps if we had a true reference, our preferences in such tests would be different. But given the impossibility of that, most of seem to fall inline with the above observations of "why."

Thank you Amir. It's now clear to me that the test subjects were indeed operating with a reference. We just don't know what it was. You call it an "internal reference." I agree, one kind of reference can be internal. It is the summation of our past experiences that inform us of our current judgments. When comparing two loudspeakers to each other, at a minimum, one speaker's quality is compared directly to that of the other. There are two references here. Our internal ones and the sound of the other speaker being compared. At more that a minimum, perhaps they are both being compared to another speaker, not there, but remembered as a reference, or even to the sound of live music. It varies from person to person, I am sure, but there is clearly some reference, and not "no reference".

To judge the tone of a loudspeaker, or to make any judgement, one must refer to something. In the Harmon tests, Speaker A was preferred to speakers B and C (or some other order) because they listened to all three and chose the one with the fewest artifacts, distortions and deviations from a smooth frequency response. These judgements could not have been made in a vacuum with no internal reference to anything. The preference relies on the subject referring back to his own ideas and experiences of relative distortions, artifacts and the sound of a smooth frequency response. We just don't know what the references the test takers used were, but it is clear they used one or many, just as you used one when you took the test. Did you not think about your own past experience with distortion, artifacts and smooth frequency response? Did you not think of the other two speakers that you just heard before you voted for your favorite? Those are references.

Perhaps you meant that the blindfolded test takers were not thinking of the sound of real instruments or voices when you wrote "no reference" were used. If so, you did not write that at the time. That is certainly possible, but I would think difficult to prove without asking them what they were using for their own internal references.

 

[amirm]

Are you a wine connoisseur? If you are, you would know that different wine glasses are commonly used for different wines and that this is related to perception and enjoyment. Why would this be any different when it comes to soft drinks?

Because it would not change the outcome of orange soda against coke. Those two are so different in taste that the glass they are served in doesn't change the outcome. At least it is not plausible that it would. Likewise, we are discussing preferences for loudspeakers in a test that I personally participated in. The argument is that should they have changed amplifiers or whatever, that outcome would change. I want to hear some plausible argument that such a change would make the outcome of loudspeakers that sound so different by themselves different. You said you are a man of your experiences. I just shared my direct experience. Tell me why I should not believe it and instead believe you with no experience with said test.

Or any other field of activity involving human sense perception, something that is by its very nature subjective and can be objectified only by turning human subjects into pigeons in the laboratory of some Harvard scientist.

Sorry, no. The test is what everyone does day in and day out. Listen to two loudspeakers and express which one they like better. The only difference was that other factors which have nothing to do with the sound as such make, model and looks of the loudspeakers are hidden. Tell me why I should not trust my personal experience in such a test. That I should like a speaker that sounds worse, better because it looks better. Is that what you are saying?