Can anyone explain the vinyl renaissance?

[MattHooper]

Haha, I got sucked in and broke my Matt rule. My bad.

If your "Matt rule" whatever it is, amounts to putting me on ignore, perhaps at least you won't misrepresent what I write. (Though whatever rule that is, hasn't stopped you).

Let’s see how my conversation with board goes, ie the one that is probably in good faith.

Good faith? Do you find lying about misrepresenting other people's views to be "good faith?" (e.g. that "Matt is in denial" about the power of cognitive bias, when any honest reading of what I wrote shows that is the opposite of the truth).

If I get it wrong and fail to reflect the audio science, then I am very open to that correction: just show me the audio science that is inconsistent with my reflection. Don’t regale me with anecdotal stories that you think have merit, littered with Whataboutisms.

Do you remember writing this: Just do the blind tests, and show me the consistency with sighted.

Two people did just that. And you've ignored this yet again. And now you are waving things off as "anecdotes and whataboutisms." So, no, you don't really seem open to amending your claims. You ask for something then ignore the responsees or move the goal posts. And people notice this.

The issue is that you have often been sloppy in what you have written, verging over the edge to dogmatic, rather than nuanced and scientifically cautious.

So, if you see other writers doing all the above with my posts, just remember that they are disputing audio science….at least, up until the point that they provide audio science to the contrary.

No, this is your blind spot Newman. (And you aren't the only one, just the most persistent).

You aren't just citing the science. You have been giving us YOUR dogmatic interpretation.

Take an example I've used: you could cite the scientific data that shows how our sight can be fooled. You can show studies that appeal to optical illusions, even produce the illusions themselves, to show how some built in heuristics/biases in our visual system can lead us to error. That would be uncontroversial.

But IF your interpretation of this data was to then claim "From this we can see that our visual system is thoroughly unreliable!" then...no...now you've overreached. Because our visual system clearly is not "thoroughly unreliable." There's endless observations showing our visual system can deliver reliable information. Your claim would have to account for all that before anyone should take it seriously. (This ringing a bell?). But if you reply "ignore those who protest, just remember they are disputing the science," ...no, you've failed to understand the problem. Specific data isn't disputed; your interpretation of the data in what you write is being disputed as often sloppy and overreaching.

This is what's happening, often enough, when you write about cognitive biases in regard to our hearing. Clearly, generally speaking, our perception of sound is not "thoroughly unreliable" outside scientific controls, whether it's in audio or anything else. You always portray sighted listening as dominated by cognitive bias such that one is not really rating the sound waves, and that clearly isn't the case in all sorts of examples. In fact, you've just ignored yet again two examples: the ones given to you by me and board. If our hearing in uncontrolled conditions is"thoroughly unreliable"...how were our bind test results consistent with our sighted listening impressions? Don't ignore counter arguments/evidence, Newman.

So your new statement you've attempted to tighten up:

sighted listening is thoroughly unreliable as a means to determine the sonic attributes or preferences of the sound waves alone due to a device under test, and the only reliable listening-based way to determine such attributes or preferences is via well-controlled listening tests where the listeners don’t even know what are the devices being tested, and ideally involving three or more devices.

Ok, you are almost there. Your "thoroughly unreliable" is still possibly misleading. If you mean "sighted listening will always lead to inaccurate perception of the sound" then you are clearly wrong. But this is what you often imply. But IF you can admit what you mean is actually "sighted listening CAN lead to inaccurate perception of the sound"....then you'd be on firm ground. Then you wouldn't be unjustly dismissing all sighted listening as inaccurate.

But then, why don't you just default to the more careful, accurate claim I already gave you before?

If you want to be confident that cognitive bias isn't altering your perception of the sound, you should use controls to minimize that variable, e.g. blind testing.

 

[tmtomh]

His point is that sighted impressions are completely unreliable. If there were practical nuance I suspect this argument wouldn’t exist.

I'm not denying that's his claim. When I ask "can't we all agree," I'm not limiting my question/request to one or the other person or side in this argument in particular.

To put it another way, I'm trying to ask anyone who's willing to respond, "if you feel we can't agree on those three basic propositions, then say which one you don't agree with and explain why you can't agree with it."

 

[JP]

I'm not denying that's his claim. When I ask "can't we all agree," I'm not limiting my question/request to one or the other person or side in this argument in particular.

Sure, those are reasonable, but they're "stricter" than the subject matter in that the discussion isn't about nailing down specific differences, rather preference for an overall character.

 

[tmtomh]

Sure, those are reasonable, but they're "stricter" than the subject matter in that the discussion isn't about nailing down specific differences, rather preference for an overall character.

Well, then I guess you and I will have to agree to disagree about whether that's a meaningful distinction.

 

[JP]

Well, then I guess you and I will have to agree to disagree about whether that's a meaningful distinction.

Let's illustrate it with real examples.

150MLX is my rig with an AT-150MLX loaded to be flat. PP200 is someone else's rig with a Phasemation PP-200, which is representative of the character of "high-end" cartridges being discussed. Here are some measurements, though not of the same copy.

 

[tmtomh]

Let's illustrate it with real examples.

150MLX is my rig with an AT-150MLX loaded to be flat. PP200 is someone else's rig with a Phasemation PP-200, which is representative of the character of "high-end" cartridges being discussed. Here are some measurements, though not of the same copy.

Sorry, I've lost track of what the "it" is suppose to be here that's being illustrated. As I clearly wrote above, I have no disagreement with the proposition that our sighted listening impressions can pick up on real sonic differences. My view is that our ability to do this is very real, and also more limited and less reliable than most of us tend to think. In my view, whether we apply this idea to detailed, very specific differences, or "overall character" is immaterial.

There's also the 1980 Toole blind-test comparison of various phono cartridges, which I know you're aware of.

Toole Blind Cartridge Comparison

From Audio Scene Canada August 1980, a blind screen comparison by Floyd Toole of 2 MCs and a MM which I decided to add under its own thread rather than at least 2 others to which I might have added it.

www.audiosciencereview.com

To me that test illustrates that the reputations of the cartridges going into the test were based precisely on sighted listening impressions of "overall character" - impressions which turned out to be wrong, nonexistent, or overstated when the comparisons were done blind.

So once again, it would seem that sighted listening indeed has some ability to detect differences - but only some. And I would maintain that the aggregate evidence shows that we tend to overestimate our ability to accurately and reliably detect differences - whether specific or "overall character" - using sighted listening comparisons.

 

[MattHooper]

Can't we just agree that our ability to hear sonic differences in sighted, uncontrolled listening is a real ability, but also an ability that most of us overestimate? (And that part of listening bias is precisely that we tend to overestimate our abilities and underestimate the power of unconscious factors?)

And can't we further agree that our ability to identify the cause or source of heard differences is even more overestimated?

Agreed. (Though...it can depend on who is doing the estimation. Some people overestimate far more than others. I'd think we would agree that the typical "golden ear" subjectivist audiophile overestimates far more than we tend to on ASR).

And can't we further agree that we overestimate our ability to predict ahead of time which listening differences or situations will be ones in which we can reliably hear differences through sighted listening?

I would probably agree, though I'm a bit unclear. Maybe an example would help.

In this regard I agree with @MattHooper on the basic principle that sighted listening is not entirely disconnected from objective sonic reality - it would be absurd if it were. But I agree with what I take to be the thrust of @Newman 's position in that the kinds of anecdotal accounts Matt raises for the purpose of making what he sees as common-sense points also tend to give an impression that sighted listening is more reliable than it actually is.

JP is right: we are mainly having this disagreement due to Newman's implied position that sighted listening is always entirely unreliable. It's simply the apparent dogmatism (and inaccuracy) that is worth pushing back on.

As to impressions from what I write, I guess "impressions" are ultimately in the mind of the reader. They may or may not be accurate to what I'm arguing. (Newman's impressions, for instance, when he tries to tell others what I believe, bear almost no resemblance to what I write).

I've argued about what conclusions we can justify (not "know" with high certainty, but justify) in regular, casual sighted listening conditions. In making the pragmatic case I believe I've been very careful in my examples no to not overreach, and to incorporate the possibility of cognitive bias.

We have to remember that cognitive biases come in all sorts of flavours. Just as a subjective audiophile can be biased to select observations that support his view, so can an "objectivist" who scorns subjective reviews and sighted listening. That person is just as likely to collect "reviewers getting it wrong" anecdotes to support his scorn and skepticism, and tend to ignore "reviewers getting it right." Which they actually do surprisingly often, at least for loudspeakers.

I happened to be reading earlier an older Fremer review of some Von Schweikert speakers. He wrote: "Pressed to find fault with the spectral balance, I'd say the top end was ever so slightly mellow and the mids a bit forward." And that's exactly what showed up in the measurements! But someone who totally dismisses someone like Fremer is never going to count anything as a "hit." In fact I've seen some here fight tooth and nail not to accept that a subjective reviewer reported a problem that clearly showed up in the measurements.

Likewise, I've described the sonic characteristics in plenty of speakers on this site, whether it's for Mo-Fi speakers, Kii audio, Joseph Audio, MBL and tons of others. There are measurements available for all these and to my memory, nobody has shown that the measurements indicate my impressions were unreliably out of whack. From what I've seen, my descriptions seem generally consonant with the measurements, whether it's frequency response, spatial/dispersion characteristics, comparison to other speakers, or what have you. (Looking at for instance Stereophile's frequency response for the KII speakers, that matches pretty much precisely with the spectral balance that I perceived). That doesn't mean sighted listening is highly reliable. But sighted listening isn't "wholly unreliable." For audible differences sighted listening is only wholly unreliable IF your bar is a scientific level of confidence. But we can - and in many cases require! - justifiably accept lower confidence bars for pragmatic use.

I will always agree that for truly reliable data we need blind testing and measurements. But I will push back on overreach that suggests sighted listening is wholly unreliable. Perhaps the fact I push back on this, and bring examples and arguments, gives the impression I downplay the existence of bias effects. But...I don't believe I do. I try to strike a pragmatic balance in terms of confidence levels.

We know that sighted bias is a thing. Tons of science supports it. It can never be ignored in ANY claim about what we think we hear. But also, one really does have a problem explaining a lot of experience and observations if skepticism is taken too far.

(Sorry for the spam...)

 

[JaMaSt]

So, if you see other writers doing all the above with my posts, just remember that they are disputing audio science….at least, up until the point that they provide audio science to the contrary.

I think you are taking this post way to seriously.

 

[mppix]

I just want to make a general comment to the general reader of this thread.

A few writers are making reference to “Newman’s opinion” and proceeding to mock or ridicule. I wish to emphasise that there is no “Newman’s opinion”. I attempt to confine my statements, on subjects such as audible preferences, to simple reflections of the best audio science on the topic, as far as I am aware. The only “Newman’s opinion” is that audio science holds true.

If I get it wrong and fail to reflect the audio science, then I am very open to that correction: just show me the audio science that is inconsistent with my reflection. Don’t regale me with anecdotal stories that you think have merit, littered with Whataboutisms. Don’t set up a straw man that I said something other than what I said, and proceed in bad faith to mock, argue and discredit that. Such bad faith argumentation usually gets short shrift from me.

So, if you see other writers doing all the above with my posts, just remember that they are disputing audio science….at least, up until the point that they provide audio science to the contrary.

Now…back to something less personal. Here is a bit of audio science, that I think is relevant to vinyl connoisseurs making statements on the sonic attributes of different expensive, or famous, or legendary, or whatever, cartridges, and they are using sighted listening: sighted listening is thoroughly unreliable as a means to determine the sonic attributes or preferences of the sound waves alone due to a device under test, and the only reliable listening-based way to determine such attributes or preferences is via well-controlled listening tests where the listeners don’t even know what are the devices being tested, and ideally involving three or more devices.

If you have an issue with that, then you don’t have an issue with “Newman’s opinion”: you have an issue with audio science. Yet, when I reflect the above by saying that cognitive biases are driving sighted listening impressions on exotic cartridges (just like they do for loudspeakers that sound clearly different under controlled conditions, as demonstrated by researchers), I get pushback. To put it mildly.

I hope the general reader sees a pattern here.

Lol.

 

[JP]

My view is that our ability to do this is very real, and also more limited and less reliable than most of us tend to think. In my view, whether we apply this idea to detailed, very specific differences, or "overall character" is immaterial.

I'm not necessarily disagreeing, but the degree depends on the nature and size of the difference. My ability to call out a 1% level change as a level change is practically non-existent. My ability to call out a 10% level change as a level change is excellent. With controls and rapid switching, while I can identify a 1% level change reliably, my ability to identify it as such is still practically non-existent.

The files are real examples of the typical "high-end" cartridge sound vs. a flat cartridge. Do you think controls are required for most of that population to identify the difference, whether they have the knowledge to identify the rising FR, or they use different words (detail, air, etc.) to describe the same thing.

To me that test illustrates that the reputations of the cartridges going into the test were based precisely on sighted listening impressions of "overall character" - impressions which turned out to be wrong, nonexistent, or overstated when the comparisons were done blind.

There's an age-old feeling amongst a group of people that MC are superior to MM - that MC have some unmeasurable or elusive property that makes them superior. That's what that test was for, which is markedly different than what we're discussing here.

There were no audible differences that could not be accounted for by differences in measured frequency response and tracking ability, and nothing emerged that seemed to be unique to either moving coil or fixed coil cartridge types. Several exhibited energetic resonances at very high frequencies, a feature that made them sound relatively bright and added "definition" and "detail"-occasionally to excess.

 

[pkane]

I just want to make a general comment to the general reader of this thread.

A few writers are making reference to “Newman’s opinion” and proceeding to mock or ridicule.

If you're talking to "the general reader" then perhaps he should be responding to your posts? Ridicule is deserved, since you keep dishing it out yourself. "The general reader" will see that, no doubt. That said, you have to love someone to make fun of them, and we do love you, @Newman! Can't speak for the "general reader", though.

 

[tmtomh]

I'm not necessarily disagreeing, but the degree depends on the nature and size of the difference. My ability to call out a 1% level change as a level change is practically non-existent. My ability to call out a 10% level change as a level change is excellent. With controls and rapid switching, while I can identify a 1% level change reliably, my ability to identify it as such is still practically non-existent.

The files are real examples of the typical "high-end" cartridge sound vs. a flat cartridge. Do you think controls are required for most of that population to identify the difference, whether they have the knowledge to identify the rising FR, or they use different words (detail, air, etc.) to describe the same thing.




There's an age-old feeling amongst a group of people that MC are superior to MM - that MC have some unmeasurable or elusive property that makes them superior. That's what that test was for, which is markedly different than what we're discussing here.

Yes, I agree our ability to accurately detect sonic differences depends on the type and degree of the difference. And I also appreciate and agree with the distinction you make between our ability to detect that something has changed (like that 1% level difference) and our ability to identify precisely what it is that's been changed.

I have no problem believing that controls are unnecessary for most folks to identify that there's a difference between those two cartridges - again with the caveat you mention that people would use different terms, with very different implications and connotations, to describe what they're hearing.

Where I disagree with you is about the Toole cartridge study - it is not at all "markedly different than what we're discussing here." To the contrary, it's part of the very same discussion. I am well aware of the reputation of MC vs MM that was a motivator of that study. Nevertheless, the different technology of each kind of cartridge does not make that study irrelevant for this conversation. The reputation of MC vs MM was shaped precisely by uncontrolled listening impressions that convinced many that MC cartridges sounded different and better than MM ones - the same kinds of impressions that led folks in the same audiophile communities to prefer some individual cartridge models over others. Some of those preferences no doubt were/are based on objectively measurable differences - and others are not.

And the claims made based on those impressions were precisely of the type you mention when you discuss how people are likely to describe the differences between the two cartridges in the examples you've provided - "detail," "air," and other common descriptors.

Your examples and the Toole study are converse instances of the same thing: yours is an example of a case where uncontrolled listening enables us to identify real differences consistently. The Toole study is an example where uncontrolled listening failed in that regard. The fact that uncontrolled listening identified differences that weren't real - as opposed to failing to identify differences that were real - does not impact the relevance of the Toole study to this discussion. It's absolutely relevant.

Taking these two examples together, we can see them as data points (albeit only two data points) in support of the IMHO totally reasonable proposition that our uncontrolled listening impressions are sometimes sufficient and sometimes insufficient or downright misleading when it comes to identifying and discerning sonic characteristics.

I also think we (collectively) are risking underestimating the importance of being able to identify the cause, source, or true nature of the differences we hear. If we hear more "air" and better ambient cues in a speaker when auditioning it with a handful of our favorite tracks, it is important for us to know if that quality is a product of low distortion, linear frequency response, a fortunate interaction (or lack thereof) between the speaker and the auditioning room, a nonlinear "bump" in the mid-treble, or a nonlinear dip in the mid-bass (which can make the treble sound clearer or more prominent). Without knowing any of that, we are unable to know how the speaker might sound in our own listening space, over the long term, with the full range of recordings we might play now and into the future. And that is why I argued above that however much we overestimate our ability to reliably identify sonic differences through uncontrolled listening, we overestimate our ability to identify the cause even more.

I think this is both important and relevant because the point of being able to detect sonic differences or characteristics is not simply to see if we can detect them. The point is to use such impressions to select gear that will make us happiest. So if our uncontrolled listening impressions turn up results that seem clear to us, but we draw mistaken conclusions from those results, that can negatively impact our success in selecting gear for our setups. So measurements are necessary - not just as a check on whether we were "right" in our uncontrolled impressions, but also to help us understand what exactly it was that we were hearing in the cases when we were "right." We can certainly end up with a very good system using purely uncontrolled comparisons, but the claim that we've consistently picked the best gear throughout all those comparisons is based on a misplaced confidence about how much we actually know based on those listening impressions.

 

[JP]

Yes, I agree our ability to accurately detect sonic differences depends on the type and degree of the difference. And I also appreciate and agree with the distinction make between our ability to detect that something has changed (like that 1% level difference) and our ability to identify precisely what it is that's been changed.

I have no problem believing that controls are unnecessary for most folks to identify that there's a difference between those two cartridges- again with the caveat you mention that people would use different terms, with very different implications and connotations, to describe what they're hearing.

Ok, so same page there.

Where I disagree with you is about the Toole cartridge study - it is not at all "markedly different than what we're discussing here." To the contrary, it's part of the very same discussion. I am well aware of the reputation of MC vs MM that was a motivator of that study. Nevertheless, the different technology of each kind of cartridge does not make that study irrelevant for this conversation. The reputation of MC vs MM was shaped precisely by uncontrolled listening impressions that convinced many that MC cartridges sounded different and better than MM ones - the same kinds of impressions that let folks in the same audiophile communities to prefer some individual cartridge models over others. Some of those preferences no doubt were/are based on objectively measurable differences - and others are not.

It broadens the scope of the discussion. I'm not, nor do I believe anyone else is, disputing that bias is a significant part of the equation. AFAIK the only part of this discussion that is in contention is the myopia/dogmatism that a preference cannot be for the signal (FR) if it was formed without controls/that perceptions formed without controls will be driven purely by bias.

I've no disagreement with the rest of what you wrote, it's just beyond the singular point ^ being discussed.

 

[tmtomh]

Ok, so same page there.



It broadens the scope of the discussion. I'm not, nor do I believe anyone else is, disputing that bias is a significant part of the equation. AFAIK the only part of this discussion that is in contention is the myopia/dogmatism that a preference cannot be for the signal (FR) if it was formed without controls/that perceptions formed without controls will be driven purely by bias.

I've no disagreement with the rest of what you wrote, it's just beyond the singular point ^ being discussed.

Okay, in that case we're in agreement, except I would like to clarify that my "can't we all agree" comment that began this exchange between you and me was not directed just at you (and Matt, and anyone else we would identify as broadly agreeing with you). It was also directed at Newman. So I was not attempting to broaden the discussion beyond Newman's particular claims, but rather to try to sum up what I take to be the consensus (though not unanimous) view at this point in the thread.

 

[deweydm]

Reasons for vinyl in this modern world
1. Full size artwork. Anyone see Squaring the Circle (The Story of Hipgnosis)? The art and packaging can’t be replicated in digital.
2. Some works just aren’t available digitally
3. Special releases and box sets (see 1 and 2)
4. Nostalgia
I gave away 90% of my “records” over the past few years, the rest are in climate controlled storage. Most CDs gone as well. What I have is either irreplaceable, was never released digitally or I have sentimental reasons to hold onto. My dad bought me my first stereo in middle school and I still have my first album that I purchased after saving my 50 cent/week allowance - Cosmo’s Factory. Sound quality, fidelity, etc. plays no part for me.

You missed at least one reason. For some of us the most important. Sometimes subjectively better mastering on LP, not readily available from streaming or on CD. Those using LPs as a work around to streaming versions remastered for listening in the car, with inexpensive Bluetooth speakers, etc.. I can’t objectively argue that LPs sometimes sound better than streaming or CD, but subjectively, this is often sadly true too often for some of us.

 

[MattHooper]

I'm very impressed with this summation by an independent artist of the history of vinyl, the current resurgence, and what it's like trying to get a record out on vinyl.
She did her research, speaks really well, and it's very succinct:

My BIG Problem with Vinyl​

Speaking of records, here's a wacky purchase (through discogs) that just showed up on my doorway, circa 1971. Sitar? Plus super early analog synths? Count me in! Looking forward to spinning it and reaching a state of dhyana later tonight

 

[Newman]

His point is that sighted impressions are completely unreliable. If there were practical nuance I suspect this argument wouldn’t exist.

Of course there is practical nuance.

…So my argument would be that the results of a sighted test are much more difficult to predict and generalize to all sighted tests than say a blind test where the psychological variables (product identity) are removed. We can repeat blind tests over and over with different listeners and get pretty consistent results. Try doing the same with sighted tests with different listeners and I am sure that this will not be the case.
In ranking objective/subjective measurements in terms of how reliable and trustworthy test they are I would say:
#1 A well-controlled double-blind listening test.
#2 Meaningful Objective Measurements that Predict #1
#3 A sighted listening test.

BUT,

There is remarkably little variation from person to person, including differences in experience, culture, country of origin, etc. The critical factor is that the listening tests must be blind or double-blind - sighted tests cannot be trusted.
…

If I had written ‘cannot’, I would stand condemned, right?

Okay, I’ll say it: sighted listening tests cannot be trusted.

Go on, shoot me.

 

[tmtomh]

Perhaps where we're collectively getting stuck or going in circles here is that two propositions that might seem contradictory, aren't.

Specifically, sighted listening impressions do not have to always be wrong in order for Toole (and Newman) to be correct that "Sighted tests cannot be trusted."

To my eyes, the real disagreement is about how often and under what circumstances sighted comparisons can be considered reliable. If someone forced me to choose between, roughly speaking, "the Newman position" and "the other position," I'd have to go with Newman, because (a) I don't believe he's actually claiming that sighted impressions are never correct, (b) I feel that we all tend to overestimate our ability to correctly and reliably discern sonic characteristics through sighted listening comparisons, and (c) pointing out individual scenarios when known differences between gear are large and obvious enough to be easily detected in sighted comparisons does not tell us anything about whether sighted comparisons are on the whole correct more often than random guesses would be, especially when we go into such comparisons with no prior knowledge of whether or not the equipment differs sonically or what type and degree of differences there are.

 

[Axo1989]

Of course there is practical nuance.

BUT,

If I had written ‘cannot’, I would stand condemned, right?

Okay, I’ll say it: sighted listening tests cannot be trusted.

Go on, shoot me.

Rubber bullets, blanks, or whatever @Sal1950 wants to use for his next hangout with Darko?

These posts of yours are rather like people who argue with snippets from scripture. I mean Toole also wrote "two ears and a brain are much more analytical than a microphone and a meter". Or is context meaningful when cherry picking sentences?

 

[MattHooper]

Sean Olive: In ranking objective/subjective measurements in terms of how reliable and trustworthy test they are I would say:
#1 A well-controlled double-blind listening test.
#2 Meaningful Objective Measurements that Predict #1
#3 A sighted listening test.

^^^ Yup! A solid inference! Wouldn't argue against that!

Okay, I’ll say it: sighted listening tests cannot be trusted.

Sure, if you mean by "trusted" one is looking for a scientific level of confidence.

In the same way "sighted cooking can't be trusted" if you are looking for a scientific level of confidence. E.g. adding more salt MAY alter the taste of your dish in the way you expect, or it could be an expectation bias that causes you to think you've altered the taste. Scientific controls would weed out bias and get more reliable results. But for every day use, so long as we aren't practising something scientifically untenable, like "homeopathic cooking," we can make reasonable inferences as we test and create recipes in informal conditions. Likewise, in many cases, for sighted listening in audio, for audiophiles.

Specifically, sighted listening impressions do not have to always be wrong in order for Toole (and Newman) to be correct that "Sighted tests cannot be trusted."

To my eyes, the real disagreement is about how often and under what circumstances sighted comparisons can be considered reliable. If someone forced me to choose between, roughly speaking, "the Newman position" and "the other position," I'd have to go with Newman, because (a) I don't believe he's actually claiming that sighted impressions are never correct, (b) I feel that we all tend to overestimate our ability to correctly and reliably discern sonic characteristics through sighted listening comparisons, and (c) pointing out individual scenarios when known differences between gear are large and obvious enough to be easily detected in sighted comparisons does not tell us anything about whether sighted comparisons are on the whole correct more often than random guesses would be, especially when we go into such comparisons with no prior knowledge of whether or not the equipment differs sonically or what type and degree of differences there are.

Fair enough!

(Though I'd disagree somewhat about Newman: his default clearly is to attribute sighted listening to bias effects, often telling people: you aren't evaluating the sound waves in sighted listening. This kind of stuff is used to poo-poo all subjective reviewers, any subjective language or reports someone doesn't care for, etc. I wouldn't disagree with anyone saying "I don't put any confidence in sighted subjective impressions or subjective reviewers; that doesn't give me the level of reliable knowledge I'm looking for. I look instead to measurements and/or controlled listening." It's only when it is implicit, or explicit, that sighted listening and subjectively described impressions can be summarily dismissed as bullshit, that things tip too far IMO).