How do you judge your system's neutrality?

When I began this thread, I advocated a strict kind of Objectivism regarding coloration and neutrality. That is to say, I argued that coloration and neutrality are characteristics that are INDPENDENT OF PERSONS. In my second post on 12/12, I relaxed my position somewhat, by defining colorations as “audible inaccuracies.” If colorations are “audible,” I conceded, then their existence is DEPENDENT UPON PERSONS. In that sense, I believe that…

(1) Colorations are SUBJECTIVE characteristics.

To say that colorations are subjective characteristics is to say that their existence is DEPENDENT UPON PERSONS. For example, one person may hear distortion when another person does not, which seems true enough. In cases where two people disagree about the presence of a coloration, a Subjectivist might conclude either…

(a) There is no fact of the matter about whether the coloration in question exists.
…or…
(b) The coloration in question exists for one person but not for the other.

I believe that (b) is correct. In other words, in cases of intersubjective disagreement, a coloration exists for one person but not for another. Put another way, the existence of a coloration is, to some extent, IN THE EAR OF THE BEHOLDER. I say “to some extent” because, to a complementary extent, colorations are also IN THE WORLD, as I will now try to show…

Recall that I defined colorations as “audible inaccuracies.” By defining colorations as “audible,” I have acknowledged that they are SUBJECTIVE. But by defining colorations as “inaccuracies,” I have subsumed colorations under a larger category of characteristics that are OBJECTIVE, namely, inaccuracies. In other words, I believe that…

(2) Inaccuracies are OBJECTIVE characteristics.

In a previous post, I defined inaccuracies as: Alterations to information in the playback system that conceal, corrupt, or eliminate information about the music. To say that inaccuracies are objective is to say that their existence is INDEPENDENT OF PERSONS. For example, jitter may exist in a playback system even if no one can hear it. The existence of jitter is therefore independent of persons, or objective. Likewise for other inaccuracies.

Taking (1) and (2) together, we get:

(1) Colorations are SUBJECTIVE characteristics.
(2) Inaccuracies are OBJECTIVE characteristics.

This raises the following question: What is the relation between colorations, as subjective characteristics, and inaccuracies, as objective characteristics? The answer, I believe, is that…

(3) Colorations CORRELATE with inaccuracies for expert listeners.

In my view, as a person develops expert perception with respect to the playback of recorded music, inaccuracies become more audible. Put another way, as listener expertise increases, the perceptibility of colorations increases. For this reason, the expert listener will hear far more colorations than the naïve listener.

In light of this, colorations are clues that help the audiophile understand his system. By correlating with inaccuracies, colorations enable the audiophile to use his ears to identify how information about the music has been concealed, corrupted, or eliminated by his system. And colorations empower the audiophile to make informed changes to his system when the inaccuracies those colorations reveal are inconsistent with his preferences.

Hi Bryon - I have been out of town again for several days, and have just read your last two posts. To respond to the 12/31 one first: evidence for too many audiophiles losing the forest for the trees is all over any audio site. I would estimate that at least one person a week posts in each one about how he drove himself crazy and is not enjoying the hobby or the music anymore, or some such story. We will have to agree to disagree about the applications of analysis to music performance vs. music playback. As you say, both result in music, and the bottom line is whether the music is enjoyable or not. To use your phrases, too many listeners to an audio system have their attention split between the music and the equipment, resulting in impaired functionality. They will completely ignore many recordings, even entire recording labels, because "they don't sound good on my system." This, to me, is a crying shame; the definition of misplaced priorities, the system becoming more important than the music. Many threads on this forum and others have discussed such issues at length. I agree completely that both the art and the science are important, and both have their subjective and objective aspects - it is a question of prioritizing all of this, which will vary with each individual, and there are a great many audiophiles who complain that they struggle with how to do this. My contention earlier in the thread was that if more audiophiles spent some time learning a little music theory and taking an aural skills class, that this will be much more beneficial to their enjoyment of their music in the short term, and for their ability to hear how better to tweak their systems in the long run as they develop these abilities. The one must be done to truly be able to do the other, as one must decide not just whether or not a coloration is there, but how damaging to the music it actually is - preferences determining priorities (the classic example would be the analog/digital debate).

This brings me to your post of 1/5. Not sure I agree with everything you say about your numbers 1), 2), and 3), but granting them for the moment, the real issue I have is with the "expert listener" concept. The fact that everyone hears differently has been much discussed already in this thread, I will only point out that this certainly includes "expert listeners." Mrtennis has made some other very good points about human hearing in this thread. Too many audiophiles are ONLY concerned about learning to listen for flaws in their systems, and this is as far as their ear training ever goes. I would never call an audiophile of this variety an "expert listener," no matter how many years experience in the hobby they have. I have talked with people who cannot identify a major from a minor chord, yet claim to hear very specific "colorations" in a speaker when in fact they are merely biased against it's design based on things they have read/been told, and probably could not tell one speaker from another if their back was turned and they were only relying on their ears, to use an extreme example. I have often read a review of one of my orchestra's concerts in the paper the next day and wondered if the reviewer was at the same concert I was. Same with the reviewer of a piece of audio equipment. And just because one is an audiophile does not mean that one has better ears than someone who is not. On this forum there is usually a new thread every couple of weeks, it seems, where some guy is posting about how his wife heard something better than he did, even though she knows nothing about the hobby, helping him make up his mind. Orson Welles' final film, F for Fake, is a hilarious send-up of the idea of "expertise," by the way. I think you would greatly enjoy it, though as an objectivist you may find it very disturbing. :)

… evidence for too many audiophiles losing the forest for the trees is all over any audio site. I would estimate that at least one person a week posts in each one about how he drove himself crazy and is not enjoying the hobby or the music anymore, or some such story.

Learsfool - I agree with you that audiophiles commonly lose the forest for the trees. What I was challenging was the idea, stated in your post on 12/31, that…

too many audiophiles lose the forest for the trees by getting bogged down in trying to eliminate various types of colorations, etc.

This makes it sound as though audiophiles are more likely to succumb to this loss of perspective if their efforts focus on the reduction of colorations. It is that belief that I was challenging. As I see it, audiophiles can lose perspective focusing on a vast array of concerns. To me, the loss of perspective says more about the psychology of the audiophile than it does the specifics of their preoccupations.

…too many listeners to an audio system have their attention split between the music and the equipment, resulting in impaired functionality. They will completely ignore many recordings, even entire recording labels, because "they don't sound good on my system." This, to me, is a crying shame; the definition of misplaced priorities, the system becoming more important than the music.

I completely agree with this. Having said that, the quality of recordings is outside the audiophile’s control. The quality of his playback system is not. Therefore, efforts to improve the playback system are rational, provided that they increase his enjoyment while not resulting in the loss of perspective you mentioned. In contrast, audiophiles who regularly avoid inferior recordings are showing symptoms of audio nervosa, and should seek medical attention.

My contention earlier in the thread was that if more audiophiles spent some time learning a little music theory and taking an aural skills class, that this will be much more beneficial to their enjoyment of their music in the short term, and for their ability to hear how better to tweak their systems in the long run as they develop these abilities.

Agreed. In my view, learning music theory is a way of increasing listening expertise, which brings me to…

…the real issue I have is with the "expert listener" concept. The fact that everyone hears differently has been much discussed already in this thread, I will only point out that this certainly includes "expert listeners."

This is the issue of how much variability in perception exists among expert listeners. My view, expressed in a previous post, is that once PREFERENCE is differentiated from PERCEPTION, the amount of variability among expert listeners’ PERCEPTION is lower than has been estimated by the Subjectivists on this thread. But this is a matter of speculation for us both.

You go on to raise other objections to the concept of an 'expert listener':

Too many audiophiles are ONLY concerned about learning to listen for flaws in their systems, and this is as far as their ear training ever goes. I would never call an audiophile of this variety an "expert listener," no matter how many years experience in the hobby they have.

Being an expert or a naïve listener is not a binary state. There are DEGREES of listening expertise, and most likely KINDS of listening expertise. The phrase “expert listener” is really just a shorthand way of saying “a person with some degree/kind of listening expertise.” In light of that, the hypothetical audiophile you mention in the passage above does have a certain KIND of expertise. Your reluctance to refer to him as an “expert listener,” in light of his failure to develop other kinds of listening expertise, seems reasonable to me. It does NOT, however, cast doubt of the validity of the concept of 'listening expertise' more generally. Moving on to your next objection:

I have talked with people who cannot identify a major from a minor chord, yet claim to hear very specific "colorations" in a speaker when in fact they are merely biased against it's design based on things they have read/been told.

These are pseudo-experts. In any field where there are experts, there are pseudo-experts. In some fields it is easier for pseudo-experts to avoid detection, but genuine experts can often tell the difference. More importantly, the existence of pseudo-expert listeners does not cast doubt on the reality of genuinely expert listeners, any more than the existence of pseudo-expert doctors (we all know them) casts doubt on the reality of genuinely expert doctors.

…just because one is an audiophile does not mean that one has better ears than someone who is not. On this forum there is usually a new thread every couple of weeks, it seems, where some guy is posting about how his wife heard something better than he did, even though she knows nothing about the hobby…

This phenomenon is real. I have experienced a version of it myself. But it does not impugn the concept of ‘listening expertise,’ because, while your wife knows THAT she heard something new or different, better or worse, she usually cannot tell you WHAT she heard, WHY she heard it, or what you should DO about it. That is to say, your wife, as a naïve listener, may have very acute hearing (particularly because she has not abused it as much), but her acute hearing does not make her an expert listener, since she lacks the knowledge and understanding that listening expertise entails. If you need a demonstration of this, the next time your wife mentions that things sound shrill in your system, ask her whether she thinks it is jitter in the source, resonance in the tweeter, or flutter echo in the room. She will remind you, in no uncertain terms, that she is not an expert listener.

Orson Welles' final film, F for Fake, is a hilarious send-up of the idea of "expertise," by the way. I think you would greatly enjoy it, though as an objectivist you may find it very disturbing.

I think questioning the whole idea of ‘expertise’ has it place, since there are certainly cases where expertise is unfounded, unquestioned, exaggerated, or fabricated. But this does not motivate the abandonment of the concept. That is throwing the baby out with the bathwater. Expertise is real, though it is also variable, fallible, and counterfeitable. I suspect that nearly everyone believes in expertise, whether they claim to or not. The Onion had an amusing article recently, satirizing the skeptics of expertise:

NEW YORK—Inside the Montessori School of Dentistry, you won't find any old-fashioned cotton swabs, or so-called periodontal charts, or even any amalgam fillings. That's because at this alternative-learning institution, students are being encouraged to break away from medical tradition and discover their very own root canal procedures.

"At Montessori, we believe dentistry is more than just the medical practice of treating tooth and gum disorders," school director Dr. Howard Bundt told reporters Tuesday. "It's about fostering creativity. It's about promoting self-expression and individuality. It's about looking at a decayed and rotten nerve pulp and drawing your own unique conclusions."

"When performing a root canal, there's no such thing as right or wrong," said Montessori educator Vanessa Perrin, who added that she doesn't so much teach her students how to treat an inflamed nerve, as lead them to an open mouth and then stand back. "Sure, we could say to our students, 'The enamel here has completely eroded and needs to be addressed immediately.' But what's more satisfying, what's more dynamic, is to just let them slowly develop an 'impression' of why a patient might be screaming."

I would encourage skeptics of the concept of 'expertise' to visit the Montessori School of Dentistry for their next root canal.

After ten weeks or so, this thread has slowed to a halt. In light of that, I would like to thank all those who participated. I, for one, learned a lot from our discussions, and I had a lot of fun too.

I have one final set of ideas that I would like to contribute. They are largely ecumenical in spirit, in that they are an attempt to reconcile the major ideological division on this thread, namely Subjectivism and Objectivism. In one way or another, that difference in attitude has affected nearly every argument we have had, whether it was about coloration, neutrality, accuracy, transparency, or whatever.

What I would like to suggest is that both Subjectivism and Objectivism are valid points of view, but that sometimes one is more warranted than the other. Which is more warranted depends on (1) the kinds of recordings an audiophile tends to listen to, and (2) the audiophile’s priorities. To make this point, I have to say some preliminary things about representations and truth. To begin with…

A FEW WORDS ABOUT REPRESENTATIONS:

(1) A representation contains information about some state or event.

To elaborate…

A “synchronic” representation contains information about a STATE, that is, a MOMENT in time. A “diachronic” representation contains information about an EVENT, that is, a PERIOD of time.

Some, but not all, representations are IMAGES of the states/events they represent. That is to say, the relation they bear to the states/events they represent is that of RESEMBLANCE.

Most, but not all, representations are INCOMPLETE. That is to say, they do not contain ALL possible information about the state/event they represent.

Putting these things together…

(2) Music recordings are diachronic, imagistic, incomplete representations that contain information about musical events.

Here is where things get important…

Some representations contain information about REAL events, that is, events that ACTUALLY OCCURRED. Think: A photograph of a landscape that exists.

Some representations contain information about VIRTUAL events, that is, events that DID NOT ACTUALLY OCCUR. Think: A painting of a landscape that is a composite of various landscapes drawn from the painter’s memory.

This bears on music recordings, in that…

(3) Some music recordings contain information about REAL events, while other recordings contain information about VIRTUAL events.

To elaborate…

Music recordings can be thought of on a continuum according to how REAL or VIRTUAL the event is that the recording represents.

At one end of the continuum is a music recording that represents a musical event that is MAXIMALLY REAL: It was a live performance of acoustic instruments recorded in a real acoustical space (i.e. not a studio) with a single stereo or dual mono microphones, it was not edited, and it was minimally altered during mixing. Some audiophile music recordings approach this standard.

At the other end of the continuum is a music recording that represents a musical event that is MAXIMALLY VIRTUAL: It contains no real instruments, it is highly edited, and altered liberally during mixing. Some electronic music approaches this standard.

In the middle of the continuum is where the vast majority of music lies: It was recorded with multiple microphones, sometimes in different spaces or different times, often edited together from multiple performances, and mixed with the use of level adjustment, channel placement, equalization, filtering, reverb, and other “effects.” In the case of popular music, many of the sounds were not sourced from real instruments at all. The more these techniques are used, the more we describe an album as “produced." The difference between recordings that represent real events and those that represent virtual events is crucial to whether a recording can be evaluated as to its TRUTHFULNESS. Which brings me to…

A FEW WORDS ABOUT TRUTH:

(4) The truth of a representation is its objective correspondence to reality.

The meaning behind the word “objective” here is that the truth or falsity of a representation depends only upon its correspondence to how things actually are. It does not depend upon OUR BELIEFS about how things actually are. A representation is TRUE when the information it contains about an event CORRESPONDS to how that event actually is (or was). A representation is FALSE when the information it contains about an event FAILS TO CORRESPOND to how the event actually is (or was).

But what if a representation contains information about a VIRTUAL event, rather than a REAL event? Then the question of its truthfulness, in the sense above, DOES NOT APPLY. That is to say, while representations about REAL events can be judged as to their truthfulness, representations about VIRTUAL events CANNOT.

This brings up back to music recordings…

(5a) The more a music recording represents a REAL musical event, the MORE it can be judged as to its truthfulness.

(5b) The more a music recording represents a VIRTUAL musical event, the LESS it can be judged as to its truthfulness.

“Truthfulness” here refers to TRANSPARENCY TO THE MUSICAL EVENT, that is, how much the information presented at the ear during playback resembles the information that was presented at the microphone during the actual performance. But “truthfulness” could also be thought of as ACCURACY TO THE SOFTWARE, that is, how much the information about the music is preserved as it passes from software to ear. (I am indebted to Almarg for this important distinction.) However, it is transparency to the musical event, not accuracy to the software, that “truthfulness” refers to in (5) above.

These observations were a preliminary to my proposal for…

RECONCILING OBJECTIVISM AND SUBJECTIVISM:

OBJECTIVISM regarding representations is the idea that representations can be evaluated as to their truthfulness. It is a justifiable point of view when representations contain information about REAL events.

SUBJECTIVISM regarding representations is the idea that representations cannot be evaluated as to their truthfulness. It is a justifiable point of view when representations contain information about VIRTUAL events.

In terms of music recordings…

(6a) Music recordings of REAL events can be evaluated as to their truthfulness. And to evaluate a recording’s truthfulness is to adopt the point of view of Objectivism.

(6b) Music recordings of VIRTUAL events cannot be evaluated as to their truthfulness, though they can be evaluated in terms of preference. And to evaluate a recording in terms of preference is to adopt the point of view of Subjectivism.

In other words, I believe that Objectivism and Subjectivism are both valid, but in different situations. The more REAL the musical event that a recording represents, the more OBJECTIVISM is warranted. The more VIRTUAL the musical event that a recording represents, the more SUBJECTIVISM is warranted. Hence, the appropriateness of one point of view or the other is largely a consequence of what kinds of recordings an audiophile tends to listen to.

It is worth pointing out that an audiophile who tends to listen to recordings of real events is still entitled to be a Subjectivist, and hence evaluate recordings solely in terms of preference. My point is not to tell anyone how they should evaluate the playback of recorded music. My point that Objectivism makes less sense as recordings become more virtual, and more sense as they become more real. Whether or not an audiophile who tends to listen to recordings of real events choses to be an Objectivist or a Subjectivist is largely a matter of his priorities, which brings me to a recent TAS web post by Jonathan Valin:

There are, IMO, three types of listeners in the high end, although these types tend to overlap. First, those who, first and foremost, want recorded music to sound as much like the real thing as possible--I call them the "absolute sound" type. Second, those who, first and foremost, want their recordings to sound exactly as good or as bad as the engineering and mastering allow them to sound (and want to hear the engineering and mastering, to boot)--I call them the "faithful to mastertapes/mike feed" type. And third those who are, primarily, less concerned with the absolute sound or the sound of mastertapes and more interested in hearing their recordings sound as beautiful and moving as possible--I call them the "as you like it" type.

The first kind of audiophile prioritizes TRANSPARENCY to the musical event. The second prioritizes ACCURACY to the software. And the third prioritizes MUSICALITY as he defines it. For those who prioritize transparency or accuracy, Objectivism may seem to be the more valid point of view, but only to the extent that recordings represent real events. For those who prioritize musicality, Subjectivism may seem to be the more valid point of view, regardless of whether recordings represent real or virtual events.

The above is my swan song - my final effort to reconcile the validity of Subjectivism with that of Objectivism. No doubt some will feel I have failed. Others may think that I have given Subjectivism a less important role than Objectivism in music playback. That may be true. Since I have been an outspoken advocate of Objectivism, it has no doubt contaminated my effort at reconciliation. But my attempt has been in earnest. In any case, other than reserving the right to clarify or defend, this is me signing off…

Bryon

Interesting. Did you take the blue pill or the red one?