Audio McCarthyism Letters

Letters in response appeared in May 1992 (Vol.15 No.5):

Credible causes & cables
Editor: Point: "As We See It," January 1992.

Take a good objective look at some of the high-end audiophile claims, explanations of some claims (if any are given), and prices charged for some high-end products. Add in subjective reviewing as a whole. Total everything up. Are you really surprised that people like Dan Dugan can find followers and believers for his "crusade"?

I'm not surprised at all.

I wonder: Who appears to have a more credible case to the lawmakers, Dan Dugan or the high-end audio industry? Something to think about, before it is too late.—Thomas L. Hauk, San Antonio, TX

Cables & the audio press
Editor: Robert Harley's report on "Audio McCarthyism" at the Audio Engineering Society's 1991 meeting left me with mixed feelings. First, I'm not an objectivist regarding audio equipment evaluation. I reject the idea that an engineer (who hasn't listened himself and who might have tin ears in any event) can tell me I'm not hearing something that I am in fact hearing. However, on the issue of audio cables, I have to admit a certain sympathy for the views of Mr. Lopez, the New York consumer cop.

The audio press simply hasn't done its job in informing the public about cables; as for the high-end industry, it seems to have an actual vested interest in spreading confusion and mysticism here.

Years ago, manufacturers assumed a responsibility to inform consumers as to the general characteristics required for interconnect cables. For example, the instruction book for the McIntosh C-11 tube preamp, ca 1962, states as follows: "The capacitive reactance of connecting audio cables should not be less than 8,000 ohms at 20,000 cycles [hertz]. This is the reactance of a capacity of 1,000 mmf [pF]. Audio cable having a capacity of 25 mmf [pF] per foot may be 40 feet long; 13.5 mmf [pF] per foot cable may be 75 feet long." This was written at a time when no audio retailer made any serious money selling interconnects.

By comparison, the manual for my Audio Research SP-10 preamp, while full of detail on most subjects, states only this regarding cables: "At the performance level of the SP-10 most audio interconnect cables will indeed degrade the sound quality. Only the very best available are really suitable for this application. Be sure to discuss this with your authorized Audio Research dealer."

My reaction to this is, what's the big secret, Audio Research? Surely Audio Research employs engineers who are able to provide minimum cable specifications. I'm not asserting that any cable meeting these minimum specs will be adequate, simply that the consumer doesn't know where to begin when trying out cables without knowing the engineers' intended minimum requirements. Without this information, a consumer may well choose a very expensive cable that pleases him only because it provides a desired alteration of frequency response, a result that could have been obtained much more cheaply. The high-end industry's argument that cable quality is more than a matter of inductance, resistance, and capacitance would be more impressive if the industry actually supplied these basic specifications to the consumer so apples could be compared with apples, etc.

One striking feature about today's high-end audio retailing is the great effort retailers undertake to sell extraordinarily high-priced cables and interconnects. The importance of these is stressed emphatically by salesmen who have only the dimmest idea of what "reactance" might be. Since these megabuck interconnects are, with all due regard for their elaborate assembly and finish, still basically just wire, you don't have to have an MBA to realize that the markup is considerable. In fact, it is obvious that high-end retailers benefit from the same sales strategy that automobile retailers traditionally relied on—that is, once bargaining has been completed for the basic item the customer came in for, and his sales resistance has been conquered, sell him some optional "accessories" with a much higher markup. There's nothing illegal or grossly unethical about this sales practice, but a consumer-oriented press ought to inform consumers to be wary of it. In the case of cables and interconnects, it generally has not done so.

For example, many highly touted interconnects feature beautifully finished, precious-metal RCA plugs. The general audio public is probably not aware that the RCA plug is inherently a bad design. It would often be cheaper, and much better technically speaking, for a consumer to hire a technician to hard-wire his cables in place, or to fit the gear for professional XLR or other connectors, than to purchase these "Rolex" interconnects. (The argument that such a modification would void the manufacturer's warranty, other than for the actual parts replaced or modified, is ludicrous, legally speaking.) And it certainly would be far cheaper for manufacturers of consumer audio gear to provide screw terminals (as professional gear does) as an alternative to RCA jacks. But, of course, screw terminals would deprive the retailer of the ability to sell RCA plug "jewelry" to the public. I am not so puritanical as to deny to any audiophile the pleasure of owning these handsome golden items; but I suspect that many buyers have thought they were buying them for sound, not looks. Where is the voice of the supposedly consumer-oriented audio press on this issue? Can one be forgiven for suspecting that your silence is related to your apparently considerable advertising revenue from interconnect producers, judging from their numerous glossy (and totally uninformative) ads in each month's Stereophile?

Before waving the bloody shirt of McCarthyism on the interconnect issue, the press and industry need to clean up their act.—John J. McFadden, Bryn Mawr, PA

The answer to Mr. McFadden's question is: no connection. And no silence either, as far as I am aware.—John Atkinson

Cables & the government
Editor: Ironically, the major threat to progress turns out to be an institution which purports to encourage progress—government. I recommend that all lovers of recorded music read (or re-read) Robert Harley's excellent verbal re-creation of the 1991 Audio Engineering Society Convention's Loudspeaker Cable Workshop in the January 1992 issue and then take some action to prevent the likes of Dan Dugan from using government to promote their special interests.

It is well-known to those who study public choice that government responds to special-interest groups (typically a small minority of the population) because they are organized and vocal, to the detriment of the unorganized silent majority. As a result, many decisions made by government are not optimal. That is, they benefit special-interest groups at the expense of the majority. Comments made at the Workshop by Wilfredo Lopez of the New York Department of Consumer Affairs reveal the impact that an organization such as the AES can have on government policy, to the detriment of progress in the audio industry.

The irony of it all is that an action taken by a government agency—for example, a ruling that if it can't be measured, it can't be heard and therefore it can't be claimed—purportedly would protect consumers from the unscrupulous practices of manufacturers and retailers. The fact of the matter is that there are differences between audio components, including cables, that can be measured by equipment, and there are differences which equipment (other than the human ear) cannot measure. This obviously makes many engineers uncomfortable. As a result, they pressure government, via an association such as the AES, into legislating toward their own interests. If manufacturers and retailers can be prevented from making claims that only trained ears can detect, then engineers have the last word on what equipment sounds best. This obviously yields economic and other advantages to engineers. But it is detrimental to progress in the audio industry because it limits claims of improvements in audio hardware and software to those that can be measured by electronic equipment. The fact that the ear is in many ways a superior piece of equipment is ignored. Progress is thus constrained and the majority ultimately bear the cost.

It is important that people recognize that government is all too often the source of problems. In its attempt to solve what for a few is a problem (engineers cannot measure some things which the ear can detect), it creates new problems (retarding progress to the detriment of consumers of recorded music; ie, most of us). We must be constantly wary of those who purport to use government to promote the well-being of those other than themselves.—Michael T. Saliba, New Orleans, LA

Two old ladies?
Editor: Now that the objectivists are trying to make you guys illegal (AES report, January 1992), I guess war is declared. The best way to win this one is to pull out.

When you devote so much space to proving that you are right and they are wrong, you sound uncertain of your position. It's interesting and informative to once in a while hear the subjectivist position explained, but when your tone gets angry or combative it brings to mind two ladies swinging purses at each other. Let the other guys be petty. Plow ahead with the truth, and your position will be strengthened.—Doug Scharf, Granada Hills, CA

The eternal subject
Editor: Now a comment on double-blind testing, discussed by Robert Harley in two excellent articles in the January issue. Double-blind testing is, of course, not new to the pharmaceutical industry. In most clinical drug-testing programs a control group is dosed with a placebo, a harmless coated sugar pill, in order to account for psychosomatic effects. The identity of the placebo group or the drug group is not known to either the clinical practitioner or the patient until the code is broken.

There seems to be a difference in the way double-blind testing is done on audio equipment and on pharmaceuticals. With drugs, a patient receives either the placebo or the active agent, never both. In audio, I have the impression that if, say, $10/foot speaker cable is being compared to lamp cord, that the source is alternately switched through one and then the other. All of the testers then try to detect a difference or state a preference.

This switching introduces a distraction that could cloud the results, particularly if the test involves musical enjoyment. A better way might be to split the testers into two groups and allow one group to audition only the premium cable and the other only the lamp cord (neither group knowing to which they are listening). The detailed subjective sonic impressions of the two groups can then be compared to detect the relative quality of the two cables.—Leonard M. Weinstock, Hilton Head Island, SC

Objectivity & References
Editor: The question of whether there can be objectivity in listening to and evaluating hi-fi components and whether such objectivity is even desirable is an interesting topic. I would venture to say that such objectivity is really just an "idea" when it comes to the actual listening experience.

The part that memory plays in listening to a musical event is at the crux of this matter. Each of us has stored in the brain and organism (it is now known that even individual cells have a kind of memory) all of our past experiences, including those of a musical nature; just how this occurs is of great interest to people of many scientific fields, but that it does occur appears incontrovertible. I would not be able to recognize the sound of a guitar, for instance, if I had never before heard what a guitar—any guitar—sounded like. And those of us somewhat familiar with the sounds of live musical instruments have probably heard many different kinds of particular instruments, so that many differing sounds might all be recognized as guitar sounds, though the guitars themselves would each have unique sonic characteristics. Of course, what I am saying could be taken to extremes, and we would begin to get an idea of how complex the system of memory and sensation is.

Let's say that I have never heard the sound of live music, other than heavy metal music played in large auditoriums at extremely loud levels. If I go looking for a hi-fi system, taking along copies of Megadeath, or whatever, then I will most likely find most "accurate" the system that most closely reproduces the live concert sound I am used to. How an unamplified acoustic guitar might sound over such a system—well, I don't know. But can one deny the point of reference which such a person brings to his or her musical experience, and say that such a system is "inaccurate"? It may very accurately reproduce the sound of distorted electronic instruments at 100dB, as this relates to the live concert event.

So we need to know where we are coming from when talking about this stuff, and this will be based on our points of reference which are a part of memory. Our experiences differ in both large and subtle ways. Many people might agree upon the type of instruments being heard over a hi-fi system (again, this points to some shared past experience with that particular sound), but each person's experience would have most likely been somewhat different, and so judgments made on more subtle levels may begin to show up those differing experiences. Where, please tell me, is the objectivity in all of this? And is objectivity even to be of concern? Or would it be more helpful, truthful, and accurate to become familiar with our prejudices, thereby gaining a better understanding of where we are coming from? This is revealed when we begin to share our personal tastes, preferences, musical experiences, etc.

What is wrong with this approach? Music is an experience involving the senses, with some aspects of memory actively engaged, and has emotional outcomes as well; not only is it not possible to become 'scopes and metered instruments, but why would we want to?

Another approach which may simplify the whole thing is to find components through which one truly enjoys listening to one's favorite recordings, and leave the technobabble, picky critiques, endless evaluations, and determinations of accuracy to those paid to do this sort of thing! There is a Wife Acceptance Factor; what about a Sheer Enjoyment Factor?—Stew Glick, Springwater, NY

The unscientific "objectivist"
Editor: While Robert Harley's "The Listeners' Manifesto" (Vol.15 No.1, January 1992, p.111) has added valuable food for thought, he has also added unnecessary complication and confusion to the issue by connecting the "objectivist" approach to supposedly flawed assumptions of Western science. Instead, the flaws in the so-called objectivist approach are actually quite simplistic, having nothing to do with the limitations of science. The objectivist approach fails on its own terms, because it is not scientific at all. The alternative to the "objectivist" approach should not be called a "subjectivist" approach, because, as Mr. Harley correctly points out, the alternative should combine objective measurement with subjective listening. Unlike the objectivist position, this alternative can reflect a correct understanding of scientific methodology.

The objectivist's position (or positions) actually rests on several unverified assumptions. One of these assumptions is that presently existing instrumental testing methods using a) presently existing test equipment and b) the presently used battery of tests, can fully determine which audio components sound best and which sound alike. The scientific approach is to accept this position as a hypothesis to be tested. To the extent any of the so-called objectivists accept this position a priori, it is they who are being unscientific. But their failure to use scientific methodology goes beyond this.

Listening tests are the only way to test the objectivist's hypothesis, for only by listening can one tell which components sound alike and which do not. Many objectivists realize that listening tests are required, so they rely upon certain limited types of "blind" listening tests. These testing methods are questionable not because they are "blind," but because they are too limited. In addition, they raise yet another question: Who determines what sounds better in listening tests? The majority? The statistical average of a randomly selected group? The experienced reviewer? The novice? The objectivists typically add another a priori assumption: If most people cannot hear it, or if I cannot hear it, no one else can. This assumption is not derived from objective evidence; on the contrary, it is very unscientific and even absurd.

In fact, we already know, and can determine as objectively as it is possible to determine, that the ability to hear varies from one person to another. Let's take the simple example of ear tests performed in the office of a typical qualified specialist, whom I'll call the "ear doctor." The ear doctor knows that some people can hear a 20kHz sinewave, and some cannot. (I know that I used to, but cannot now.) Similarly, some people can hear a particular level of sound at 1kHz, some cannot. Note that the ear doctor can only determine whether a patient can hear a particular sound through a listening test and by asking the patient what he or she can hear. No qualified ear doctor will tell a patient, "Everyone else can hear this, so you must be hearing it regardless of what you say." Conversely, no such doctor could reasonably tell a patient, "You did not hear that. No one else I've tested has."

Obviously, the ear doctor's reliance on listening tests does not make that approach unscientific. The listening test just happens to be a necessary part of the inquiry.

The point here is not that an ear doctor can determine who can best test audio equipment, but that to test a person's hearing, we have to conduct a listening test and ask the person what can and cannot be heard. In this respect, we cannot do better than the ear doctor. Just as an ear doctor would be committing malpractice if he or she told you what you hear based on what others have heard, the "objectivist" cannot say that "no audible differences exist" if anyone claims to hear them.

Let's look at this from another angle. Suppose a wind quintet plays before a small audience, and only one person in the audience is an experienced musician. If the oboe is slightly out of tune, it is not unlikely that only the musician would know it. Some of the audience may vaguely sense that something is not quite right, but most of them would probably not know that the oboe was slightly out of tune, and many might hear nothing wrong at all. If the same piece were repeated with the oboe in tune, it is quite likely that only the musician could positively state that he or she could hear the difference between the in-tune performance and the slightly out-of-tune performance. If someone took a poll and found that only the musician noticed the difference, are we to conclude that the musician was wrong? Obviously not.

Listening to audio equipment is not unlike this. Many types of distortion consist of "out-of-tune" noises added to the music. This is not to say that musicians would generally make good reviewers of audio equipment, but the more attuned you are to how the music should sound (whether you are a musician or an acute listener), the more likely you are to notice a difference if a small amount of "out-of-tune" distortion is added. [See my Test CD article elsewhere in this issue.—Ed.]

The problem with listening tests to "prove" that audible differences can or cannot be heard lies in devising an adequate test, just as the problem in testing any hypothesis is finding an adequate experiment to test it. The "objectivist" is theoretically correct in insisting that such listening tests should ideally be "blind." The "objectivist" is wrong, however, to assume that any particular "blind" test is adequate. Typically, "blind" testing to date has involved listening under the confining conditions Mr. Harley has described. The underlying assumption in these blind tests is that significant differences among audio components can be heard quickly under pressure. The flaw is not in the concept of blind testing, which is theoretically correct, but in its limited application.

For example, every experienced reviewer and listener knows the phenomenon of "listener fatigue." Sometimes it takes several hours, or even days and weeks, of listening to notice that more irritations are present in reproduced sound when listening to one component than are present when listening to another that is comparable. In view of this, one "blind" test to consider performing would be to give an experienced equipment reviewer two different amplifiers contained in identical boxes. The reviewer could compare them in the environment he or she is used to and listen to them over a long period of time. Then the boxes could be passed on to another reviewer. I have no doubt that such long-term blind testing would produce consistency among many trained reviewers, not only in showing that Amplifier A is different from Amplifier B, but also in defining exactly what those differences consist of. Pure subjectivity is transcended by agreement among subjects about what they have heard.

In principle, obtaining agreement among subjects is no different from objectivists agreeing that the needle on the dial or the digital display "says" a certain thing. You have to look at the meter to read it. Is what I see the same as what someone else will see? Objectivity, even in science, is based upon agreement among perceiving subjects. Doesn't the objectivist claim boil down to a dubious assertion that what one "sees" is more objective than what one hears?

Thus, the flaws in the objectivist positions are really quite simple. The objectivists fail to be objective in acknowledging and testing their own assumptions. Mr. Harley's article misses the point when he criticizes the objectivists' "false premise" that "audio equipment quality can be reduced to a series of mathematical representations" (p.119). First, scientific inquiry does not necessarily require mathematical representations or models, but it does seek models which have some predictive value, and mathematical models are paradigm examples of such models. But more importantly, just as with the objectivist positions, we must ask Mr. Harley, "How do you know that no such mathematical representations can be discovered?"

The answer is that you cannot know (unless you have access to a deity no one else can access) without experimentation and testing, and even then you can only prove the affirmative (that there are such a set of mathematical representations and here they are), not the negative (that no such representations can be discovered). It is one thing to say that no currently known mathematical models can correlate to audio equipment quality (which is probably true); quite another to say that no mathematical models can ever be found to correlate to listening ability. The latter is unprovable. If you use such an unprovable assertion as your thesis, you have entered the realm of religion, not of scientific inquiry. In this respect, Mr. Harley's anti-mathematical position and the "objectivist" position rest on similar ethereal and a priori ground.

An important goal of component testing (and here, Mr. Harley seems to agree) should be to discover the instrument tests which best correlate to the listening experience of those who can hear a difference. It is a goal, in keeping with scientific inquiry, of finding the means of predicting results using the appropriate models. This goal appears to be in keeping with the usual approach of Stereophile's best equipment reviews, which: 1) combine extensive listening tests with instrument tests, 2) attempt to correlate the two, 3) determine when the two do not seem to correlate, and 4) incorporate new means of instrument testing when they become available and appear promising. Eventually it may be possible to find the battery of instrumental tests which can predict which audio components will sound best to acute listeners. But acute listeners, of necessity, remain the final judges.—John L. Hodge, Montpelier, VT

sudont's picture

Clearly there are differences in sound between various components, even cables. What disturbs me, is that over the last couple of decades, reasonably priced cable has all but disappeared from the market. If there's a justification for the ridiculous prices of cable, I haven't heard it. It's certainly not in the cost of the materials.