God is in the Nuances Letters part 2
Editor: I read Markus Sauer's "God is in the Nuances" articles with great interest. I've believed for several years that listeners' emotional reactions to differences in reproduced sound (as opposed to their perceptions of sound quality) could and should be investigated experimentally, just as are other psychological phenomena.
Jürgen Ackermann's experiment seems to be a successful example of this. While no doubt many will argue with them, his findings---that listeners responded more positively emotionally to tube and analog components than to transistor and digital ones---are certainly consistent with the persistence of enthusiasm for LPs and tubes, despite their technical shortcomings. This is in contrast to experiments focusing on sound per se, which have become notorious for producing results---"all amps sound the same"---that are flatly contradicted by the "direct experience" of countless listeners.
Whatever the reasons---listening is too much of a skill, short-term switching is too confusing, too much individual, subjective taste is involved---experiments of this type are apparently a dead end, while experiments like Ackermann's may hold real promise.
In one way, there's nothing unusual about rating components on their ability to emotionally involve a listener. I've read many reviews, for example, wherein the reviewer not only described what he heard, but remarked very positively on the fact that he became so involved in the music he completely forgot about the sound. Some writers advise listeners to specifically look for this---and if it doesn't happen, to keep looking.
As Sauer and Ackermann would obviously agree, that's excellent advice. However, to accept that the listeners' emotional reactions should be explored with formal experiments is rather different. First, you have to be fully persuaded that emotional reactions are neither necessarily conscious nor peculiarly individual; that is, people can't help reacting in a given way when their buttons are pushed, and their reactions are more similar than different. Or, if you prefer, emotional reactions are objective phenomena that are to a greater or lesser degree independent of conscious control, and are susceptible to quantifiable measurement; given samples of adequate size, conclusions can be drawn from data that are of general validity.
Meaning, for example, that if Ackermann's results hold up, LPs push buttons in most all of us that CDs just don't, just as most of us are hurried along unconsciously by music in supermarkets and restaurants, as well as in those epitomes of applied industrial psychology, gambling casinos. If they decide to add sound quality to their bag of tricks, Harrah's might one day try tubes in their music system to see if that raises slot receipts. This would be no stranger than their putting different scents in the air conditioning (vanilla works best); of course, the lab rats at the slot machines won't notice---they're not supposed to.
But, even given all the above, should emotional communicativeness be "the ultimate yardstick"? Sound quality per se does matter: I'm sure I'm like many in that I first bought better equipment to hear more of the details and structure of music that already gripped me emotionally (my assumption being, of course, that the more I heard, the more I'd feel). Further, we're talking of reproduction: It's the job of a system to reproduce as closely as possible what's fed into it; it's the job of the artists, engineers, and producers who make the recording to ensure that the sound is emotionally communicative. However, it seems undeniable that a system that doesn't adequately reproduce the emotionally communicative aspects of a recording fails in a critical respect, making its other strengths almost irrelevant.
And there is not now any reliable, systematic way of knowing what makes a sound emotionally communicative and what doesn't. Since reproduction is necessarily imperfect by any standard, all designs are trade-offs; and designers really should know how their choices are affecting the EC (emotional communicativeness) of their equipment. But what if the greater EC of, for example, an LP or single-ended triode isn't the result of something it gets right, but of something it gets wrong---ie, distortion? If so, it properly should enter the chain only at the production end.
However, no one really knows. All the more reason, therefore, to test this and any other proposition or controversy that seems useful at the moment. Evaluating the EC of the new digital formats would be an obvious priority. Which is best? Does it suffer in comparison with any analog formats? Once it can be reliably established that CE-X is greater than CE-Y, the next step is to discover why; in the process, entirely new design criteria may be developed.
Assume success, and you have a revitalized High End driving the rest of the audio industry to make sound better instead of, arguably, worse, and enriching the spiritually impoverished lives of listeners everywhere---in all seriousness, an exciting possibility.---Robert C. Thweatt III, Bronx, NY
...to the original emotion?
Editor: I think we have to accept the fact that Stereophile is no longer committed to high-fidelity sound reproduction. That conclusion is inescapable after reading the first of Markus Sauer's "God is in the Nuances" articles (January 2000, Vol.23 No.1).
Sauer argues that, since most music listeners prefer distorted sound, and since these listeners should choose audio components that distort the sound in the way they find pleasing, it is regrettable that "You cannot tell what your emotional response to a component's sound will be from a description by a critical listener..."
This carries us back to the days of the console-model and table-model radio/phonos with distortion engineered in to create a characteristic "sound" for the brand line: customers shopped for the Stromberg-Carlson "bass," the "mellow" sound of the Philco, the varying frequency ranges and the distortions therein of the Zenith, Silvertone, Majestic, Magnavox, RCA Victor, Emerson, ad infinitum, sets. The playback equipment was lousy, so we shopped around to find sets that sounded good to us when we played the recordings (also, for the most part, lousy) that were available to us.
High fidelity is the opposite of that approach. The goal of hi-fi equipment is accurate reproduction of the signal fed into it. In the beginning, that meant live sounds, usually orchestras and bands. But since some of the early adherents were interested in sound but not necessarily music, they tested their systems with everyday sounds---jingling door keys, the snaps and crackles of a Volkswagen motor cooling, even the sound of rain, captured when a hi-fi radio station hung a microphone out a window one rainy day. High fidelity meant fidelity to actual sounds.
That changed. More people became interested, bringing in more of those who listened to the highly processed, distorted, never-heard-in-the-real-world sound of bop, bebop, rock'n'roll, rock, heavy metal, synthesizers, et al. The initial population of sound nuts, classical-music freaks, and jazz freaks enlarged to include people who, knowingly or not, wanted the sounds coming from their "stereos" to produce the same effect in them as did recreational drugs. More and more representatives of this mindset became reviewers of audio equipment.
J. Gordon Holt warned us about this. He insisted that evaluation had to be based on real sounds---did the oboe sound like an oboe, the trombone like a trombone, the orchestra like an orchestra, the hall like the actual recording venue? For this position, the old coot was derided, and Stereophile proceeded to publish reviews describing how sounds that had never existed in real life sounded when played through certain components. But, since there is no reference sound to compare these sounds against, reviewers talked about how a certain component "kicked butt," unaware that what they were really praising was the way the equipment distorted the sound in ways that they found pleasing. They stopped comparing sound systems against reality and instead compared sound systems against each other.
Markus Sauer's article is an inevitable result of the movement away from high fidelity. In the article, Sauer establishes premises, supports them by citing a study conducted by Jürgen Ackermann, and comes up with the conclusion quoted (in part) above. But I find many of Sauer's premises questionable and/or insupportable, and the Ackermann study poorly designed (although rigorously conducted) and fatally flawed.
I would dearly love to go into detail about my problems with Sauer's article and Ackerman's study, but I doubt that I would ever be allotted the space necessary to support my slanderous accusation. So I will confine myself to addressing what I consider the "fatal flaw" in Ackerman's study: Ackerman compares the "emotional" reactions of test subjects to three varying sound systems, telling us which of the three systems that particular test panel preferred. That's all. We can't extrapolate from that, or draw any general conclusions, or decide which set is best, or even how other test panels would rate the same systems.
Ackermann's study, as described by Sauer, is fatally flawed because it did not establish a reference standard, such as comparison of the three systems against live music or a reference system of known quality. Without that, his study has no relevance to high-fidelity reproduction.
Sauer's self-described "far-reaching" conclusion is that "you cannot tell what your emotional response to a component's sound will be from a description by a critical listener." Why would any music-loving audiophile want to? What we want are components with the least possible sound of their own; components that will pass through, with the least possible distortion, the music we feed into them. If there is distortion that arouses emotion, let it be in the source itself---the tape, the CD, the LP, the DVD, whatever---but not in the equipment.
It is true that, in the old days, we compared the "sound" of the components---the HH Scott against the Fisher against the Marantz against the McIntosh against the Dynakit against the Heathkit against the Bell against the Bogen against whatever---and we made our choices on the basis of what "sounded best" to us. But the choice was based on the absence of sound created by the components themselves: We picked equipment that we thought imposed the least distortion on the signal fed through it---the equipment that gave us the closest illusion of real music coming from real performers on real instruments in real acoustic spaces.
Now, on the rare possibility that some reader has stuck with me this far, I want to make it plain that I do not want to deprive people of their right to listen to any kind of music or any kind of sound system. I do not want to stop reviewers from writing for those consumers who want their components to have characteristic sounds. I do not look down on magazines that cater to that taste---after all, such listeners make up the majority of listeners, and catering to them makes good economic sense. I do not in any way feel superior to people who want to play distorted music through sets that impose more distortion. I'll let them ride their horse if they let me ride mine.
But I think Stereophile is at a crisis point and needs to make a decision. I see three courses of action: go back to dealing with high-fidelity reproduction; stick to the present course into the sound-for-sound's-sake camp; or split into two separate publications. I strongly doubt that one publication can serve the needs of two such disparate populations.
Let me make that even stronger: It is impossible for Stereophile to keep a foot in each camp and remain the highly regarded (although frequently attacked) and respected publication of record that it has been.---Paul A. Alter, Pittsburgh, PA, firstname.lastname@example.org