SIGNALS & NOISE (Letters to Editor) (Dec. 1989)

Sound Argument

Dear Editor:

I am a little confused about how audio equipment is evaluated by magazines, audiophiles, or anyone else. The reviews of the Linn LK1 preamp and LK280 power amp in the April issue provide a good example. One source of confusion involves the evaluation of banana plugs. According to author Leonard Feldman: "The friction fit of Linn's special plugs made far better contact than would be obtained with most banana plugs." It is my understanding that Pomona Electronics, a manufacturer of standard banana plugs, specifies the resistance of their banana plugs at the contact point as less than 0.001 ohm. This seems low enough to be significant to me. Am I missing something in not seeing the value in a better fitting banana plug? If not, why would a manufacturer go to all the trouble of providing non-standard plugs to achieve an insignificant improvement in performance?

Further, Feldman states that "Linn does not put too much stock in lab measurements." As anyone can see by the measured parameters of these products, there does not seem to be anything in the measured performance that is noteworthy. The distortions (those that we know how to measure) introduced into the signal chain by these components are somewhat higher than in most other products on the market. In fact, the noise peak in the preamp that resulted from the proximity of the power amp is a serious design flaw, in my opinion.

So what is it that makes these components so good? There is little said about the construction, reliability, or any other design attribute that would indicate high quality, except for r.f. shielding and microprocessor control.

If these products had been far lower priced (e.g., $500 total rather than $2,500 total) and from a different company, would Feldman have likened them to a Rolls-Royce or Mercedes? I doubt it. This suggests that the author's opinion of the quality of the products might have been based more on price and brand name than on actual performance. Similarly, would Feldman's response to the sound of the components have been just as favorable? I certainly have my doubts.

Now, I know that many audiophiles have been claiming that there is little or no correlation between measured performance and sound quality. However, the trend to ignore measured performance is becoming more prevalent all the time. In its place, unfortunately, evaluations are based increasingly on cost, construction, design philosophy, and uncontrolled subjective listening tests. Feldman even has to apologize to some portion of his audience for measuring the equipment and presenting the results of those measurements.

It seems to me he was probably afraid of offending and/or boring some portion of the audience and the manufacturer by measuring (God forbid!) the components.

Audio is, more and more often, providing readers with opinions instead of facts. Yes, Audio measures the component (except in the "Auricle" reviews) but, in this particular case, then proceeds to ignore the results of those measurements. There is no statement which summarizes and draws conclusions about any of the measurements.

Why not? Feldman is in a far better position to draw conclusions about the measurements than most of his readers. This is sending a message to the readers that the staff of Audio believes such measurements are meaningless.

Does Audio really believe that uncontrolled subjective listening tests are unbiased? Is it really possible that measured performance is totally irrelevant to sound quality?

Audio began life almost as a professional journal, with a staff of people who had careers in engineering, music, and broadcasting. The staff, today, is still composed of many who have achieved recognition in their respective fields. However, the policies of Audio reflect a movement away from the style of the professional journal and toward a consumer-oriented magazine. The consequence of this is that readers of Audio hear, more and more, what they want to believe rather than what is factual. A true journal would be more interested in publishing data and experiments rather than conjecture and opinion. Maybe Editor Eugene Pitts should spend some time reading real journals as a model for Audio.

-Chuck Butler; Kalamazoo, Mich.

Author's Reply:

Over the years, I have had countless quarrels with the "subjective listener" champions who insist that most audio measurements are meaningless. I have maintained (and still do and always will) that if measurements don't tell the full story, it is simply because we haven't learned which measurements are the right measurements to make or our measurement techniques and instrumentation are not sensitive enough to yield the correlations we seek.

I have been attacked ad nauseam by the so-called "underground press" for my steadfast insistence that measurements are important in evaluating audio, video, or, for that matter, any other electronic product I test. That's not to imply that I don't also listen to the products I test. Of course I do, and it is the reasoned combination of listening and measurement that should yield a test report that can be trusted and believed. So to have you imply that I am anti-measurement really struck me as rather odd. Had I been anti-measurement, I certainly would not have invested several thousand dollars in the Audio Precision test system I now use. I purchased this equipment because it dramatically increases the level and sophistication of the measurements I am able to perform. I am about to invest several thousand dollars more, within the next month or two, to upgrade that equipment so it will be able to measure performance digitally as well as in the analog domain. Does this sound like the actions of a person who is anti-measurement?

As to the specific product report you mentioned, please be assured that I never worry about offending any manufacturer--whether that company's products are middle of the road, low end, or high end. I do, however, take cost into consideration if I feel it has a bearing on the overall evaluation of the components I test. In short, Mr. Butler, we are really on the same side. The levels of distortion exhibited by the Linn products were a bit higher than the "triple zero after the decimal point" numbers other amplifiers and preamplifiers boast, but this did not seem to alter their superb sonic quality. Total harmonic distortion really becomes aurally significant only at about 0.5% or so.

-Leonard Feldman

(Adapted from: Audio magazine, Dec. 1989)

= = = =