Stereophile's Writers on an Audio Quest Page 4
Sommerwerck: If you listen to one speaker which is low impedance and you listen to another speaker which is high impedance, and they sound different with a speaker cable, you tell the readers that. That's a fact. But you don't draw the conclusion that it necessarily sounds different with different speaker impedances because you don't have enough of a database to draw that conclusion.
Olsher: You mean that I cannot offer any conjecture? What's wrong with conjecture?
Sommerwerck: Because you don't want to be making generalities that you don't have evidence for.
Robert Deutsch: There are two issues here. One is what's known as internal validity of measurement. That is, if you're going to compare cables or amplifiers or speakers, you should do everything within your power to ensure that you're not unduly influenced by what you've heard about the product or because it comes second in a sequence of measurements or that you've connected something wrong. And it's certainly the responsibility of every reviewer to go through whatever they can—and that's, if you like, professional responsibility—to make sure that they're doing all of that correctly. Now once that's done, and once you've decided that a certain product has certain characteristics or it sounds better, whatever, then the next issue is what is known as external validity. Does it apply to situations outside of here?
If we said that a certain cable sounds better in a system that involves tube components, is it then going to sound better in other systems involving tube components? And there may be a problem with a generalization, but what the public wants to know is not necessarily does it work with the Air Tight amplifier and some other tube preamp that they may not be considering purchasing, but is it going to have the similar characteristic with another set of components?
Low: They do want to know this. It's an important subject.
Deutsch: So at some point you have to view external validity. And the more experience you have with other products—you can call it prejudice, perhaps you can call it experience, knowledge, conjecture, or whatever—the better you're going to be at it. But you're not going to be perfect in making those predictions. And sometimes the reviewer is going to be wrong. Because there could be a tube product out there that simply will interact in a negative fashion with the particular product that you've reviewed and very honestly and very carefully and thoroughly came to a correct and valid conclusion about.
Low: It does happen, although I think that the reviewing community has partaken in what I call the Band-Aid effect much too often, which is when they use the term "compatibility." I'd like to see this applied to a high-mass tonearm and a high-compliance cartridge, or a moving-coil cartridge into a phono input that needs 2mV, as being "incompatible." It happens occasionally with cable: if you use a Naim amplifier with a cable that is not highly inductive and sufficiently long, the amp will be unstable. These are genuine cases of compatibility. But the word is 99% of the time used for aesthetic compatibility compared to electronic compatibility. If components were more rigorously evaluated for their absolute quality and not how they integrate into a single system, everyone would be better able to predict whether it will work in their system.
Atkinson: But you don't have access to absolute quality. Last night, Tom Norton, Bob Deutsch, and I were listening to the Signet SL280 speakers, and the sound was quite hard, quite bright, not very pleasant. Tom then substituted a different pair of interconnects, the system relaxed a little, the soundstage deepened a little, the singers moved back a bit, it became more enjoyable while remaining the same basic sound. Now, you don't have access to the absolute knowledge of the sound of those two cables we used. All you can do is report on the anecdotal differences they produce. Now that's exactly the use of the word compatibility which you decry, but I think that's the only word you have. If you have a system with a certain kind of balance, if you add this cable rather than that one, the sound'll go even more in that direction. It's incompatible not aesthetically but musically.
Low: Right. But it's not as though the performance of that cable changed because of what equipment it was hooked up to. It's that the problems with that cable—everything is distorted, everything is wrong—it's all a question of damage control, all the way through a hi-fi system. This is a fact of life. And accumulating too many sins of the same type, too many pieces that are subjectively irritating—we have this language problem where we tend to talk in terms of amplitude response and it's almost never amplitude response. So we imagine ourselves with an equalizer with this piece going up and this piece going down, and you put them together and it's somehow livable.
Holt: That's the way they often sound...
Low: So it's subjective in interpretation. It may be a valid language, but it implies things that are not true.
Atkinson: You implied something earlier which disturbed me a little which was that manufacturers can influence the outcome of reviews by tailoring their presentation to a particular reviewer. Do you think that happens a lot?
Low: Did I say that? I think it's true, but...
Sommerwerck: He said some reviewers were susceptible to being influenced. The thing is, for example, I am an engineer and I consider I know what I'm talking about: I feel I'm totally immune to anything but the truth.
Low: You have to have some prejudice to be in this, I agree.
Olsher: I don't think that being an engineer is necessarily a help in this business.
Sommerwerck: I think it's a tremendous help...
Lipnick: I don't agree at all.
Olsher: Only if you're going to be swayed by the sweet-talk of a circuit...
Low: Knowing nothing is sometimes better than knowing a little. It's knowing a little but not knowing that you know only a little...
Olsher: Bill [Sommerwerck], as an engineer, how would you approach a cable?
Sommerwerck: The only cable manufacturer that has written anything that was vaguely convincing, one way or another, was OCOS.
Low: You just haven't read my literature yet.
Sommerwerck: The point is that looking at it from the point of view of an engineer, I do hear differences between cables. I do not feel that I understand what those differences are. If a manufacturer like OCOS comes along and says, "Well, we've shown that matched impedances are necessary and so forth," and they show the mathematics of it, that might bias me a little bit toward liking the product. But if the statement is true, I would want to prove it by testing it, by getting multiple amplifiers, multiple speakers, and trying a bunch of speaker cables and seeing if the OCOS is consistently better than the other cables.
Atkinson: But doesn't what you've just said reveal that the elegance of OCOS's snowjob has actually been quite effective in influencing your opinion?
Sommerwerck: No, no, all it is is that I'm looking at their claim that the characteristic impedance has to be matched. We all know that that's true. The math is right, but we don't know if the original assumption is correct.
Olsher: Whereas if they said, "Gee, you know, we made our cable out of Swiss cheese, we really don't know why it works, try it," you'd say "Oh, come on, guys."
Sommerwerck: Well if it sounded good, I'd say it sounded good.
Richard Lehnert: But you wouldn't spend as much time and energy exploring that as you would an argument that seemed more elegant, more reasonable to you.
Low: You're actually being set up for another conclusion that itself will be irrelevant and, if made, not responsible. Which is that if you do like that cable, you'll be tempted to say impedance matching has been vindicated.
Sommerwerck: No. I would not say that. What I said was that let's say I had an amp and a speaker. Okay? And I happen to listen to that cable, and I thought it sounded superior. What conclusion can I draw from that? No conclusion. I can't draw any conclusion except that I think that cable sounds better than the other cable with that amp and speaker. In order for me to even begin to draw any conclusion about whether or not their original assumption is correct as opposed to just their mathematics, I'd have to get a whole lot of amps, a whole lot of speakers, and start doing a whole lot of listening.
Atkinson: I think that the point here is that any specific product story is tailored, coincidentally, for a particular reviewer's belief system. Hence the susceptibility Bill Low was talking about.