John Dunlavy On "Cable Nonsense"

Plasma, take a look at my post on the "Are Audiophiles Obsessive Nuts?" thread here on Agon, and I won't bother to repeat it here. Thanks for the link to the essay by Dunlavy; he sure speaks more authoritatively than do I. Thank God there are some, albeit precious few, who are willing to speak the truth about the widely accepted, yet critically unexamined high-end orthodoxy.

Why is it that several of the above posts hurl aspersions at John Dunlavy, yet not one post has hinted at any semblence of concern or displeasure with the cable manufacturers? Can someone explain the psychology at work here?

Leafs, I thought that you didn't believe in tests? So now why are you slamming MIT based on the result of a pernicious and feared test?

Oh, never mind. I now understand that if it is a controlled test conducted in a systematic, unimpassioned, objective way then the results are suspect. But throw together 5 audio dweebs and "Katy bar the door."

Uh-oh Plasma, you're up to it again :-)

Rgd, forgive me for my skepticism, but have you (or anyone here) actually observed or been involved in the manufacture of cables? I would think that the extrusion of the jacket/insulation over the cable would be very difficult to perform by hand and my understanding of cable/wire manufacturing is that it is a very highly automated and capital intensive industry. Production economies dictate that a relatively low value-added, commodity type item such as cable (remember, audiophiles are unique in what they are willing to pay for wire) must be manufactured in large quantities to justify the capital intensity I have mentioned.

In such situations, small players that cater to limited volume niche markets (like the market for high-end audio cables) must rely on outsourcing or contract manufacturing arrangments. Given their low volumes, they simply could never recoup their investment were they to attempt to manufacture the cable themselves. Perhaps there are some exceptions to this? Anyone???

I've done a bit of poking around the Net and looked at the North American Industrial Classification System maintained by the Bureau of the Census. The industrial structure is pretty much as I painted it in the last post, only with more levels in the supply chain. For the trivia hounds amongst you, here is the series of handoffs (NAICS text in quotations).

#1 --- wire is drawn

"NAICS 33142: Copper Rolling, Drawing, Extruding, and Alloying
This industry comprises establishments primarily engaged in one or more of the following: (1) recovering copper or copper alloys from scraps; (2) alloying purchased copper; (3) rolling, drawing, or extruding shapes, (e.g., bar, plate, sheet, strip, tube, and wire) from purchased copper; and (4) recovering copper or copper alloys from scrap and rolling drawing, or extruding shapes (e.g., bar, plate, sheet, strip, tube, and wire)."

#2 --- drawn wire insulated by firms at next level in supply chain

"NAICS 33592: Communication and Energy Wire and Cable Manufacturing
This industry comprises establishments insulating fiber-optic cable, and manufacturing insulated nonferrous wire and cable from nonferrous wire drawn in other establishments."

#3 --- insulated wire sold to audio OEM's and turned into audio goodies

"NAICS 3343: Audio and Video Equipment Manufacturing
This NAICS Industry Group includes establishments classified in NAICS Industry 33431, Audio and Video Equipment Manufacturing."

I would venture that most, if not all, of the audio cable manufacturers pop up at the third/last level rather than the second.

The community should understand the difference between "statistical" and "practical" significance. Consider the following example.

Through omniscience, we know that:

Car A gets 20.1 mpg.
Car B gets 20.2 mpg.

Obviously, the TRUE difference between the fuel efficiency of cars A & B is 0.1 mpg. Sampling a sufficiently large sample size of cars A & B, we would be able to detect a "statistically significant difference" between the mpg for the two cars. But once we do detect a "statistically significant" difference, what does it mean?

In this case, it means precious little in terms of "practical significance." That is, how much is a 0.1 mpg difference really worth in the real world? Not much! And certainly, consumers would not be willing to pay a significant amount of money for such a trivial difference in fuel economy.

Now to come back to the issue at hand in this thread, let's assume that it really is possible to find a "statistically significant" difference in the sound of cables (this is a leap of faith given JD's blind testing results, but references to a "J. Peter Moncrieff of IAR" keep cropping up and I'll accept this single result as valid). The central question becomes -- Is this "statistically significant" difference a "practically significant" difference. I would argue that it is not, especially in the context of the obscene prices charged by high-end cable manufacturers.

Understand that when statistical testing ends in non-significant results or mixed resutls (i.e., sometimes significant, sometimes non-significant) the most frequent reason is that the difference that is being observed (i.e., the "effect size") is so small. I suspect that it is the case here. Any true differences in the sounds of cables are so subtle and small that they cannot be detected with any consistency if at all. So a rational person must ask, is it really worth thousands of dollars to achieve such a subtle and small difference? Only the individual can answer that question for themselves. But as far as the audiophile cable manufacturers' claims of vast differences in and superiority of sound, I maintain that they are hooey at best.

Jay, I think that you may misunderstand what I said. In your example, the gain in power may be real and thereby detectable as "statistically significant". But it is the practical significance of such a trivial gain that is in question. And I agree with you that people who are willing to go to such great lengths, both financially and otherwise, to exact such trivial differences do indeed "look crazy" (your words), or at least a little neurotic.