Speaker Cables: Can You Hear the Difference? Tables With Data

Tables With Data


Table I shows the results of the listening panel’s pre-test evaluations of the three cables tested (left part) and their ratings of the cables during the open listening tests (right part), together with a statistical analysis. The ratings in the upper half of the chart are mean scores derived by averaging the questionnaire responses of all the panelists; the numbers shown in parentheses below them are their standard deviations—that is, the statistical spread around each mean score. Thus, the higher the standard deviation, the wider the range of scores.

The lower half of the table indicates whether the differences in the ratings of the three cables are statistically significant. The method of comparison used, the Student’s paired test, estimates the probability that an observed difference in preferences could have occurred by chance. The number given for each comparison in each category or test represents the probability that the difference did not occur by chance alone. Probabilities below 95 percent are considered statistically insignificant and are indicated by “NS.”



Table II. The two charts on this page show the results of the controlled, double-blind listening comparisons of the three cables. At right are the results for the panel as a whole, together with a statistical analysis of the comparisons. The scores for each member of the panel in each comparison are shown in a separate chart.

The ABX comparator system is statistically similar to flipping a coin and predicting whether it will come up heads or tails. Each of the two results is equally probable, so random predictions are likely to be correct 50 percent of the time. Since each of our comparisons comprised fifteen trials, a listener would have gotten a score of 1.5 if he could hear no differences between the cables and were just guessing. Any score much above 1.5 is thus better than chance and might be significant.

Using published tables of the binomial (“bell-curve”) statistical distribution, we calculated the likelihood of correct scores in each comparison for the entire panel. We found that 91 or more correct answers out of a total of 165 in each comparison gave a probability of 95 percent or more that the results were not due to chance. Results meeting this criterion are thus indicated as statistically significant.

We also used, however, a stronger criterion: psychoacoustical significance. In psychoacoustical testing, it is generally accepted that the threshold at which a phenomenon can be considered definitely audible is when listeners are aware of it at least 75 percent of the time. This is the basis for our definition of a “hit” as at least twelve out of fifteen correct answers. Applied to the scores of the whole panel, this meant that 124 answers out of 165 trials for each comparison had to be correct before we concluded that the differences between the cables were indeed audible.

Results that meet the 75 percent rule are due to more vivid effects than those producing a merely statistically significant result. An additional clue to the magnitude of the differences between the cables tested is the number of listeners who scored a hit in each comparison; the more hits, the more striking were the audible differences.

anmpr1's picture

I give Noel Lee credit for realizing that a lot of dollars can be made by exploiting audiophile neurosis. That said, I've used his less expensive speaker wire, and can reliably report that it works.

The most ridiculous are expensive "super high end" Ethernet cables used to connect your music server. Some have "directional" connections. Evidently people that buy these things have no idea how Ethernet works.

I knew a guy who bought a set of Sound Lab electrostatic speakers (good product), but the dealer sold him some ceramic thingys (looked like insulators from old telephone poles) that were used to keep the expensive wire off the carpet. It was supposed to reduce some sort of "floor induced" electrical interference, an artifact no one could ever explain coherently.

All this stuff is just a way for the dealer to make extra dollars from a high profit item. Now, where did I put my Valhalla Edition Hosemonster Cables--the ones with the silver plated Gotterdammerung spade lugs? My dealer said he'd take them in on trade for the new and improved Red Kryptonite lined Phlogiston free cable, featuring phase coherent bi-directional Interocitor connections. I read in a magazine somewhere that they "blow away" the old model. Can't wait.

Deus02's picture

An age old argument in which I think Mark Twain's words are more relevant than ever, "It is easier to fool those than convince them they have been fooled". The believers will believe because they have to justify their outlandish expenditure and probably, before too long, the "fire and brimstone" of audiophiles with their $300 "audiophile fuses" in tow will emerge.

Of course, there is also the issue of differences showing up on measuring devices that don't necessarily emerge in a real world listening/viewing environment. A friend of mine, who has been involved in the a/v retail business for over 20 years, told me years ago that they push so-called high-end audio/video cables because, of all of the equipment and accessories they sell in their store, these cables provide them their biggest profit margin and that same scenario continues to today.

Old Ben's picture

The expensive cable gambit is total fraud. If it weren't, then there would be verifiable proof of the superiority of these cables. The design of the experiment ought not be too difficult. Take a high-quality audio source and connect it to an oscilloscope using an inexpensive cable and an expensive cable that are otherwise identical (e.g., length, gauge) and compare the signals at the downstream end of the cables. Are they different? If so, then publish the results!

drny's picture

I remember the original article on SR.
I won't argue the merits for or against so called high end speaker wire.
After 40 years of using all types of wires (24-10 gauge, including Monster), there are several variables. Such as speaker Impedance, run distance, contact with any other signal inducing or carrier product.
A general catch all basic for speaker wire is copper material, high level of insulation, 16-12 gauge.
For the past twenty years I've negotiated my speaker wire into the purchase of the actual speaker system. You will find that the dealer will throw in the wires, if you are buying at retail, or if its a discontinued, close out, demo speakers (does not apply to box store such as Magnolia/Best Buy).
In a Home audio world where a High End Speaker line can go over $50,000 a pair (basically for status symbol or 'fools gold'), the speaker wire debate is almost esoteric.

brenro's picture

Thirty foot runs of Monster Cable didn't sound any different than lamp cord. Earth shattering.

William Lee's picture

Should I conclude that using expensive cables would make everyone hears Yanny?

Tumara Baap's picture

Audio Critic was the other magazine that was relentless in its criticism of woo woo. It is sad that even proclaiming the most ordinary position of common sense caused such consternation. In theory, even if cables differed a little in capacitance or impedance, any effect would have been eclipsed by the Lord of Distortion and Sloppiness, the loudspeaker itself. The typical loudspeaker with its slovenly passive crossover is a terribly crude device. There simply exists no pathway for cables to audibly assert themselves in this context.

PunchyRedcrown's picture

Do we throw this in the same bucket as line conditioners and biwiring? I liked the comment about Mark Twain, which particularly holds true when reading the user reviews on Audioquest and Monster speaker wire- something like 5x the price. Let's face it, folks. They're really reviewing their brand new $2000+ 5.1+ home theater.

rgsalinger's picture

If you think that wire makes a difference, hire an elf. Put two lengths of wire between your amp and your speaker one is expensive wire and one is proper gauge zip cord. Have the elf switch them back and forth on a random basis when you are not home. The elf needs a key. Listen intently for 90 days, every day and write down which cables you are listening to. Compare your results to what the elf did and see if you could actually hear the difference.