What if I told you that someone had invented a new machine that could take any music signal; split it; run it through a series of processes, both in- and out-of phase; put the signal halves back together so that any possible differences would cancel out; and present you with either a graphic, electronic, or written read-out as to whether or not it was worth listening to?
Even if it worked, would you want it?
I wouldn't, either: Music is something to listen to, not to measure. It serves any number of purposes – anything from, like a military march, motivating us and keeping us in step, to (as it may do at a concert or in church), capturing our attention, raising our spirits, and expressing, focusing, and even directing our emotions. Music can truly be all things to all people; entertaining us, enlightening us, making us want to dance, and providing the background for anything from romance to working, to shopping, to riding in an elevator.
What it can't do is to be measured.
Oh, of course we can determine its frequency content, its tempo or changing tempi, its overall length, its average or instantaneous volume level, and any number of other things about it. But only by measuring the response of the people listening to it and not the music, itself, can we determine if it's any good – either on some absolute scale or in performing whatever task we have set for it.
Obviously, no such machine has ever been built or is ever likely to be. But, once again on the internet, a controversy is raging about measuring versus listening as the proper way of determining the relative value of the equipment we play our music on. The source of it all is someone who claims to have invented a machine to, once and for all, prove whether cables make a sonic difference. Of course, this is not the first attempt to do such a thing: Remember the old ABX Comparator, which was said to allow for true "double blind" testing – a type of comparison where neither the tester nor the testee knows which of two or more things is in use at any given time and can therefore impose no bias in judging its results.
What (to my admittedly limited knowledge, but as the inventor has written) this machine does is – just as I suggested earlier – split the signal; reverse the phase of half of it; and put the two halves back together so that they will theoretically cancel, leaving, in the case of the theoretically perfect cable, nothing but silence because the half of the signal carried by the cable would have been carried perfectly, and thus canceled perfectly by the non-cable-carried half.
Without going into why this seems to me not likely to work (far too technical for these pages or this article) let me simply say that the process proposed seems to be nothing either more or different than the huge quantities of "global" feedback used by solid-state mid-fi gear of the last century to produce receivers and other audio components that measured brilliantly and sounded less so.
This was brought-up by a number of people contributing to the thread, but the "inventor" replied that it was irrelevant; that if the gear measured good but didn't sound good, there was obviously something wrong with the measuring process; that anything at all can be tested if the right test is devised; and remained adamant in his insistence that his machine was the "right" test.
But is it?
Cables are far from the only things of which people have refused to accept the validity or sonic differences without double-blind testing. At least two such tests that I know of were conducted on audio amplifiers, which the Audio Engineering Society (AES) insisted must all sound the same if they measured the same within some set of established parameters. Both of these tests were, as I recall, conducted in the mid-to-late 1980s. And in both of them, multiple testers listened to multiple amplifiers (ten amplifiers, in the first case, as I recall) in double-blind mode and were supposed to tell, by listening blind (after one exposure of listening to each while knowing what they were listening to), which one of the amplifiers was being played.
The odds of anyone being able to identify all ten of the amps under test in that first case were in the range of statistically impossible. Even so, out of 200 testees, two (1%) – a number that simply could not have been achieved at random – were able to identify all ten. When the second test also showed a much higher than statistically probable result, the AES, as I recall, declared the test to have been "flawed" in some way, and thus invalid.
Whether or not the testing was correct in its outcome seems to me to be irrelevant: Perhaps manufacturers should try in every way they can, including the very best measurement and testing techniques and equipment possible, to make their products as good as they can. But for consumers – the people who actually buy the stuff – testing of practically any kind other than just listening seems to me to be nothing more than a distraction.
Remember what I said earlier about making "...a sonic difference." Sonic is the key word. Music is an artform for the ears, and our systems have delighting our ears as their main purpose. How, then, can anyone say that our ears should be supplanted by a machine to make our decisions for us? For one thing, while I believe that a cable should do nothing at all other than convey signal from Point A to Point B, and should make no change of any kind to the music signal that it passes (and while I have always designed cables to achieve that goal), other people may think something different. So what if they want a cable that acts as a filter or a tone control or that adds or subtracts something from the music that they listen to? Isn't it their music? And their system? And their ears to be pleased?
That's why I‘d never rely on a machine – not even one that worked – to choose my music, and why I'd never rely on anything other than my ears in choosing the equipment I play it on.
Forget your machine; all I want to do is...