As I do about once a year, I am vacationing with my sister and my brother-in-law George S. Louis in their house near San Diego, CA. Technically it is in El Cajon, but on the north side of Mount Helix. Mount Helix on the south side is a famous and prestigious San Diego neighborhood known for being the home of many successful creative people like artists and writers. The north side is similar, or maybe even nicer, though the far away ocean view is replaced by a view of El Cajon and the surrounding mountains. One of the great virtues of George's home is lack of sound interference or complaint from neighbors since there is considerable distance from them. It is also very large, 3500 sqft, with an open floorplan that virtually eliminates room modes. I claim that much of the good and sometimes wonderful sound that George achieves stems from those two hard facts.
He claims that it doesn't matter that he has a large listening room, instead ascribing his good sound to his astute choices of equipment, positional and level adjustments, and the combined effects of all of his audio tweaks, which are now nearly countless. Such tweaks include custom cables and connectors of outlandish design, equipment covers made of felt (I agree with that one actually because equipment itself resonates mechanically), ionizers, alleged energy field generators or absorbers, black lights, tourmaline stones, damping pads on the top of CD's, CD chemical pre-treatments, 5 second pause protocol ("CD's always sound best when paused 5 seconds first"), polarity correction, teflon or polypropylene film capacitors (I agree with that) having chemical, thermal, or mechanical treatment (but I don't agree with that), completely unplugging major appliances and computers before listening, choosing certain display modes such as minutes-into-track display (not necessarily display off), avoiding room damping but using limited constrained layer damping on equipment and speakers. George insists that everything except the brand of batteries in the remote control makes a sonic difference, and he has proven all these tweaks to himself and others in relatively short listening tests (mostly without listening to tracks all the way through). He also insists that measurements are most often misleading. "Listen to the notes, not the measurements"(tm) is his trademarked motto.
George and I have been friends and card carrying audiophiles since the early 1980's, even before he met my sister. While we initially collaborated on the use of Linkwitz-Riley crossovers (which he subsequently abandoned) our approaches to achieving audio bliss have diverged more and more over time. Since we met, he has also gone into the manufacturing and sale of audio enhancement products, sometimes both affectionately and dismissively called "tweaks" by other audiophiles and other recorded music listeners. (This has nothing at all with the use of the word "tweak" in other contexts; in the drug use world, the word "tweaker" has been appropriated for those who use Heroin. Well maybe there is some similarity in the fact that both heroin users and audiophile tweakers are addicted to their respective means of achieving bliss.)
George usually has an extremely good sounding audio system usually--better than the average audiophile (I felt it was not up to his usual level of excellence this year, though still good in many ways), as well as a the most extensive collection of recordings of anyone I know. He also has an extensive network of audiophile friends and customers, and one of the frequent delights of my vacations is visiting other audiophiles with very different audio systems--one of the best ways to learn.
And one of the shocking aspects of George's living room audio system is that despite the usual high level of sonic excellence, at first glance it appears to be little more than cast off audio junk. The stereo speakers are a two-piece system consisting of a pair of old obscure and unimpressive looking speakers made by AR turned upside down, with another pair of speakers originally sold by Radio Shack on top. However, a closer examination reveals tweaks and more tweaks everywhere, and most of the tweaks are home spun inventions. I know from my sister that George spends endless time listening to his system in brief spurts and adjusting it, as best he can, mostly with homespun tweaks based on his own iconoclastic theories.
I concede he has good sound, which has been either better sounding overall or at least better in a few ways than mine. But I am not particularly interested anymore in trying his approaches in which purely empirical audiophile tweaks with little or no understood audio engineering basis are believed by him to play a huge role. I contend that his good sound results not from the inexplicable voodoo tweaks per say, but other adjustments and alignments he has also done in his countless hours of adjusting. And it's not like the actual parts he uses are not fairly good to begin with. While many tube-o-philes and others dismiss early transistor amplifiers, his Sony 3200F amplifier got a rave review in Audio Magazine when it was introduced in the late 1960's, AR made consistently good speakers while it was still an American concern, and the Radio Shack RCA Pro-LX55 is a descendent of an almost identical speaker which got a surprisingly favorable review by John Atkinson in Stereophile. These are, in my view, adequate components which can very well if well set up in a wonderful sound room. Rather than saying, as he does, that the Sony is far and away the best sounding amplifier ever made, a claim that is both iconoclastic and outlandish, I say that amplifiers of adequate quality, and there are many to choose from, sound basically the same so it really doesn't matter which one he has chosen, so long as it is sufficiently powerful etc. for the speakers used.
He says he must have good hearing if he has chosen and adjusted a system to sound so good. He believes his system to have the best sound anywhere, and he has haunted many stereo shops, visited countless audiophiles, and attended a large number of hifi shows.
I concede he has good hearing. He has not settled for bad sound, which could easily have resulted from his iconoclastic and penny-pinching tendencies. But I believe his hearing, like that of most audiophiles, can also be easily fooled into experiencing sonic differences even when don't really exist. There are tiny random effects, mainly on perceptive attention, that can seem like sonic differences even if the exact sound were reproduced each time. And the use of memory, since two sonic performances cannot be compared side by side like video screens. That he can be so easily fooled at the margins is not at all inconsistent with being able to eventually hear real differences when they really do exist. As intelligent as he is, and his is very intelligent, somehow he can never follow this argument. Of course he also claims that the improvements that result from his tweaks are huge, huger than those that result from component swaps mainly. Actually I see his main error not in hearing, which must be good, but in ego, believing in his hearing and every brief listening test he has performed, and not accepting that any of this is random or beyond his control.
One of the best demonstrations of how random our attention and therefore perception of sound can be is given when some orchestras play world premiere music. Sometimes they play the same work twice. I have always been amazed how the second playing sounds entirely different. The first time this happened, I was convinced the conductor was playing a trick on the audience. But the trick is this: we never hear the same thing exactly the same way.
For a number of reasons, I am not particularly interesting in trying things when I have not a clue as to how they work. Most often, audio tweaks have either no effect on the measureable performance of an audio system, or a measureable effect so small as to make it implausible that they have an audible benefit, or even a measureable effect going away from the assumed direction of audible improvement. Quite often, audio engineers would be unable to explain how they have any benefit or even audible effect at all. I concede that these engineers could be wrong, and there could be many things that improve audio which are either not understood or actively dismissed by them. However, over time, it would seem to me that if real methods of improvement had been found, these benefits would eventually be confirmed by scientific methods such as Double Blind Testing, and ultimately adopted and explained by audio engineers and scientists. In my 45 years of audio play, investigation, and reading, I have not seen this happening, and have become sympathetic more to the argument of audio "objectivists" (another word used differently in other contexts; an audio objectivist is not likely interested in Ayn Rand) who insist that audio benefits be proven by scientific methods and measurements. A famous audio objectivist is Peter Aczell, publisher of The Audio Critic. I agree with much of what Peter Aczell says nowadays. While most often I think his rhetoric is too harsh, I am also sympathic with his "no more Mr. Nice Guy" feeling. Especially each time after visiting George, who cannot help but constantly cross examine my audio thinking in a way that seems aimed to find fault with it, and to confirm his own ideas. I enjoy my visits, but this aspect is annoying. I would be comfortable being non-judgemental, but when my ideas are constantly called into question, it is hard to be that way.
Audio tweakers and their sympathizers always insist that one should "listen and decide for oneself". I see many problems with that approach. In my view, human hearing is not a mechanical process but a creative one. It is true we can often hear tiny things that would have seemed to be implausible on the basis of common sense. But at the same time, we never hear anything exactly the same way in two successive listening sessions, even it it is a physically identical playback. As such, we can easily be fooled into believing that something sounds better or worse, when in reality it has been unchanged. So if an audio tweak has exactly zero effect, something like half of the time we will find it to make things sound better, and the other half worse.
Phenomena like this, and various psychological tendencies such as "confirmation bias" are well known to scientists, and the reason for controlled methods such as double blind testing. Non-audiophiles might be amazed at how loath audiophiles are to use blind testing methods, particularly double blind testing. We haven't actually proven humans can hear the benefit of far out tweaks. When an audiophile says some tweak makes a huge difference, you should take that with a huge grain of salt if swallowing it at all.
The audible benefit of such tweaks (having no engineering explanation) has not been established by double blind testing. To establish such benefit requires more than one successful double blind test result (though that would be suggestive) because that could happen as a matter of chance. If many such tests are performed, but not made known to others, the few successful results that would happen as a matter of chance might get more attention than they deserve. But in the case of most audio tweaks, except for those now accepted by objectivistic audio engineers, I am not even aware of ANY successful double blind tests, after 55 or more years of audio tweakdom. If successful results have occured a few times, in all that time, they have not been replicated or published in the audio engineering literature. Yet they continue to be promoted, sometimes on very bogus grounds. Thus, Peter Aczell's "No More Mr. Nice Guy."
Many audio "subjectivists" or near-subjectivists declare that they believe double blind testing is an invalid approach, and so they aren't interested in performing them. Even for those which believe in the utility of such tests, they are very very hard to do rigorously.
To his credit, George does not dismiss and even actively promotes the use of double blind testing. But does he use them himself? To his credit, he has tried. On two separate occasions done my best to perform two double blind tests with George as the subject, on tweaks he insisted had such great impact he could not err in identifying it. In both cases, my interpretation of the results (and I have quite a bit of experience and even some qualifications in the use of statistical methods) is that they showed no audible significance, and with near-random results to boot. George's interpretation of the first test we performed in the 1980's was different, but he agreed about the result of the test we did around 2005. Though rather than change his thinking, he mainly seems to have forgotten about these tests. Another common approach is to always find fault with such results. These kinds of responses fit the pattern of confirmation bias.
I well understand that the failure to show statistical significance in any one tests does not prove anything. The test could be flawed, the listener not sufficiently trained, etc. This is even more true than in the case of a single "good" result. What is most damning about audio tweaks is not the failure to find benefit in particular tests, but the failure to find benefits after decades of audio tweakdom. And also in light of the wonderously huge benefits claimed by tweak sellers and fans. If a benefit is indeed truly huge, as often claimed, it should be easily proven in double blind tests.
My ultimate feeling about the "listen for yourself" doctrine is that I don't have enough time to try everything claimed to improve audio performance. I don't even have enough time to try all the things based on well established audio engineering, so why should I waste a lot of time pursuing voodoo tweaks? Even many tweak fans admit that most tweaks are useless (the countless ones they have not chosen to use or had later abandoned).
Furthermore, I fear that such tweaks, whose method of operation or measureable effect is usually unknown, it could actually be making matters worse. And yet, at the same time, just as a matter of chance and the variability of auditory perception in listening to music, it might occasionally seem to make things sound better. If such a tweak is chosen, it may stay in place for years until later found to be sonically objectionable, after years of bad effect and badly influencing other audio tests. When this is found to have happened, an upheaval results.
Lacking both measurements and scientific validation through double blind testing, audio tweakophiles are hunting in the dark with no compass. At great effort, in principle (if not merely 45 years of experience) they might find some useful enhancements. Some fine day. And just as often, steps may be taken backwards.
I have tried a few voodoo tweaks myself, and I have often let George and others demonstrate them for me. At times, I have experienced perceived improvements, but none so huge as to make me interested in going down the tweakophile path given the issues and risks described above.
While I don't go for the kinds of tweaks that have no engineering explanation, I do try to use good equipment--low noise, low distortion and wide flat frequency response, high intrinsic linearity, possibly better than necessary based on the limits of human perception. I seek out low dielectric absorption capacitors, low capacitance or low inductance wiring (the former for interconnects, the latter for speaker cable, there's engineering wisdom behind that). While I can't or haven't proven these to be necessary, their method of operation is known, and they can be measured and/or categorized. For example, teflon capacitors are the best because they have by far the lowest dielectric absorption. So with tweaks like these, there is a visible path forwards. More Teflon! Mistakes may be made, but not so much upheaval.
Many might find highly questionable the way I buy most equipment unheard, based on reviews and measurements (and my own thinking about them, of course). But most often, I have not been disappointed for a long time afterwards, if ever. I think I would not do as well following the "listen for yourself" doctrine. Most often in such listen-for-yourself situations, I have found very shortly afterwards I made the wrong choice. Short listening tests are a very unreliable way of moving forwards. And that variability leaves a wide opening for voodoo tweaks.
On any given listening session, a tweak that does nothing at all might have a 50% chance of sounding like an improvement, perhaps more if the test is non-blind and one has faith in the people endorsing it. And if it doesn't sound like an improvement, a seller has another chance by going back to a A after doing A and B. So it turns out that short listening tests are more often than not a way to sell nonsense, like voodoo tweaks, and not really a way to critically evaluate equipment. Critical evaluation requires long term listening, measurement, technical and scientific analysis, comparative evaluation, more listening, and ultimately blind testing. I can't claim to have done much of this critical evaluation. But I try to move in that general direction, and not waste much time with short listening tests to hear the value of voodoo tweaks.
