You are reading the older HTML site

Positive Feedback ISSUE 30
march/april 2007


Myth, Mirth or Magic? - The Peter Belt 'Snake Oil' Fallacy
by May Belt

You know the kind of thing. A new audio tweak surfaces and minutes later, the internet hums with cries of 'Snake Oil' from its detractors, often claiming deep knowledge of physics. Directional cables, fuses, equipment supports, Shakti Stones, or whatever—we've heard it many times.

No one has suffered longer than qualified audio engineer (and my dear spouse), Peter Belt of PWB Electronics in Leeds UK. For more than forty years, he has been saying simple and demonstrable things about audio improvement but still he's howled down. Year after year, day in, day out, new generations of self-styled experts demand 'proper evidence' for his claims, usually in terms of 'double-blind' testing. It's tedious of course, but it goes with the terrain.

Sometimes things get nastier however. Between 2005 and late 2006, a section in the news group in Stereophile called Why do so many buy into the "cons" in high-end Audio (link) ran an entry from 'Buddha' who actually wrote this:

The Peter Belts of this world are sly, like the serpent. They are driven off, but then always find ways to slither back into to hobby to suck the green life blood from the uninitiated…. We, as ethical audiophiles, have a duty to remember the past, so that others aren't doomed to repeat it.

Buddha’s comments are confirmation (if ever confirmation was required) of the intolerance to new ideas described by Dave Clark in his article

"Audio Ramblings - Faith and Belief in Audio" see: audioramblings

You know, stuff that defies Science and rattles the people who refuse to listen (let alone think) with an open mind. You know their mantra, "Ha! If you believe that pile of crap, do I have something to sell you! Hoax! Snake oil! Dupe! Scam artist! Rip-off! Charlatan!”

Leaving aside the thought that implying that Peter Belt 'sucks the green life blood from the uninitiated' might well be defamatory under English law, let's jog 'Buddha's' memory by recalling what Peter Belt actually says. His claims have always been these:

1. There is much more information in most recorded material than often seems to be retrieved. Even modest equipment can perform far better when the causes of its apparent limitations are identified.

2. The failure of accurate retrieval seems likely to be due to influences other than electronic engineering.

3. Increased accuracy of retrieval can be obtained by manipulating aspects of the listening environment, some of which seem unlikely according to standard audio engineering practices.

4. Finding a relevant theoretical base to explain these phenomena is particularly difficult so that even partial explanations require innovative thinking, some of which seems strange to the audio community.

Contention No 1 is clearly a problem to the hi-fi industry. Its solution is to make ever more costly equipment which accounts for the ever-expanding 'high-end' market. Though contentions 2 and 3 are easily demonstrated up to a point—there are obvious physical properties of buildings and environments that make differences to audio, too many in fact to list in a short article—these aspects of audio design have never been investigated systematically or with any rigor.

We regard contention No. 3 as extremely important and I would like to expand further on what our thinking is in the area of manipulating the listening environment to improve sound. We believe, simply but firmly that doing this, demonstrates something very important indeed: that there must have been a wealth of information already in the room, which the listener had not been resolving correctly.

When someone does something to a room, and gains an improvement in the sound, this means that they are hearing additional information. Since it is only the room that has been treated, then logically this means the additional information they hear must have been not only on the disc but also that it had already been retrieved from the disc, having been 'handled' perfectly adequately by the audio equipment and presented into the room via the loudspeakers. The problem of incorrect resolution is with the listener affected by something in the room, rather than with the equipment.

We believe that the audio industry often looks in the wrong places to develop  continuous improvements to audio equipment. The more that people report doing something to their rooms and gaining improvements in their sound, the more seriously the audio industry should be taking notice. Our contention is most audio equipment is (and has always been) perfectly capable of 'handling' far more information and presenting it into the room than most people realize. 

This way of looking at things does not sit comfortably with ‘traditional audio magazines’, particularly those in which new equipment is advertised and reviewed. The idea is unwelcome because it means that the perpetual urge to upgrade domestic equipment is less important than the audio trade would have magazine readers believe.

What Peter Belt has shown (to his own skeptical satisfaction and that of many others in professional audio and journalism) is that apparently very unlikely aspects of both equipment design and environment do turn out to interfere with information retrieval from recorded and broadcast sources. Two elements of evidence reinforce this proposition: first we have a very extensive list of published articles saying that the improvements we claim for relatively cheap equipment can be demonstrated and secondly, all of the methods we use to clear up information retrieval problems are wholly reversible. If something we do produces an improvement in perceived sound, we can always return to the previous 'untreated' conditions.

Dave Clark’s article referred to earlier, describes his experiences beautifully.

“First time I experienced Belt treatments. Had no clue what Carol was up to as I was in the other room, though I did know she was playing around with something audio. She asked me to come in and give a listen, but with the following constraints: 1) I had to play the same track three times in a row, and 2) I could not open my eyes during the session. Easy as pie! Out came my favorite Lambchop disc What Another Man Spoils and on went track 1. Hit play and… "Sounds fine Carol, sounds like it usually does; great recording, fun music, love it!" She hits Stop and after a few seconds, she hits Play again, and like "WOW, this sounds much better! Way more musical and golly-gosh-gee, this is really cool!" She hits Stop and a minute later, she hits Play again, and shit! I mean SHIT, this sounds like shit! Not the shit, but shit. Not more than 5 seconds into the track and STOP IT NOW BEFORE I TEAR MY EARS OFF! What did you do? "Well, the first time nothing, the second time I wrote Reimer > O.K! with this red pen on a piece of tape and placed one on the top of each speaker, and the third time I replaced the tape with another that said… Reimer > BAD." Huh, sounded as it usually did with nothing on the speaker, sounded really great when you wrote O.K, and sound really bad when you wrote Bad. Weird.

No implicit beliefs or faith on my part as I did not consciously know that Belt (device) was being used in the room, yet it worked anyhow. On the other hand, I did find much of what Belt was suggesting to make sense from earlier readings—the issues with us being at odds with the world, the stress, and all that it brings into our lives.

Problems in theorising about product Development

Forty odd years is a long time in any business and Peter Belt's experiments have been continuous. So contention 4 (above) is also self-evidently true for him. At PWB Electronics, we now have upwards of a hundred 'devices' all of which are capable of improving the perceived quality of sound from even high-end audio and video equipment. In the past, Peter Belt has manufactured audio equipment himself but does not do so now, simply because of commercial viability: high production costs and intense competition make life difficult for the small manufacturer.

But the real challenge in our work is presented by stumbling on repeatable and reversible effects from our products and then having to explain why and how they happen. Orthodox deductive reasoning based on Newtonian physics and electronic engineering practice fails to generate hypotheses for our discoveries and we rely on a different approach. We reason inductively instead, asking the question 'What is happening when we try this?' Then we perform experiments, notice patterns, form tentative hypotheses about what's going on and look for explanatory theory. It's thinking 'outside the box'—a perfectly valid method for exploring new ground, though hardly amenable to either 'sound bite' discussion or to fully 'accepted' theory.

At PWB Electronics we are completely sure about the following:

  • Our products and ideas can improve perceived performance from all grades of audio-visual equipment.

  • The effects are sufficiently great to convince significant numbers of people.

  • The effects are independent of 'belief' in the products or personal self-deception.

  • We have never 'conned' anyone, ever.

'Joe E' replying in the same Stereophile news group to the singularly uncompassionate 'Buddha', comments that:

It still comes to the fact that my buddies and many other listeners hear differences that strict objectivists say should not be there. If you, and at least one other contributor here can't hear those differences then don't spend your money on what you can't hear. Just don't get indignant when we report hearing things you can't or won't. If we spend money on expensive wire or cables or wood blocks just remember it's not your money. This is a hobby for most of us. We are allowed to have fun any way we want.

PS: You could be wrong about what you think. Maybe there is no "con."

We agree with 'Joe E' wholeheartedly, which is why (though our critics rarely mention this) we offer a money back guarantee to customers dissatisfied with their purchases. 'Buddha' knows us better of course and feels able to write (after I had pointed out to him that although he had the luxury of ‘freedom of speech’, with that luxury also came responsibilities!):

Ms/Mrs Belt, you really do piss me off with your righteous indignation.

I take my First Ammendment [sic] responsibilities seriously.

I can only wish that you and Peter belt did the same.

Quite extraordinary, some might think.

Helping PWB Develop

We are as interested as anyone else in finding a comprehensive theory to explain our discoveries and will happily respond to ideas that help us think harder about why our products work. We are equally interested in finding out more about instances where our products fail to deliver what we say they do. No cost need be involved, since many of our suggested 'treatments' are free or, alternatively, we will supply a sample of our Rainbow Foil free for anyone to experiment with.

All we ask in return, is that failing experiments are reported to us carefully, and yes, controlled conditions—'double-blind testing' and statistical data for example—would be useful in these circumstances. We have often done these ourselves (though 'Buddha' and others like him, may happily believe otherwise) and we are certainly concerned about accuracy in our reporting.

There is a list of experiments available for anyone who is "up for the challenge'!