Peter Walker (from Quad Electroacoustics, apparently) when he did the double blind test with “Golden Ears” who denied that solid state amps sounded ad good as tubes, did something naughty the Golden Eared crowd never forgave him for.
He used his fine Quad II tube amps, the solid sate 303 and the solid state 405. He then told them he had a new experimental amp. Peter had resigned the amp to add 2% THD across the board. The Golden Ears doubled blinded could not distinguish any of the amps.
In Kessler’s book the Closest Approach, he is asked why bother to make amps with 0.1% or less THD. He said, because it does not cost any more to make an amp with 0.1% THD, and he could not sell an amp with 2% THD!
This is a fantastic story. Really says a lot about our senses and our subconscious mind…
Anyone who says "I trust my ears/senses/memory, "or “I know what I heard/saw” is absolutely denying their own humanity. Over and over again our senses have been shown to be subjicated to the power of our mind.
Hence blinded testing to verify. Double blinded testing had been the only legitimate method for medical research for decades. I don’t know why audiophiles think their minds work differently and better than medical researchers or doctors.
Wasn’t there a movie about wine and blinded testing?
OTOH if you don’t understand very well what hypothesis you’re testing, what each testing method can and can’t prove etc., you can just as easily get fooled by tests. And that tends to be even worse, because then you’re running around saying “I’ve got science on my side” and refusing to hear any counter-arguments, when you really don’t have science on your side, you’ve just misused or misunderstood the science. Kinda like the “vaccines cause autism” crowd and their bogus and eventually discredited Wakefield study. Or kinda like vegans. (Had to say it, sorry.)
For example, in the scenario above I have to wonder how wide a variety of music they used in that blind test, because music masks distortion and different music can have different masking properties. Or even more to the point: did they use specially chosen “worst offender” tracks that are especially likely to reveal distortion when it’s present? And then I could go on to comment that just because 2% “across the board” THD wasn’t audible in a blind test it doesn’t mean you can take that exact number and apply it directly to the specs of any amp you’re looking to buy - random amps can have various distortion profiles, some more audible, some less. It’s not always linear.
Objectivism is hard. There’s always that one more variable you haven’t thought of that can invalidate most or all of your conclusions. That’s why science can only be done by multiple independent teams, checking and re-checking eachother’s work.
Couldn’t agree more. You lay out the need for rigor very clearly. In the world of HiFi audio appreciation a multi site double blinded randomized placebo controlled trial is unlikely, though.
You must agree that for our purposes even the most basic blinded listening test is better than plugging headphones from one amp to another and saying “X sounds better than Y.”
True, true, but of course in the car scenario it might be just as often the sound of rattling windows or door components rather than the speakers’ THD.
But yeah, looks like a good confirmation of why NwAvGuy was recommending 0.05% (-66 dBr) as the threshold for THD and IMD.
(Had to say it, sorry.)
