Demagnetising LP's & cd's

ISTM that the detractors of blind testing do so largely by pointing out that the testing is never 100% perfect, and cannot account for every eventuality.

Nothing is perfect. All blind testing does is remove most of the obvious sources of potential bias and barriers to allow a better assessment to be made.

It has absolutely nothing to do with preference - only identifying differences.

 

Sorry but it just introduces a different set of barriers, plus I don't believe your initial set of barriers exist, you have to be pretty bloody daft to allow those other factors to influence what your ears hear. But you have to train yourself with experience to adjust to the barriers that are created for newbys to blind testing. I saw it over and over again in nearly 10 years of it on Hi-Fi Choice panels, plus when there was the last fad for it in the mid late 70's. Those that set such store on blind testing as the arbetter of truth are just plain naive.

Just because your perception is shut down doesn't mean there is no difference, but the people who experience it are *conned* into believing that is the case, so they become part of the vocal crowd who try to push the concept onto everyone else.

 

That is a very "arm-waving" rationale, Richard.

What could be simpler? you listen to some kit while a mate swaps the cables/majiked CD/whatever a few times. If you accuratel discern a difference, there is a good chance the difference is real. If you don't, the oppsite applies.

Chris

 

Richard,

Why do you think the medical profession insists on blind testing for drug trials?

 

It is far simpler just to listen to it, the rest is just pointless unnecessary complication.

 

Because it is different.

Basically it has nothing to do with listening to music as far as I know.

 

The medical profession insists on moderated, randomised double-blind, placebo-controlled trials, conducted with suitably large samples. This is in order to minimise the effect of false positive and false negative errors.

In most cases, the DBTs conducted in audio could have been constructed to introduce error, rather than eliminate it. Their sample sizes are an order of magnitude or two too small to be anything else. The "it's only audio" argument doesn't hold water, either. Unfortunately, statistics doesn't really care whether it's being used to save lives or test whether things sound different - the number-crunching remains the same.

 

They can be proven to exist very easily.

You yourself even admit this - all that Linn/Naim brainwashing back in the 80s and 90s - many of the same techniques used.

 

Still better than the alternative, and it can be shown time and time again that people make judgments based on what they are told, see or expect.

I can and have made people believe some very strange things with a little suggestion here and there. Put someone in front of a salesman and anything is possible.

 

That was different, that was people not being *allowed* to compare products on a level playing field, or the retailer controlling the dem in terms of set up.

The curious thing was that given a level playing field and an honest retailer other competitive products sold in very acceptable if not greater quantities, but the retailer then could lose his Linn / Naim agencies.

There was a need to stop the industry corruption not making the customer blind.

 

This is still outside of my experience unless you went to a corrupt or agency protecting retailer. *Never* in my experience has anyone including myself made a decision to purchase not made on the music, unless they were a complete newby wishing to be led. But even they when given the confidence and told their ears are as good as anyone elses will still take the better (for them) musical performance.

 

I disagree that it's better than the alternative, because instead of relying on observation, you are simply rolling the dice. That die is loaded heavily to produce the desired result, which is why Hi-Fi Choice's blind tests produce identifiable differences between CD players and cables and yours don't.

There really shouldn't be any dichotomy here. Qualitative and quantitive data go comfortably together in most scientific disciplines, and that's really what we are talking about here, before everyone started name calling. I just find the idea of promoting one data-gathering system and negating another on principle just so 19th Century.

 

Not so.

If I have three listeners sitting here and I play them two amplifiers, for the sake of argument a Cyrus and Naim integrated, you can be certain of two things.
The Cyrus will be described as crisp, rather lightweight and a little brittle, but with good clarity. Put the Naim in the system and the comments will shift towards things like 'communicates better' but sounds rather two dimensional.

Its all bollocks.

Neither amplifier has sufficient difference in the design to cause the above observations. They will spec and measure differently, but by tiny amounts and well below the threshold of detectability for the vast majority of people.

You can demonstrate this by repeating the test with the identity of the amplifiers hidden, and the same listeners will now not be able to form the same strong opinion as to any difference. Or if you want to be really cheeky, power both up, have them on show but swap the connection so that when the Cyrus is playing, people assume it is the Naim........ they hear the Naim.

If you want to test blind testing as a test regime it isn't difficult.
Compare the Cyrus to a Prima Luna - you will get the listeners identifying clear differences under the blind condition.

Why?
Well, because the difference is now of sufficient magnitude to be consistently audible, and more importantly, to matter.

You are correct, it is possible to assemble a system in order to create differences under blind conditions, or not reveal them, but that implies that the tester has deliberately set out to deliver a result. Doesn't mean this has to be the norm.
On cables, pick a source or pre amp with a tube output stage, or use a passive system in which source and load impedances have not been carefully considered and you can get cables to sound different under blind test conditions. There is no mystery to this and it boils down to basic electronics and the effects of filters formed at various parts along the chain.

Similarly you can select a speaker system to stress amplifiers, and thereby magnify differences but these will be largely limited the effects of output impedance and current delivery.

Everything hinges on one question - what are you testing for?
If you are testing for the sonic differences between silver and copper conductors, grain structure, enclosure materials, different types of capacitors or resistors, one brand of tube against another of the same spec etc, then the system used for the test cannot skew the result because it doe not alter any of the above. They are entirely independent of system matching considerations.

So again it all comes back to what causes differences to be audible, and in my view there is always a sensible, rational and technical explanation.

 

There should be an absolutely obvious answer to this, that by placing the listener under blind conditions it robs him of some of his perceptive faculties. I could do the same to you under a situation of physical conflict as I am trained in it.

The human energetic systems are linked, remove one and you dull the others (our subject) control one and you disable the others (my subject).

As I have said I have seen it over and over again, even with designers, who when placed in a position of stress and threat to their perceived greatness make idiots of themselves and don't even recognise their own product. The eyes are part of our perception, if you want to test blind then get blind men to do it as they have learned to compensate.

 

How can dulling a sense, in this case the eyes, possibly dull the hearing ability?

If anything it will make the hearing more acute.

I could understand how at a concert the senses would be linked, you are looking at the musicians, the venue, the crowd etc and reacting to that along with the sound. That is an entirely positive link.

At home you sit and look at two speaker enclosure and a rack of kit.
It neither moves nor reacts in any visual way. The only visual influence here is a potentially negative one.

The reason you've seen designers make idiots of themselves under blind conditions and be unable to identify their kit is simple. The equipment has no identifiable characteristics!

 

No, it implies that (at least) one set of test results is returning false results, you just presume that it's not you. It also suggests that if the test protocol is so easily undermined by the tester, it is just as invalid as a sighted test in a room full of high-pressure sales people from a scientific standing.

Except that, some of the HFC tests I mention did precisely none of these things and still presented differences. The test system remained boringly constant for years, with an active solid-state preamp and power amplifier, with very carefully considered input and output impedances and a benign loudspeaker load, chosen because everything from a 2W SET to a 1kW solid-state monoblock would be put through the same test system.

Even the music and volume level for every product under test remained the same for almost a 10 year period.

And no, the results didn't fall into 'this silver cable sounds bright and silvery' or 'Naim=Naim sound' results. But they did get results that you claim are impossible to achieve. How?

 

There is absolutely no evidence for this myth. The brain doesn't work that way. There's a possibility that unformed cognitive pathways in people blind from birth would be diverted to different brain functions, but the visual cortex and primary auditory cortex are separated by a lot of brain mass.

Including loudspeaker designers listening to their own designs? Surely these have recognisable and measurable characteristics? You'd have thought a designer would be able to spot his own design in a line-up. And yet, there's a lot of evidence of designers failing to identify their own products in a blind group. Even in a group of wildly different designs.

 

It isn't a false result - it is a result that is correct for that particular set of circumstances. If you understand the circumstances you can better interpret and understand the result. The danger comes when the listeners are not kept in the loop.

As for the HFC cable tests, I simply do not accept that you could insert electrically similar cables into the system you describe and get differences under blind conditions. It is not possible, and unless I see this test for myself I will question the procedure and the findings.

I have no such issue with differences being shown under blind testing with amps, dacs, speakers or indeed any electronics, but as I said earlier, there will always be a perfectly rational explanation.

 

It may well be a myth, which is why I said 'if anything'.
You choose to explain away this myth in a very plausible way, explaining the distance between the various parts of the brain and doubting that there could be any link between them. Yet, in audio we are expected to accept many myths, despite similarly convincing argument and evidence to the contrary. I'm afraid you cannot have it both ways!

As for speaker designers not recognising their own designs, again it simply shows that their design is not sufficiently distinctive within the test group.
The situation you describe should only occur where other speakers offer a similar performance or have similar character.

Another explanation is that perhaps their speaker isn't doing quite what they believe and expect. It is very easy to spend time designing something and becoming convinced in the process that what you are achieving is unique, or doing something exceptionally well. In a group, and up against peer products that might not be the case.

Lastly, if a speaker truly is very 'different' and the designer cannot identify this, I'm afraid I'd have to conclude that the designer doesn't understand what he has created, or is deaf to what he has created. Stress and pressures from blind testing come way down the list of probable reasons.

 

You are unlikely to find this now. It was done in the mid-1990s and not published online. It was published in two booklets over successive months, one for digital and analogue interconnects and one for speaker cables.

AFAIK, the procedure was identical to the CD and amp test procedure of the time. Cables were evaluated blind and random samples from one session were reinserted into the test in another session to ensure consistency of results. Cables were also measured for their resistance, capacitance and inductance, and but there were no great correlations between measurement and sound quality, or materials used and sound quality, or price and sound quality. The differences between the cables were not considered 'substantial', but were 'observable' and repeatable.

I distinctly remember a 99p/m Gale speaker cable doing better than basically anything apart from a £30/m Tara Labs cable (including several £100/ft cables), and some £50 interconnect made from bits of old tin helmets and railway sidings from Sonic Link outperforming some extremo-connect from the States that was the first of the gaspingly expensive stuff. Which is probably why nothing of its kind has been repeated.