Weltanschaung and HiFi (warning, contentious and wordy)

Joe - It would be proof that an effect exists in the absence of a theroretical explanation which would then prompt the search for a theory to underpin it. As say quantifying "different or not" is much easier than "how much different and in what ways". In your example you can of course run a test to determine a hypothesis on the propeortion of listeners that will prefer one piece over the other. Cables may have the worth that people attach to them - that is the definition of worth after all - but it does not mean that worth is based in actual value.

3d - science (and abx / db) is based on hypothesis, experiment, repeatability and falsification. Therefore it is not a cult. You seem to be something of a romantic! No its relevent. Pro-cable makes claims apparently based on performance but which are not based in fact. Therefore it is valid to look at the real reason the performance claims are made as an alternative to those claims being based in fact. Yes SOME people perceive a difference (but then SOME people also respond to placebo) but its anaedotal - when test conditions are imposed and psychological effects removed then tthe effects vanish. In all of science only cable claims seem to be exempt from normal scientific method that everyone else accepts. Just like homeopathy - they use the same arguments, as do paranormal investigators. There is a pattern - when rational enquiry fidns no evidence then invoke some mysterious effect to create a fictitious loophole. Your proof of god argument is very revealing, as you lump it together with cbales then it reveals that both are a matter of faith. Which is of course beyond verification and not scientific.

It only passes if you find scientifically rigourous evidence - it doesnt pass by default thats the point . Where is the evidence?

If you make cable claims as beleif then fine - but if you claim it as science then you are subject to scientific scrutiny - in good grace without accusing anyone of heinous injustice. This thread is afterall about exactly this debate so claiming the high ground on being attacked is hardly consistent.

I have contended no such thing

 

There is no such thing as 'actual value'.

 

substitute efficacy if you want to be picky

 

Hi,

There is non either way (as the currently extant testing is sufficiently questionable to allow outright dismissal).

In such cases I suggest to people to make up their own minds.

I am perfectly happy with you saying "I will not buy your cable (or Amplifier, CD Player et al) unless you give me proof of a rigerous DBT showing it to be different to another cable (or Amplifier, CD Player et al)."

I am equally happy with you saying: "I auditioned your cable in my system and I it sounds different and I like/Do not Like (delete as apropriate) the way it sounds different and I like/Do not Like (delete as apropriate) to buy this cable but it's too expensive."

In fact I do not give a flying french coitus what people believe about cables (or indeed any other gear) and if they buy any or not.

What does interest me is the philosophical aspect of this and the human dynamics. As for being a romantic, I used to be, I'm more of a "professional cynic but my heart's not in it" type now. Hence I am not only cynical about the Pro-Cable crowd but also about the ABX Cult.

Ciao T

 

Is one tedious thread about cables not enough guys?

 

Such an attitude is commendable! Be skeptical about everything - especially practical abx (though in theory its fine practice oftens falls short!). There is some non-questionable evidence around - the fact noone has claimed that £1M for a start . even fermats last theorem got caught by that!

Fine but dont call it science call it faith.

 

ahem yes how did this get distracted from the original topic... thorsten?

 

This one has me in it though, so the tone has been raised considerably.

 

Hi,

I think it is still on topic.

The point of this thread was to highlight how much the Worldview people hold influences what they hear, how they argue etc., to introduce such niceties as moral affordability and so on.

Ciao T

 

Well, quite. Indeed, the selfsame arguments about perceived differences in quality apply to any piece of hifi equipment; it's just that cables polarise opinions more radically than anything else (and, unless I've missed it, no-one's offering £££s for successful ABX-ing of amplifiers or speakers).

 

Yes but it didnt have to get onto cables to do that

Well speaker abx has been done and surprise surprise it shows significant subjective differences for speakers - regardless of all the issues T has raised which are just as applicable to speakers as cables. So why not for cables?

 

You've offered no argument to support this claim other than suggesting that a biased single blind non-test would produce non-results.

Paul

 

Hi,

ABX with speakers has shown a lot of interesting things, including the fact that under anechonic conditions speakers that measure reasonably flat on axis appear indistinguishable from real music (something I will also dispute as poor experimental conditions BTW).

Now we need to note that with speakers response differences of +/-3db between different models, on axis, in the midband are not uncommon. Distortion in the 0.5%-1% range at normal levels is not uncommon either, By comparison the differences in cables are a decade or two lower in magnitude at least. this suggests a much lower audibility and for any otherwise identical test (including sample size) a much lower audibility (or value for .p in our case).

That is BTW WITHOUT accounting for the added randomising factors from expectation....

But you know this of course, so one would question what particular reason you have for bringing this up, as again it has a connection in the region of asking if the designated hitter rule is more or less green than the fatness of the pig.

Ciao T

 

Hi,

I referred to biased single or small sample blind tests.

And guess what, ALL the "cable" ABX etc. Tests I am aware of (which are certainly all the most commonly quoted ones) where single/small sample size, biased (eg a "challenge" envoironment in most cases) and returned of course, as they should, null results.

So I dismiss them as being useless to answer the question:

"How confident can we be that a difference was or was not heard!"

Where is your problem with that?

Should accept useless, biased and fatally flawed data simply because this data happens to suit your Weltanschaung?

Past that, I think I have given you the rest of your answer in the other (cable) thread and promptly, choosing to wear the cap that fits, you choose to insult me, merely confirming and illustrating my point.

L8er T

 

3d - thats somewhat fanciful. People can discern about half a db on single tones and even less broadband. I dont see that you have answered anything I wrote. Green pigs or otherwise not withstanding . This randomising factors issue is a misrepresentation of statistical arguments, as detailed before. All cable tests as you described should return null results? Thats a pretty bold statement and one that appears to be something of a self fulfilling prophecy. Its not science - if the test returns a postive result you are vindicated and if its negative then thats to be expected and therefore dismissed. I hate to tell you this but you dont appear to be wearing any clothes...

 

Hi,

Sure. So what?

Nope. If you have a strong believe on the item under test I can demonstrate to you (but only - sadly - if you are not prewarned like you are now) that testing blind a readily and drastic audible difference stimulus becomes inaudible, as long as you have sufficient strong belief on the outcome.

Let us for now give drastically different cables (such as RG-214/Mil vs. RG-58 steel core) a value of .p of .6, that is the difference would be correctly identified in 60 out of 100 trials.

This is arbitrary but likely generous.

What is the chance that such a definite but weak difference stimulus would NOT be identified in a test demanding 12 correct identifications in 16 (that is significance of .05)? The answer is .83.

In other words, out of 100 such trials conducted under ideal conditions only 17 would return a result of "different" and 83 would return "same" Now this is of course statistical and in reality we are even less likely to see any correct identification.

Forgive me, but if we then factor aural/mental fatigue over all these trials and expectations we can conclude that that the likelyhood of the test returing a "null" result DESPITE the existence of a definite and real stimulus approaches for all intent and purpose near absolute certainty.

We could calculate this more and we would likely find that the chances of "passing" the SP £ 1,000 challenge are better than winning the lottery, but the degree of difference would most likely fall under the term "academic".

It is not a prophecy, it is merley an observation of fact, an observations of the methodology employed and the results obtained.

It is, more-so than these tests we are discussing at any extent. :JOEL:

I hate to tell you, but in Germany we do not have this morbid fear of showing our naked body, it is called "Freikoerperkultur" (FreeBodyCulture).

But worse, I may be comfortably naked, but you don't have a leg to stand on....

Ciao T

 

By definition an ABX isn't a 'challenge' environment. I don't know what you mean by 'sample size', it only takes one person who can hear the difference to show they can hear the difference to be valid.

But the data is only 'useless, biased and fatally flawed' in your 'Weltanshaung'.

Paul

 

You appear to have calculated that using cables that sound the same will make it difficult to pass the Pinkerton test. Which is as it should be.

Paul

 

Mr T,

You feel your dignity may be intact but if you beleive you are wearing a fine robe and swagger thusly then you will be foolish despite your personal comfort .

First of all you dont need to tell the listeners they are listening to cable changes. Thats an example of removing bias. That being the case almost all of your argument evaporates. The rest fizzles to ash under statistical examination. As I pointed out earlier you match the number of tests to give the desired degree of confidence from the observed variance and also the magnitude of the effect if it exists. In addition you can compare 10 cables instead of 2 and examine the ranking structure which gives you additional constraints (effectively by compounding probabilities) that give you a higher degree of confidence from a lower number of overall tests and listeners. By constructing an arbitrary scenario for a simple significance test you can of course choose parameters in the assessment that match any purpose - you are reverse engineering for an equation that encompasses every scenario. With all due respect your statistical understanding isnt in the same league as your knowledge of audio engineering - which I am certain is without peer on this forum. Consequently your arguments dont hold water, though I am still enjoying this immensely. Clearly you are too. Have you fought with windmills recently?

Dont fall into the trap of premature congratulation

N.

 

Dear Nick,

No I tell you. Freikoerperkultur. It's intentional. Guess why. Try pick a naked mans pocket!

That is how I did my own cable tests. And it the polar opposite of what is done in all the published tests normally referenced. Hence one of my criticisms of said tests.

Remove Bias, leave p @ .6 and you still have .83 against. Add listener fatigue at 16 or 20 Trials and you have .95+ (estimated) against.

Dear Nick, this indeed I DO. The ABX cult does not, however and for good reasons. And it is not as if all these criticisms are unknown to them, on the contrary. Instead they deliberatly rejected these improvements to their tests, claiming that "they would compromise the significance", in other words and translated to english, would result in outcomes undesirable to them.

I have done that when I did my blind test that formed part of the data for the 300B grouptest I published a while back in Enjoythemusic.com.

As said, I AM quite aware of the tests, the menthodology and how to maximise the results, as well as the limitations of them.

However I level this exact criticism against the perpetrators of the ABX and DB Tests commonly quoted, referrred and widely published. In fact, I am UNAWARE of any DB or ABX test that actually cannot be subjected to the criticisms I leveleed and improved by applying the suggestions you repeat and which where made to said individuals as much as 20+ Years ago!

I will readily admit that. I merely took stats as added module. I understand enough however to know BAD applications, especially if they are as bad as wat I object to.

Hmm, let me think. Human Resources. Temp Bank. IT Depatrment. Yup, enough for today.

Look, if I subtract from the sum total of all DB and ABX Tests formally conducted and documented and published within my ken those to which we can apply substantial, drastic and at leats potentially falsifying criticism the remainder gives the same result as the tests usually do, namely NULL.

Care to answer THAT charge?

I do not dispute that well implemented DB Tests can give us answers.

I merely observe that no-one has done any, on the subject of cable differences, to date, that I aware of.

Ciao T