IPB

Welcome Guest ( Log In | Register )

2 Pages V  < 1 2  
Reply to this topicStart new topic
Need help with a Stereophile article, An ABX debate that I want to understand.
Audible!
post Aug 5 2004, 04:37
Post #26





Group: Members
Posts: 523
Joined: 28-June 03
From: CA, USA
Member No.: 7426



QUOTE (Pio2001)
If the goal is to test modulation cable worth 1000 $ per meter, it rules out the use of the ABX box, unless all its component are themselves of much higher quality that 1000 $/m cable, in order for it to be transparent compared to the tested cable (or plugs, or other devices).


Certainly, assuming the measures of "quality" are in some way individually quantifiable with or without the use of an ABX-like system. Of course, I would be quite interested to discover interconnect that is indeed "worth 1000 $ per meter" rather than interconnect that is merely priced at that point wink.gif

It seems to me that it would be well within the realm of possibility to build such a box using individual components that bench tested as perfectly as possible (in terms of electrical characteristics/tolerances), and then test the assembled result, modifying it if necessary to achieve the absolute best results.

QUOTE (Pio2001)
Testing an expensive cable, we are testing the hypothesis that normal cables and interconnects, like an ABX box, have an effect on the sound.


As long as the hypothesis is falsifiable, of course!

QUOTE (Dunlavy by way of krabapples link)
So why does a reputable company like DAL engage in the design and manufacture of audiophile cables? The answer is simple: since significant measurable differences do exist and because well-known and understood transmission line theory defines optimum relationships between such parameters as cable impedance and the impedance of the load (loudspeaker), the capacitance of an interconnect and the input impedance of the following stage, why not design cables that at least satisfy what theory has to teach? And, since transmission line theory is universally applied, quite successfully, in the design of cables intended for TV, microwave, telephone, and other critical applications requiring peak performance, etc., why not use it in designing cables intended for critical audiophile applications? Hmmm! To say, as some do, that there are factors involved that competent engineers and scientists have yet to identify is utter nonsense and a cover-up for what should be called pure snake oil and buzzard salve - in short, pure fraud. If any cable manufacturer, writer, technician, etc. can identify such an audible design parameter that cannot be measured using available lab equipment or be described by known theory, I can guarantee a nomination for a Nobel Prize.


The bolded section should be read and reread by every subjectivist "audiophile", five times daily. Mr. Dunlavy is absolutely not joking about the Nobel prize.

This post has been edited by Audible!: Aug 5 2004, 04:38
Go to the top of the page
+Quote Post
2Bdecided
post Aug 5 2004, 09:51
Post #27


ReplayGain developer


Group: Developer
Posts: 4945
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



I'm going to read through the whole article again.

However, I have a feeling that these extra statistics (type 2 errors) are simply telling us what we already know - i.e. that the statistics aren't really proving anything. What's more, to use these extra statistics properly, it requires an a-priori knowledge of what we're trying to find out! The false proven statistic (type 1 error) is OK, but to accurately calculate the chance of a false unproven (type 2 error) you have to know the chance of hearing a difference given infinite trials. Well, if we knew that, we wouldn't be ABXing - because if it is more than 50% than that proves there is an audible difference!


So that means we should assume the worst. We should assume the difference is barely audible (i.e. 60% chance of picking it out). However, I think that makes the statistics even less useful. If the chance of hearing a difference is only 60%, then the gap between type 1 and type 2 errors - the region of useful results where you can be fairly certain to avoid both types of errors, is very small. But that's just common sense - if it's very hard to hear a difference, then it'll be very hard to ABX. On average, you won't get a statistically significant result until you've done a very large number of trials.


I'll read it again...

Cheers,
David.
Go to the top of the page
+Quote Post
2Bdecided
post Aug 5 2004, 10:37
Post #28


ReplayGain developer


Group: Developer
Posts: 4945
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



I've read to the end (the print view makes life much easier!). The last letter is very sensible. It makes the point about probabilities better than I did.

Earlier in the debate, John Atkinson says how insensitive ABX tests are, and mentions some stereophile reviewers whose subjective reports we should trust more.

If I were to tell you the widely held beliefs about one of those trust-worthy reviewers, I'm sure I would be sued. All I can say is that many people in the industry choose not to trust him, allegedly with very good reason.


I can repeat an anecdote (probably legend!) about a different (anonymous) reviewer from the 1970s. This was an era where the "big" Japanese manufacturers were producing popular mass market equipment, but this equipment didn't always receive good reviews! Their latest amplifier was driven around to the reviewer's house in a Porsche. A note was attached: "Return the amplifier, keep the Porsche"

Cheers,
David.
Go to the top of the page
+Quote Post
Pio2001
post Aug 5 2004, 12:11
Post #29


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



Maybe he wrote a bad review, returned the amplifier and kept the Porsche smile.gif
Go to the top of the page
+Quote Post
jaustin
post Aug 5 2004, 13:04
Post #30





Group: Members
Posts: 39
Joined: 26-July 04
Member No.: 15785



QUOTE (Audible! @ Aug 4 2004, 10:37 PM)
QUOTE (Dunlavy by way of krabapples link)
To say, as some do, that there are factors involved that competent engineers and scientists have yet to identify is utter nonsense and a cover-up for what should be called pure snake oil and buzzard salve - in short, pure fraud.


The bolded section should be read and reread by every subjectivist "audiophile", five times daily. Mr. Dunlavy is absolutely not joking about the Nobel prize.
*


Good to know we have someone on the board that has pull with the Nobel committee.

As an erstwhile audio writer trained as a physicist, I agree that we probably have a solid theoretical grasp of the basic laws that govern the operation of cables (and to a lesser extent more complex audio devices) but I worry that the quote is being used here to imply more than it should. Of course we know the basic theories that apply to cables, component design, etc. But anyone who has ever done scientific research (and I suspect anyone who has ever done much engineering) knows that it's one thing to understand the basic theories, and quite another to properly apply them. Indeed, much of my research career was spent trying to figure out WHICH theory was the most appropriate for a certain (experimental) context, and how it should be applied. Often careful modeling is necessary. As far as measurement goes--yes, I think we can probably measure everything that's relevant; what's not possible, currently, is to directly relate what we measure to what we hear. These issues are far from trivial.

I would translate this quote, and the message that follows, as follows: "If a subjectivist audiophile's assertions don't correspond to expectations derived from a simple-minded, naive (mis)application of (poorly, by the writer) known laws of physics, then that person is dishonest and trying to rip you off." It's a mistake to slander those who have the good sense to believe more in their own perceptions than they do their own (inadequate) understanding of science. With respect, some people on this board would do well to take a similar approach.

It's true that there's little statistical proof of much of what the subjectivist audiophile camp asserts. Objectivists would do well to be measured in their response and stick to this lack of evidence in rejecting subjectivist's claims. When you get into the realm of theory you're on much less solid ground. I have never met an engineer who's insight into the laws of physics was deep enough to effectively rebut any but the most obviously dubious claims of the audio subjectivists. Furthermore, most experienced engineers know that much design--including electronic design--is based on phenomenology, convention, or tradition, and not on the direct application of theory. Theoretical understanding often lags engineering wisdom. That's not to say that there's any Nobel-inducing radical new theory involved. There's plenty of work to be done in figuring out the proper way to apply existing theories.

Jim
Go to the top of the page
+Quote Post
jaustin
post Aug 5 2004, 13:08
Post #31





Group: Members
Posts: 39
Joined: 26-July 04
Member No.: 15785



QUOTE (2Bdecided @ Aug 5 2004, 04:37 AM)
If I were to tell you the widely held beliefs about one of those trust-worthy reviewers, I'm sure I would be sued.
*

I think you're wrong about getting sued--they hear this stuff all the time, and are quite used to it, though I'm sure that doesn't take the sting out completely. Still, I trust that people who are sketpical and scientific enough to reject subjectivist claims made without statistical evidence will also be skeptical enough to reject "widely held beliefs" and urban legends presented without evidence.

Jim
Go to the top of the page
+Quote Post
2Bdecided
post Aug 5 2004, 14:36
Post #32


ReplayGain developer


Group: Developer
Posts: 4945
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



QUOTE (jaustin @ Aug 5 2004, 12:08 PM)
QUOTE (2Bdecided @ Aug 5 2004, 04:37 AM)
If I were to tell you the widely held beliefs about one of those trust-worthy reviewers, I'm sure I would be sued.
*

I think you're wrong about getting sued--they hear this stuff all the time, and are quite used to it, though I'm sure that doesn't take the sting out completely. Still, I trust that people who are sceptical and scientific enough to reject subjectivist claims made without statistical evidence will also be sceptical enough to reject "widely held beliefs" and urban legends presented without evidence.
*



I labelled the Porsche anecdote as a legend. I didn't label what I referred to in the earlier paragraph as such, for several very good reasons.

Cheers,
David.
Go to the top of the page
+Quote Post
2Bdecided
post Aug 5 2004, 14:41
Post #33


ReplayGain developer


Group: Developer
Posts: 4945
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



jaustin,

You're right about the misapplication of theory (a little knowledge is a dangerous thing etc etc).

However, are you of the opinion that, if two devices display identical measured performance, then they must display identical subjective performance?

Or are you in the camp that believes there are differences that we can hear, but not measure?

I realise I'm making a grey issue black and white, but take it at face value: are there audible differences that are due to things that we cannot measure?

Cheers,
David.
Go to the top of the page
+Quote Post
jaustin
post Aug 5 2004, 15:59
Post #34





Group: Members
Posts: 39
Joined: 26-July 04
Member No.: 15785



QUOTE (2Bdecided @ Aug 5 2004, 08:41 AM)
jaustin,
However, are you of the opinion that, if two devices display identical measured performance, then they must display identical subjective performance?

Or are you in the camp that believes there are differences that we can hear, but not measure?
*

It depends on whether you make the right measurements, and if anybody knows what the right measurements are, I haven't heard about it.

I believe that there are audible things that we do not yet know how to measure--or if we know how, that no one has yet made those measurements. Establishing a correspondence between perception and measurement isn't trivial. As an example, I think that we cannot rule out the possibility that particular distortion "signatures" are responsible for perceived differences between amplifiers and other audio equipment (including, eg, output stages of CD players). But the problem of relating a particular distortion signature to perception is very difficult because it's very hard to identify specific aspects of the sound that relate to those subjective perceptions. And, yes, whenever this is the case--that perceptions are difficult to specifically identify, it is appropriate to be skeptical--but not closed-minded--about whether those perceptions are real. (ABX testing is desirable under those circumstances, but can be difficult to perform with real-world equipment). Limiting yourself to statistically valid data is a legitimate, even essential, for many purposes, strategy, but in the interest of clarity and certainty such an approach limits the information that's considered.

Any statisticians out there? What are the chances of replacing the binary ABX test with a Bayesian, inductive approach that would be just as valid, statistically?

Cheers,
Jim

This post has been edited by jaustin: Aug 5 2004, 15:59
Go to the top of the page
+Quote Post
ff123
post Aug 5 2004, 16:25
Post #35


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



QUOTE (jaustin @ Aug 5 2004, 06:59 AM)
Any statisticians out there? What are the chances of replacing the binary ABX test with a Bayesian, inductive approach that would be just as valid, statistically?

Cheers,
Jim
*


I'm not sure what is meant by this. Bayesian implies to me use of current knowledge as an aid in making a decision.

This can be applied, for example, in a sequential test, where we are allowed to see the results of each trial. There is a Bayesian method (Wald's ratio test) which can be used to determine an optimum stopping point for an ABX test given prior test parameters (type I and II error risks, probability of detection).

There are also several types of double-blind tests to determine whether or not a difference exists, of which ABX is just one. However, I think ABX is probably the most robust.

ff123
Go to the top of the page
+Quote Post
MugFunky
post Aug 5 2004, 17:53
Post #36





Group: Members
Posts: 493
Joined: 28-December 03
From: Melbourne, Aus
Member No.: 10767



QUOTE
I have never met an engineer who's insight into the laws of physics was deep enough to effectively rebut any but the most obviously dubious claims of the audio subjectivists.


you could say exactly the same for atheism vs religion. i know of no scientists that can demonstrate successfully that God does not exist (there was an amusing Simpsons episode that dealt with this), but OTOH, what proof of God's existence can possibly be offered? certainly nothing but a few books of highly questionable origin.

but i wont get into that as i'll needlessly start offending people. i think it goes someting like this - you can't convert a hardcore sceptic, and you can't convert a hardcore audiophile, no matter what evidence you offer.

subjectivism and objectivism use completely different (and probably mutually unintelligible) languages. the issue will probably never be resolved.

but that "like a veil has been lifted" really gets up my objectivist nose (i can give proof of this too...just have a beer with me some time) smile.gif
Go to the top of the page
+Quote Post
krabapple
post Aug 5 2004, 17:54
Post #37





Group: Members
Posts: 2159
Joined: 18-December 03
Member No.: 10538



QUOTE (jaustin @ Aug 5 2004, 04:04 AM)
QUOTE (Audible! @ Aug 4 2004, 10:37 PM)
QUOTE (Dunlavy by way of krabapples link)
To say, as some do, that there are factors involved that competent engineers and scientists have yet to identify is utter nonsense and a cover-up for what should be called pure snake oil and buzzard salve - in short, pure fraud.


The bolded section should be read and reread by every subjectivist "audiophile", five times daily. Mr. Dunlavy is absolutely not joking about the Nobel prize.
*


Good to know we have someone on the board that has pull with the Nobel committee.


It does not require 'pull' to know what constitutes Nobel-quality discovery. It requires no pull for me to accurately assert that anyone who could prove the claims of creationists and thus overthrow the theory of evolution, would be awarded a Nobel prize.


QUOTE
As an erstwhile audio writer trained as a physicist, I agree that we probably have a solid theoretical grasp of the basic laws that govern the operation of cables (and to a lesser extent more complex audio devices) but I worry that the quote is being used here to imply more than it should. Of course we know the basic theories that apply to cables, component design, etc. But anyone who has ever done scientific research (and I suspect anyone who has ever done much engineering) knows that it's one thing to understand the basic theories, and quite another to properly apply them. Indeed, much of my research career was spent trying to figure out WHICH theory was the most appropriate for a certain (experimental) context, and how it should be applied. Often careful modeling is necessary. As far as measurement goes--yes, I think we can probably measure everything that's relevant; what's not possible, currently, is to directly relate what we measure to what we hear. These issues are far from trivial.



Dunlavy himself has been advertised as an engineer/physicist, so it's possible his training has made him aware of these issues.

He has had much more, and in more detail, to say on cable design:

http://home.austin.rr.com/tnulla/dunlavy6.htm


QUOTE
I would translate this quote, and the message that follows, as follows: "If a subjectivist audiophile's assertions don't correspond to expectations derived from a simple-minded, naive (mis)application of (poorly, by the writer) known laws of physics, then that person is dishonest and trying to rip you off." It's a mistake to slander those who have the good sense to believe more in their own perceptions than they do their own (inadequate) understanding of science. With respect, some people on this board would do well to take a similar approach.


It's a mistake to believe in your percerptions sans an awareness of the known flaws in perception.

QUOTE
It's true that there's little statistical proof of much of what the subjectivist audiophile camp asserts.


And little scientific or engineering basis for it either. It may be unclear to subjectivists, but there are *two* pillars upon which skepticism of audiophile claim rests: 1) measurements which indicate only small differences between A and B, and 2) what we know about human psychology.

QUOTE
Objectivists would do well to be measured in their response and stick to this lack of evidence in rejecting subjectivist's claims. When you get into the realm of theory you're on much less solid ground. I have never met an engineer who's insight into the laws of physics was deep enough to effectively rebut any but the most obviously dubious claims of the audio subjectivists.


Please visit rec.audio.high-end, where I suspect you may be satisfied in that regard. Thena gain, I'm not sure which claims you are classifying as 'obviously' flawed.

QUOTE
Furthermore, most experienced engineers know that much design--including electronic design--is based on phenomenology, convention, or tradition, and not on the direct application of theory. Theoretical understanding often lags engineering wisdom. That's not to say that there's any Nobel-inducing radical new theory involved. There's plenty of work to be done in figuring out the proper way to apply existing theories.


If theoretical understanding lags behind engineering widom, and engineering wisdom tends to find audiophile claims absurd -- then what are we to conclude?

This post has been edited by krabapple: Aug 5 2004, 18:10
Go to the top of the page
+Quote Post
krabapple
post Aug 5 2004, 18:03
Post #38





Group: Members
Posts: 2159
Joined: 18-December 03
Member No.: 10538



QUOTE (jaustin @ Aug 5 2004, 06:59 AM)
QUOTE (2Bdecided @ Aug 5 2004, 08:41 AM)
jaustin,
However, are you of the opinion that, if two devices display identical measured performance, then they must display identical subjective performance?

Or are you in the camp that believes there are differences that we can hear, but not measure?
*

It depends on whether you make the right measurements, and if anybody knows what the right measurements are, I haven't heard about it.



For cables, the measurements have been known for some time. If you assert otherwise, it up to you (or someone) to show that those measurments are inadequate for predicting audibility of cable difference....some nicely controlled listening tests would be a start.

QUOTE
I believe that there are audible things that we do not yet know how to measure--or if we know how, that no one has yet made those measurements. Establishing a correspondence between perception and measurement isn't trivial. As an example, I think that we cannot rule out the possibility that particular distortion "signatures" are responsible for perceived differences between amplifiers and other audio equipment (including, eg, output stages of CD players). But the problem of relating a particular distortion signature to perception is very difficult because it's very hard to identify specific aspects of the sound that relate to those subjective perceptions. And, yes, whenever this is the case--that perceptions are difficult to specifically identify, it is appropriate to be skeptical--but not closed-minded--about whether those perceptions are real. (ABX testing is desirable under those circumstances, but can be difficult to perform with real-world equipment). Limiting yourself to statistically valid data is a legitimate, even essential, for many purposes, strategy, but in the interest of clarity and certainty such an approach limits the information that's considered.


I would ask, then on what basis you are skeptical of any reported phenomenon? Because everything from
lucky numbers to ghosts to God could be plugged into your paragraph above in place of 'audible things'.

To me, the 'I believe there are things we can hear but do not know how to measure' is a sort of special pleading. The belief exists to justify a view of anecdotal experience for which *other* , far more well-founded explanations -- psychological bias, in the case of audio -- exist. It never enough to say 'there's lots we don't know' -- it's important to show why what we *do* know is probably inadequate.
The audiophile argument seems to boils down to 'lots of us hear it'. Do you consider that sufficient reason to believe that current knowledge is inadequate?

This post has been edited by krabapple: Aug 5 2004, 18:11
Go to the top of the page
+Quote Post
Otto42
post Aug 5 2004, 19:55
Post #39





Group: Members
Posts: 1075
Joined: 15-October 03
From: Memphis, TN
Member No.: 9323



I admit that I'm not big on testing methodologies and such. I don't fully understand ABX testing and so forth. At least, I don't understand how to read the results, though I do understand the purpose behind it.

So, it seems to me that "hearing" is just another form of "measurement". If you can really hear a real difference, then you should be able to hear that difference in a controlled situation, like an ABX test. The ABX test simply quantifies the difference that you are actually hearing and proves, to some desired statistical level, that you are actually hearing that difference.

In other words, the question seems flawed. If you can perceive a difference, then you can always measure that perceptual difference via some method or another. So there's no such thing as a difference that you can hear but not measure because the question is silly. Hearing is a form of perception, which can always be measured, really.

Or am I missing something here?


--------------------
http://ottodestruct.com
Go to the top of the page
+Quote Post
JSonnabend
post Aug 5 2004, 22:21
Post #40





Group: Members
Posts: 21
Joined: 20-January 04
Member No.: 11399



QUOTE (Audible! @ Aug 4 2004, 10:37 PM)
QUOTE (Dunlavy by way of krabapples link)
To say, as some do, that there are factors involved that competent engineers and scientists have yet to identify is utter nonsense and a cover-up for what should be called pure snake oil and buzzard salve - in short, pure fraud.


The bolded section should be read and reread by every subjectivist "audiophile", five times daily. Mr. Dunlavy is absolutely not joking about the Nobel prize.
*


That's all well and good until the paradigm shifts. Our understandings of the physical world are merely models, and models, even long standing ones, are all subject to change. Remember Newtonian physics?

Still, if a validly conducted ABX test statistically demonstrates that someone can hear the difference between cables, doesn't that answer the question of whether there is a difference in sound between cables for that person, regardless of why they sound different?
Go to the top of the page
+Quote Post
jaustin
post Aug 5 2004, 23:43
Post #41





Group: Members
Posts: 39
Joined: 26-July 04
Member No.: 15785



QUOTE (JSonnabend @ Aug 5 2004, 04:21 PM)
That's all well  and good until the paradigm shifts.  Our understandings of the physical world are merely models, and models, even long standing ones, are all subject to change.  Remember Newtonian physics? 
*

Switching sides here, but just a little, and for only a moment. Newtonian physics is alive and well and does an outstanding job of describing the mechanics of the everyday world. I don't think we need to turn to new physics/paradigm shifts to explain why there are things about audio people don't understand.

There's a series of physics textbooks by a pair of Russian authors, Landau and Lifshitz. They described it as the "minimum of material" that a physicist must master to know his (or her) shit. I know more physics than most folks, but after ten years of training I probably knew ten percent of what's in those books.

Like I've said before, folks think they know theory, they think they know how to apply it, but much of what they think they know is freshmen-level stuff, zero-order approximations to reality. It's not that physics is just a model of reality--true enough, but irrelevant. The real issue is that the models most people have are simple-minded, even if they have technical training (like me). We've all got a long way to go in understanding before we can legitimately start claiming to be approaching the limits of known science.

Cheers,
Jim
Go to the top of the page
+Quote Post
Pio2001
post Aug 6 2004, 00:38
Post #42


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



QUOTE (JSonnabend @ Aug 5 2004, 10:21 PM)
That's all well  and good until the paradigm shifts.  Our understandings of the physical world are merely models, and models, even long standing ones, are all subject to change.  Remember Newtonian physics? 
*


The bold part also states :

QUOTE
If any cable manufacturer, writer, technician, etc. can identify such an audible design parameter that cannot be measured using available lab equipment or be described by known theory, I can guarantee a nomination for a Nobel Prize.


Mind the two sides : to measure using available lab equipment, or to describe with known theory.

Paradigms shifts only affect the second side. Newtonian physics said (theoretical side) that masses generated a field force. In practice, it makes all objects fall toward the center of the earth with an acceleration of 9.8 m/s2.
Then Einstein published the Genreal Relativity. The theoretical side of Newton's theory was comlpletely removed. Masses generate no such thing as a field force, but their presence can be described as a curvature of space-time. In practice, it makes all objects fall toward the center of the earth with an acceleration of 9.8 m/s2.

So after a paradigm shift, maybe we can describe theoretically a phenomenon that is now unexplained. But we should already be able to measure it with available lab equipment, thus John Dunlavy won't give the Nobel so easily smile.gif !
Go to the top of the page
+Quote Post
JSonnabend
post Aug 6 2004, 02:51
Post #43





Group: Members
Posts: 21
Joined: 20-January 04
Member No.: 11399



QUOTE
So after a paradigm shift, maybe we can describe theoretically a phenomenon that is now unexplained. But we should already be able to measure it with available lab equipment, thus John Dunlavy won't give the Nobel so easily smile.gif !

Pio (and JAustin) -

I believe we're all on the same page, really.

With regard to the "we should be able to measure it with existing lab equipment" thing, though, that's not always true. We measure only what we look for, and what we look for is governed by our model of the physical world. We generally don't build equipment to measure things we don't think exist.

Audio cables can be measured six ways to Sunday, but I bet no one has measured them for "pink interactions". What are "pink interactions"? Well, they don't exist in our current model, so we don't try to measure them. But when we start observing differing behaviors of different cable systems that shouldn't behave differently (because our models say they're functionally the same), we may need to change the model and start trying to measure something new (ok, so "pink interactions" is a lame one, but you get the point). That new measurement, of course, may require new equipment. Newton couldn't have measured E-M field strength or alpha-decay, but that doesn't mean those things didn't exist.

FWIW, I'm not saying that there is a noticible difference between different cables. I was just a little bored at work today, hence the James Burke stuff.

- Jeff
Go to the top of the page
+Quote Post
Audible!
post Aug 6 2004, 02:52
Post #44





Group: Members
Posts: 523
Joined: 28-June 03
From: CA, USA
Member No.: 7426



QUOTE (jaustin)
Good to know we have someone on the board that has pull with the Nobel committee.

huh.gif
What an amusing comment, truly.
Please reread.

QUOTE (jaustin)
I would translate this quote, and the message that follows, as follows: "If a subjectivist audiophile's assertions don't correspond to expectations derived from a simple-minded, naive (mis)application of (poorly, by the writer) known laws of physics, then that person is dishonest and trying to rip you off." It's a mistake to slander those who have the good sense to believe more in their own perceptions than they do their own (inadequate) understanding of science. With respect, some people on this board would do well to take a similar approach.

laugh.gif
How very droll.
You have failed to understand the comment, and given your undoubtedly thoroughly prestigious background in physics I'm quite shocked.

Blind, preferably double blind controlled studies are the only effective way to circumvent bias effects, like the "placebo effect". This of course, has little to do with physics, but I'm still quite taken aback. If you knew a thing about drug trials you would know why it is necessary to have double blind trials in order to determine whether the experimental product is superior to the placebo. Otherwise you are prescribing Ginkco to cure cancer based on the "flows of chi".

The point is absolutely not in any way, shape or form that when assertions not derived from theory are made that they can be instantly rejected. Exactly the opposite is the case if the assertions carry with them statistically viable evidence.
Where is this evidence?

It is when assertions are made without a single, solitary shred of statistically viable evidence that they can be readily doubted until such evidence is presented.

As to my knowledge of "science", I would think any scientifically knowledgable individual would reserve comment until such time as an adequate sample set of my writings on scientific subjects became availible.
Apparently I am proven incorrect yet again. In point of fact I am a professional scientist in the chemical sector.

QUOTE (JSonnabend)
Still, if a validly conducted ABX test statistically demonstrates that someone can hear the difference between cables, doesn't that answer the question of whether there is a difference in sound between cables for that person, regardless of why they sound different?

Of course.
The problem is, where are these tests?
This is the entire issue, the subjectivist crowd largely refuses to conduct them, and certainly the manufacturers of high end cable barring a few ("objectivist", you'll note, most of whom I've read of claim that their cable performs exceptionally when measured, but are largely if not totally inaudibly better than 12 Gauge zip) exceptions. Mr. Austin has suggested that bias effects can be somehow eliminated by ocasionally "calibrating" ones mind rather than always performing rigorous (and tedious) testing to check ones subjective impressions. I am aware of no evidence to suggest this is the case. The human mind is not a pipetteman or a sensor.

QUOTE (Pio2001)
So after a paradigm shift, maybe we can describe theoretically a phenomenon that is now unexplained. But we should already be able to measure it with available lab equipment, thus John Dunlavy won't give the Nobel so easily

Precisely.

I know I won't have to remind Mr. Exhalted Professor of Great Science JAustin of this, but you must satisfy the correspondence principle in order to have a "paradigm shift". Hence Newtonian Mechanics being superseded due to the inaccuracy of prediction by the theory versus measured data.

This post has been edited by Audible!: Aug 6 2004, 05:25
Go to the top of the page
+Quote Post
2Bdecided
post Aug 9 2004, 12:39
Post #45


ReplayGain developer


Group: Developer
Posts: 4945
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



QUOTE (jaustin @ Aug 5 2004, 02:59 PM)
QUOTE (2Bdecided @ Aug 5 2004, 08:41 AM)
jaustin,
However, are you of the opinion that, if two devices display identical measured performance, then they must display identical subjective performance?

Or are you in the camp that believes there are differences that we can hear, but not measure?
*

It depends on whether you make the right measurements, and if anybody knows what the right measurements are, I haven't heard about it.

I believe that there are audible things that we do not yet know how to measure--or if we know how, that no one has yet made those measurements.
*



Well, as you might guess, I disagree!

It's certainly very difficult (some would say impossible) to properly correlate measured performance with perceived performance: you can't always say what measured property causes such-and-such a perceived subjective effect?

However, if we just look at the "there's no measurable difference" vs "I can hear a difference", it doesn't necessarily mean that we failed to measure something. If you do specific measurements looking for X, Y, or Z then you might miss A, B or C. However, where ever it's possible to do a difference analysis between two devices, if that difference analysis reveals a difference that's lost in the noise of the experiment itself, and if the noise of the experiment is sufficiently low in sufficient dimensions to allow us to conclude that any difference that exists is just far too small to matter, then we can conclude that there is no difference worth mentioning between the two devices.

Subjectivists say "ah, but I can hear a difference". This is a difference that they can't ABX, and that can't be measured. I'd suggest that, in truth, there's no audible difference! The reason the pretty amplifier sounds nicer to them than the plain amplifier is the very well known phenomenon of what they see effecting what they hear.

When someone says "no, there is a difference, it's just that you can't measure it" (note, not "didn't measure it" but "can't measure it") then I simply don't believe them.


Of course, in the real world, the situations where you can carry out such a difference analysis are limited, and the equipment pairs which exhibit a null measured difference are rare indeed. However, you can do a fairly good job with loudspeaker cables.

With loudspeakers, you can measure faults more accurately than you can correct them, yet subjective reviewers often praise loudspeakers with measured 10dB frequency response faults!!! Yet I'm supposed to believe these same reviewers when they claim to hear a difference between devices that measure the same, and that they can't ABX.

I don't. The burden of proof clearly lies with them.

Cheers,
David.
Go to the top of the page
+Quote Post

2 Pages V  < 1 2
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 20th April 2014 - 22:19