In article , Rob
wrote:
Jim Lesurf wrote:


Rob wrote:




OK. Do you accept:


1. Some people can differentiate between cables


Depends on the cables and their conditions of use. e.g. Some cables
have a large (relatively speaking) series resistance or inductance, so
with some speakers would produce a change in response I'd find
audible, and assume others would find audible, in appropriate
circumstances. Hence this isn't a matter of blanket "accept" or not.
Depends on what details


I'll take that as an 'OK' :-)

It is actually "In some cases, yes, in others, I doubt it, and in others I
have no way to tell." :-)




; 2. DBT doesn't reliably support that finding;


Depends which "that finding" you are referring to. :-)


The finding that "Some people can differentiate between cables"

See above. Can't give a yes/no answer to such as question as it would vary
from one case to another.


3. DBT, for this purpose, is a flawed method?


Define "this purpose" in this context. :-)

The aim of explaining why some people can differentiate between cables.

This assumes there is a genuine difference as distinct from an error. :-)


The aim of Stewart's test is to see if *you* can provide any
substantiation of your claim/belief. If you find the differenced
"night and day" I'd assume you would think the difference was so great
as to be easily revealed by a simple test. Thus giving you a chance to
pick up some case and show you are correct. However I don't know if
this is the "purpose" you have in mind.

Er, steer clear of 'me' for the moment - I'm keeping well clear! The
bottom line with this is that DBT doesn't explain why people *can*
distinguish between cables.

That presupposes that they 'can'. What if they are simply mistaken?
Before seeking an explanation it would help if we could establish, by
suitable evidence, if any (alleged) effect is arising in reality
or in imagination.

I'd also be interested in any lateral thinkers out there who could
explain why it's not, as a methodological point, a substantial method in
any event.

The main criticism I have seen is as follows:

People who dislike ABX/DBT methods invoke a 'new mechanism' by which the
test process 'masks' or 'confuses' the ability to distinguish.

The problems with this are as follows:

Firstly it proposes an alternative hypothesis for the inability of people
to hear "night and day" differences in an ABX/DBT test.

The default hypothesis/explanation is that any actual difference is either
non-existent or tiny. Hence cannot be replicated as it either isn't
"obvious" or perhaps even exists.

The new hypothesis is that the test includes a mechanism which 'masks' what
was previously "obvious".

To distinguish between these competing hypotheses we would require a test
or experiment whose outcome would provide evidence which undermines
(falsifies) one hypothesis and supports (consistent with) the other.

It would also help if the details of this new 'masking' process were
explained in terms of the relevant physics, etc. This would aid in
assessing if it were plausible in terms of what is already well
established are regarded as reliable in terms of other evidence, etc.

However so far as I know, no-one who has argued for such a 'masking'
mechanism has been able to propose an adequate test for distinguishing
between the competing hypotheses, nor a plausible, testable, mechanism.

Hence until such time as someone does, it seems reasonable to doubt such a
mechanism *is* causing a 'masking' effect. This leads to it being
reasonable to proceed on the basis that any claimed differences are either
tiny or non-existent, despite the assertions to the contrary. But also to
be open to someone proposing a suitable test or new evidence that would be
a basis for reconsidering this. Occam's Razor. Proceed on the basis of new
experimental observations *regarded critically*.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article , Rob
wrote:

What I am (or am not) is irrelevant. The fact remains that quite a few
people spend a lot of money on wire.

Many people also pay money to cover their gardens with wooden 'decking' as
well. However I am less than convinced this is a good idea. :-) Seems to
me to be a 'fashion' promoted by gardening programs on TV.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

Level-matched to +/- 0.1dB at 100Hz, 1kHz and 10kHz, pass criterion 15
correct out of 20 trials, your choice of ancillaries, of music, and of
location.

That is clear subject to some clarification of music.

That's pretty much it, unless you want to introduce
something very unusual and not relevant to a music system, in which
case a little negotiation might be in order.

Although I think I understand the gist this is a potential source of
problems. For example, is an ultrasonic tweeter (which appears to be
one of the current fashions) very unusual?

Indeed - and it would not be applicable to listening to music. I can
very easily show differences among cables by introducing ultrasonic
(or of course video) signals, but this test is about *audible*
differences, not mere trickery.

One of us may be losing the plot here. If you require the cables to
have no significant difference over the audible frequency range then
they must interact with a nonlinearity in the range where they are
different in order to generate something audible. Meeting the challenge
surely rests on this?

You guys claim 'night and day' differences when listening to music,
so why suddenly so coy?

I think you may be confusing me with someone else.

How about the source being in the conventional audio range,
and would typically (but not essentially) be music?

That would rest on what is meant by "conventional audio range". If
there is to be no signficant difference in the cables over the audible
range and no signficant signal out of it the challenge would appear
almost impossible by definition. It might be possible to do something
with instabilities in the power amplifier but that side of things is of
no personal interest.

If you're not happy with that,

Happiness comes from tough but achievable.

might one enquire the source of these 'night and day'
opinions about 'cable sound'?

I suspect we are in full agreement about the source. See elsewhere in
this thread.

There is a somewhat deceptive tone in your suggestions.

Far from it. I am being quite open about how one could go about meeting
the challenge.

I want it tackled in the spirit of claims of 'night and day'
differences heard when *listening* to these cables.

Not sure about 'night and day' since this would imply to me the
difference between loud and quiet. I presume an audible difference
acceptable? Does the non-standard cable have to sound better or just
different?

Not sure about *listening* to the cables either. Cables are relatively
insignificant passive components what would seem much more important is
the response of the active system to the differences in the cables.

Generating special
signals which have no relation to audio, typified by high-power
ultrasonic intermodulation, proves nothing.

Unless you explicitly forbid it in the rules then it is a very simple
way to demonstrate audible differences between cables and meet your
challenge.

On Wed, 1 Jun 2005 08:56:32 +0000 (UTC), Rob
wrote:

Don Pearce wrote:
On Tue, 31 May 2005 21:41:09 +0000 (UTC), Rob
wrote:



Rob, I've already supplied an email with plenty of evidence - which
you are free to follow up if you wish.


Sorry Don - I didn't get the email, could you send again please?
(gramsci at btinternet.com). Cable hegemony ... mmm :-)

Rob


Sorry - didn't mean email, meant usenet post.

d

Pearce Consulting
http://www.pearce.uk.com


Mmm - I've just been through the whole thread(!). Your evidence appears
to hinge on:

1) An 'after hours' experiment that you carried out;
2) No known participants in SP's DBT;
3) A technical reply to Jim Lesurf's post.

I'm not sure if you've misunderstood, so I've posted my original request
he

The reason you give (IIUC) for the observations of claimed cable
superiority and sound quality

is vested interest - people pay a lot of money for a cable and fool
themselves that it sounds better.

Correct? Any other reasons, or any evidence, to back up your claim?

Rob

Look, Rob. It is known, for sound technical reasons that competent
cables can not impart a sound to a signal. The extraordinary claim
here is that they can. That is the claim that requires proof.

Also, as I have said before, it is ridiculously easy to demonstrate
that cables impart a sound (if they do). You just show that you can
hear that sound. If I were to blind test two cables and say that I
couldn't hear a difference would that prove anything? Of course not.

So the burden of proof must lie with those who claim a difference. It
really is time to put up or shut up.

d

Pearce Consulting
http://www.pearce.uk.com

In article , Signal
wrote:
"Stewart Pinkerton" emitted :

I wonder why SP has not issued the challenge to the cable
manufacturers?

Already been done by Tom Nousaine, and they welched. I've tried it with
Ecosse cables in this country, and they ran away and hid. I've tried it
with Randy at BEAR Labs, and he ducked out as well. Basically, they're
just conmen, and they'll always duck a real test.

Yeah, but you and Tom are not scientists.

There is no requirement I am aware of for someone to wear a white coat, or
have a PhD, or belong to the IoP or IEEE to engage in a defined or
controlled experimental test of some kind. :-)

A 'scientist' is someone who employs the scientific method(s). Mind you, if
the IoP were as good a 'union' as the lawyers or medics have, things might
be different. 8-]


On the contrary, the both of you think you already know all the answers
already and are far from open minded or impartial to the results.

But neither Stewart not Tom (?) would presumably actually be the subject of
the test. Or is your argument that simply being in the same room as a
'sceptic' removes the ability of someone to hear "night and day"
differences?

The involvement of money or betting means it's unscientific from the
outset - basically it's a dick measuring contest, not the honest persuit
of science.


I'm afraid that you have an unrealistic view of 'scientists' if you think
they work without any money being involved. Mind you, their employers, and
grant-givers, do often seem to hold this view... :-)

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

On 01 Jun 2005 08:44:18 GMT, John Phillips
wrote:

On 2005-06-01, Don Pearce wrote:
On 31 May 2005 20:17:52 GMT, John Phillips
wrote:

So in brief, if one argues that cables sound different, you need to
consider the capacitance in analysing the claim. Fred did so argue,
but failed to consider capacitance. So his paper was flawed.

I have done some calculations based on the fact that a loudspeaker
terminates the cable and an amplifier drives it.

In the case of the two most capacitive cables of the 12, I agree that the
error due to ignoring the capacitance may be of the order 0.2 dB and 0.1
dB (the highest and second highest respectively). For all other cables
the likely error is smaller. (I did the maths quickly and I admit I
will have to check again to be absolutely sure.)

Even if the error is of the order of 0.2 dB and makes a bigger difference
(actually I think you are right that it reduces the differencce) then
the differences shown still do not stray into the audible regions of
the graph whose URL I posted.

From that point of view the paper seems to be rather less flawed than
you propose.


So your proposition is that because the errors in his paper are not
bigger than the differences he is actually reporting, then they don't
matter?

No Don. That's not what I wrote. Please take note of the "From that
point of view" and what it refers to.

You have taken a narrow argument, broadened it ad absurdum and thrown
it back at me as thoughh it were my own. That form of rhetoric is
frustrating, especially so when I expected better of you.

All models have errors. A model is fit for purpose when the errors,
if present, do not change the conclusion. To be clear, I wrote that
for the purposes of concluding that the deltas shown in the paper did
not indicate that cable differences are audible, the maximum errors
I calculated (very conservatively) did not change that conclusion.
I repeat "From that point of view the paper seems to be rather less
flawed than you propose."

Do think again.

Indeed I have. I have made a better estimation of the errors due to
the omission of the cable capacitance. Based on the worst case damping
factor for the driving amplifier (amp A - DF=25 at 20 kHz), the worst
cable (6,000 pF - that's -j.1300 Ohms at 20 kHz) and the worse of the
two loudspeaker loads at 20 kHz (LS B).

The omission of the capacitance actually makes 0.01 dB difference.
Yes, my previous estimate was really too high, even if I have not
actually found the worst case point this time. Thus I agree with the
proposition in the paper that the capacitive component is indeed to
small to materially influence the result for just about all purposes.

Leave this with me. I will do the sums and come back.

d

Pearce Consulting
http://www.pearce.uk.com

In article , Signal
wrote:
"Stewart Pinkerton" emitted :

In this context I argue that it *might* be. Under normal (for now,
non-DBT) conditions some people claim difference.

So what? Once they can *prove* that they hear a difference when they
don't actually *know* what's connected, we'll continue the debate.

Fast-switching DBTs, the type practiced by some of the more outspoken
members of these groups, cause confusion and are based on techniques
which bear a remarkable affinity to established methods of trance
induction.

So far as I know, it would be easy enough to arrange with those involved
that the 'ABX' be switched as and when suited the person doing the
listening. Thus if you dislike "fast switching" this could be avoided.

Can you give some references to journal papers that show how quickly
switching between very similar musical sounds causes trance induction? I've
not encountered that claim before.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article , Signal
wrote:
"Don Pearce" emitted :


The *question* may be simple, but if you think getting reliable data is
simple then you are approaching the subject too simplistically. Removing
bias without impinging on sensitivity is FAR from trivial. Drawing
conclusions from null results is a fallacy. Drawing conclusions from
results without paying due consideration to Type 2 statistical errors is
a fallacy.

It is reasonable to argue that a null result does not 'prove' that there
are no differences. However a null result also undermines any claim that
the differences are "large" or "obvious" or "night and day". It is also
consistent with there being little or no audible difference.

I am not sure what 'bias' you are referring to with a DBT/ABX since no-one
in the room should know from prior/external information if 'X' is 'A' or is
'B' at any time during the actual test run.

Don, I worked for a major Pharmaceutical company on proper scientific
double blind drug research trials, and by comparison the intellectual
effort... the scientific scrutiny.. the size and scale of the so-called
audio DBTs discussed in these groups are *laughable*. These guys aren't
grade A scientists, they are over-opinionated closed minded wannabes.
Stewart even thinks he can make global pronouncements on audibility
based on tests he's set up with his mates, but in reality.. a few hours
later he's doing what he *really* knows best - maintaining a print room
;-)

Can't comment on Stewart. Nor do I know of any "Grade A" accreditation
schemes for 'scientists'. :-)

I do have a white coat, though. Does that help in my case? ;-

The problem with the analogy with medication testing is that in the case of
cables we seem to find that the tests all produce nulls or results with no
satisfactory level of statistical significance. To paraphrase someone else
- to null one test seem unfortuante, but to null them all seems careless.
:-)

Any poor design in the test will result in making it easier - not
harder- for them to hear a difference.

A presumption... pretty far fetched IMO. A poorly designed test could
mask differences.

Please suggest a *testable* mechanism for this 'masking'. Otherwise your
assertion is a personal belief, and not testable by the usual
scientific/academic methods. If you can suggest a plausible, testable,
'masking' mechanism we could try to deal with it in any proposed future
tests.

The money, far from invalidating the test, makes for even greater
incentive to hear a difference than personal pride alone.

"Incentive" = increased sensitivity?

Another presumption.

Tom Nousaine recently said in an ideal world test participants should
have a gun to their head, or words to that effect. Do you agree?

In fact the test regime bends over backwards to accommodate the
triallist.

That's debatable - but even if true, it doesn't substantiate the
unscientific position that a null results prove anything. That's just
opinion - an opinion that indicates a very clear bias on the part of
those making such statements.

The 'scientific' position is that a (series of) null tests fail(s) to
support the hypothesis that a difference in sound is so large as to be
"obvious" or "night and day". The normal scientific method is then to
proceed on the basis that there is either no difference, or one so small as
for it to not be reliably detectable. *Until* someone can carry out a
better experiment that may show otherwise, and/or make a new hypothesis
which is *testable* in an experiment whose results have the ability to
falsify/support the hypothesis.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article , Signal
wrote:
"John Phillips" emitted :

In this context I argue that it *might* be. Under normal (for now,
non-DBT) conditions some people claim difference.

So what? Once they can *prove* that they hear a difference when they
don't actually *know* what's connected, we'll continue the debate.

Fast-switching DBTs, the type practiced by some of the more outspoken
members of these groups, cause confusion and are based on techniques
which bear a remarkable affinity to established methods of trance
induction. Normal humans are not immune to this. ,,,

I thought the listener was in control of the ABX switch so I can't
understand why they would want to switch it fast if they thought that
would be detrimental to their ability to hear the difference.

The scientific community says audio memory is short

I'm afraid that in this context that is a bit of a vague and sweeping
generalisation.


- a few seconds at best. So you might say a properly designed test
shouldn't afford participants the opportunity to switch slowly, such
that their memory fades.

That would not be a requirement so far as I can see. The purpose of the
test (AIUI) is to see if the participant can - as they believe - hear a
difference and reliably identify 'X'. If they feel they can do that more
easily using longer periods I can't see that being a problem. What matters
is if they can reliably identify 'X' by switching as *they* choose.


Or do you think they should be allowed to modify the test such
that it no longer conforms to scientific principles and is thus invalid?


I don't see where you get the idea that "scientific principles" rules out
allowing the participant to switch with longer periods. I'm not aware of
any organised union of 'scientists' who impose this as a legal requirement
on any tests. I suspect that Stewart and others who propose such tests
would be happy enough to allow longer listening if the subject wanted this.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article .com,
andy
wrote:
That's already been stipulated.

Can you please point me at where.

And no, you don't get to pump 100 watts of 60kHz and 63kHz down the
cables

So the subject cannot choose the source?

This is supposed to be about the sound of cables in a high-fidelity
music reproduction context, so obvious and totally unrepresentative
cheats will be excluded.

I am not sure I fully understand the idea of cheating. If you require
the cables to have the same response from 100Hz to 10kHz then it is hard
to see an audible change coming from anything other than the cables
interactingly differently with a significant nonlinearity above or below
this range.

I am inclined to agree. :-)

However the key point here is that it is generally accepted that changes in
frequency response or level are audible. Hence the real question arises
when people say they can hear differences when the system changes involved
do *not* change the response or gain by an audible amount. Thus Stewart is
trying to exclude non-contentious 'differences' which he, and most others,
would probably accept as being audible.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html