Kiang

The Great Cable Debate

Recommended Posts

Guest earl of sodbury

Biscuit wrote:

maths makes everything seem so terribly dull :?

So very, very true: it kills the joy in everything it touches.:^

Maths is evil.

Share this post


Link to post
Share on other sites

Well fuck me,

I've just been looking at this thread and if I'm going to be honest, it's this sort of crap that makes me feel like a right sad bastard in admitting that I've got a passing interest in hifi, (well, goodmusic reproduced nicely to be more exact).

I'm all for discussion and some testing and don't want to be overly critical of anybody at all but.........

Pio me boy, those equations :shock::shock::shock:, do you have a girlfriend/wife?

Christ almighty, it's a HOBBY isn't it? a test here, a test there, done like this, done like that. It's meant to befun I'm sure, and, interesting also perhaps, with some degree of accuracy maybe thrown in.

But FFS, no-one's getting sent into space, get a grip you :geek:out there. :peace:

Share this post


Link to post
Share on other sites

mosfet wrote:

Pio wrote:
To make it short, for every million tests the formula predicts in average 48,050 results with at least 12 right answers, and the simulation produced 47,835 ones.

Yes, I understood the reason for running a simulation and the significance of the formula predicted results being more or less the same as the simulation results. This proves the correctness of the formula to a good degree . The mathematics behind the probability calculations themselves still baffle me a little - but I’m happy to remain slightly baffled :) because the formula is proved.

I think that it is perfectly safe to reduce N to the number of explicit answers instead of the total number of participants. This will give new p values for each possible score. The comments will remain the same for similar p values, and of course not for similar numbers of right answers. Good idea.

I see.

Sums up my take on it perfectly.

The maths have been proved, now lets apply them... where are the results? :D:whistle:

Share this post


Link to post
Share on other sites

pissed off trev wrote:

Pio me boy, those equations :shock::shock::shock:, do you have a girlfriend/wife?

Christ almighty, it's a HOBBY isn't it? a test here, a test there, done like this, done like that. It's meant to befun I'm sure, and, interesting also perhaps, with some degree of accuracy maybe thrown in.

Kiang has spent a lot of time masking the cables and preparing the test. People have been waiting for the results for monthes. If the maths don't come and validate the test results, then what purpose would all this work have served ?

If listening to the cables, appreciating them, and comparing people's impressions is all that counts, then why mask the cables ? Why hide the results for monthes ? Why use a form in order to give results ? Just type your power cable's model in google and look for people's opinions.

Listeners invested time in this test also because it was a blind test, scientifically designed. Let's get the stuff done properly until the end, in order to get a result that's worth the effort.

meninblack wrote:

The tricky bit will most likely be deciding which hypotheses to test - the test for "A is better than B" being numerically different from the test for "A is the same as B" as I'm sure your aware.

Oh I see. I had not read the test form for a long time. So for every participant, we should have :

-Which is the cable that sounds most like cable D ? (A, B or C)

-A ranking of the cables A, B and C in sound quality. Here, as I see it, we should not discard the answers to the first question if all cables are ranked the same, as suggested by Mosfet, because the listeners may want to tell that the cables sound different to them, but that none of them is altogether better than the others. Just different. Perfectly identifiable from each other without superiority from any of them.

-The last question (which one is the best, which one is the worst) should give the same result as the ranking.

So we can also perform a significance analysis on the rankings of the cables. However, this takes the maths on a higher level !

I'm currently sorting out the details, but I'll post them later. There are several probabilities interfering with each other and getting the matter straight is not easy.

And, Pissed off Trev, maths are a hobby too ! :)

Share this post


Link to post
Share on other sites

Pio wrote:

-Which is the cable that sounds most like cable D ? (A, B or C)

-A ranking of the cables A, B and C in sound quality. Here, as I see it, we should not discard the answers to the first question if all cables are ranked the same, as suggested by Mosfet, because the listeners may want to tell that the cables sound different to them, but that none of them is altogether better than the others. Just different. Perfectly identifiable from each other without superiority from any of them.

example one

The cable most like cable “D†was [discard any answer as counting toward x]

Between cables A, B and C: (use Options: better than, worse than, same as)

The cable “A†was same as cable “Dâ€

The cable “B†was same as cable “Dâ€

The cable “C†was same as cable “Dâ€

example two

The cable most like cable “D†was A

Between cables A, B and C: (use Options: better than, worse than, same as, different to)

The cable “A†was different to cable “Dâ€

The cable “B†was different to cable “Dâ€

The cable “C†was different to cable “Dâ€

...because the listeners may want to tell that the cables sound different to them

Yes, but there is no option on the questionnaire form for the listeners to say this; only better than, worse than, same as.

What you say makes sense Pio but the questionnaire options do not allow for this. It perhaps would have been better to include the option different to for the reasons you say.

For example one, discarding of answers as counting toward x is, I think, still necessary for the reasons previously given.

Share this post


Link to post
Share on other sites

mosfet wrote:

example one

The cable most like cable “D†was [discard any answer as counting toward x]

Between cables A, B and C: (use Options: better than, worse than, same as)

The cable “A†was same as cable “Dâ€

The cable “B†was same as cable “Dâ€

The cable “C†was same as cable “Dâ€

example two

The cable most like cable “D†was A

Between cables A, B and C: (use Options: better than, worse than, same as, different to)

The cable “A†was different to cable “Dâ€

The cable “B†was different to cable “Dâ€

The cable “C†was different to cable “Dâ€

If as in example 1

The tester means that atleast 2 cables are different but not better or worse than D

This is so as it would not be possible to answer the first question "the cable most like D" If he truly thought they were all identical. He is not blind to the fact that there is an audiophile cable , DIY cable and kettle lead.

If as in 2

This is not possible as atleast one cable has to be the "same as"

The tester is not blind to this knowledge.

Share this post


Link to post
Share on other sites
Moderator

I would separate it into two questions:

1. Could the testers correctly identify the kettle lead?

This would be a yes/no outcome for each tester (null result = no) so a simple one-sided t-test would suffice (with Yates's correction applied to the chi-squared calculationbecause the response data has only one degree of freedom.)

2. Which cable did the testers prefer overall?

In this case the cables could be "scored" 0 - 4 and aone-wayANOVA test (variable = true cable type, response = preference) would determine significance. Null results could be discarded, or treated as "draws" and scored equally.

Share this post


Link to post
Share on other sites

Kiang wrote:

If as in example 1

The tester means that at least 2 cables are different but not better or worse than D

If same as means sounding different but not better or worse then you have not given the tester ANY option to say the cables sound the same.

When comparing any two cables in the test the options given are therefore:

P sounds better than Q (the better than option)

P sounds worse than Q (the worse than option)

P sounds different but not better or worse than Q (the same as option)

This is forcing the tester to answer that a difference of some measure was heard.

This means even the man from Switzerland will say he heard a difference! Which I suppose will be a remarkable result in itself!

Share this post


Link to post
Share on other sites

mosfet wrote:

Kiang wrote:
If as in example 1

The tester means that at least 2 cables are different but not better or worse than D

If same as means sounding different but not better or worse then you have not given the tester ANY option to say the cables sound the same.

When comparing any two cables in the test the options given are therefore:

P sounds better than Q (the better than option)

P sounds worse than Q (the worse than option)

P sounds different but not better or worse than Q (the same as option)

This is forcing the tester to answer that a difference of some measure was heard.

This means even the man from Switzerland will say he heard a difference! Which I suppose will be a remarkable result in itself!

To comment on this Mossy would mean to quote an example , this might/or not bias our esteemed last tester;) . Will hold off for now.

Share this post


Link to post
Share on other sites
Guest

Kiang wrote:

mosfet wrote:
Kiang wrote:
If as in example 1

The tester means that at least 2 cables are different but not better or worse than D

If same as means sounding different but not better or worse then you have not given the tester ANY option to say the cables sound the same.

When comparing any two cables in the test the options given are therefore:

P sounds better than Q (the better than option)

P sounds worse than Q (the worse than option)

P sounds different but not better or worse than Q (the same as option)

This is forcing the tester to answer that a difference of some measure was heard.

This means even the man from Switzerland will say he heard a difference! Which I suppose will be a remarkable result in itself!

To comment on this Mossy would mean to quote an example , this might/or not bias our esteemed last tester;) . Will hold off for now.

for all is worth guys, i have tested a lot of cables throughout the 30+ yearrs in retail,"no longer in the high street" and all cables can influence your system but they are also system dependant, it's up to the individual to choose wich suits his or her's ears most, in my opinion the best is up to you. :cool:

Share this post


Link to post
Share on other sites

Kiang wrote:

To comment on this Mossy would mean to quote an example, this might/or not bias our esteemed last tester

Of course. ;)

When I downloaded the questionnaire form to have a look and read the options better than, worse than and same as I took same as to mean sounding the same as!

Not meaning different to but not better than or worse than.

In any case, if someone gives an explicit answer to the question

The cable most like cable “D†was …

then it’s fair to assume they understood this question. And the results of this question I think are by far the most important to the debate.

As MIB says:

1. Could the testers correctly identify the kettle lead?

This would be a yes/no outcome for each tester (null result = no) so a simple one-sided t-test would suffice

Share this post


Link to post
Share on other sites
Guest

mosfet wrote:

Kiang wrote:
To comment on this Mossy would mean to quote an example, this might/or not bias our esteemed last tester

Of course. ;)

When I downloaded the questionnaire form to have a look and read the options better than, worse than and same as I took same as to mean sounding the same as!

Not meaning different to but not better than or worse than.

In any case, if someone gives an explicit answer to the question

The cable most like cable “D†was …

then it’s fair to assume they understood this question. And the results of this question I think are by far the most important to the debate.

As MIB says:

1. Could the testers correctly identify the kettle lead?

This would be a yes/no outcome for each tester (null result = no) so a simple one-sided t-test would suffice

agree , absolutlly, noting to hide have we?:cool:

Share this post


Link to post
Share on other sites

The important thing for artists, creative right brain thinkers, poets and music lovers everywhere who (for whatever reason) care about this test...and WE DO...is to just let the left brainers play here for a while. I'm sure they love what they do too and in the end may even contribute to validation if that's important (Its not to me, but there were 23 of us and some observers). Everyone in the wigwam will getr something from these results and no-one need feel that attitude 'a' is more valid than randomly generated attitude 'x' (where x= a number of small frogs launched into the van halen belt by elastic bands of various tensions by blindfolded robots programmed to repond to the release word "UGG")

Wait for the results. Then let the fun begin!:)

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.