New Power Cable test proposal

Hi all,

The previous test was a great effort, only really let down by a poor headcount on the day, of course this can't always be helped and I think we can say we learn't a lot about the test procedures.

The results lean in favour of standard power cables, but ulitmately many are not satisfied with the results. A shame as everyone worked really hard to get the most accurate result possible.

Anyway, I am not sure how many of you are aware, but there is now an idea about another test. One of our members is planning to take three seperate cables, of varying cost / construction and disguise them. To overcome any effect the disguise might have they will all be disguised using the same materials etc. The cables will be marked A B C and posted to a forum members in a round Robin style. There will be an online questionare for volunteers to fill in and record their findings.

The adjudicator will then colate the results and report back once "enough?" people have had a listen.

Right now I don't think the organiser is a member here, but I will suggest he comes over to get your ideas and suggestions / comments etc on methodology.

Some one asked on here why we need a test! Good point. However there is so much circular debate on the effectiveness or perceived effectiveness of power cables in particular that we decided to try and do something about answering the question once and for all.

Of course we may ultimately have flaws in our test procedures, but a big enough sample group and a decent consultation period should iron out the larger majority.

If you would like to get involved, please post here and I will work a way (perhaps with the help of your mods here) of managing a cross - forum test.

It may be that we have two sets of test cables in the end. They would need be the same of course... you get the idea I'm sure.

Anyway, let me know and we'll see how we can play it.

Regards,
James.

Second paragraph edited. Please re-read.

 

If there's a test near Edinburgh I would take part. I could bring along my homemade one.

 

Hi,

The idea is that the cables will be posted round for as many people as possible to test at home. So the test will be in Edinburgh as long as someone will vouch for you to take delivery of the cables.

Cheers,
James.

 

we have a veritable hornets nest here in Auld Reekie ...

 

Well, the Glasgy lads n the Smellytown lads can join forces across forums and across the M8. :MILD:

 

The cable test - wigwam

Hi guys

Check this out
http://www.hifiwigwam.com/forum1/1113.html
Suggestions and volunteers welcome.

Volunteers please indicate your location in the U.K. if not in your profile already so I can plan the sequence in which these cables could be posted.
I am new to this forum and am not familiar with members and will accept members with positive feedback or as recommended/approved by the moderator.
Cables are not inexpensive - I hope you understand.

This is the current layout - to be confirmed , suggestions please
http://www.auricles.com/new_page_41.htm

 

Hi,

Sorry to be so negative, but this will be even more amusing than the last one.

In order to make this test meanigfull as an "absolute" reference with any reliability you need to do more.

One possibility may be to include a CD which includes a set of tracks, slected for discriminative power and processed to give a reliable reference set of "audible phenomenae". Any participant should be required to first take the "goldenear" test that makes sure they can actually hear something AND failure to pass with reasonable marks shoud disquality the "tinear" from inclusion in the final analysed dataset.

Secondly, given that many possible mechanisms for "cable sound" are well understood it may be worth to select the three cables (again, it may be worth limiting this to two) strictly on the basis of design so as to maximise design differences.

But then again, I suspect that this whole malarkey is merely an attempt to find another dataset that can be sufficiently dressed up and massaged to yield a predetermined conclusion (hence I do not trust stats I have not faked myself), so I shall not loose my genteele naivity if I see no attempts made to make the data more meaningful or give it context.

Ciao T

 

Regrettably I have to agree with 3D. Without a control element to the test the reliability of the test results can only be guessed at outside of listener bias.

Apart from..

which is the sort of cynical holier-than-thou crap I've now come to expect from you Thorsten.

 

Oddly enough, at the moment, we're currently selecting new taste panellists for flavour ingredient testing. And this is exactly what we do - the would-be panel members are checked out with standardised flavours, to make sure that they have the necessary palate perception, before being allowed on to a panel.

However, given the big range of ages, hearings and perceptions of hi-fi owners, would it not be better to let participate anyone who wants to participate, but require them first to take the "golden ear" test and submit the results to a referee, who would keep them confidential until the testing of the cables themselves is finished, and then release them along with the results?

 

Hi,

All of this raises only one fundamental question.

Do you want a test to make a test, with no further consequences, or do you seek a result from the test that allows you draw conclusions from the test that hold up to reasnoble scrutiny and are not easily criticised?

If the first, do as you like, if the second then you MUST take into consideration the fundamental critiques and issues with such tests. I have touched on some in this post:

http://www.zerogain.com/forum/showpost.php?p=141774&postcount=42

I have NOT really touched on issues such attention span isuues and listening fatigue, which impact on sensitivity of tests, though these are obviously relevant to, but much harder to "codify".

Lastly a note on the approach of sending out three cables that look identical but are not (or maybe are). You must make sure that two cables are in fact identical and the testers should be required to find the "odd one out", after they have illustrated that they, in the context of their own system can distinguish small but significant differences.

As test signals for weeding out tinears I suggest the following signals (always three tracks are needed, two identical ones and a third different one):

1) Tracks where one includes the polarity inversion of one stereo-channel.

2) Tracks where one shows a level difference of 1db (it is essential that all tracks are attenuated by several db in order to ensure that all are treated equally by the DSP routines and are different materially only in level).

3) Tracks where one is equalisied by +1db @ 5KHz Q=1 (it is essential that all tracks are equalised, the two identical controls should have the equalisation reversed by applying susequently -1db @ 5KHz Q=1).

Note: the above items are generally agreed to be audible with music. Anyone failing single or all the above tests blind must be considered somewhere between Brass-, in- or Clotheared in the context of a DB Test (which measn that under testconditions they are de-facto of diminished hearing acuity) and should be excluded from all data anlysis.

4) Tracks where one includes polarity inversion in both stereo channels. Any music piece used must have been recordedd using minimal microphone arrangements and without mutitracking to ensure sensitivity to polarity inversion is not diluted by having different polarity tracks in the same track.

Note: Audibility of polarity inversion has been shown definitly audible with suitably constructed test-tones, but is considered questionable with music. Anyone passing this one blind should be considered an excellent DBT Subject for audio on subjects they have no strong opinion on or which they can be kept "completely blind".

Well, so much for theoretical and practical issues I can note and advise on.

Ciao T

 

If all of this is necessary hasn't the conclusion already written itself?

If indeed the questionable differences are so questionable that all these requirements need apply?

Notwithstanding someone with a specific hearing impairment outside of what is normal hearing acuity for their age, the audible differences attributed to aftermarket power cables (by the hi-fi press and so on) should be audible on more or less any system to anyone. Shouldn't they?

If not the conclusion is already apparent.

 

Hi,3Dsonics, I have said all along that we need to keep it simple the test that is.Reading your post started to give me a headache some I took a rest and came back to it. It seems that yourself and others want to turn this little,easy test in to a Oxford degree.
It is a shame that you have to mention different frequency and loudness.
You may not like my comments but this test of power cables could be so easy.And it seems people are afraid to put a test into practice, they seem to scared of trying something to help themselves in the future.
Apologies if you take offence at the above this not just a dig at yourself but to the others ready to slam if a cable may or not work.
Lee

 

Hi,

Look guys, I am not telling how to your run test as such. I AM telling you guys merely what needs to be done if you wish to illustrate that the test results are relevant and significant.

Otherwise you will likely find yourself with another pre-programmed null result.

BTW, have you guys considered approaching Tony Faulkner about taking part, this guy is a great blind tester if you want to know if something is really audible, he seems rather more able than most to ignore the actual test stress.

At any extent, do your "little, easy test" any way you like, but if you fail to demonstrate that your test would have reliably revealed known audible phenomenae you must accept to be told that it has no consequence whatsoever on the subject of suspected or alleged differences and hence should be sumarily ignored.

It's your call, do you want to know or do you merely (mis)use ABX/DB Testing to support the usual "debunker" mentality. If you really want to know, the requirements are set out partially in my various notes, I can suggest added sources if needed.

If all you want is to obtain another "there was not enough valid date to draw any conclusion and therefore the null hypothesis was not rejected" result which you than actually claim as "suggests no differences exist" you are taking the right approach in keeping everything very casual.

L8er T

 

Hi, If our little test worked and every person said the after market power cable made the most difference and they gave you a 5000 word document to prove this .By the wording of your reply, to me you still would not be interested.This is a test to try and satisfy everyone. It is a shame that people can not see this and do the test. Oh and every test know to man can be taken apart and put back together again,and faults will be found.
I can see that you must be a sceptic or you would put your name down on the list. I am not being nasty to you, this test could work if everyone opened their minds instead of typing away trying to make the test far to complicated .

Lee

 

Hi,

IF every person said this it would be rather interresting, HOWEVER, you might find that the numbers are smaller, much smaller than 100%.

Hmmm. I still note that we have a few fundamental issues here....

Funny, I already DID that sort of test (actually we compred the RG-213 version of the UBYTE-M against the RG-214/Mil one, with the main difference being that one cable has silverplated conductors and screen, the other is plain copper).

My point exactly. Thus if you wish to conduct a test you wish to reference publically, YOU MUST ensure you can answer criticism and suggests of faulty implementations.

I am a sceptic only insofar as I generally subscribe to the notions of Bishop Berkley and Hume.

In cables I actually happen to know a thing or two about what makes them tick.... You may find some of my articles useful:

http://www.tnt-audio.com/clinica/mains_e.html

http://www.tnt-audio.com/clinica/intere.html
http://www.tnt-audio.com/clinica/spkcbl_e.html

http://www.tnt-audio.com/clinica/diyce.html
http://www.tnt-audio.com/clinica/ubyte2e.html
http://www.tnt-audio.com/clinica/ffrc_e.html

Note the dates and sadly most of the parts/materials have become unavailable or have changed so they are not terribly useful.

I have organised quite a few blind tests myself. I am merely pointing out the issues you should consider. BTB, I have just done a similar "send out a bunch of different cables" test, for market research in the context of my own interconnect designs. The results suggest not only that people could hear a difference, but also that they preferred in general the same design. But you will not find me publishing the results, I had enough running battles on that subject.

Ciao T

 

of course the results will be less than 100% in favour of one cable. Here again you are picking holes again .I have posted on Hifi Wigwam That if some people had to change a light bulb some would sit it the dark first wondering which hand to use,stand on a chair or get a ladder.I am sorry you feel the need to read to much into what could be a simple productive, informative test.If it goes ahead or not least this idea has rubbed some of the letters off my keyboard.

Please this is not just a dig at you it is for everyone who has posted about the two different test.
Lee

 

I'm affraid 3dsonics has this about right............

You can only get peoples opinions from such a test, by posting it round. I can't see any way to corelate any results obtained.

I'll repeat get a uni involved they have the numbers of people, and an openness of mind . Have them set up a system were people can come and go after hearing a couple or three tracks played with or without the or with a control cable. The changes to be made behind the scenes at random.....you don't even need to tell people that they are looking for a change in sound. Just get them to say which they preffered.

Thats just off the top of my head.....so probabley has a number of flaws, but any results will be statistically significant.

 

I agree we need plenty of results.But all the extra's 3D's mentioned puts more people off than it attracts. That is why I have a problem with his idea's. Keeping it simple attracts more tester,which in turn gives more results to compare and you get the best conclusion in the shortest time scale.

Lee

 

Don't go anywhere near Thorsten's suggestion to build mains hi-fi cables out of aerial cable Lee.

Despite what he says no competent electrician would approve or recommend this much less this design having met the safety requirements of BS6500.

This cable from Supra is approved to HD 21.5.3 “Polyvinyl chloride insulated cables of rated voltages up to and including 450/750 V†under the low voltage equipment directive.

http://www.jenving.se/lorad.htm


..please continue