LCR and cables

Although Measuring LCR, did they ensure that LCR was the same amongst the cables tested?

 

Absolutely Nick, that says it all.

 

Hi Nick,

My point exactly.

Again, my point as well, in fact it is something along the lines that we use internally.

One may contest that some differences between cables can also be very obvious, a favourite one of mine was to use the cheap cables that are included with any budget CD-Player and to compare it to a cable made from CT-100 Coaxial Satellite Cable (or RG-214/Mil) with high grade plugs.

The difference is quite obvious, even though the electrical parameter differences between the cables are minimal, virtually same capacitance, inductance and resistance are a little lower for the CT-100, but in a line level connection with an output impedance of a few 100 Ohm and at least a few KOhm input impedance the differences in L & R fall below any aknowledged treshold of hearing.

To take another extreme, if we are comparing several different grades/types/makes of PTFE insulated computer ribbon cables (and commercial audio cable derivatives) fitted with WBT or Neutrik plugs the differences become very small and appear below the noisefloor of any experiment I have been able to construct.

Yet some individuals at least appear to hear substantial differences and think them worth the extra money, who am I to argue with them that they should not hear the difference or not spend their money?

So, at the bottom level of quality at least (and also with cables that deliberatly manipulate the signal using networks etc.) differences between cables can be much larger than they should be, once a certain level of quality is reach in "non-manipulative" cables differences become very difficult to spot reliably.

Oh, to get back on topic, I do happen to know the "ideal" LCR Parameters (also conductance G and dielectric absorbtion DA) for interconnects, speaker cables and mains cables, but:

A) I am not telling

and

B) They are anywhere between somewhat and heavily context/equipment dependent, so one set of L/C/R/G/DA parameters may not be ideal for some equipment, but can be ideal for other equipment

and

C) While the LCR parameters have a significant impact on the cable's behaviour (I called them "first order effects" in my Cable articles on TNT-Audio) other effects do exist (yes, I know what they are but I do not shouting into the wilderness either - read my articles for a starting clue)

Kind regards Thorsten

 

Hi,

I remember the cables where basically comparable in parameters, but depending on what degree of precision instruments you employ arguably no two cables (even from the manufacturer and type) have EXACTLY the same LCR parameters.... Where do you draw the "not the same" line?

Ciao T

 

Thorsten,

I stand (well sit really) corrected. Do you have any info on the JAES article and I try and get a copy to peruse?

 

that comment shows no comprehension of what sample size in statistical analysis is for....

DBT are a waste of time with fewer than 150 AS AN ABSOLUTE MINIMUM thats why they are a waste of time and effort...

if you find cables sounds no different it just means your not hearing it, full stop ....it doesn't mean there are no differences.

 

Actually it does - significance is in interaction between size of observed effect and sample size. if you use students t and have a large enough effect then a sample size of 20 is certainly feasible, as would rank correlation testing on say 20 cables with 20 observers. I am sure a test of this type would conclusively show the difference between say AM and FM radio.

You are speaking as if it is proven. It isnt. If it was as you say then there would be no contention just as in the above example.

The effect - if it exists - is extremely small and for the vast majority of over priced cables (say anything above qed cable) the money would be better spent elsewhere. Much to the detriment of the retailers margin. The only thing worth spending money on is good connectors.

 

Actually guys IMO sample size does not matter. If we have a 1000 people testing 2 cables and only 1 is able to consistently/reliably identify the cables in a DBT then there must be a difference. Even if the difference is so small that 999 others don't hear it. The trouble is that we haven't yet been able to conduct such a test for all the arguing.

VFM and dealer margins are another topic.

 

Just for the hell of it I spent yesterday afternoon swapping around digi cables between a Benchmark DAC1 and Teac T1.
The DAC1 is alledgedly cable agnostic but I know from experience it ain't.
The cables were ;
Cambridge 75 ohm
NVA 75ohm soundpipe
Belden 75 ohm
Sound tech video(75 ohm)
VanDamme 75 plasma grade
Ixos(the red stuff-about £30)
The differences wree VERY easy to hear.As a rough guide the differences were rather like swapping different amps into the system....
The Cambridge had such a pronounced presence boost I had to drop the gain on my HF power amp, it was in yer face and sounded like it was clipping.
The Belden produced another bright presentation false like a 70's SS 'hifi'amp with little or no depth.
The SoundTech was surprisingly well balanced and even response with good stereo and depth-things felt under control.There was a slight grain to everything-a veil to it but it only costs a tenner.
The VanDamme(my present choice)had a deeper bass more detail and the veil/hf grain had dissappeared.
The NVA offered an edited version of the VD presentation with a soft wooly sound to the bass-it was however pleasant to listen to even tho' it had shortcommings.
The IXOS was like dropping a valve amp into the system in terms of freq balance silky sweet treble, huge deep sound stage very clear mids and a slightly soggy bass-this sounded analogue and incredibly easy to listen . It was a if the extra space was paid for by the removal of some things from the mix though.
FWIW The GHz nature of digital signals means mere LCR isn't enough, there are other parameters that come into play on how a cable will perform-I'm sure these differences are measurable given the right kit so no dispute with the numbers boys here.
Oh and I did return my HF amp gain to its original setting after the cable talk
I've left the IXOS in for now as it is a rather pleasant change safe in the knowledge I'll return to the 'honesty' of the VD by about wednesday!
Some may label this post as 'boring' and of no scientific value (like I give a shit)but so what?
If you have a spare afternoon and some old (cheap) digi cables lying round it's a fun thing to try.

 

Dev,

Actually it does matter - out of 1000 people one might do it just by chance. Thats the point

Nick

 

a better example would be dab and fm ....and not with so called audiophiles, just general run of the mill joe public.

Theres no fudging the issue ...the bigger the sample size the more accurate the analysis will be or "Larger sample sizes confer greater power on statistical tests"

A recent so called dbt test [published on the web and linked to on this forum] carried out using if memory serves 8 people and a number of questions is a prime example of how a little knowledge can be very dangerous. At the time I did back of the envelope check and found that there sample needed to be over fifty not 8 to be able to draw any conclusions.

Its totally impractical to do this unless your a university or a research establishment where you can organise the required sample size, setup the filter questionnaires prior to listening and the the listening results each conducted singly [so to avoid peer pressure] ....the list is endless any one item if not carried out will make the test statistically inviable.

As to your quote of any effect being "extremely small" your position already skews the argument by making the claim appear as "a given fact" which plainly it is not, regardless of anybody else opinion.

Its as simple as this .....if you can't hear any difference then you have no reason to buy anything more exotic than bell wire....if conversely you have bought exotic cables you must be able to hear the difference otherwise you would have taken them back as being ineffectual and got a refund. There can be no other position there is no grey area...there either is an audible difference or there is not, but remember this is unique to each person.

So really this argument is totally superfluous if you can't hear the difference cables make no amount of hyperbola will change your mind. Don't forget that works both ways.....

one other point cables believers never become cable skeptics...but cable skeptics do become cable believers [imo]

I know because I was one for my first 10 years as a hifi enthusiast !

As to plugs again I'd have to disagree expensive plugs ie silver bullets do make a small difference but are not cost effective If you'd read the backoff write up in the articles section you will discover why myself and those present thought that way ...my opinion has not changed since.

 

If he/she can do it successfully, say 20 times, he/she is either psychic or extremely lucky. I wouldn't mind that gift (I think). IOW the sample size of the testers doesn't matter but the number of tests does. Sorry, I should have clarified.

 

Before you conduct any DBT how do you ascertain wether the system/speakers/music used in the test can differentiate/resolve sound differences between cables?(if they exist, obviously)

 

IMO it doesn't matter. If a "believer" believes he/she can identify a cable in a given system, let him/her decide which system to use. The simpler you make the test the easier it is to implement and draw conclusions.

 

Rubbish.

I think you mean:

"if you find cables sounds no different it could mean your not hearing it....it doesn't necessarily mean there are no differences."

 

Hi,

This depends on what the purpose of the test is. As "pro cable" posters are often reminded, extraordinary claims need extraordinary proof, this blade cuts both ways.

If we desire to demonstrate inaudibility of differences other than LCR in cables we must first demonstrate that all "small but agreed to be audible" differences are reliably detected, without this "calibration" any test cannot be considered to offer a wide applicability (that is one outside the exact experimental setup).

If all we do is to make a personal purchase choice, we may do whatever we please.

BTW, I currently do not have access to a library of the JAES nor do I have the references immediatly handy, I'll try to dig when I have time.

Kind regards T

 

Or there is a bias or giveaway.Anyway the point is even in the large sample tests - there is no significant audible difference.

Sample size most definitely does matter. However this is in conjunction with the size of the observed effect. To capture the significance of a very small improvement you would need many more observations. These two factors always exist in conjunction and cannot be taken in isolation.

The fact that an issue is in contention for cable but not FM v AM says something as it is itself a type of very loose significance test - but one using many thousands of opinions. This places limits upon the size of the effect - otherwise no-one would contend it.

If one uses 100 quid cables throughout ones system then maybe the total cable cost might be a grand. In almost every case (whether or not effect exists) the money would be better spent on better components - until the point where the cost of incremental improvements in components matches the increase in cable cost going from something like QED to something like Nordost et al. Given that good quality cables with excellent connectors like neutriks can be had for £30-40 then I would venture this would not occur in any system that was under £50k or so, even if the effect exists.

Buying something because you thought you heard a difference does not mean that difference exists. Thats the crucial flaw in your logic. It either does or does not make a objective difference - its not dependent on who is listening, though some might not be able to hear it which is why hearing tests and calibration would also be necessary in a rigorous test (I have done so in the past) to see if there is a correlation with some other factor to explain why if there is a difference then some hear it and some doubt.

However cable sales have all the hallmarks of every snake oil scam which is also a powerful argument.

Also consider that there is no scientific evidence to back this up and nothing has been measured. Some of the claims like cable directionality disagree with every known law of physics in fact, so the onus of proof is therefore on the believer.

 

I agree, thats the moot point isn't it? I described earlier clearly audible differences with different digital cables(a less controversial area IMO) and would expect any doubters of my sanity to come and hear for themselves-the percieved differences might not be apparent over a different system (and so the DBT would be pointless).

 

Show me the maths.

If you personally can get 14/16 in a trial of cables then I think you'd have made your point. I think 12/16 would be looking pretty good.

Paul

 

Hi,

What about 4 or 5 in 5 trials?

Most people have serious attention span limitations, I find 5 trials in a row is the max I find to give good results, repeatably.

And if the same person under the same conditions scores 4/5 four times I would suggest that the result becomes 16/20 and thus is quite significant, statistically speaking.

Ciao T