Speaker cable blind listening tests

[shokhead]

Only real test is without any devices and only your ears.

[musicoverall]

Fascinating. There is a 'prize' of several thousand dollars waiting from rec.audio.high-end to anyone who can pass a DBT of a pair of cables that have been vetted to be within 0.01 dB of each other at three FR test points. I urge you to consider taking them up.

http://groups-beta.google.com/group/...e=source&hl=en
.

I didn't see all the rules of the offer but in reading the earlier posts leading up to your link, it appears to be the same faulty test I've read about elsewhere. The participant travels to a foreign location and listens to an unknown system in an unfamiliar room where two sets of cables are swapped.

I'll pay you $10,000 if you successfully pass an eye exam that I will give you. You must simply detect which of two different shades of blue is which under an ABX format. You must score 15 corrects out of 20. You must also wear my eyeglasses during the test... and I'LL adjust the room lighting as I see fit. And since the test must be repeatable to prove anything, you must pass it 5 times for a total of 75 corrects out of 100 trials.

[krabapple]

Only real test is without any devices and only your ears.

Wow...no *devices*? Do live musicans play in your listening room?


Seriously, leaving aside your use of 'devices' to play back recorded music (these devices are OK...but a passive switch *isn't*?) you actually have hit on the reason double blind testing is used to substantiate claims of audible difference: it's the only method that ensures you're using *only your earss* to make the call. The typical 'audiophile ' method, on the other hand is readily 'contaminated' by what the listener believes about the devices. It happens even if you present the *same* device to the listener, twice. It's trivially easy to generate reports of 'difference' --even *vast* difference -- simply by leading a listener to believe that the first device is a high-priced boutique brand, and hte second is a mass market brand. Or simply by leading the listener to believe that that two are different devices.

Given how easy it is to 'fool' the ear in a *sighted* comparison, why treat them as 'informative'?

[krabapple]

I didn't see all the rules of the offer but in reading the earlier posts leading up to your link, it appears to be the same faulty test I've read about elsewhere. The participant travels to a foreign location and listens to an unknown system in an unfamiliar room where two sets of cables are swapped.

Nope, I don't know where you got your ideas, but in fact the test can be with the user's own system, using their preferred listening material...an important stipulation , though ,is that a mutually agreed-upon third party has to be there to proctor the test and its results, assuring e.g. that the setup was as per the 'rules'. Would Tom Nousaine do, for you?

I'll pay you $10,000 if you successfully pass an eye exam that I will give you. You must simply detect which of two different shades of blue is which under an ABX format. You must score 15 corrects out of 20. You must also wear my eyeglasses during the test... and I'LL adjust the room lighting as I see fit. And since the test must be repeatable to prove anything, you must pass it 5 times for a total of 75 corrects out of 100 trials.

Funny, but not apropos to the actual cable challenge. Having allayed your fears about the actual cable test, I urge you to go for it. Or at least do some bragging on rec.audio.high-end first; it would be most amusing.

[krabapple]

Carlstrom's data is from the seventies and Self's is from the eighties.

So..cables *now* sound different, but didn't *then*? Wow.

The NYT article was lacking in any details other than cables can offer the last two percent of performance.

..but without offering any evidence for that...

Got any real tests using something like Valhalla more recently than 15 years ago?

rw

Not me, personally...you? Got any data about Valhalla frequency response? If so I'll make a guess as to whether they'll sound different from other cables. For that matter, got *any* data showing speaker cable differences that can't be explained by simple RCL parameter differences?

I at least know where to find some measurement data about 'modern' cabling:

http://www.audioholics.com/techtips/...ewsfaceoff.php

Note the disturbing disjunctions between the claims the manufacturers make about their cables, and either scientific likelihood, or alas actual measurement.

Got any data showing that sighed comparison of speaker cables is a reliable way of evaluating their 'sound', and that DBTs *aren't* necessary to circumvent bias effects when comparing cables?

Really, do you think Elliot and Self are just *ignorant* of some new development in cables, cable-testing, or what? How about John Dunlavy?

How about interconnects? Here's a recent shootout:
http://www.hydrogenaudio.org/forums/...howtopic=33951


Lastly, audioholics has kindly provided a one-stop site for all cable nonsense debunking. Let's all thank them .

ttp://www.audioholics.com/techtips/audioprinciples/cables.htm

[musicoverall]

. Having allayed your fears about the actual cable test, I urge you to go for it. Or at least do some bragging on rec.audio.high-end first; it would be most amusing.

I may just do that - after I finish digging into the basics of DBT for music listening. So far, it's not looking too good for DBT as a viable means to discern subtle differences. There's really no good test so far as I can tell at this point. DBT looks good for drug testing; not so good for subtle differences in audio gear.

Do you have a link showing the official rules of this test? Anyone tried it as far as you know?

And I'm always happy to amuse people! I never fail to get tickled by two things from the naysayers. First, their zeal in arguing over something they believe is non-existent and second, their willingness to believe only when someone else proves something. I think the best form of humor is two people making each other laugh!

[mystic]

I may just do that - after I finish digging into the basics of DBT for music listening. So far, it's not looking too good for DBT as a viable means to discern subtle differences. There's really no good test so far as I can tell at this point. DBT looks good for drug testing; not so good for subtle differences in audio gear.

Do you have a link showing the official rules of this test? Anyone tried it as far as you know?

And I'm always happy to amuse people! I never fail to get tickled by two things from the naysayers. First, their zeal in arguing over something they believe is non-existent and second, their willingness to believe only when someone else proves something. I think the best form of humor is two people making each other laugh!

After reading krabapple's link to re.audio.high-end, it looks to me like they are saying they can make a zip cord cord sound like any audiophile cable by altering the zip cord (e.g., "add a few passive components to some zip cord to achieve the same FR"). It isn't clear whether they also would want to be able to alter the audiophile cable.

Requiring you to identify the cable correctly 15 times out of 20 to win the money seems to set the p value bar a little high(i.e., 0.02). Usually in hypothesis testing a p value of 0.05 is chosen, which I believe would be 14 out of 20. It's not much difference, but I guess they want an edge in case you get very lucky or hear a difference in the cables.

Without seeing a full description of their rules, I can't comment further on the cable challenge. Before participating in such a test, however, I would want everything in writing and in the form of an enforceable contract. Otherwise, you may have a difficult time collecting the money if you win.

It's hard to beat someone at their game.

[hermanv]

Fascinating. There is a 'prize' of several thousand dollars waiting from rec.audio.high-end to anyone who can pass a DBT of a pair of cables that have been vetted to be within 0.01 dB of each other at three FR test points. I urge you to consider taking them up.

It sort of unclear if the pool exists. The author says he has contributed $200 and others would contribute. He also says that he'll pay $1,000 if you go to England , I didn't see an actual address, bank account nor a enforcable contract.

ps. He is asking for +/-0.1dB, sort of reasonable where +/-0.01 dB is not. It is possible that two identical cables would not pass the +/-0.01dB hurdle especially if they are hand made. It also unclear where on could find an instrument that was 0.01 dB accurate. The best I've used has a resolution of 0.01 DB but nowhere near that accuracy (that's about one tenth of 1%; if you read the fine print 2% is common 1% rare for RMS meters).

According to his rules, one could easily cheat. He checks at 100, 1000 and 10,000Hz. A notch filter at 200-800 Hz would pass his rules yet any fool could hear it was there (you'd put it in one of those little boxes MIT cables seem to like so much).

This offer may have been made in the passion of cable discussions, it seems a little questionable.

[E-Stat]

So..cables *now* sound different, but didn't *then*? Wow.

You're right. Technology hasn't advanced at all in the past quarter century.

Really, do you think Elliot and Self are just *ignorant* of some new development in cables, cable-testing, or what?

We'll never know until they do so and update their comments.

Lastly, audioholics has kindly provided a one-stop site for all cable nonsense debunking. Let's all thank them .

They've done a most thorough job of quantifying LCR parameters. Yawn.

rw

[FLZapped]

Very well. Engineers have determined what the answer isn't.

rw

No, audiophiles just don't want to hear the truth. It doesn't justify their wasting of money.

[FLZapped]

My case:
1. Several people have repeatedly "passed" the double blind cable listening tests. Either they are psychic or there is in fact an audible difference between cables. Once any one person can do this repeatedly the argument that there is no difference is dead.

Is this published? Has it been independantly verified?

2. The mathematics of reactance calculations for cables that are a meter or so long and conventionally constructed show that the reactance components are tiny with respect to the audio band and the circuit impedances, this makes the reactance components of a cable extremely unlikely to cause the "cable's sound signature" under discussion.

Then you probably need to look for another variable outside the cable.

3. I have thrown away nothing of engineering knowledge; I have only said that the RLC model by itself does explain how cables work. (Example: it is possible to build a cable with identical RLC values to your twin lead example that is effectively worthless for transmitting RF signals.)

This is just another strawman. We aren't talking about rf cables.

In other words there is a fact that cables sound different.

Where? Citation of said fact, please.

There is a second fact that RLC does not seem to explain why they sound different. So the simple fact is that most people, and possibly no one, knows what causes the cable effect. I say most, because as I have said in other posts, there seem to be a small number of cable companies (examples like Kimber and WireWorld) whose cables get high ratings at many different price points. Even though they are built of different things someone there knows how to get good performance out of quite different ingredients.

Again, you should then suspect a variabe outside the cable. as for the commercial cable companies, they are in business to make money, the more, the better. Marketing rules the roost when it comes to maximizing profits.

So you seem to be saying that in spite of absolute proof that cables do sound different....

Where is your proof?

As an electronic engineer I was a powerful skeptic because I did the math. I concluded that a cable sound signature was an extremely unlikely phenomenon, but I listened and became instantly convinced that the RLC model couldn't begin to explain the differences I heard.

Have the results been published, reviewed and verified independantly?

-Bruce

[FLZapped]

It's curious that no audio cable manufacturer (and certainly no *high-end* cable mfr) has ever published any DBT results in scientific literature..or even in promotional literature, AFAIK. Apparently that includes Wireworld. Wireworld is wrong about ABX cable tests...they *can* be set up so that the user does the switching. Contrary to what Wireowlrd calims, they also certainly can be and are used in mp3 development (see www.hydrogenaudio.org). So Wireworld is a rather suspect source to me.

Yes, isn't it also amazing that not a single one has ever williningly participated in one either.

-Bruce

[musicoverall]

No, audiophiles just don't want to hear the truth. It doesn't justify their wasting of money.

LCR parameters being the whole story behind cable sonics isn't necessarily the truth and better sound via higher definition cables or any other means certainly isn't a waste of money. If you disagree, perhaps better sound isn't your goal.

[FLZapped]

LCR parameters being the whole story behind cable sonics isn't necessarily the truth

Prove it.

and better sound via higher definition cables

Prove any such thing actually exists.

certainly isn't a waste of money.

Prove it.

If you disagree, perhaps better sound isn't your goal.

Yawn......prove it.

-Bruce
(Astounded at just how idiot final statement is)

[hermanv]

Prove it.
Prove it.
Prove it.
Prove it.

Wow, boy, sure hard to argue with that. One can only hope you joined the school debating team. With such stunning repartee, with such carefully constructed and considered logic you must have been devastating to the opposition.

I've heard it said that ignorance is bliss; in that case one might fear you're getting too close to comatose (that might explain the yawn).

So what do you think? Tens of thousands of people have bought expensive cables and all of them are fools?

Every time someone does in fact prove that cables have a sound of their own, the opposition just moves the goal posts. First it was a probability of .05, then, if that is met, suddenly only .02 will do. If that is met, now we need an outside observer because we all know that audiophiles lie. Then we need it to be published and last we must have equations because the opposition has taken on all the aspects of a 6 year old saying; is not, is not, is not, is not.

The mathmatical proof for the 4 color map theory took 600 years to be developed, but anyone could easily prove by experimentation that 4 colors always worked.

Anyone with a decent system can easily hear differences in 2 cables; one generic copper, Radio Shack style and the other, one of the sneered at high price models. The really amazing thing is that the difference will be there, whether you can explain it or not.

Just listen, it's not that hard.

[Resident Loser]

...a fool and his money are soon parted...

So what do you think? Tens of thousands of people have bought expensive cables and all of them are fools?

Soooo...Most of the retail folks I'm acquainted with think it's one big hoot...put some Chinese wire in a fancy skin, stick it in a stained walnut box with some laser-etched logo, maybe a little velveteen, drawstring snood and yer good to go...

...The really amazing thing is that the difference will be there, whether you can explain it or not...

Did anyone say there couldn't be a difference? The only folks who seem to think that's what has been said are the aftermarket afficionados...countless times...Just like "...play it again, Sam..." No one, in my recollection at least, has ever said there can't be a difference....question is, is it improvement or merely difference?

Despite all the talk of "difference", the salient point is, under controlled conditions, I haven't seen any evidence that anyone can distinguish OEM stuff from the high-priced spread with any degree of statistical relevance...bada-bing, bada boom.

jimHJJ(...wire, it seems, is wire...)

[jneutron]

The mathematics of reactance calculations for cables that are a meter or so long and conventionally constructed show that the reactance components are tiny with respect to the audio band and the circuit impedances, this makes the reactance components of a cable extremely unlikely to cause the "cable's sound signature" under discussion..

What are you basing this on? Why are you saying the reactance components of a cable are extremely unlikely...to have an effect? Please explain yourself.

I have thrown away nothing of engineering knowledge; I have only said that the RLC model by itself does explain how cables work. (Example: it is possible to build a cable with identical RLC values to your twin lead example that is effectively worthless for transmitting RF signals.)..

You can? Please provide an engineering example to explain this..I of course, assume that you actually meant "carrying RF signals from a source to a load", as opposed to the transmission into free space.

In other words there is a fact that cables sound different.

You have not demonstrated, nor pointed to information which would support this statement.

There is a second fact that RLC does not seem to explain why they sound different. .

No, you have provided no facts in support of that statement.

So the simple fact is that most people, and possibly no one, knows what causes the cable effect. .

You would be incorrect.

I say most, because as I have said in other posts, there seem to be a small number of cable companies (examples like Kimber and WireWorld) whose cables get high ratings at many different price points. Even though they are built of different things someone there knows how to get good performance out of quite different ingredients..

"Someone" knows squat about engineering a cable. They do not know how to get good performance out of a cable. They know only how to put stuff together in a random, haphazard, illogical fashion, call that sillyness "engineering", toss some marketing crapola into a farcical "white paper", and charge big bucks to pay the bills..

So you seem to be saying that in spite of absolute proof that cables do sound different, if I can't replace the RLC model with another science model then suddenly cables can't sound different anymore?

I would LOOOOOVE to see that "other science model". Mind you, it cannot violate physics, it must be consistent with observation, it must be internally self consistent, it must be EXTERNALLY consistent with the known universe. It cannot use itself as a basis of justification...it must provide predictions which can be measured and verified..Failure in any regard there trashes the model.

As an electronic engineer I was a powerful skeptic because I did the math. I concluded that a cable sound signature was an extremely unlikely phenomenon, but I listened and became instantly convinced that the RLC model couldn't begin to explain the differences I heard.

I've no idea what math you are talking about. Please detail this.

I have discussed this with other engineers and the theories proposed slowly moved far out of the mainstream. I'm not willing to mention them here because the likely hood that they are the correct explanation seems pretty small.

You would be correct in that if the explanations are moving out of mainstream, they are incorrect.

Please elaborate, and perhaps I can explain why they are incorrect.

John

[jneutron]

The mathmatical proof for the 4 color map theory took 600 years to be developed, but anyone could easily prove by experimentation that 4 colors always worked.

NO. Nobody can prove by experimentation that 4 colors always works. They can only prove that they are incapable of devising a scenario that requires more than 4..

The mathematical proof does that.

John

[hermanv]

to jneutron:

With a know source and load impedance, the capacitance and inductance of a cable cause predictable deviations from a straight line frequency response (I left out resistance because in a simple model the volume control neatly compensates for that). I hope we agree up to this point.

When you calculate the amount of deviation caused by most cables, or use equipment to measure this frequency response deviation, the numbers come out in the range of 0.01 to 0.05 dB at 20kHz. There is very little musical information at 20kHz. In the mid-band, results of between 1/2 and 1/10 of that reduction are more typical. Those dB numbers calculate to ratios of around one to five hundreths of a percent mid-band.

There are two reasons why I believe that this level of differential attenuation can not explain the cable issue; First; 0.025 dB is not audible except under very special conditions certainly not with a psuedo-random signal like music. Second; Since a speaker that is flatter than +/- 2 dB is an extremely rare animal, it would appear to follow that attenuation due to cable reactance would be lost in the speakers frequency response inaccuracy.

So my problem is that I easily hear differences between one cable and another but the math says changes due to reactance are very small indeed (and also vary rather smoothly mostly at the ends of the frequency response curve). This is why I believe that the math of RLC on cables doesn't support the differences I and others hear easily.

NO. Nobody can prove by experimentation that 4 colors always works. They can only prove that they are incapable of devising a scenario that requires more than 4.

There are a finite and limited number of possible edge interactions of two dimensional shapes on a two dimensional surface. It is perfectly posible to prove empirically that every possible combination is covered with 4 colors. The number of permutations just isn't that great.

It is falacy and psuedo science to believe that only mathematical proofs are valid, graphical or logical proofs are equally valid. It was known, not suspected, that 4 colors would always work long before the mathmatical proof was developed. The mathematical proof was an interesting exercise not a requirement to validate the use of 4 colors.

I can prove a square peg will not fit in a round hole without any use of math whatsoever.

[jneutron]

With a know source and load impedance, the capacitance and inductance of a cable cause predictable deviations from a straight line frequency response (I left out resistance because in a simple model the volume control neatly compensates for that). I hope we agree up to this point.

The resistive component affects how well the amp controls the load, so cannot be eliminated by simple gain adjust. It varies with frequency. And, amp control also depends on the damping factor during all 4 quadrants of operation into a reactive load, as well as feedback loop coupling to the supply rails, this being different depending on the half plane of operation. Series resistance softens the high speed current demands.

The inductance is a frequency dependent entity, it changes 30 nH per foot from DC to infinite frequency. This is the internal inductance of a conductor, and is geometry dependent.

The capacitance of some dielectrics varies over frequency.

Some geometries, ribbon for example, starve the dielectric as a result of proximity and skin effect.

Did your calcs include these?..My guess is no.

Where did you accomodate the lagging storage mechanism of L and C, for an impedance mismatched cable driving a load?? What effective ITD vs frequency boundary condition did you use? What overall localization bound did you set for ITD and IID limits?

When you calculate the amount of deviation caused by most cables, or use equipment to measure this frequency response deviation, the numbers come out in the range of 0.01 to 0.05 dB at 20kHz. There is very little musical information at 20kHz. In the mid-band, results of between 1/2 and 1/10 of that reduction are more typical. Those dB numbers calculate to ratios of around one to five hundreths of a percent mid-band..

I would be interested in hearing how these were measured. Measuring .01 or even .05 dB accuracy levels of the current driven into an 8 ohm pure resistance at 5 to 20 khz is beyond the vast majority of electrical engineers...they do not even understand that there are errors. Mag field time derivative errors bollux both current and voltage reads. And, that isn't even getting into the acceleration or acceleration derivative mass loading changes on the drivers, or the derivative flux eddy current reactions within the pole piece and driver ring (did you forget Lenz's law??) (I hate it when that happens)

There are two reasons why I believe that this level of differential attenuation can not explain the cable issue; First; 0.025 dB is not audible except under very special conditions certainly not with a psuedo-random signal like music. Second; Since a speaker that is flatter than +/- 2 dB is an extremely rare animal, it would appear to follow that attenuation due to cable reactance would be lost in the speakers frequency response inaccuracy...

hmmm....025 dB is not audible....with monophonic reproduction, that is a true enough statement. I do not believe we are discussing mono jnd's..so, your calcs and their application do not coincide.

So my problem is that I easily hear differences between one cable and another but the math says changes due to reactance are very small indeed (and also vary rather smoothly mostly at the ends of the frequency response curve). This is why I believe that the math of RLC on cables doesn't support the differences I and others hear easily....

If you use monophonic soundfield analysis, yes, your belief could be valid. Unfortunately, you are using the incorrect analysis tool for the problem of stereophonic soundstage image generation.

You did not use the correct model, and as such, you could only come to the conclusions you have arrived at. Reasonable ones, but incorrect.

Proving that RLC metrics do not predict the cable "sound" cannot be done with the rudimentary model you discuss..You need a better model...

The fun off topic stuff:

There are a finite and limited number of possible edge interactions of two dimensional shapes on a two dimensional surface. It is perfectly posible to prove empirically that every possible combination is covered with 4 colors. The number of permutations just isn't that great.

Playing with anything, maps, whatever, and being unable to prove something incorrect, is not proof that it is correct. Shame on you..you know better.

It is falacy and psuedo science to believe that only mathematical proofs are valid, graphical or logical proofs are equally valid. It was known, not suspected, that 4 colors would always work long before the mathmatical proof was developed. The mathematical proof was an interesting exercise not a requirement to validate the use of 4 colors..

Playing with map colors for years only shows that, for years, nobody has needed 5, not that it is impossible..stating that because it has not been done, it is impossible, is not a proof that it cannot be done...merely, that it has not yet..

I can prove a square peg will not fit in a round hole without any use of math whatsoever.

No, you could only prove that you cannot do it..

But I can..two inch round hole, 1 inch square peg.. ...ya gotta tighten those sentences up..that hole was mack truck sized...had ta take it, ya know..

A topologist would also have fun with your statement...

T'was a pleasure..thanks
Cheers, John

[hermanv]

The capacitance of some dielectrics varies over frequency.

Some geometries, ribbon for example, starve the dielectric as a result of proximity and skin effect.

Whoa. Many of the discussions on these threads center around RLC models and cables. I have always maintained that these simplistic models do not explain what is going on. In particular a mathmatical model that uses only pure R, L and C such as a first order simulator, will quickly lead you to believe that the whole cables have signatures thing is nuts.

Once you add second or third order effects such as dieletric absorbtion vs frequency, or current hysterysis of magnetic domains in cable alloys or skin effects you are no longer discussing the simple RLC model. The fact that a capacitor has a dielectric still doesn't excuse the generalization. People speak here in terms of asking for proof that a mathmatically perfect capacitor could cause the effects under discussion and it can not. So to many people, the fact that the simple capacitor model can not thereby means that the effect also can not exist.

I am firmly convinced that just the first order RLC effects are not what I am hearing, but I can easily believe that other more exotic effects that you start to list could be what's going on. That was absolutely not the context in which the whole RLC issue was being discussed on this thread.

There are people who essentially claim that if a pure 100nF inductor in series with a 1K Ohm resistor and 100pF to ground capacitor doesn't really affect the sound then neither can a cable with the same measurements. I keep trying to say that's true as far as it goes but it doesn't explain the issue. The cable may have those same parasitic reactances but it also has more complex things going on. It was always my contention that the other effects might explain the auditory differences.

In discussions with other engineers we have considered things such as differential thermionic heating in connector junctions, the dielectric absorbtion you cite and things such as low quality diode junctions at grain boundaries. One big problem is that there is little test equipment that measures Micro effects while in the presense of Macro effects. i.e.what happens to tiny signals while significant current is flowing to the woofer, does it cross modulate? Its not all that hard to do, just very little commercial equipment does it.

This is a discussion group, not a sceintific journal. Short hand is acceptable, spending hours looking for holes in logic while ignoring the main subject benifits no one except the oversized ego. Example; a 1 inch square peg falling through a two inch round hole simply does not meet the definition of fit as used in this context. Shame on you.

Playing with map colors for years only shows that, for years, nobody has needed 5, not that it is impossible..stating that because it has not been done, it is impossible, is not a proof that it cannot be done...merely, that it has not yet..

You insist on missing the point, you don't need years, less than an hour will exhaust every possible combination, there are none that can not be thought of. People knew with absolute certainty that no other answer was possible. It was accepted as a proven long before the math, protest if you wish, but that will not change the history or the fact that forms of proof other than equations exist.

[jneutron]

Whoa. Many of the discussions on these threads center around RLC models and cables. I have always maintained that these simplistic models do not explain what is going on. In particular a mathmatical model that uses only pure R, L and C such as a first order simulator, will quickly lead you to believe that the whole cables have signatures thing is nuts. .

If you use the human hearing model that we can only hear 20 to 20K, with monophonic JND parameters, then your statement is entirely correct.

You have missed one of my points..the criteria for judging the audibility of any effects is incorrect. You did not consider at all, the human hearing sensitivity to either ITD or IID, and it is these parameters which we humans use to discern the direction and distance of a sound source..

So, it is moot to declare even simplistic RLC parameters as not having an effect, as your model for what we hear is so inaccurate..Once the criteria for what we hear w/r to soundstage is firmly established, then simple RLC can be re-calculated for the possibility of exceeding the threshold of audibility.

Once you add second or third order effects such as dieletric absorbtion vs frequency, or current hysterysis of magnetic domains in cable alloys or skin effects you are no longer discussing the simple RLC model. The fact that a capacitor has a dielectric still doesn't excuse the generalization. People speak here in terms of asking for proof that a mathmatically perfect capacitor could cause the effects under discussion and it can not. So to many people, the fact that the simple capacitor model can not thereby means that the effect also can not exist..

To me there is no distinction when it comes to the RLC model, as I do not consider it as a simple frequency or impedance invariant descriptor. Again, the bulk of discussion here is attempting to fit the wrong human hearing model to the problem..You have fallen into the same boat, as you have also erred in this fashion. I like the fact that you seem to be attacking the issue scientifically, but you have not realized that the hearing model is not up to the task you are working on..

I am firmly convinced that just the first order RLC effects are not what I am hearing, but I can easily believe that other more exotic effects that you start to list could be what's going on. That was absolutely not the context in which the whole RLC issue was being discussed on this thread. ..

Perhaps, perhaps not..until the hearing is characterized, that statement is unconfirmed. The context of this thread is that we cannot hear RLC issues.."proven", by the application of circuit equations to a wholly inaccurate hearing threshold set..

There are people who essentially claim that if a pure 100nF inductor in series with a 1K Ohm resistor and 100pF to ground capacitor doesn't really affect the sound then neither can a cable with the same measurements. I keep trying to say that's true as far as it goes but it doesn't explain the issue. The cable may have those same parasitic reactances but it also has more complex things going on. It was always my contention that the other effects might explain the auditory differences...

I've not stated either way on this issue. I am more interested in applying the equations to the real criteria for hearing, not the old one..

In discussions with other engineers we have considered things such as differential thermionic heating in connector junctions, the dielectric absorbtion you cite and things such as low quality diode junctions at grain boundaries. One big problem is that there is little test equipment that measures Micro effects while in the presense of Macro effects. i.e.what happens to tiny signals while significant current is flowing to the woofer, does it cross modulate? Its not all that hard to do, just very little commercial equipment does it....

Hmmm...cool. I'm really beginning to like you...

Peltier and Seebeck effects are not symmetric..so, from physics, if this reaches a high enough level of conversion, there should be some signal non-linearities caused by non linear heat loss... Peltier defines the heat produced as proportional to the current, while Seebeck has the voltage proportional to the temp difference..for high impedance signal runs, the heat across the junction will be miniscule, therefore the seebeck conversion will be very very small, while peltier heat generation will be maximized due to little heating and small thermal gradient...

Again, both effects are incredibly small, and subsequent non linear loss is well below the threshold of detection for audio rate signals...not zero, but many orders of magnitude below the equipment.

Diodes at grain boundaries...hmmm..so far, the test guys here (I'm part of that group) haven't seen any such effects, but we only go up to 30 Kiloamps and look at the microvolt level. As such, only a upper limit of effect can be spoken of..I do feel, however, that this span of range is a tad beyond "high end audio"..

If you wish to discuss "micro-diodes", you have to first get around the fact that grain boundary interfaces are simply lattice discontinuities and the dimensions of the discontinuities are well within the mean free path of the "electrons". IOW, they really don't see the boundaries per se, but will only suffer a relatively minor increase in the number of collisions as a result of lattice defects. Given that at 300K, the mean free path for copper is about 6 orders of magnitude below the grain size, this incremental increase in resistive loss is, shall we say, of little concern?

This is a discussion group, not a sceintific journal. Short hand is acceptable, spending hours looking for holes in logic while ignoring the main subject benifits no one except the oversized ego. Example; a 1 inch square peg falling through a two inch round hole simply does not meet the definition of fit as used in this context. Shame on you. .

This is the AUDIO LAB (boomy echoey voice)..scientific discussion is encouraged...go as deep as you wish...
You mentioned the hole and peg..it took about two seconds to come up with a response to that...

As I stated, that was the off topic fun stuff, and I did not ignore the main subject..so I kept it as last..

What you have missed, is WHY I responded to the hole and peg statement..obviously, it was too subtle..my fault..

You have calculated what simple RLC does to the signal, and then applied that result to the current outlandishly incorrect model of human hearing, and decided there was more to "cables" than can be explained using the RLC metrics..

You are guilty of trying to fit your square peg of RLC calculations, into the round hole of human hearing capabilities..ok, I was too subtle...shoot me...

This discussion is not over, not by a long shot...I hope we can continue this, I have good feelings about this..

Thank you.

Cheers, John

[FLZapped]

Wow, boy, sure hard to argue with that. One can only hope you joined the school debating team. With such stunning repartee, with such carefully constructed and considered logic you must have been devastating to the opposition.

HAHAHHHAHAHHAHAHHAHAHHAHAAHHA

This guy has been going on and on without any evidence to back his assertions up: even after multiple requests. At some point, one just has to be blunt.

-Bruce

[FLZapped]

Anyone with a decent system can easily hear differences in 2 cables; one generic copper, Radio Shack style and the other, one of the sneered at high price models. The really amazing thing is that the difference will be there, whether you can explain it or not.

Uh, huh, under what controls?

Just listen, it's not that hard.

Maybe you should go over to this site and investigate Sheperd's Tones and the Tritone Paradox, and see just how hard listening can be. (requires JAVA)

http://www.cs.ubc.ca/nest/imager/con...Illusions.html

-Bruce

[FLZapped]

Whoa. Many of the discussions on these threads center around RLC models and cables. I have always maintained that these simplistic models do not explain what is going on. In particular a mathmatical model that uses only pure R, L and C such as a first order simulator, will quickly lead you to believe that the whole cables have signatures thing is nuts.

I don't think anyone has made any mention about how simple or complex a particular model of L, C, or R may or may not be.

Once you add second or third order effects such as dieletric absorbtion vs frequency, or current hysterysis of magnetic domains in cable alloys or skin effects you are no longer discussing the simple RLC model. The fact that a capacitor has a dielectric still doesn't excuse the generalization. People speak here in terms of asking for proof that a mathmatically perfect capacitor could cause the effects under discussion and it can not. So to many people, the fact that the simple capacitor model can not thereby means that the effect also can not exist.

And that has to be counterbalanced against the likelyhood of their ability to effect a circuit, when the circuit impedance is extremely low. In the case of an audio pre-amp, many are going to be 50 to 100 ohms output impedance. An audio power amplifier, will be orders of magnitude below that. One has to decide if it would be worth the effort to include secondary effects when they will most certainly be swamped out by the circuit impedance.

Jung in his paper on picking capacitors made the observation that as the circuit impedance went down, so did secondary effects.

This site will give you a general idea of how that works:

http://members.fortunecity.com/flzapped/ouch_intro.html

You will notice that the signal level created by piezo-electric effect decreases from nearly a volt peak-to-peak to about a half-millivolt peak-to-peak as the load is decreased from infinity to 100 ohms.

An example of a mylar capacitor model for dielectric effects can be found here:

http://www.national.com/rap/images/capsoak4.gif

Please note the extreme impedances involved and ask if you think it has much chance of impacting a circuit who's primary impedance is 50 ohms or less. Also consider this is for a 1uF (1 E-6) part and the capacitance of the average 6 foot interconnect cable will be 180pF (180E-12)

Certainly if I were designing a pre-amp for a piezo-electric guitar pick-up, or an instrument amp with 1,000,000 ohms circuit impedance I would be very careful about my choices. At 50 ohms(or less), I would hardly give it much thought.

-Bruce