Objectivist or Subjectivist? Give Me a Break

[JackD201]

Well, actually, that's only true if the cable is poorly designed.

The claims made for many varieties of audiophile cable would imply that the wire between your telephone (assuming you have a land line, yes, I know, that's old-fashioned) and a central office 2 miles away could not support communication. I refer specifically to a large variety of "grain" and other misunderstandings in how a signal passes down a cable.

Let me ask you something: How far do you think an electron moves down a 12 gauge wire conducting 10 amps DC?

Dear Sir, since you've already been kind enough to establish my point that not all cables are the same by providing the example of poorly designed cables, would you be so kind as to please define what a "poorly designed cable" is and in so doing define what a "properly designed cable" should be.

 

[JackD201]

It is true that there can be small differences between cables due to easily quantifiable and measurable differences in resistance, capacitance and inductance. If the reviewers were required to establish their credentials by demonstrating an ability to discriminate between cables with different R, C, L characteristics in double blind tests, then their reviews of specific cables might be of more worth. And it would add to the value of the review if they (or the manufacturer) supplied data on the measured characteristics of the cable.

Another aspect that I find puzzling is that if different cables really do have audibly different signatures, then it is surely obvious that one should try to correlate these differences with the measurable quantities, so that one could try to home in on "better sounding" cables by selecting/designing them to enhance the measurable characteristics that correlate with the improved sound? And yet many audiophiles seem to be remarkably incurious about *why* their prefered cable sounds better. I suppose this comes back to an earlier comment I made, that some audiophiles have an almost theological need to believe that these questions cannot be adressed by science.

Chris

My mantra is stop spending at the point where you can't honestly tell the difference. Unless of course it just looks cool but that is a totally different subject.

 

[cnpope]

My mantra is stop spending at the point where you can't honestly tell the difference. Unless of course it just looks cool but that is a totally different subject.

Mine too. Which is why I use cables from the local hardware store.

Chris

 

[Groucho]

Dear Sir, since you've already been kind enough to establish my point that not all cables are the same by providing the example of poorly designed cables, would you be so kind as to please define what a "poorly designed cable" is and in so doing define what a "properly designed cable" should be.

He was onto a loser there: if he hadn't said "poorly designed", someone would have said "but what about a cable without a screen, or 1 km long?". Because he did say it, it opened him up to the charge of admitting that not all cables are equal. He couldn't win.

 

[JackD201]

It's all good then right?

 

[JackD201]

He was onto a loser there: if he hadn't said "poorly designed", someone would have said "but what about a cable without a screen, or 1 km long?". Because he did say it, it opened him up to the charge of admitting that not all cables are equal. He couldn't win.

I'm not in this discussion to score points. I was never into pissing contests and never will be. I know the limitations of my own knowledge and I am not ashamed to say these limitations are great. Even if I were into forum sports, I would not count J-J out. He knows the answer to my question.

 

[Atmasphere]

It is true that there can be small differences between cables due to easily quantifiable and measurable differences in resistance, capacitance and inductance. If the reviewers were required to establish their credentials by demonstrating an ability to discriminate between cables with different R, C, L characteristics in double blind tests, then their reviews of specific cables might be of more worth. And it would add to the value of the review if they (or the manufacturer) supplied data on the measured characteristics of the cable.

Another aspect that I find puzzling is that if different cables really do have audibly different signatures, then it is surely obvious that one should try to correlate these differences with the measurable quantities, so that one could try to home in on "better sounding" cables by selecting/designing them to enhance the measurable characteristics that correlate with the improved sound? And yet many audiophiles seem to be remarkably incurious about *why* their prefered cable sounds better. I suppose this comes back to an earlier comment I made, that some audiophiles have an almost theological need to believe that these questions cannot be adressed by science.

Chris

Touchingly naive. The data would all look the same, with minor variations, and anyone could come along and duplicate those characteristics at about $0.10 per foot.

From what I have seen, the cable sector seems to have the best markup.

I think that the real world changes of the signal due to R, L and C in a normal audio system would be tiny - we can simulate it to confirm that - so no one could learn to distinguish between them, anyway. The cable is just a tiny element in a chain of hundreds of components, electrical connections, and $0.10-per-foot cable in the recording studio, pressing plant etc. Can we find an authoritative audio design text book that teaches prospective audio designers what to look for in a cable, beyond the obvious shielding and mechanical aspects? I don't think so. Audio cable 'design' lives in a technical vacuum, because there is nothing to know or learn about it, beyond the obvious. Everything else is just flummery.

The R,L and C come together to create something called Characteristic Impedance. This is a property of all cables as a transmission line. It is also something that can be measured, using a Time Delay Reflectometer (TDR). Essentailly, the CI has to do with the impedance wherein a signal is not reflected at the output back to the input but is instead completely absorbed by the load. People familier with RF transmission lines will be very familier with this concept. The TDR, BTW, does its work at RF frequencies. The popular notion is that whatever is going on at radio frequnecies has nothing to do with audio.

I was involved in a rather ambitious cable project back in the early 1980s where we explored this issue. What we got from the study is that CI has an influence at audio frequencies, like it or not (imagine the signal being reflected back to the source as a noise component...). However the influence is not nearly so profound as what one encounters in the RF realm! But it might explain why different cables sound different; to control a transmission line you need a termination load and this is something that is lacking in the audio world- sometimes we load the cable at 100K ohms which is like no load at all; sometimes it is at 10K ohms which is still outrageously high. For several decades the pro audio world did their terminations at 600 ohms which is still on the high side (the standard was based on an air dielectric and the spacing you see on telephone poles) but despite the error it does seem as if cables terminated at 600 ohms are essentially inaudible without regard to the cost of the cable.

We were able to measure differences in CI with various cables available at the time. Otherwise with regards to the audible artifact of cables here are some interesting tidbits:

The telephone company is the one that really explored the effects of cable on sound. Prior to the introduction of balanced lines, the longest 'phone calls possible were only across the state, and such calls required some serious yelling to be heard at the other end. After some research on the subject the 'phone companies began using balanced lines and trans-continental phone calls became possible and routine. It was at this point that the technology gained serious use in the recording industry- resulting in the beginnings of the hifi era. Single-ended cables do indeed have artifact, chief amoungst them typically being a high frequency rolloff. How much of this can you hear with shorter cables? It does seem that a lot depends on your system. The higher impedance the preamp output and power amp inputs, the more you will hear it. This does mean that if you run entirely solid state you are less likely to hear artifact from the cables as the impedances involved are much lower and interact with the cables less.

You can measure the effects of audiophile cables- the trick is to measure the system spectral bandwidth in the room rather than the cable itself. On the bench it seems harder to quantify the effects of one cable to another like you can measure in a functioning system, and the effects of a given cable will vary from system to system. A simple trick- make the cable in question longer and see if its easier to see its effects.

Electronics engineers go to college and university, pass exams, get qualifications. As far as I am aware, there is no literature on the audbilitiy of cables, and no electronics engineer could refer you to a text book on the subject, nor find any information on how to design a cable that sounded better than another. For unqualified people to come along and tell the engineer that there is something about electronics that he doesn't know, yet for there to be no way he can learn about it - because there is no literature on the subject, nor any teacher of the subject - is a bit of a conundrum.

One of the difficulties in obtaining a doctorate is finding a subject that is in some way original and can contribute to the sum of human knowledge. So has anyone, ever, gained a doctorate through the study of the audibility of cables? Why not? It's a wide open field.

I think we know the answer.

 

[Groucho]

The R,L and C come together to create something called Characteristic Impedance. This is a property of all cables as a transmission line....The popular notion is that whatever is going on at radio frequnecies has nothing to do with audio.... But it might explain why different cables sound different...

Not exactly what it says here:
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf

...at audio frequencies, a cable less than 2,000 ft long is no more complicated than its series resistance and parallel capacitance. As the cable becomes longer, or as frequency increases, the cable will begin to behave as a transmission line.

We could put some figures on it: would the lack of a proper termination on a 0.5m cable at audio frequencies account for more than 1 LSB of error in the received signal (e.g. 16 bit)?

 

[Atmasphere]

Not exactly what it says here:
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf

Yes, I did use layman's terms- that document goes into it with a bit more depth. What we were able to measure though flies in the face of some of what Jim states there, although the gist is the same. More to the point we were not expecting any unusual results either, so our measurements surprised us as well.

We could put some figures on it: would the lack of a proper termination on a 0.5m cable at audio frequencies account for more than 1 LSB of error in the received signal (e.g. 16 bit)?

It might. When I worked at Sperry Corp on something called the Eagle Project (a math co-processor for the 1190 mainframe) we easily ran into problems like that on boards that had much shorter traces, if the termination was lacking (the termination resistance was 2 ohms!). I suspect our clock frequencies were a bit higher though

 

[Ethan Winer]

The industry, by and large, does not seem to want good measurements, and frankly I'm not surprised

Nor am I surprised. If consumers had accurate specs showing how well their proposed purchases actually perform, they wouldn't spend $10,000 for a power amp or pay more than $1 per foot for wire etc.

If you think you can get industry to walk the walk, tell me how.

I'm presenting a workshop at this year's AES convention in New York City on exactly this topic. The abstract is below, with the relevant portion in bold.

--Ethan

-------------------------------

The fidelity of audio devices is easily measured, yet vendors and magazine reviewers often omit important details. For example, a speaker review will state the size of the woofer but not the low frequency cut-off. Or the cut-off frequency is given, but without stating how many dB down or the rate at which the response rolls off below that frequency. Or it will state distortion for the power amps in a powered monitor, but not the distortion of the speakers themselves which of course is what really matters. This workshop therefore defines a list of standards that manufacturers and reviewers should follow when describing the fidelity of audio products. It will also explain why measurements are a better way to assess fidelity than listening alone.

 

[Ethan Winer]

what I can never understand is why some *consumers* ("audiophiles") are also happy to go along with this.

Yes, that's the more interesting question. One reason is someone who already spent a lot of money on wires doesn't want to admit to himself or others that he overpaid. Another reason, I believe, is that some people prefer to believe in magic. Anyone with a grade school education should understand that astrology is nonsense, but many eat it up anyway. Same for buying lottery tickets. It's trivial to show that putting $5 aside every week will eventually be worth real money, yet some people spend that on the lottery that they have zero chance of ever winning.

--Ethan

 

[Ethan Winer]

For unqualified people to come along and tell the engineer that there is something about electronics that he doesn't know, yet for there to be no way he can learn about it

Nailed it. May I steal that and use it for my forum sig?

--Ethan

 

[audioguy]

This workshop therefore defines a list of standards that manufacturers and reviewers should follow when describing the fidelity of audio products. It will also explain why measurements are a better way to assess fidelity than listening alone.

Won't make any difference. That would be the equivalent of pulling back the Green Curtain at the Wizard of Oz Castle

 

[Groucho]

Nailed it. May I steal that and use it for my forum sig?

I would be honoured!

 

[rbbert]

... It's trivial to show that putting $5 aside every week will eventually be worth real money, yet some people spend that on the lottery that they have zero chance of ever winning.

--Ethan

I hope you're not counting on compound interest (at a 0.1% annual rate) to demonstrate this.

 

[Soundproof]

One of the difficulties in obtaining a doctorate is finding a subject that is in some way original and can contribute to the sum of human knowledge. So has anyone, ever, gained a doctorate through the study of the audibility of cables? Why not? It's a wide open field.

I've attempted this argument. No show ...
As you write, it's very hard to find areas where one can do truly groundbreaking work in academia, and one would think that the remarkable properties of HiFi-cables would have people swarming the field.

When I began noticing that cables were able to adjust a setup so as to improve a weakness perceived by the owner, and that the same cable was capable of delivering contrary effects for different owners, I realized that the effect was in the listener, independent of the connect.
I spent some time going through cable reviews, forum comments and manufacturer's listings with customer feedback, and the Audience Au24 page with review quotes was just one of many amusing finds. I was somewhat surprised that the manufacturer would list that many reviews with contrarian cable effects - a little bit as if someone selling an approved medicine claimed it could cure anything the user wished to have cured.

And that, of course, goes to the topic of the thread - which is not cables, but subjectivist/objectivist appreciations of what it takes to reproduce music in a pleasing way.

 

[j_j]

Dear Sir, since you've already been kind enough to establish my point that not all cables are the same by providing the example of poorly designed cables, would you be so kind as to please define what a "poorly designed cable" is and in so doing define what a "properly designed cable" should be.

Is there some reason you're trying to pick a fight? A cable has one job, which is to provide at one end what went in the other. This is something that is measureable to fair-thee-well.

A poorly design cable is one that does not do that job. There are also things like poorly designed equipment that does not have proper ground or neutral immunity, and outputs from equipment that can not drive a cable of a meaningful length.

These are, however, things that are clearly and simply understood in the world of electrical engineering, and should not be a problem. Note, "should", you may be sure I've found equipment that does not necessarily implement this.

Designing around such problems is not particularly expensive, nor is it particularly hard.

Having said that, you may be sure I've found problems here and there, but price has not been correlated beyond saying that the ultra-cheap should be subject to careful examination.

In other words, how to do it right in analog terms is well established, for wires and circuitry. There is no mystery, and no lack of engineering knowhow. Naturally, that does not prevent bad things from happening, and where it happens is not as correlated with budget equipment as it should be.

 

[j_j]

You can measure the effects of audiophile cables- the trick is to measure the system spectral bandwidth in the room rather than the cable itself. On the bench it seems harder to quantify the effects of one cable to another like you can measure in a functioning system, and the effects of a given cable will vary from system to system. A simple trick- make the cable in question longer and see if its easier to see its effects.

There are several points here:

1) Some cables deliberately add an R/C rolloff. Of course this is audible, but this is not doing what a cable should do, it's what an equalizer or tone control is for.
2) You ***MUST*** test cables in-situ, or you have zero idea how they will interact with other equipment, like driving impedence, current sourcing, etc. While those should not be a problem, the word "should" is there for a reason.
3) As somebody pointed out, the phone company has worked on this a lot. If the tomfoolery about quantum effects, "grain rectification" and the like were true, cross-town telephony would have never worked, let alone cross-continent telephony. It does, ergo silly tomfoolery is tomfoolery.

Now, I have heard cables make a difference in several ways. The first was when somebody got one of the first audiophile speaker cables (that had mind-boggling capacitance), and connected them to a high-end amplifier. The "pop" as the output transistors died (the amp was not stable into such a huge capacitive load, please feel free to point fingers at whichever you like, I'm not going to even name the components) was quite audible as was the "dead silence" that resulted, followed by the sound of windows opening to let out the smell of burning resistor.

The second was in a lab, a very carefully set up lab that thanks to equipment design required some RCA single-ended connections. Ok, something sounds wrong. Hm. I did some measurements, and sure enough something was center-clipping. Removed cables, etc. No evidence. Put them back in, no center clipping. Examining some of the connectors (tin on gold bad) showed some corrosion, and removing and reinserting the connectors scraped through it. So MOVING the cable made a difference. The cables should never have been tin on gold (jacks) to begin with, and you may be confident that was resolved promptly, of course.

I wonder some days if that isn't the cause of "improvements" attributed to cables, and if reseating all cables would have a similar effect.

Connectors do matter, there is no real doubt, but all they have to do is get it right. RCA connectors, by their design, kind of suck, but it's hard to get around them.

 

[JackD201]

Is there some reason you're trying to pick a fight? A cable has one job, which is to provide at one end what went in the other. This is something that is measureable to fair-thee-well.

A poorly design cable is one that does not do that job. There are also things like poorly designed equipment that does not have proper ground or neutral immunity, and outputs from equipment that can not drive a cable of a meaningful length.

These are, however, things that are clearly and simply understood in the world of electrical engineering, and should not be a problem. Note, "should", you may be sure I've found equipment that does not necessarily implement this.

Designing around such problems is not particularly expensive, nor is it particularly hard.

Having said that, you may be sure I've found problems here and there, but price has not been correlated beyond saying that the ultra-cheap should be subject to careful examination.

In other words, how to do it right in analog terms is well established, for wires and circuitry. There is no mystery, and no lack of engineering knowhow. Naturally, that does not prevent bad things from happening, and where it happens is not as correlated with budget equipment as it should be.

Pick a fight? Whoah. Slow down J-J. Besides where did I ever bring price into the picture. Groucho categorically said that cables all sound the same. I made no claims as to what may or may not sound better. Just that this is not the case. When I say cables I mean all cables and that does include poorly designed ones as you yourself pointed out. Now there are cables that work and cables that work better and I'm not even talking about audio quality just plain working. Like for example using an underrated power cord borrowed from an appliance on an amplifier causing the amplifier to eventually shut down (not work). No sound is not good sound. That power cord is not poorly designed (it worked for the sewing machine) the guy who used it (me in my misguided youth ) was just not very prudent.

So we've got IEC and CE to mandate specs and AES and EBU that do so for other specs. One could say that if cable manufacturers were to follow these then the cables are properly designed for their intended uses. That would be a fair assumption. I'm not in the ultimate fidelity camp but how many of these cables regardless of price would actually pass let's say, a 1kHz square wave null test? As you say easy enough to measure and prove. The contrary was never my assertion.

 

[DonH50]

Typo: TDR = Time Domain Reflectometry (D = Domain, not Delay)

Works fine at audio.

Back to your regularly-scheduled arguments...