Audio Cables
In reality this would not be possible but if using very good cables (such as Audioquest) that are designed well that induce very little distortion, that is when you could compare it to a cable such as transparent audio and then if you have access to that specific list of components then and only then will this post make any sense.
In reality this would not be possible but if using very good cables (such as Audioquest) that are designed well that induce very little distortion, that is when you could compare it to a cable such as transparent audio and then if you have access to that specific list of components then and only then will this post make any sense.
The one that absolutely makes me giggle uncontrollably is the "audio" power cables! Rather than making any reasonable argument about the power lines coming to your house being absurdly noisy, I'll simply point out something that I see every day in my lab (FYI, I'm a chemist): absolutely ZERO high end instrumentation of any kind comes with exotic power cords! You will find every variety of exotic shielded signal cables and super spiffy double plated connectors on these million dollar instruments, but almost every one of them comes with a cheap, mass produced power cable! (the exceptions are typically the ones hard wired into high voltage lines) If there was ANY REASON WHATSOEVER to put an exotic cable on their instrument, you better believe Bruker, Thermo, Agilent, Perkin Elmer, or anybody else would JUMP at the chance to extort extra money from corporations and universities! But with all of their R&D money, they all have reached the EXACT SAME conclusion--as long as it can handle the current, power cables are all the same.
Does anybody here really want to make the claim that all of the scientists and engineers developing these instruments are naive and that a bunch of DIY'ers have more data than they do?
Anyway, on to the discussion. . .
Since my day job revolves around chemical instrumentation in which I have to deal with measuring and amplifying some pretty minuscule electronic signals, here are the questions I would immediately ask if I got your random piece of data in a paper that I was reviewing:
1) Are you making these measurements in a Faraday cage?
2) How are you generating a waveform repeatable enough to guarantee the difference in intensity between runs is not accountable for your reported differences in intensity?
3) What is the standard deviation between experiments, IE error bars on your graphs?
4) What does your control data look like (this ties back to 3, but I would like to know how you verified your instrument prior to making these measurements, and what your controls were!)
There are dozens more, but if you can start off with those, then you will at least be able to convince some of the people around here that you should be trusted to take measurements correctly (something I am not convinced of in the slightest at this point)
I know that AudioExpress isn't in the business of peer reviewing submissions like scientific journal are, and I have no idea what sort of review went on during your submission process, but I am disappointed that you aren't addressing these questions prior to a random proclamation. Having said that, I have not read your Audioexpress article. If you do include answers to these questions in that article, I will definitely take a look before commenting further on your methodology.
John, I know you have built some pretty good gear, and I am certain that your knowledge of analog audio circuitry is pretty darn good (admittedly far better than mine), but if you are arguing that Simon's claims and methodology should not be subjected to rigorous scrutiny, then I would argue that your personal ideas are getting in the way of an objective analysis of this question. Of course, I fully recognize that I may be misunderstanding your position, so please let me apologize in advance if I am misconstruing your intention.
Random on-topic side point, but I just finished putting together a set of my old standby cables: guitar cord and decent connectors from parts express. Practically speaking, they are really flexible, make fine connections, and are really cheap (the 100' spool of cable I bought 5 or so years ago for ~$30 is still about half full). Considering the inordinately larger impact that new drivers or even room dampening would have for the same price as exotic cables, why would you worry about spending money on such a minute aspect of a system as your interconnects?
Does anybody here really want to make the claim that all of the scientists and engineers developing these instruments are naive and that a bunch of DIY'ers have more data than they do?
Anyway, on to the discussion. . .
Simon, do you actually understand what you wrote? You really think that your electronics measurement gear is accurate and repeatable enough to measure low PPB numbers? That's measuring the difference between 1.00000000 V and 1.00000001 V.
Since my day job revolves around chemical instrumentation in which I have to deal with measuring and amplifying some pretty minuscule electronic signals, here are the questions I would immediately ask if I got your random piece of data in a paper that I was reviewing:
1) Are you making these measurements in a Faraday cage?
2) How are you generating a waveform repeatable enough to guarantee the difference in intensity between runs is not accountable for your reported differences in intensity?
3) What is the standard deviation between experiments, IE error bars on your graphs?
4) What does your control data look like (this ties back to 3, but I would like to know how you verified your instrument prior to making these measurements, and what your controls were!)
There are dozens more, but if you can start off with those, then you will at least be able to convince some of the people around here that you should be trusted to take measurements correctly (something I am not convinced of in the slightest at this point)
I know that AudioExpress isn't in the business of peer reviewing submissions like scientific journal are, and I have no idea what sort of review went on during your submission process, but I am disappointed that you aren't addressing these questions prior to a random proclamation. Having said that, I have not read your Audioexpress article. If you do include answers to these questions in that article, I will definitely take a look before commenting further on your methodology.
John, not to be the grammar police here, but I am not really following that statement. I am assuming that you are encouraging people to look beyond textbooks and look to new research to broaden their knowledge. I would definitely agree with that outlook, if that is indeed what you meant. Would you not agree, however, that overturning "textbook" precedent requires solid evidence? All I ask for is scientifically rigorous measurements and meaningful interpretation, which we don't currently have from Simon7K.
John, I know you have built some pretty good gear, and I am certain that your knowledge of analog audio circuitry is pretty darn good (admittedly far better than mine), but if you are arguing that Simon's claims and methodology should not be subjected to rigorous scrutiny, then I would argue that your personal ideas are getting in the way of an objective analysis of this question. Of course, I fully recognize that I may be misunderstanding your position, so please let me apologize in advance if I am misconstruing your intention.
Random on-topic side point, but I just finished putting together a set of my old standby cables: guitar cord and decent connectors from parts express. Practically speaking, they are really flexible, make fine connections, and are really cheap (the 100' spool of cable I bought 5 or so years ago for ~$30 is still about half full). Considering the inordinately larger impact that new drivers or even room dampening would have for the same price as exotic cables, why would you worry about spending money on such a minute aspect of a system as your interconnects?
Sir, I beg to differ. Someone being as daft as to buy a megabuck power cord treats it the way power cords are treated in lab tools - they don´t shrinkwrap them together with unshielded cables carrying milliwatt information signals.
I am being deliberately vague in my comments here. I don't dare tell you what my associates and I have heard in the differences between cables. I would be run off the website. Of course, it is because my associates and I do hear differences, that we can optimize a design to make it a cut above the average. This is why many good engineers fail to make a really successful audio design, no matter how they try.
You're right. Phrased that way it does sound ludicrous. Bring forward the results of their examination of power cables and we can wrap this up. 😀
I disagree, I think that power cables make a difference. I have heard the differences, myself. Of course, I had to use my ears to detect the difference. I have test equipment in my lab that, at audio frequencies up to 100KHz is equal to SYN08's, and I can measure at least 1 part in 1 million in harmonic distortion. Still, I have never measured a power cord to any significant degree.
you can simulate the whole chain. you need the output impedance of your wall outlet, the impedance of the cable and the input impedance of the supplied circuitry. that is the way we designed power supplies in the high tech industry and we used surprisingly small values of capacitance even in 90kW DC power supplies because you may find a major resonance quite high up in frequency that has to be suppressed.
there is also a lot of "blind" current that has to be taken care off unless it is fed back into the wall outlet and cost you money that you do not really use.
that combined is one the reasons one power cable or power conditioner will work well in one system and fail in the other. i trust all you say here but i miss a little science from time to time.
We all like 'science' Joachim. It is just that sometimes, conventional 'science' doesn't measure some factors in audio design.
i agree totally but pure subjectivism does not cut the mustard ether. i think a good measure of practical theory can help to save us time and money so can be of great advantage for the development of our art without any disadvantage that i can see. the ear is the final arbiter, even for me that has got some hard scientific training and i would always trust someone like you more then a kid with no experience but great talent for science or mathematics. you proved that you can make good sound.
Agreed.
No, see previous posts where I stated that different types of claims could be made and the test results need to be commensurate with those claims.
Sure, there are always opportunities to improve a test protocol. Happy to hear them (no pun intended). How, specifically, would you test two cables for audible differences? What would be your protocol?
I’m interested in what appears to be consumer fraud, undertaken by manufacturers in conjunction with subjective review magazines and the retail chain. Exploring cable audibility in the lab is fine. When claims are made for audibility and they come with high (some would say outrageous) prices that we are in rip off territory. You may think its OK for consumers to be ripped off, I don’t. Consumer protection legislation was created just for this reason, to dissuade manufacturers and retailers from engaging in deceptive conduct.
You misunderstand the test I would support. The test results will either be: “After statistical analysis the listening panel could distinguish cable A from cable B.” or “After statistical analysis the listening panel could not audibly distinguish cable A from cable B.” That’s it. If there is a difference then its on to which cable “sounds better”, which is just an aesthetic judgement. If there is no audible difference and the cables are significantly different in price, then buy the cheaper one.
The other thing worth stating is that those that criticise the people who have gone to the trouble of conducting proper double blind tests are dumping on people who could have well said: “Look, the physics and engineering just doesn’t support that claims that these two cables could sound different.” But in spite of this they went further, did the proper listening tests anyway just in case they might have missed something in the physics or the metallurgy. That’s a service to the audio community that really, as said by Syn08, the manufacturers should pay for.
I guess we all agree with that, it however has nothing to do with the topic of this thread. Also, who is going to decide what other might hear or not?
If a customer is foolish enough to pay a ridicilous price for cables that have no influence on the SQ of his system, perhaps he deserves to be ripped off. If he can however hear a difference in SQ, I would say it is up to him to decide whether it is worth the extra money, not a decision for anybody else to make.
I´m still wondering about this question. Is there a sort of misundertanding what the term "positive control" means?
A positive control wrt to scientific testing simply means to present something that is known to be detectable and that has to be detected to show that a specific test is sufficient sensitive for the task.
Please remember that a test has to valid, that means it must test what it pretends to test.
As the differences measured between two cables under test usually are very small it seems to be wise to use positive controls requiring the most sensitive listeners.
That´s why i suggested that positive controls should be selected for example from the list Paul Frindle told about in his paper held at the AES convention `The Measure of Audio`; i think his results (resp. the results of his team) were quite impressive even using an ABX test protocol.
Another thinkable control could be something like the difference between two cd players as mentioned as detected at the ABX homepage (old Philips vs. and slightly newer Sony, models as mentioned).
If you establish controls on different sensitivity levels you´d have additional data, could see if a learning curve exists while training your participants and so on.
A negative control is needed to show that the whole test setup and procedure does produce positive results (means the rejection of the nil hypothesis) only due to the EUT.
And not due to any other effects like switching noise, hum or buzzing, or soft facts like presentation order or any other possible confounder.
May i renew my question which way you´d be able to show that listeners in a dbt would have been able to detect a difference if you wouldn´t have used positive controls?
The phrase "trained listener" means wrt to our discussion topic, that the participants of the test would be used to listen under the specific double blind test conditions and are able to produce reliable results.
I think together with the explanation posted above it is quite selfexplanatory which way to choose whether they are trained enough. 🙂
I think we are running in circle on this specific point; as you might remember i cited the laste time some manufacturers that were (partly quite strong) promoting double blind tests like wireworld (afair Joachim Gerhard could tell more about these attempts and the reactions they got).
As you speak german i´d be happy to email you a master thesis paper which was based on a public double blind test and was supported by a german cable manufacturer. I think after reading it will be clear why the results weren´t used in marketing 🙂
Other than that i don´t feel really obliged to speculate about the motivation of cable manufacturers, but would ask you which benefit a manufacturer would have from a positive test result?
Would he suddenly be able to sell cable to those who believed before that all listeners of differences must be deluded? 🙂
Would any of the others, who choose cables due to own listening tests, don´t do that but rely instead on a dbt done by other?
Wishes
@ dfdye,
if you visit a certified lab for emc testing you´ll usually see a lot of fancy accessories to make everything work together as expected and to provide the best measurement accuracy possible. 🙂
A lot of this accessory isn´t sold by Agilent and the other usual suspect.
Manufacturers of test equipment specify the framework conditions under which the gear is within the specifications; part of the framework are the regulations from IEC.
That doesn´t mean that the gear couldn´t do better if you undergo extra measure for better conditions. 🙂
Wishes
if you visit a certified lab for emc testing you´ll usually see a lot of fancy accessories to make everything work together as expected and to provide the best measurement accuracy possible. 🙂
A lot of this accessory isn´t sold by Agilent and the other usual suspect.
Manufacturers of test equipment specify the framework conditions under which the gear is within the specifications; part of the framework are the regulations from IEC.
That doesn´t mean that the gear couldn´t do better if you undergo extra measure for better conditions. 🙂
Wishes
I have to disagree. In the USA and most of the rest of the industrialized world, we have consumer protection laws. Is it OK for someone to sell me junk that does no good - and at a high price? This is a grey area. If it might do harm, that's a problem. If it might do harm by false claims (cures cancer!) that's also a problem. Other false claims such as "faster, bigger, better" can also be pursued because the can be disproved. The term "Best" means nothing in ad speak. It is allowed.
Should Bernie Madoff's investors have no recourse just because they were "foolish enough" to invest in his funds?
So if someone is selling expensive cables that they expressly claim are "better" than others, then they would have to prove it. But who is going to push the claim? Who would find it worthwhile? If big guns, big money wanted to take on Monster Cable to actually prove their cable is actually better than another, that would be great. But who is going to do that?
In audio is is so easy to think you hear a difference, or to make other people think they hear a difference that fraud is very difficult to pin down. That does not make it right, however.
Scott, being 'vague' and being 'dishonest' are two different things. THIS THREAD is laden with many 'sin bins'. I have to tread my way, carefully. At this very time, it appears that I can be more forthright on this thread, and not get my hand slapped. I hope that it continues.
- Status
- Not open for further replies.