If I understand the mechanism correctly 'expectation bias' acts as strongly on those who don't think cable differences are audible as on those who do. It's a factor but not a deciding factor. At best it's a strong argument the answer lies with properly designed (i.e. by auditory specialists) and implemented (by engineers?) tests.
hermanv said:
Well, I don't know about that. The only attempt I found was in a magazine, I think it was HiFi choice, that started out on a clear DB way, and then afterwards fudged the results because to their horror a cheap cable came out on top of the expensive ones. That was corrected speedily!
There is a thread about that on this forum that spells out the details if you are interested.
Edit: http://www.diyaudio.com/forums/showthread.php?postid=507234#post507234
..and here: http://www.diyaudio.com/forums/showthread.php?postid=507497#post507497
But maybe you have examples of successfull DB tests you can point us to?
Jan Didden
I can only conclude that you simply didn't carefully read what I wrote.
It *is* likely that your unconscious bias matched, because as I said, the general tendency of human beings is to 'hear' difference when they are in a situation where it appears difference *could* exist.
Somewhat similarly, we are all subject to many optical illusions whether we *consciously* acknowledge that or not. It is not something to be ashamed of, or to deny. It is likely a consequence of our wiring, and hence evolution. Here's one of my favorites:
http://web.mit.edu/persci/people/adelson/checkershadow_illusion.html
Who among us would not SWEAR that the squares are VERY different shades of grey? Yet they aren't. Due to mechanisms entirely unconscious to us, we perceive them as different.
I do wish audiophiles would stop wishing away, or softpedalling, or
special-pleading their way out of one of the least controversial findings of human psychology: that we are all subject to biases of which we may not be actively aware.
PS: all of the best audio gear designers know this too. So do the best marketing people.
Where? I have been researching this issue for years now, and I can count the number of published cable blind tests on one hand. Perhaps the most famous was Greenhill's Stereo Review article from
the 80's. There, differences were statistically significant only when there were large differences in gauge or when one cable had a measurable dip in frequency response compared to the other -- in other words, nothing that couldn't be easily explained from known parameters.
In the tests I'm aware of, in no case were the routine audiophile reports of 'vast' , jaw dropping difference between cables EVER supported once the comparisons were blind.
Participating in 'tests' such as you conducted leads no one any closer to the truth. And by the same logic you use, true believers will always believe. If I had a dime for all the times I've been told 'I trust my ears', by people who always use SIGHTED comparison, I'd have a nice fat piggybank from it.
krabapple said:
In the context of listening tests outside a lab then what impact does expectation bias have on those who are convinced audible differences between substantially similar cables do not exist? Is it:
- they hear no difference because an 'unconscious bias matched', or
- hear a difference because they're placed in a 'situation where it appears difference *could* exist'?
I hope we agree that in the context of investigatory auditory testing the answer 'hear no difference because there isn't one' is begging the question.
And I'd be a billionare If I believed (and got a dollar from, etc) all people who constantly think that measurements are the be-all end-all.
I'd be a billionare, but I sure would have shitty souding audio gear.
So ,yeah, measurements aren't the be all end all. People come up with measurements, to correlate or attempt to corelate what they hear.
If you can't correlate what you hear to a measurment, then you aren't measuring for the right thing, or your equipment does not resolve to a fine enough level. maybe a bit of each.
And don't be fool enough to believe that your particular physical capacity for auditory discernment (it has aspects of a learned skill to it- this is well known, as point made by science and research)is the same as everyone else's. That would be patently ridiculous. And as you just said - beware your own bias.
I'm jumping into this one a bit late, so pardon me if those particular swords have been crossed before. 🙂
I'd be a billionare, but I sure would have shitty souding audio gear.
So ,yeah, measurements aren't the be all end all. People come up with measurements, to correlate or attempt to corelate what they hear.
If you can't correlate what you hear to a measurment, then you aren't measuring for the right thing, or your equipment does not resolve to a fine enough level. maybe a bit of each.
And don't be fool enough to believe that your particular physical capacity for auditory discernment (it has aspects of a learned skill to it- this is well known, as point made by science and research)is the same as everyone else's. That would be patently ridiculous. And as you just said - beware your own bias.
I'm jumping into this one a bit late, so pardon me if those particular swords have been crossed before. 🙂
I agree, KBK. I can consistently measure to -120dB, 1 part in 1 million, or .0001% , but I can't find a consistent correlation with what I measure, and what I (or just about anyone else in the universe) hear. Darn, so much invested, still no real solution. When I do actually measure DIFFERENCES between audio cables, I am told that it is my equipment that is the problem, because cables should not measure significantly different with with single tone running through them. Still, they do, and they sound different as well. So much for measurement. 
Good point rdf,
Perhaps tests should be conducted with alot of bonered up wires thrown into the field... just to "loosen" the bias I get people really listening.
As long as there is one dollar and two people on the planet... this hooey will rage on...
😀
Perhaps tests should be conducted with alot of bonered up wires thrown into the field... just to "loosen" the bias I get people really listening.
As long as there is one dollar and two people on the planet... this hooey will rage on...
😀
john curl said:I agree, KBK. I can consistently measure to -120dB, 1 part in 1 million, or .0001% , but I can't find a consistent correlation with what I measure, and what I (or just about anyone else in the universe) hear. Darn, so much invested, still no real solution. When I do actually measure DIFFERENCES between audio cables, I am told that it is my equipment that is the problem, because cables should not measure significantly different with with single tone running through them. Still, they do, and they sound different as well. So much for measurement.
John,
If you are referring to the measurements you reported and that were discussed a few years ago, I think you are oversimplifying things. Correct me if I remember wrong, but as far as I remember, you did distorsion measurements om some cables, and found a trace of distorsion in them. Scientifically this is not sufficient proof unless others can repeat the experiments, but it would nevertheless be most interesting if this were indeed correct measurements.
The problem seemed to be that somebody else with more modern test equipment with better resolution could not find any distorsion in the same type of cables (if it was the same type). That suggests that either of you must have done something wrong. However, we only have two conflicting experiments, and that is not sufficient to make any claim about which one of you messed up the measurements. It could be that your measurements were the more correct ones. We simply cannot know from just these two experiments. If you still believe in you measurements, you really should try to convince a number of other people with suffcient test equipment to try repeating you results. Before doing that you should however scrutinize if there might be something wrong with the way you set up the experiment, but didn't SY come over to help you have a look at it? What came out of that?
That there is no obvious reason why cables should cause distorsion is no excuse for not trying to measure distorsion, although it is a good reason for not being disappointed if finding no distorsion. Sometimes in physics, theory predicts later empirical verification, and sometimes empirical studies necessiate revising or refining the theory.
There is a fair share of snake oil sold in our hobby. The ineffective tweaks mostly dissapear never to be heard from again. The marketplace has voted and voted with their wallets.
Cables now however have indeed withstood a test of time and this in the presence on an endless number of people SHOUTING fool, snake oil, prove it, if you can't measue it, it doesn't exists etc etc.
Yes I agree, the better cables seem and are insanely expensive, only you can decide if the effect is worth the money to you. As far as the degree of improvement, there indeed a subjective result is pretty much guaranteed. There is no such thing as an instrument that measures the physical quality of music and sine waves often provide non useful results.
For my personal financial situation, cables over $500 are just not in my future or in my system. To many, the more expensive cables might in fact be worth it, for me the financial limits make the point mostly moot.
I too did a search for DBT results, most are so old that there are still flame marks but little original material. Many of these tests when traced, lead right back to people who have been in the high end industry for a long time, those whom you might find it possible to believe. Those who are still in the business presumably because the one thing they aren't selling is snake oil.
Cables now however have indeed withstood a test of time and this in the presence on an endless number of people SHOUTING fool, snake oil, prove it, if you can't measue it, it doesn't exists etc etc.
Yes I agree, the better cables seem and are insanely expensive, only you can decide if the effect is worth the money to you. As far as the degree of improvement, there indeed a subjective result is pretty much guaranteed. There is no such thing as an instrument that measures the physical quality of music and sine waves often provide non useful results.
For my personal financial situation, cables over $500 are just not in my future or in my system. To many, the more expensive cables might in fact be worth it, for me the financial limits make the point mostly moot.
I too did a search for DBT results, most are so old that there are still flame marks but little original material. Many of these tests when traced, lead right back to people who have been in the high end industry for a long time, those whom you might find it possible to believe. Those who are still in the business presumably because the one thing they aren't selling is snake oil.
Christer said:
I also remember someone using newer (better, and properly calibrated) AudioPrecision gear measuring the same set of cables and not finding any residuals down to 150dB or thereabouts.
Listen, it is this simple.
If we want to talk about engineering and measurable results.
There ARE areas where cable distortion would be plain catastrophic. Astronomers and particle physicists routinely measure things that are far, FAR beyond the requirement of even most a#al golden eared audiophile. The amount of energy released by a pass of a single neutrino for example is so unbelievably small that noise and distortion of even elite audiophile gear would be simply laughably inadequate in comparison with amplification devices used there. Same for radio signals - mapping cosmic microwave background radiation for example required technology and devices that did not even exist until recently - the signals are that weak. Detecting micro Kelvin or two variation on top of 2.7K "noise" is simply a heroic task.
Yet, no place I ever saw uses any of the audiophilobsessive cables.
Sound engineering, careful impedance matching, yes;
Golden Ratio (tm), X Multipole Technology (tm), New Ratio (tm), cryogenic treatment, Mpingo Magic (tm), Echo Wave (tm) - no.
And why would be this the case ? Astronomers and physicists being so anti audiophile that they would deliberately avoid known superior cabling just to prove a point ?
Hmmm.
Bratislav said:
I am no physicist so my knowledge could be dated, however as far as I know there are several misconceptions in your post.
1. Most of these tiny numbers are extrapolations of much larger scale events. They are calculations (probably accurate) scaled down from far larger events not direct measurements.
2. The trick is the detector used, not the cable. the signals for these kinds of measurements are almost always scaled, amplified and only then driven through coaxial cables, no they do not attempt to measure a nano volt or two directly.
3. Huge amounts of statistical results are sifted ever so finely for consistent offsets, super computers are common.
I believe your analogy fails. However I recommend you visit the National Instruments test and measurements site. There they explain quite carefully how improper use of cables can indeed make measured results meaningless. National Instruments speciallizes in tools and hardware for converting measurement to computer values often used to run statistical analysis. Most of their stuff is lmited to 16 bit accuracy so if cables can cause problems at those resolutions, imagine the problems for signals as tiny as those of which you speak.
ps. Oh, yeah, CD's are 16 bits too, damn. 🙂
Cable drama doesn't persist because of validity... it is purposefully moot (look it up)... and this is well fed by the purveyors. They persist because they are basically one of the only things a novice can tamper with safely. Think about custom wheels and tires, mufflers, fiberglass spoilers. The average audiophile is no more inclined to break into his gear with nippers and a soldering iron, than the average auto enthusiast will change his ring and pinion.
Do the cables work as well as the mufflers?
The auto guys have a thing or two going in this regard... they are less inclined to argue with the stopwatch; and most will admit that the spoiler is just for looks.
😎
Do the cables work as well as the mufflers?
The auto guys have a thing or two going in this regard... they are less inclined to argue with the stopwatch; and most will admit that the spoiler is just for looks.
😎
"1. They are calculations (probably accurate) scaled down from far larger events not direct measurements"
Wrong. Astronomers and physicists measure DIRECTLY. Day in, day out. Visit your friendly astronomer today and see for yourself.
"2. The trick is the detector used, not the cable. the signals for these kinds of measurements are almost always scaled, amplified and only then driven through coaxial cables, no they do not attempt to measure a nano volt or two directly."
So what makes audio different ?
And you are wrong. They are measuring nanovolt (in fact MUCH lower than nV) directly.
"3. Huge amounts of statistical results are sifted ever so finely for consistent offsets, super computers are common"
You have to use simlper English for me for this explanation (English is my third language). You have lost me there completely.
How does any of this invalidate need for distortion free cables in scientific applications, yet audio needs better?
Anyway, if you really think that audio is the most critical and demanding application human kind knows of, and really warrants 15,000$ interconnects and speaker cables, be my guest. They need to make living too, I guess. If they can't take money from scientists, engineers and other skeptics, they will be alll too glad to meet you 😀
Wrong. Astronomers and physicists measure DIRECTLY. Day in, day out. Visit your friendly astronomer today and see for yourself.
"2. The trick is the detector used, not the cable. the signals for these kinds of measurements are almost always scaled, amplified and only then driven through coaxial cables, no they do not attempt to measure a nano volt or two directly."
So what makes audio different ?
And you are wrong. They are measuring nanovolt (in fact MUCH lower than nV) directly.
"3. Huge amounts of statistical results are sifted ever so finely for consistent offsets, super computers are common"
You have to use simlper English for me for this explanation (English is my third language). You have lost me there completely.
How does any of this invalidate need for distortion free cables in scientific applications, yet audio needs better?
Anyway, if you really think that audio is the most critical and demanding application human kind knows of, and really warrants 15,000$ interconnects and speaker cables, be my guest. They need to make living too, I guess. If they can't take money from scientists, engineers and other skeptics, they will be alll too glad to meet you 😀
ya know?
it's okay if someone, be they an industry guru or otherwise, states that THEY hear differences. Really...
What's not okay is stating that they're experience and opinion is fact without passing the muster of review (be it peer or otherwise)
If you don't believe in measurement, then why bother doing them? just state something like:
"in my opinion, cryofrozen cables with intermolecular diode compensation and revelatorical impasse annihilation superseding phenomenological anomalies induced by monodirectional inhomogeneities, sound better to me than those that don't"
...and leave it at that...
some of us know better, and work in fields where such proposed causes would create problems that are magnitudes greater than the signal levels of interest
Bratislav is absolutly correct here...
auplater
it's okay if someone, be they an industry guru or otherwise, states that THEY hear differences. Really...
What's not okay is stating that they're experience and opinion is fact without passing the muster of review (be it peer or otherwise)
If you don't believe in measurement, then why bother doing them? just state something like:
"in my opinion, cryofrozen cables with intermolecular diode compensation and revelatorical impasse annihilation superseding phenomenological anomalies induced by monodirectional inhomogeneities, sound better to me than those that don't"
...and leave it at that...
some of us know better, and work in fields where such proposed causes would create problems that are magnitudes greater than the signal levels of interest
Bratislav is absolutly correct here...
auplater
Just for the record, I could measure 'something' with the identical cables that I was given to test, BUT I have measured far worse cables than these 'examples'. It is true that the 'latest' piece of test equipment did not measure anything, but I still could measure differences between many cables. Why? Nobody knows. I dropped the subject, because even though my test equipment cost just as much as the 'latest' measurement equipment when it was new, it was at least 10 years old, and therefore subject to criticism, and there was no way to argue the point further. I don't make cables, so what is my point in further trying to measure differences? Also, the 'differences' that I appeared to measure, did not necessarily correlate with the best 'sounding' cables (if there is such a thing) so I could not use my 'results' to necessarily infer which cable was 'better' with my measurement method.
Hi Bratislav, what are the typical centre frequency, bandwidth and dynamic range requirements of the astronomical measurements you're discussing? The game changes completely at high RF frequencies.
It varies enormously. Some things are changing so slowly that they can be safely taken as static (say microwave background radiation map). Most of the radio telescope measurements are in order of few Hz or lower (as antenna scans the sky), on which a signal of high to VERY high frequency (MHz to GHz) is superimposed.
On the other end of the scale gamma ray burtsts require better than microsecond resolution (as some complete events may last only a few milliseconds), and of course subparticle physics operates with femto seconds routinely.
Dynamic scale ? How about same radio telescope measuring radio signals from the Sun (radiating >1kW/m^2) while a typical BRIGHT signal from a distant object will be measured in Jansky's (10E-26W/m^2/Hz).
Dynamic enough ? 😉
Bandwidth ? SETI used to search for signals from 1Hz to 10GHz. They seem to have concentrated lately around 1.4GHz (21cm), as far as I know with sub-Hertz resolution. They may use several million channels though 😀
On the other end of the scale gamma ray burtsts require better than microsecond resolution (as some complete events may last only a few milliseconds), and of course subparticle physics operates with femto seconds routinely.
Dynamic scale ? How about same radio telescope measuring radio signals from the Sun (radiating >1kW/m^2) while a typical BRIGHT signal from a distant object will be measured in Jansky's (10E-26W/m^2/Hz).
Dynamic enough ? 😉
Bandwidth ? SETI used to search for signals from 1Hz to 10GHz. They seem to have concentrated lately around 1.4GHz (21cm), as far as I know with sub-Hertz resolution. They may use several million channels though 😀
- Status
- Not open for further replies.