Actually, I suspect that some cable deniers have never actually conducted a comparison. I think some cable deniers adopt their view purely on theory or dogma.
Does Everything Make a Difference?
- Thread starter Ron Resnick
- Start date
Indeed, dogma in the service of "science".
Or rather, dogma in a disservice to science.
Science is not dogmatic.
Because that's a misunderstanding of how bias effects work, and the range of bias effects.
"Expectation Effect" or "confirmation bias" - forms of expecting something to sound better - is not at all the only way you can misperceive things. Look at this list of known bias effects:
List of cognitive biases - Wikipedia
en.wikipedia.org
Our conscious attention is nowhere near some strict, reliable machine or interpreter. Perceptions often arise based on how we feel, or how our attention happens to be directed, consciously or unconsciously, at any given time.
So let's take, for the sake of argument, that you have two different AC cables. Both are exactly the same design, but one has been made to look more expensive, and it has a higher price.
As you listen, you may experience something like the primacy, or recency effect: you may be more cognizant, or remember details when you first began to listen better than subsequent listening. So if you listened to the "cheaper" cable first, you may perceive or remember detail that was different from listening to the "expensive" cable next. And then "hey, what do you know? I thought the cheaper cable sounded better! Clearly it must be true, because it wasn't what I was expecting!"
But there's all sorts of ways to misperceive. Merely setting out to compare the sound of two different items already sets up a bias for hearing differences. Our attention is not completely under our control, and just in how you find yourself concentrating on one vs the other, can lead you to "hear" differences. You can have the experience when you are least expecting it. One morning you are listening to a piece of music and based on your mood, or just how your attention happened to be directed at that time you think "Wow, I've never heard that horn part in the back of the orchestra so clearly, or with such burnished beauty..." and then the audiophile impulse is to think "well, what did I change in my system that might explain this?" And you attribute it to the new cables you bought, or that your amp is finally "burning in" or whatever.
There are just so many ways your perception can be influenced, to misunderstand the basis for an impression, that go beyond mere placebo/expectation effects.
Nobody is immune, nobody can totally predict it, that's why science involves controls for bias, even for the scientists who are acutely aware of such biases.
And the "measurement aficionados" - I'm guessing the ASR crowd would be implicated - already know this. It's usually the purely subjective based audiophiles who are under the misunderstanding and make assumptions like "I didn't expect what I heard, therefore it wasn't a bias effect."
Cheers.
*(And none of the above, the mere discussion of bias effects, means you therefore are NOT hearing real differences. Of course we hear real things all the time. But IF one is trying to get to the truth of such matters, it will involve acknowledging that there is "bias noise" to deal with as a variable, in one's method of investigation).
Great post. I think many underestimate the power of the human mind/brain at work. Just as the auditory system is able to arbitrate and compensate for contributions from room acoustics, so does cognitive hearing make corrections and adjustments based on “Expectation Effect" or "confirmation bias" - forms of expecting something to sound better. This is what often results in misleading or misplaced accounts of sound improvements attributed to “audiophile” tweaks, products upgrades and perceived improvements that are nonsensical, nonlogical, or those not supported by science.
Let me give an example that I personally dealt with recently. I was tweaking the crossovers of my new reference horn system. I experimented and made a final adjustment by raising the gain of the mid-bass crossover by 2dB on both channels. The change brought about the results that I had targeted and wanted. A few days later I listen to the system again and I appreciated the results of the adjustments. I recorded a couple of system videos that night and the next day while listening to the videos, I was pleased to note that the adjustments to the crossover could be clearly heard on the videos as well. Mission accomplished right?
A few days later I went back into the HYPEX Filter Designer software to make another adjustment and to my surprise I noticed that the gain level on the mid-bass had reverted back to the original gain setting’s. How could this be? I had been listening to the system and videos on two subsequent listening sessions And had heard the “desired” changes. What had happened?
In the HYPEX software you can change the overall gain of the module on the main window as I had done originally by raising the gain in each channel by 2dB. What I had forgotten was that in the “Device Setting” window there is a setting for “Gain At Startup” that overwrites or resets the gain to the specified gain each time the module power is turned ON. So the changes to the gain setting on the main window of the HYPEX software are not “Sticky” as the “startup Gain” setting overwrites the gain each time the module is powered ON. This is done to prevent damage or unexpected loud levels.
So while I heard the positive improvements the first time I made the adjustment, every subsequent time after that, and even on the videos, what I heard could not had been different than what I heard prior to making the adjustments, yet I continued to hear the “desired” results of the adjustments.
I would venture to say that most, if not all, of what “Audiophiles” report as positive or impactful changes brought about by tweaks and components that are nonsensical, nonlogical, or those not supported by science, are a result or a product of “Expectation Effect" or "confirmation bias" - forms of expecting something to sound better.
Last edited:
In the spirit that everything matters; this is true for not just cables.
The ears:
Assuming the person doesn't have physical hearing damage then I believe everyone is capable of hearing differences. But some listening training may be necessary. For example, in another thread there was a discussion about timing where I was talking about differences in the low single digit microsecond range. People questioned this and were basically telling me I was wrong by a couple orderes of magnitude. I posted a scientific article where this was thoroughly studied. One of the many conclusions of the article is that untrained people could hear timing differences in the 18 microsecond range. Trained people could hear timing differences in the 8 microsecond range with some individuals hearing timing differences below 5 micro seconds. The point is that there is a "training" aspect to be able to hear finer differences in sound.
(This is a very good example of people who thought they knew and understood the science of hearing but in reality they did not know what they thought they knew. Me posting the article did not change anything except their knowledge. I could hear changes in timing just as accurately before the paper was published, after the paper was published and after the other person knew the information in the paper)
The Setup:
The first step has to be to get the speaker integrated as best is possible with the gear at hand. If there is too much acoustic noise/distortion then trying to listen for changes to something else (Like cables or footers) makes zero sense. It will simply be overwhelmed by the acoustic noise + distortion of the speaker/room.
The equipment:
I have setup some pretty meager/budget systems. (<$10K). Even in these systems once the speaker is integrated differences between things can emerge. So a person doesn't need $100K+ system to hear these differeneces but the system needs to be well sorted. However, I do think there is a limit. For example, I am not sure what is possible with a $500 pair of speakers and a Sony receiver. Put any cable you want on that system and they will likely all sound the same. My suspicion is that the noise floor of the reciever (and speakers) is too high to hear anything we are talking about.
Ron, I think you answered one of my questions with this post about trying to devise a protocol. It is still unclear to me the purpose of the protocol and why you want to devise it and why it needs to be acceptable by “both sides“.
Some who conducted a comparison and do not hear a difference between a certain make of high-end cables and their stock cables are claiming only that. Does such automatically make them “cable deniers”, or just folks who are making an observation about one comparison they conducted.
Likewise, when you were an admitted cable denier that auditioned another set of cables that sounded better, does that mean that everytime one compares cables one set will always sound better (or worse) than another set of cables? No, there can be no audible difference between two sets of cables.
Analogous, something that everyone should relate to regarding “Expectation Effect" or "confirmation bias" is with the human visual system. You type a note and it looks perfect no matter how many times you read it. Go away and come back to it later and you will notice all the grammatical mistakes, misspellings, and missing lettters.
I m a cable denier , the only cable i would spend money on is when its made by aliens out of extraterrestrial material.
Having heard many shows with 50 K or more cable set ups still not convinced .
A lot use TORUS power conditioners which gives a overly smooth sound.
The best cable is the ALLEGO DC charge cable i use for my car , for audio ....... couldnt care less just good quality cables nothing fancy
Having heard many shows with 50 K or more cable set ups still not convinced .
A lot use TORUS power conditioners which gives a overly smooth sound.
The best cable is the ALLEGO DC charge cable i use for my car , for audio ....... couldnt care less just good quality cables nothing fancy
Of course it will.
Perceptual biases are well known, well studied, and can be invoked reliably via experiments.
If you have someone listen to an audio device, say a cable, and then have that person listen to EXACTLY the same device, but tell them it's a different device, the listener will often report a sonic difference between the two.
What would a more plausible explanation be for this than that it's a perceptual bias/imagination doing the work?
Likewise for when someone feels the hear differences between cable A and B when they know which is being used, but can not reliably detect between them at all when the identifies of the cables are hidden from them? The most plausible inference is a perceptual error, imagination, bias effect for the sighted listening. How else could we ever discriminate between imagination and reality?
Have you ever had a hearing test? The Audiologist is administering a blind test. For good reason. The Audiologist only plays the tones, with no other cues allowed - there is no "12kHZ" sign that goes on when the 12kHz tone is played, because then you'd know it was being played and could just say you heard it, or imagine you heard it. This is why it's a blind test, and you learn what you REALLY can hear and REALLY can't hear using such blind tests.
Hi Lee!
Sure, but remember...as I pointed out earlier in the thread...if we are really bringing science in to this, then the scientific method is to first have a reliable way of demonstrating X exists, before you have to go explaining X. That's why science isn't busy "trying to explain why astrology works" because there has been no scientifically reliable demonstration it does work to begin with! You need to first demonstrate there is a "there" there, to explain.
So if we are talking about being scientific about audio...
1 and 3 seem anecdotal. Where these claimed audible differences actually demonstrated reliably, in tests controlling for sighted biases? Not that I'm aware of, but I'd like to see whatever experiments/studies in support of such claims.
#2...same thing. Audiophiles reporting differences in CD transports doesn't mean there were sonic differences. Even IF jitter is a real thing, you'd have to have done controlled tests to take such reports out of anecdote to reliable data. I'm not aware Harley did this, are you?
Yes, apparently jitter can be audible. But that doesn't actually validate all the claims for sonic differences audiophiles make for digital devices. Not if we are thinking scientifically.
That wouldn't be scientific. Science does not "trust"...it is a system of skepticism - it attempts to verify (actually, in a way, see if a hypothesis survives attempts to disconfirm), not trust.
Cheers.
Sure, but remember...as I pointed out earlier in the thread...if we are really bringing science in to this, then the scientific method is to first have a reliable way of demonstrating X exists, before you have to go explaining X. That's why science isn't busy "trying to explain why astrology works" because there has been no scientifically reliable demonstration it does work to begin with! You need to first demonstrate there is a "there" there, to explain.
So if we are talking about being scientific about audio...
1 and 3 seem anecdotal. Where these claimed audible differences actually demonstrated reliably, in tests controlling for sighted biases? Not that I'm aware of, but I'd like to see whatever experiments/studies in support of such claims.
#2...same thing. Audiophiles reporting differences in CD transports doesn't mean there were sonic differences. Even IF jitter is a real thing, you'd have to have done controlled tests to take such reports out of anecdote to reliable data. I'm not aware Harley did this, are you?
Yes, apparently jitter can be audible. But that doesn't actually validate all the claims for sonic differences audiophiles make for digital devices. Not if we are thinking scientifically.
That wouldn't be scientific. Science does not "trust"...it is a system of skepticism - it attempts to verify (actually, in a way, see if a hypothesis survives attempts to disconfirm), not trust.
Cheers.
Ron, I agree that can be the case. Dogmatic thinking can be found anywhere.
On the other hand, a lot of people using the Golden Ear "trust your ears" approach don't seem to recognize the dogmatism of their own approach.
At least with the "objectivist" take, the point is to actually be open to evidence, so you can find out you are mistaken and change your view. In other words "Here is why I think the claim under consideration is highly unlikely, BUT...HERE are ways we can test this and HERE is the type of evidence that would demonstrate the claim, and change my mind." So they look at how to make claims falsifiable, and demonstrable.
On the other hand, quite a number of purely subjective audiophiles operate under a more intractable approach. They Trust Their Ears To Tell The Truth. If for instance the Golden Ear (GE) says "I hear a difference between cable A and B" and I express skepticism, the reply will be "you need to hear this for yourself then." Ok, so I try the cables and don't hear any difference. Have I, using the method suggested by the GE, disconfirmed his claim? No. Instead I will be told "Well then your system isn't resolving enough." Ok, I listen on a resolving system - no difference. Then the reply is "Well then your hearing isn't good enough, like mine, to detect these differences."
Ok, so now the ONLY focus of the claim is the GE's claim to hear the differences. I suggest "Ok, let's test if in fact you can hear differences between the cables, but WITHOUT knowing which is which so you are only using your hearing." Then the excuses will come about why "blind testing is an insufficient method" to vet such things.
So where does that leave us? With absolutely no way to falsify the belief, no way to demonstrate it's truth. The GE has insulated his belief from falsification, he can never be "wrong" in that case. He could just as easily be claiming to hear angels singing through certain cables, and it could no more be falsified on this mind set.
So it is often the Golden Ear audiophile who has set himself up as the ultimate I Know What I Heard arbiter of reality, in a way that can't be challenged. Which is a form of dogmatism.
And ironically, for bringing a skeptical approach, open to evidence and falsification, it's the "objectivist" who is often accused of the dogmatism!
As I said before, science gets murky when dealing with human psychology, and blind test therefore are *not* scientific. "The most plausible inference" -- yeah, right.
As much as possible, let's leave real science to realms outside human psychology, shall we? Hard data are science's domain. And hard data is where human psychology interferes with.
Bad analogy. Hearing a test tone requires nothing, no training, just hearing. Proper listening does require training.
Please, stop that blind test nonsense in the name of "science", when it comes to audio. It offends me as a scientist.
I'm pretty amazed to see a scientist rejecting the relevance of blind testing, just because "human psychology" is involved.
Blind testing is a standard method in science BECAUSE of human psychology!
To take any of countless examples: My son was part of a study concerning a new treatment for peanut allergies. It was a peanut protein-based oral immunotherapy treatment - ever increasing doses of the peanut protein to get the system used to the protien without overreacting.
It was a double-blind placebo-cotrolled trial, which as a scientist I'd expect you'd know is a recognized standard.
Well...why was it double-blinded placebo controlled? Because of HUMAN PSYCHOLOGY FACTORS. When people THINK they are being given it can provoke responses, like "side effects" or cause messy inferences about "getting better" etc. And the researchers are blinded to who got the placebo or active ingredient because we know that due to human psychology, they can misinterpret the results when they know who got placebo vs active ingredients. (In fact I was told by a couple of the researchers that they were making informal bets and inferences on who was on the placebo vs active ingredient, just based on the ongoing symptom reports. And when the results were unblinded they got it quite wrong, quite often, which only shows the need for such rigour).
When human psychology is a variable...in science you try to control for that variable!
And since we KNOW that the very same psychological/bias factors are in play with any of our senses and inference making, including auditory, it makes sense to try to control for those variables.
Which is what researchers like Floyd Toole and others did. (And it's done for testing audio codecs etc).
So this line you are drawing just where it seems to let audiophiles off the hook seems inconsistent scientifically, to say the least.
That misses the point.
The reason an audiology test is blinded is because the subject having KNOWLEDGE of when a tone is playing can influence the results - you wouldn't get reliable results. Because they could imagine they are still hearing a tone, and report that, when they are told the tone is playing...even though they can't actually hear it. So you are controlling for human perception/psychology, and allow only the use of their hearing to determine when they detect a tone.
And you can't "train" yourself out of having no bias effects. That's why the scientific method arose as it has: scientists recognize that even they, people most cognizant of bias effects, are human and prone to bias effects, to the method builds in ways of trying to control for that problem.
I'm wondering why, as a scientist, you don't seem to recognize this and it's relevance to understanding audio? There is no magic line dividing audio, and human perception of audio, from anything else in the empirical world we can study.
Do you realize the Toole trained many of the listeners. Just like the study I pointed out there were completely untrained and trained listeners. The trained listeners correlated better.
You make it sound like people are completely incapable of making obective observations using their senses. I will admit that the senses can be fooled but we can still make observations that correlate with reality.
I have a relatively unique way of evaluating the sound. I only listen to the left speaker. In this way I am not distracted by the right speaker and imaging and soundstaging etc. The right speaker is playing but it detuned. I only care about noise and distortion. Did the "thing" lower the noise floor? Did the "thing" lower the distortion? There are many cases where I expected the "thing" to help the sound quality but in reality it hurt the sound quality. And it was repeatable. How is that expectation bias? Here are a couple examples.
1. New power cords. I bought some rather expensive power cords. The dealer burned them in for a month (Roll your eyes all you want) and brought them over. Expectation is that these are going to sound great. We plugged them into my amps and I played Andreas Vollenweider's "the white winds". (I have listened to this 100's of times on multiple systems and through headphones. I know exactly what it sounds like.) At about 24 seconds there is a crescendo where we have a wide frequency band being played. WIth the new power cords all of the high frequency information was missing. This is bad. It took a couple weeks but the sound slowly recovered and now it is very well balanced. Expecation bias would have been that the heavens would have immediatly opened with the new cables. Well, that did not happen immediately as badly as I wanted it to or tried to will it into being.
2. Ground cables. a while back I bought a CAD GCR. I was plugging in ground cables and assessing the impact ot each component. I connected the GCR to one of the chassis screws on the clock. Expectaion would be that this is going to lower the noise floor and sound great. Well, that didn't happen. The sound was very edgy and sibilant. I moved the ground connection to a different location and the noise floor did drop and improve the sound quality. Completely repeatable.
Learn to listen and be objective. Press play and ask "What did I just hear?" No expecation, just the result. Describe it in words. There are many more examples where I did not like what the "thing" did even though I might have expected it to work great.
1 and 3 were proven. There was a race to the bottom (literally) of THD. Amplifiers with very low THD often did sound horrible. It was so notable that it launched the high end audio business. Shunyata Research and others have shown scientific evidence of the noise on power lines.
As for #2, this was one of the major things that brought Robert Harley notoriety back in the day. I was working on Chesky Records sessions at the time and digital audio understanding was moving fast. Chesky was one of the very first teams (actually I believe the first on 24/96) to produce commercially available dvd-audio discs which had the ironic title of "Super Audio Discs". I kid you not, lol!
Of course double-blind tests are necessary in medicine because of psychological bias. Yet the effect on the cohort, after controlling for placebo, is physiological. A hard data scientific readout.
Whereas in audio the readout is not physiological. It is an opinion by the observer. And there psychological influence comes to bear big time. A hard data scientific readout is elusive, if not impossible.
You should be able to appreciate that analytical difference in output.
I feel stressed when undergoing an audiophile blind test, yet I have nonetheless reliably found differences in blind tests because I am an experienced listener. Others report no stress, that's fine. But what if a less experienced listener is stressed and because of that cannot hear differences in a blind test, or if stress affects an experienced listener differently in a blind test than it does me? And how do you establish "scientifically" the experience or lack thereof of a listener? As you can see, you quickly land in murky waters where a hard data scientific output becomes next to impossible. (And again, sighted tests are not scientific either.)
I am very picky about this sort of stuff, because I want to separate real science from something that just seems that way. For the same reason I also have a beef with multiverse theories. They are *not* science, since in principle other universes are not observable (because of the size of our universe vs the speed limit of particles carrying information). The multiverse is at best philosophy based on science, but not actual science based on observation and experiment, i.e., real science, as I also practice every day at the lab bench. Contra a number of their colleagues, to their merit several prominent cosmologists have pointed that out.
Last edited:
Not at all. That's why I said early that obviously we hear real things all the time.
Exactly. That's the point.
IF we are looking for higher confidence levels, then a scientific approach makes sense, in which you control for things like imagination. You could be hearing something real, but in some case you might be imagining it. Controlling for bias helps weed out that variable.
It doesn't mean we can't play around, test things ourselves and (sometimes at least) come to reasonable conclusions. But if you REALLY want your confidence to be justified, then a more rigorous method than the "just trust your ears approach" would be suggested.
Ron is trying to come up with a protocol, presumably for testing whether or not everything makes a difference.
Could you walk us through such a protocol for how you think Ron might determine whether he should have absorption diffusion or neither on the front wall of his listening room?
But, again, that is because the psychological variable is controlled for. And the psychological variables can be in the form of manifesting as actual "physical" feelings/symptoms (your stomach really can get upset) and/or incorrect inferences "the pill caused my stomach upset last night" (where maybe something else caused it) etc. That's the important part. And the researchers are blinded to, to remove biased inferences.
I don't think there is this hard line you are trying to draw because at bottom we are talking about "having an experience" not merely "having an opinion." It's an attempt to understand the experience.
The subject in the medical study may "have an experience" of feeling better, or feeling more anxious, or feeling pain in their stomach. The question is what caused the experience? An expectation effect for instance? Or the drug in question.
Likewise when an audiophile tries two different sets of cables and "hears a difference" that is not just an opinion, it's an experience. The question becomes what is the rational conclusion to draw from that subjective experience. Is the rational conclusion that it was caused by a real sonic change that occurred between the cables? Or could it be a bias effect?
That's what blind testing helps weed out. There's no "get out of bias free" card to be played in audio perception.
I'm not making the case that blind testing is the be all and end all in terms of designing or evaluating audio gear. Only that it is far from irrelevant or useless for audio.
Here's the thing though: if you think THOSE variables are a problem, you should recognize that the well known variables in sighted listening should be a problem (more of a problem, because they can include everything you just wrote, and more issues).
Similar threads
