| Columns Retired Columns & Blogs |
| Loudspeakers Amplification Digital Sources | |
Analog Sources Accessories Featured | |
Music Columns Retired Columns | |
Show Reports | |
Features Latest News Community | |
Resources Subscriptions |
© 2024 StereophileAVTech Media Americas Inc., USA
All rights reserved
> The differences are audible and measureable, see:
> http://www.analysis-plus.com/design_whitepaper.html for just
> one example.
Can you point where in this example it shows the differences to be audible?
Unfortunately this would not work. Preconceptions and listening bias go both ways; those that are convinced they will not hear a difference will not.
So if those who want to hear a change do, those who don't won't....that means it's all basiclly nonsense. cus' a difference that is real is audible to all. Play a Cd play the DSD/SACD of that same recording only one is DSD the other PCM, the audible IMPROVEMENT is audible to all. And it scienctifally measurable, quantified, numerically formulized. And mostly plainly audible. And to get this easily audible improvemetn doesn't cost a lot on money anymore, since it's rreal electroncis, and liek all consumer electroncis, it gets cheaper better smaller. But why do some of teh cables sold by the cable method of teh month mfg. association always sell stuff for absurd prices, yeah $1000/ foot for some magic wire, that has no industry specs, no standards labeling, matter o rfact doesn't even have type or ratings stamped on it like legit insudstry wires used in electronics, electrical or any other legit business. Take a standard L5 cord cap, wrap it in shrink, call it some unknown bizzare concept price it at ridiculous levels, advertise in magazines, and ya sell a few hundred, and move to the next month design. The only thing inovative is the line of nonsense written up in this month's ad copy.
First of all, thanks to everyone for keep this thread and this discussion flame free. The threads on the topic of DBT all too often end up as simple flame wars from which nothing useful is gained.
Now onto the business at hand.
To clear the air, I never stated that DBTs are used to "measure" differences in the sound of audio equipment, rather I stated that DBTs are "used as a method for testing the differences in the sound of audio equipment". Testing for differences and measuring differences are two completely different things.
I applaud Andy for his efforts in trying to get us thick headed audiophiles to understand where the pro-science crowd is coming from but I still don't understand why he and the other members of the pro-science crowd have so much trouble understanding where it is the the audiophile crowd is coming from.
So let's start back at the beginning for just a moment.
What is it exactly that these DBTs are trying to accomplish? As I understand it these tests are designed to eliminate the bias of knowing which brand of equipment is playing since this "brand knowledge" may cause serious problems with one's ability to remain neutral in any given listening test. Okay, I understand the desire to eliminate this bias from a listening test and DBTs are indeed a method for eliminating this "brand knowledge" bias.
So far, so good. So DBTs' are not flawed as a method of eliminating the "brand knowledge" bias but the fact remains that as a listening test DBTs are still seriously flawed. Why do I state this? For one thing, how does one determine just what sonic differences to test for? In other words, the DBT may or may not ofer the proper equipment or audio samples to fully test, or more to the point, to fully reveal the differences between any two given pieces of equipment. All that a DBT can ever possibly hope to prove is that for the pieces of equipment under test there is no perceived sonic differences under the that DBT's parameters. In other words, in the case of a test on cables, the differences between cable A and cable B, if there are any differences, could not be heard when playing - insert musical selection from DBT here - on - insert all the other audio equipment used in the DBT here - in - insert room parameters of the DBT here.
So basically all that a DBT does is 1) eliminate the "brand knowledge" bias from a listening test and 2) control only one variable (the pieces of equipment under test) of the many, many variables that exist when it comes to the reproduction of sound and music via an audio playback system. And yet this obviously flawed "scientific method" is held up as the final word in audio testing.
Change any one of the other variables involved in the DBT and results may change as well. A different musical selection or a different set of speakers may reveal differences in the equipment under test that had previously been masked.
So much for the high holy status of the DBT.
Now onto cables and other sticky subjects.
I don't quite hold with the cables make a huge difference crowd but I do feel that cable can and do make somewhat of a difference. I feel that within any given class of cables there are not very big differences but between the different classes of cables the differences can be quite noticeable. However, differences in cables are often very subtle and not very easy to hear under DBT conditions. I know that this sounds like a cop out but the fact remains that DBT conditions are just that - DBT conditions - and there is no guarantee those conditions are adequate enough to reveal subtle differences in audio equipment.
As for the subject of lossy codecs versus lossless codecs, in addition to the shortcomings of DBT there is the very unscientific belief that one can get something for nothing, in other words, that a data set with less data is the equivalent of a larger data set. I don't know about your own engineering and science education, but one of the first things that I learned in engineering school is that one cannot get something for nothing, however, if you still chose to believe this then I know of a great investment opportunity in this perpetual motion machine, just PM me and I'll give you instructions on where to send the money.
That would be my ears....
Nice summation jazzfan.
Dave
Unfortunately you missed the tongue firmly planted in cheek.
My point is that those who crusade against cables are so determined to deny even the possibility that wire can sound different that they will never hear nor admit to any difference - mo matter how great that difference may be.
Not only do many not hear a difference or improvement with any higher resolution formats, those that can hear an improvement debate whether high resolution PCM or DS sounds better. Moreover all released DSD based recordings have been transcoded to PCM for editing (as I have pointed out before.) Thus, if DSD is inherently better than PCM the DSD stream we are listening to has has already been "damaged" by this foray into the evil world of PCM.
Nor is DSD measurably shown to be better, just different. For example it exhibits a great deal of noise not present in PCM (high or CD resolution). Now whether this is a problem or not is up to the individual listener to decide.
> Testing for differences and measuring differences are two completely
> different things.
I am not sure what you think measurement means (or testing for that matter) but it is a requirement not an option when trying to make any scientific statement. If you cannot measure it then it lies outside the scientific domain and your stated objective in the previous post was to show that DBTs were scientifically invalid. Asking someone if they can hear a difference is a measurement of a difference.
> but I still don't understand why he and the other members of the pro-
> science crowd have so much trouble understanding where it is the the
> audiophile crowd is coming from.
As far as I am aware, I am having no difficulty understanding the substances of your arguments and responding to them by pointing out where your reasoning is at fault when it comes to convincing those that hold scientific beliefs rather than audiophile beliefs. I am having more difficulty getting you to pick up on my points but that is not unexpected given the difficulties you would have in putting forward reasoned arguments.
> So basically all that a DBT does is 1) eliminate the "brand knowledge"
> bias from a listening test
If implemented correctly that is what blinding is supposed to do. Nothing more and nothing less. So no problems here.
> 2) control only one variable (the pieces of equipment under test) of the
> many, many variables that exist when it comes to the reproduction of sound
> and music via an audio playback system.
Blinding performs the function you stated in point 1). What you have chosen to measure in the experiment determines what is to be held constant and what is to vary. This is why you have to know what you are measuring before you can design an experiment to measure it. This point would seem to be wholly irrelevant to the ability of blinding to perform its function.
> Change any one of the other variables involved in the DBT and results may
> change as well.
That would be a different experiment and, depending on the variable, it would be measuring a different form of sound perception and so would be expected to be different. This point would seem to be wholly irrelevant to the ability of blinding to perform its function.
> And yet this obviously flawed "scientific method" is held up as the final
> word in audio testing.
This might be Catch 22 because the scientific method is the foundation of all scientific knowledge. If you wish to show that a DBT audio test is not providing valid scientific information you would have to use the scientific method to show it.
Unless you make an effort to get your ducks in a row I doubt you will be able to convince anyone with even a mild scientific inclination of audio DBTs problems by this sort of argument. It may help if you recognise that sound perception is not constant and what you perceive as sound depends not only what is impinging on your ears but a wide range of other factors particularly visual cues and past experience. As a consequence, perceiving sound under blind conditions means just that and perceiving sound under other conditions is measuring a different quantity.
>>> The differences are audible and measureable, see:
>>> http://www.analysis-plus.com/design_whitepaper.html for just
>>> one example.
>>
>> Can you point where in this example it shows the
>> differences to be audible?
>
> That would be my ears....
? I presume that is a no.
Because I am intentionally stacking the deck against a null result, attempting only to prove that cables can sound different, to those who assert that this is impossible.
Incorrect.
I feel comfortable speaking for Dave in stating that objective observation is as valid is any other test and, in fact, is part of the scientific method.
"Asking someone if they can hear a difference is a measurement of a difference."
-andy19191
> Because I am intentionally stacking the deck against a null
> result, attempting only to prove that cables can sound
> different, to those who assert that this is impossible.
I am not sure you would be addressing a problem that actually exists. I have seen audiophiles make the claim of others but have never seen anyone actually make the claim. What I have seen are claims along the lines of normal cables under normal conditions sounding the same. I find it hard to believe that anyone would expect a cable with a box of components attached to sound neutral.
>>>>> The differences are audible and measureable, see:
>>>>> http://www.analysis-plus.com/design_whitepaper.html for just
>>>>> one example.
>>>>
>>>> Can you point where in this example it shows the
>>>> differences to be audible?
>>>
>>> That would be my ears....
>>
>> ? I presume that is a no.
>
> Incorrect.
Somebody is losing the thread here. Dcstep posted a link claiming that differences were audible. I looked at the link and saw nothing to support the statement and asked where. To which I got the apparently meaningless response "that would be my ears". Your response also does not address the question of where these audible differences are reported. Are you answering this question or another one? If another one, can you please tell me what it is and where it came from.
Andy,
My ducks are all in a very nice row. It's the damn elevator that won't go to the top that has got me worried.
After going several rounds of exchanges I think I'm beginning to understand just where you're coming from. Here's an example of what, I think, it is you're trying to tell me. As I'm understanding it, a DBT is good for testing the sonic differences between two given pieces of the same type audio equipment when 1) the subject does not which sample is being used at any given time and 2) with all other variables fixed, i.e. nothing else in the system is being changed.
If the previous statement is true, which I believe it is, then that's exactly why many audiophiles feel that a DBT is worthless. Worthless because it tells me next to nothing about how the tested pieces of equipment may sound in my system. I said "next to nothing" on the remote possibly that my own audio system exactly matches the DBT setup. Please remember that the audio system is just one of the many variables that will be different when using the "tested" equipment at home. But I suppose that since at home I can actually see the brand name of the equipment my judgment would be so clouded that I would believe just about anything someone happens to write in a high end audio magazine.
No, that would be a yes. I've heard the differences so I don't need you or some anyone else to comfort me. The article talks about audible differences and what we can hear. It's objective isn't to "prove" that we hear differences, it's just seems to be explaining that the ear is sensitive to certain things. The fact that it didn't elect to prove that, doesn't invalidate what it says.
There's an interesting thread over on head-fi.org where a guy posted some mp3 vs. lossless of the same data. About ten people took an ABX test and eight or nine only got 20% to 35% correlation and one guy got 95%. Lots of people thought that proved that there was no difference. As a scientist you'll see clearly that there was in fact a difference. The test minimum to record a score was at least 10 cycles, so the guy with 95% did over 10 cycles and only missed one or two (depending on how many cycles he took). To get 95% I'm thinking that he missed one out of 20. I suspect this guy had trained his ears.
The same head-fi test results clearly show that the average listener can't hear even the difference between mp3 and lossless. I would agree that the differences between cables are even less gross; therefore, MOST listeners will not hear the differences between cables.
So, testing, whether DB or ABX, needs to include at least some "experts" or "golden ears". I think that the Fremer challenge would qualify for that, if Randi would belly up. I wish that it would happen, but it probably will not.
My advice to the average audiophile is not to spend lots of money on cables and interconnects. There IS a lot of mumbo jumbo out there, but there are some truly good products for those that do think that they hear differences. If you do hear differences, don't worry about the scientific comunity and just buy what you like. If you're uncertain and/or don't hear a difference, then don't worry and buy what you like.
You know what I hear, I just changed a rectifier tube in my Woo headphone amp. It made a big difference that was pretty easy to hear. I thought the Woo was very nice before the tube swap, now I think it's incredible. Some of you may think that I wasted $250 on that new tube, but I don't care. Even if I just imagined it, it was worth it to me. I'm still going to update the thread where I mention the Woo and you guys can say I'm full of sh*t all you want, but I'm going to say what I hear.
Dave
It came from our ears. We hear the differences, so we don't need you or anyone else to validate it. I'm not here to "prove" anything for you. If I did "prove" something to you, that there IS an audible difference, then I'd still recommend to you that you not spend money in this area because you cannot hear the difference. Most people can't hear the difference, so most people shouldn't spend money in this area of their systems.
I know that you're arguing about the validity of DB and/or ABX testing. I actually believe this type of testing is valid, but difficult to do correctly. Also, even when done correctly it needs to include some trained listeners. Otherwise, it's like having a freethrow or threepoint contest and excluding NBA players and then concluding that "nobody" can hit nine out of ten.
There is a difference between average listeners and trained listeners. We can all agree, if there is a trained listener, then Michael Fremer should qualify, despite his old ears, he still spends a lot of time listening to system differences. The conditions of the proposed test, as he's outlined them, should satisfy both the scientific and non-quantitative communities.
I'm wondering if Stereophile shouldn't sponsor the test. Use the NAS as a moderator, to validate the test and have Mikey go at it. Since Mikey has already proposed to Randi that he'll accept the challenge, how about preempting Randi and performing the test under supervision using the guidelines proposed by MF?
Dave
Great. For a minute there I thought you were saying that there are no audible differences
between cables. It seems we have nothing to dispute, then. What exactly are you trying to
discuss?
Great couple of posts, Dave. It's good to see that at least someone did not get confused with my post.
You are absolutely correct that a DBT test of cables must eliminate the variability of testers as an unknown.
In psychoacoustic DBT testing the difference between the inputs is known by the researcher and usually by the subject as well. What is blind is which is which. What is being tested is the audibility of the difference. That is, we know there is a difference - can it be perceived by humans, how is it perceived and under what conditions?
In the BT proposed by cable skeptics there is an attempt to prove that no difference in cables can be perceived, but without knowing whether the listener can perceive differences. That is, under this scenario we don't know whether we are testing the differences in cables or the differences in listeners.
This is what scares Randi. In Mr. Fremer he has a listener who states "I can tell the difference". "I know what cables sound like and will recognize when they have been changed. Give me an system/environment that I know well and I will demonstrate this." Randi will not accept this challenge as he will now lose.
When we change a cable in our home systems we can perceive whether there has been a change as we know the system well. The more often one tries different cables the easier this becomes; we train our ears through experience to recognize the types of changes that cables make.
These changes can be subtle. In fact, at least for me, it has been rare that the changes are worth paying much for and I return the cable. This doesn't mean however that the cables don't make a difference. I just don't find the yuck/buck ratio great enough for me to get excited and would rather spend the money on a set of race tires.
I have no problem accepting that some cannot hear these changes, that some don't find the changes significant enough with which to bother and that some refuse to even accept the possibility. But to deny the existence of repeated observation by many people is ludicrous and indeed unscientific. All science begins with observation.
BTW, I continue to be fascinated by the observation of many that not only does changing the rectifier in the Woo amp makes a difference, that this difference is greater than changing the output tubes. What fun!
OMG, you race too? How many places will you find two guys that play trumpet and race also.
I was pretty serious in SCCA and club autocross in the central Texas region. I never went to the nationals, but did some regionals and raced against several national champions. I did F-stock and E-Street Prepared in a 5.0 Mustang LX.
I've laid off for over ten years except for an occasional marque club race ('vettes and Porsches). I've gotten the wild hair again and dreaming about a GT3 or ZO6, but I'll probably end up with a Honda S2000 or Mazda RX8 to compete in either AS or BS. Still, I may just blow off being super competitive and get an M3 as a more practical daily driver/racer. (I drive what I race).
What racing do you do?
BTW, I haven't even tried rolling the power tubes on the Woo. The guys over at Head-Fi.org seem to think the most bang per buck is in the rectifier. I'm certainly happy with my new GZ34. I just wish it looked sexy like the Sovtek.
It's plain-Jane.
Dave
http://stereophile.com/asweseeit/894awsi/
http://stereophile.com/asweseeit/107/
Mostly club events now and I do some instructing. My last "serious" stuff was SCCA T1 in a Z06. I still have the car - as you know, it's perfectly street legal as long as I use the OEM lap/shoulder belt rather than the harness.
> After going several rounds of exchanges I think I'm beginning to understand just
> where you're coming from.
I am afraid I can see little evidence of this. Without a recognition of differing sets of belief and therefore perceiving the world working in different ways progress in this direction will be difficult.
> Here's an example of what, I think, it is you're trying to tell me. As
> I'm understanding it, a DBT
This is perhaps the first place we part company. To audiophiles DBTs would seem to be a great big bogeyman that needs to be mightily opposed. This binds the group as can be seen in the emoton of the posts around but is a bit of a problem for those with more mainstream beliefs that cannot see the bogeyman. Yet you continue to tell me to look at the warts on his nose, greasy hair and bad breath. Tell me about an experiment and what is being measured before having a go at the blinding component. I can see this.
> is good for testing the sonic differences
What is a sonic difference? Does it exist as vibrations in the air that can be measured with microphones or only in the brain of a listener. The former is to be preferred because it is constant (repeatable) and much easier to reason about. The latter can also be measured and reasoned about but this is more difficult because it varies with a wide range of parameters and is a lot less repeatable.
> between two given pieces of the same type audio equipment when 1) the subject
> does not which sample is being used at any given time and 2) with all other
> variables fixed, i.e. nothing else in the system is being changed.
This would seem to be an outline of an experiment where what is being measured is the perception of a difference under blind conditons between two different audio components. This makes sense to me if we set aside questions about what "same type" means.
> If the previous statement is true, which I believe it is, then that's exactly
> why many audiophiles feel that a DBT is worthless. Worthless because it tells me
> next to nothing about how the tested pieces of equipment may sound in my system.
If you measure the height of a chair with a tape measure it is quite ludicrous to criticise the tape measure because it has not provided you with the weight of a bag sugar. If you want to know the weight of a bag of sugar then measure that and not something else. And don't blame the tool.
> I said "next to nothing" on the remote possibly that my own audio system exactly
> matches the DBT setup. Please remember that the audio system is just one of the
> many variables that will be different when using the "tested" equipment at home.
We would seem to have returned to belief systems and how to reason about the physical world. This is where the real difference lies between audiophiles and those with conventional scientific beliefs. The rest is just applying those beliefs.
> But I suppose that since at home I can actually see the brand name of the
> equipment my judgment would be so clouded that I would believe just about
> anything someone happens to write in a high end audio magazine.
It is not to do with judgement in the sense of sitting there working things out but to do with what the brain does automatically to aid perception of what is going on in the world around. It uses all the senses together with past experience unless you block it by, for example, blinding.
But what you choose to feed your brain builds your past experience and can affect this automatic aid to perception as vividly demonstrated by the McGurk effect. Your beliefs, regardless of whether they are "true" or not in the physical world, are true to you and your brain. So what you read and believe in high end audio magazines can influence your perception of sound if they reinforce your experiences and follow what you believe to be going on.
For example, it may seem obvious to you that a big expensive cable performs better than a normal cable because of what your believe, what those around you believe and what your experiences of casual listening suggest. Whereas in my case, it seems obvious to me that they will sound the same (given the usual caveats) because of what I believe, what those around me believe and what my experiences of casual listening suggest. So if we sit down together and you swap some cables there is every chance you will perceive differences and I will not. No problem.
Unless, of course, you are unwise enough to project your experience into the physical world (the world of science!) believing that the sound impinging on your ears has been modified by the cable an audible amount rather than your perception of it has been modified by an amount that is audible to you (and not me). Or you believe that your perception of the sound will not follow the visual cue.
Andy,
I think that we part company well before the subject of DBTs versus subjective listening tests is even reached, namely at the assumption that science is capable of accurately measuring the subtle sonic differences between audio components, especially cables.
This is really the crux of the matter. The pro-science crowd wants desperately to believe that today's science is so advanced that something as simple as testing for the differences (if any differences exist) in the sound produced by two different audio cables, whether they be interconnects or speaker cable, can be easily tested. And since what we presently know of the science of cables, as understood by the cables simple physical and electrical properties, declares that there cannot possibly be any difference between two cables with similar physical and electrical properties then we go about structuring a testing method to prove that this knowledge is sufficient to know everything there is to know about the "sound' of a given cable. And yet many, many people can tell which cable is which in a DBT and can even describe the differences between the cables.
To put it another way, human emotions exist but I never seen a way to accurately measure love, except for those wonderful coin operated carnival machines which can predict one's entire love life with a simple squeeze of a hand lever. About as useful for measuring "love" as a DBT is for measuring sonic differences.
You know, I always thought that the trumpeter personality and the racer personality (showoff, risk takers) were one and the same, but, until now, I never met another one. NONE of my trumpeting friends race and vice versa. (Maybe it's the general lack of money of musicians).
I once won a set of autocross tires at an SCCA regional event, playing my trumpet. It was West Texas, so I played I'm An Old Cowhand with the Jamey Aebersold CD backing me. It blew the crowd away. I think they were totally taken by surprise that a racer could play ANYTHING. (In fact, around the same time I think that Jamey Jr. won a national SCCA talent contest).
Dave
Did you have to play while blindfolded?
Interesting point. It wouldn't have mattered, it was all aural.
> namely at the assumption that science is capable of accurately measuring
> the subtle sonic differences between audio components, especially cables.
Unless you define what subtle sonic differences means in a form suitable to be measured in the physical world then science can say nothing about it and nobody with any knowledge of the basis of science would claim or assume it could despite your protestation to the contrary.
Audiophiles often use a whole host of words that do not map to anything identifiable in the physical world, "musicality" is my current favourite, which is fine so long as the limitations are recognised when it comes to reasoning about matters in the physical world when they have to be replaced with something that is measureable.
> The pro-science crowd wants desperately to believe that today's science is
> so advanced that something as simple as testing for the differences (if
> any differences exist) in the sound produced by two different audio
> cables, whether they be interconnects or speaker cable, can be easily
> tested.
The advancement of science is wholly irrelevant to conducting an experiment to determine the audibility of differences between cables. What is relevant is the methodology and it's relationship to the scientific method should one desire the results to be recognised as having scientific validity.
> And since what we presently know of the science of cables, as understood
> by the cables simple physical and electrical properties, declares that
> there cannot possibly be any difference between two cables with similar
> physical and electrical properties then we go about structuring a testing
> method to prove that this knowledge is sufficient to know everything there
> is to know about the "sound' of a given cable.
Again, knowing about the "science of cables" is wholly irrelevant to conducting an audibility experiment which is about method.
> And yet many, many people can tell which cable is which in a DBT and can
> even describe the differences between the cables.
I would agree that many audiophiles believe they can (usual caveats about normal cables and conditions) but this is not the same as demonstrating that they can. An excellent illustration of this occurred recently with Mike Lavigne who had the grace to honestly report what he perceived when it all unravelled in a check prior to going off to collect Randi's million.
Just so they don't expect him to drive blindfolded. Then again, some of the racers out there drive as if they can't see . . .
(Great story. I make sure all my car buddies know I play as they are a great source for gigs.)
Okay, so was the crowd blindfolded? And if so, did you switch mouthpieces just to mess with them?
I used my plunger on the second chorus, but alas their eyes were wide open; however, they might have been drunk.
Dave
Who's this andy guy and why does he keep showing up in "our" thread??
Dave
I don't know but I rather enjoy these exchanges with him. He never resorts to flaming and always presents his arguments in a very clear and concise manner plus he spells "the" as the and not "teh". However, I do get the feeling that we are just going around in circles. The DBT crowd are the ones who claim that the results of a DBT can prove that no audible differences exist between two cables. To which I say, that's true but only for the given test conditions and the people who took part in that test. Furthermore, in no way does a DBT prove anything that can be taken as a universal truth and I state that because good science practice agrees with me and not them. Key word: repeatability - since using a different set of people or a different equipment setup may yield different results.
Such a test can only prove that differences exist, or fail to prove that differences exist. It cannot prove that no differences exist. You can't prove a negative.
True. However, if the same test is conducted many, many times with many different people under the same conditions we start to approach a certainty as to the characteristics of the bell curve. We accordingly possess a well-developed understanding of the overall characteristics of human perception. This does not mean however that each of us, as an individual, possess that "average" characteristic.
With respect to cable differences, some people will perceive them but the majority of people will not. Similarly, the majority of people cannot tell whether a perfect fifth is in tune. This lack of common average experience does not prove that cables cannot make a difference however.
What I don't understand is why some people make such efforts to argue against cable differences on paper. If you cannot hear a difference, consider yourself lucky; you need spend little on cables.
Perhaps some day we will learn how to measure that which makes some cables sound different. Or why some CDPs sound so different even though the frequency response and other measurements are the same.
Pages