Separates vs. receiver confirmation (1 Viewer)

[Terrell]

Larry, Terrell has changed his mind and is trading his Proceed gear with my Denon 3801. Isn't that right Terrell

Right on mark! But since Larry has Proceed's big brother, Mark Levinson, you might as well trade with him.

 

[Larry B]

John:

i think this slightly misstates the matter.
i think the point is NOT whether individuals can detect differences in components, but whether ENOUGH individuals can - whether there's a statistically significant number of individuals that can differentiate between audio omponents.

I think most of us (maybe even Jaleel? ) would agree that there is considerable variation in the ability of individuals to distinguish sounds. As such, the ability to distinguish subtle differences could easily be overlooked in population studies if the "better listeners" comprised only a small fraction of the test population.
However, if a carefully controlled DBT identified even ONE individual who could consistently distinguish between 2 amplifiers, the point would be proven that all amplifiers do not sound alike. Whether or not most people can do so is an interesting, but different question.
Larry

 

[Norm Strong]

I can't believe that this argument has gone on for 7 screens. This is idiocy. Enough already!

 

[RicP]

I can't believe that this argument has gone on for 7 screens. This is idiocy. Enough already

Thanks for that most erudite evaluation Norm. If this thread is idiocy, then what does that say about you for contributing to the idiocy?

There's a great solution for threads you don't care for...don't read them. I don't believe that you are the arbiter of idiocy on this forum, are you vying for the position?

 

[John Doran]

larry,

I think most of us (maybe even Jaleel? ) would agree that there is considerable variation in the ability of individuals to distinguish sounds. As such, the ability to distinguish subtle differences could easily be overlooked in population studies if the "better listeners" comprised only a small fraction of the test population.

that's an excellent point, one i hadn't really considered.

but it still seems to me that much of this thread has been concerned with the legitimacy of DBT's - that is, do DBT's prove anything? but DBT's are statistical in nature, and thus deal with collections of numbers and with the question of what beliefs or claims are made reasonable by those numbers - i.e. for which claims or beliefs can those numbers be used as the best evidence.

as such - and you are right about this - locating one individual who can always distinguish between audio components WOULD easily be overlooked in DBT's, precisely because DBT's aren't looking for one person.

perhaps this a(nother) fault of DBT's. i don't know.

but, again, i agree with you - if there WAS such a person, it would really put a whole segment of the acoustic physics world on its ear.

but probably only for a while, since no doubt an argument would spring up as to whether the existence of ONE such individual was sufficient or conclusive evidence that there are audible sonic variations in audio equipment; perhaps (i can imagine the suggestion being) that individual is just picking up on OTHER non-auditory, sensory cues - perhaps tactile or visual or olfactory (???). or maybe that person is just the best guesser in the world...

and so on.

 

[Larry B]

Ric:

Thanks for that most erudite evaluation Norm. If this thread is idiocy, then what does that say about you for contributing to the idiocy?
There's a great solution for threads you don't care for...don't read them. I don't believe
that you are the arbiter of idiocy on this forum, are you vying for the position?

Not the most diplomatic, but you certainly made me laugh
Larry

 

[John Doran]

an afterthought:

DBT's can be statistical or non-statistical depending on what they're being used for.

for instance, if one is concerned with attempting to establish generally that people claiming to be able to recognize differences in the acoustic signatures in different audio equipment, in fact can't, then you would use the statistical results of DBT's.

if, on the other hand, someone was concerned to show that some individual person was unable to pick out sonic variations in audio hardware, then one would just use the DBT on that person, statistics-be-damned.

- jd

 

[Larry B]

John:

If, on the other hand, someone was concerned to show that some individual person was unable to pick out sonic variations in audio hardware, then one would just use the DBT on that person, statistics-be-damned.

With all due respect, I disagree. Statistics are equally important when testing an individual subject, as when testing a population. Allow me to explain.

Suppose I claimed that I could distinguish between Parasound and Rotel amplifiers, and agreed to a DBT. A reasonable way to conduct such a test would be to initially allow me to become acquainted with the sound of both units; then the test would begin. Suppose you first played the Parasound (naturally, without either of us knowing which unit was playing), and I correctly identified it. Would you conclude that I really could do what I had claimed? Probably not, because I had a 50/50 chance of being right. In other words, I might simply have guessed correctly. How about if the test consisted of two trials, and I was correct on both attempts? You would (or at least, should) still be skeptical of my abilities, because there was still a pretty good chance that I had simply guessed, and nothing more. But how about if I did this correctly 98 out of a 100 times? Using simple statistical analysis we would calculate a "p" value; for arguments sake, let's say this turned out to be 0.0001. (Actually, it would probabaly be more like 0.000000000001; you get the idea). The meaning of this number is that there was less than a 1 in 10,000 likelihood that my peformance was due to chance. This would be deemed "statistically signficant," and we would therefore conclude that in fact I could distinguish between the two amps.

Larry

Edited for clarity

 

[John Doran]

larry,

you're absolutely right.

i was using "statistical" to refer to collections of data about different individuals, not about one individual.

statistics ARE critical in the way you mention.

thanks for pointing out my mistake.

- jd

 

[Larry B]

John:
Glad I could be of help
Larry
P.S. Do you think this thread is finally over?

 

[John Doran]

man, i hope so.
i really, really do...:rolleyes

 

[RicP]

Ha ha! The thread lives!
Seriously, here's some interesting reading (not on DBT specifically, but similar points)
http://www.high-endaudio.com/magaz.html#Rel
Here's my favorite quote:

who claims that there has been no true progress in home audio reproduction, and that all components have an equal potential to reveal the reality and essence of "live music", subject only to "taste" and matching, is either ignorant, a liar, a coward or an incompetent.

 

[John Doran]

ric,

you're right - that's an incredibly interesting site.

whatever the merits of his arguments, though, i can't say i agree much with making statements like :

Anyone who claims that there has been no true progress in home audio reproduction, and that all components have an equal potential to reveal the reality and essence of "live music", subject only to "taste" and matching, is either ignorant, a liar, a coward or an incompetent.

i mean, it's just name-calling, plain and simple, and, far from adding anything to his position, it actually subtracts from it. i mean, i can understand how profoundly frustrated he must be, but that's just no excuse, if you ask me. even if he's right, there are just some things that shouldn't be said.

anyway...

incidentally, how do you see his synopsis of what he calls "audio relativism" relating to the debate at hand? which side of the fence do you think he's on?

 

[JaleelK]

The comparison between DNA testing and DBT of audio equipment is a FALSE ANALOGY.

DNA testing involves the probablistic matching of two strains. It is proven mathematically and through statistics. The sample size is small and the confidence interval is exteremely high. The samples are nucleotide pairs and the nucleotides have a definitive state. DNA testing is a science.>>>>>

That's not my argument. My issue is, does one need to have actually conducted a DNA test to advocate its validity, I don't think so.

DBT of audio equipment uses a relatively very high sample size and the confidence interval is not as high. The samples are human beings who by nature is subjective and does not have a definitive state. Statistics in this method carry little value and are misleading. DBT of audio equipment is NOT a science.>>>>>

Do you think sighted non-level matched listening test is better way of comparing audio rather than Double blind and Level matched.

What you are effectively saying is that margarine and butter both taste the same. When in fact they don't.>>

What I'm saying , because I get attacked for not having conducted my own DBT, do I really need to for me to advocate the use of DB(level matched)T, once again, I don't think so, I'm not a scientist or an audio expert, I'm advocate for the use of DBT based upon the expertise of those who are authorities on this subject. I have cited statements from experts/scholars and have posted links to some of them and they all agree that at least blind testing necessary in obtaining some type of fair way of determining if audible difference exist between audio components. To me its logic and commonsense to know that you must have some kind of controls in listening test keep a persons bias from being entered into their perceptions.

 

[AjayM]

And the people who have done their own DBT's that were level matched and fall within your own range of acceptable you immediatly said were not real tests.

Andrew

 

[RicP]

I'm not a scientist or an audio expert

Finally, Jaleel says something we can all agree on.