Propeller Head Plaza

Technical and scientific discussion of amps, cables and other topics.

Return to Propeller Head Plaza

DBT's in 2007 Prove People Can Detect Differences In Audio Components

74.235.213.10
Posted on October 16, 2009 at 15:54:52
thetubeguy1954
Audiophile

Posts: 3149
Location: Orlando, Fla
Joined: January 7, 2001
 Two years ago Telarc stated online: First, in a blind A/B comparison using their existing ultra high-end converters, the analog output of their custom console, and the DAD AX24 in a session with the Atlanta Symphony Orchestra, Bishop, Telarc producer Elaine Martone, assistant producer Thomas Moore and two assistant engineers all independently chose the output of the AX24 as obviously closer to its input. In fact, if there was a difference between the AX24's output and its input, it bordered on imperceptible, even in Bishop's carefully aligned ATC monitors.

Ever critical, Bishop repeated the blind A/B test when he returned to Telarc's home studios in Cleveland. Three producers and four engineers evaluated the DAD AX24 and all seven independently selected it as the best converter from among Telarc's cream-of-the-crop converters. "Ten sets of ears, five of which have earned Grammys and the remaining five of which have Grammys in their futures, all independently selected the DAD AX24 as the absolute state-of-the-art," Bishop summarized. "The AX24 is a truly exceptional piece of electronics." That's 2 seperate DBTs and at least 10 audiophiles who were able to detect when the DAD AX24 was being used!

Thetubeguy1954

~Rational Subjectivism. It's An Acquired Taste!~

RE: DBT's in 2007 Prove People Can Detect Differences In Audio Components, posted on October 17, 2009 at 09:49:48
Dan Banquer
Manufacturer

Posts: 7440
Joined: November 13, 2002
 I'm not surprised.
d.b.

RE: DBT's in 2007 Prove People Can Detect Differences In Audio Components, posted on October 16, 2009 at 17:35:25
tomservo
Manufacturer

Posts: 2517
Joined: July 4, 2002
 I am not sure the ability to hear the difference between some components has ever been an issue has it?
An ocean of space is between this and saying all components (of equal values) have there own sound however.
To be fair, what you’re talking about isn’t a resistor or capacitor or inductor, or even a transistor type or IC but an entire processing system.

In any case, if one has done testing “without prior knowledge” of which is which, in the most “normal like” listening conditions possible and you get a result, you have shown your hearing what ever it is that is different between A and B.
Best,
Tom
RE: DBT's in 2007 Prove People Can Detect Differences In Audio Components, posted on October 17, 2009 at 17:27:00
thetubeguy1954
Audiophile

Posts: 3149
Location: Orlando, Fla
Joined: January 7, 2001
 Tom,

As always you're a beacon of honesty and integrity here. You sir are intellectually honest as they come. I've never seen you looking for a way out when what passes as being "objectivist" beliefs here are in question. For Example: One typical "objectivist" mantra is: 3 decades of tests have proven not one audiophile can detect differences in components when brand names are hidden and spls are matched. However let a blind A/B test "prove" people can detect differences and then these so called "objectivists" start talking about details like statistical significance, which controls are in place or the possibily of biases coming into play.

However when a blind A/B test "proves" people cannot detect differences, these same "objectivists" express no concern over details like statistical significance, which controls are in place or the possibily of biases coming into play. Now all they do is start chanting their 3 decade mantra over & over. To me that's being intellectually dishonest. I hope you understand when I speak about objectivists here on PHP I'm not including you, just the type of "objectivist" I spoke of above.

Thetubeguy1954

~Rational Subjectivism. It's An Acquired Taste!~


RE: DBT's in 2007 Prove People Can Detect Differences In Audio Components, posted on October 18, 2009 at 07:33:27
tomservo
Manufacturer

Posts: 2517
Joined: July 4, 2002
 Thanks for the kind words tubeguy but I don’t think one can go very far coming up with something new, if they believe “all is known”.
I have worked on this stuff for a living most of my life, it has been important to draw my own conclusions and then solutions.
There is no point in belonging to one side or the other if your trying to make something new, you use all the available tools including ears.

On one hand, test equipment and measurements are crucial but at the same time they only tell you about the question the equipment asks, does nothing to tell you what sounds like what or why. Like how a micrometer can tell the thickness with high precision, but tells you nothing else and makes a lousy C-clamp.

I do find it odd that this whole thing is so partisan, like politics in a way where in reality uncertainty rules most of the time.
Like politics, you can pretty much toss out the extremes in both directions as that usually represents people who want / need to see with certainty –or- have a financial interest in things being one way or the other.


At least working with the stuff, the conclusion I came to was that there are relatively fewer hard and fast rules, circuit theory is valid but has layers of complications, there are many things which are true most of the time but not always (conditional truth) and always there are things you don’t understand or where a mental connection isn’t made.

You can look at a schematic, see a symbol for a capacitor and think of it as a capacitor, or, you can think of it as a capacitor that has a series R, a series L, a parallel R and some properties which are level dependent.

How much those bonus items change the imagined operation, depends on how large they are relative to the governing conditions if that makes sense.
For example, in crossover design, at least the second order effects are needed to model and predict something close to what you measure.

In any case I have heard amplifiers sound different in blind testing, speakers always sound obviously different one brand to another.

Take three of our best speakers, put them side by side and you can still tell differences between them even though the measurements are similar, even though the goal is identical and were designed by the same person.
Personally I applaud your inventor like efforts to bridge the gap, if this stuff was obvious, the gap wouldn’t exist nor the discussion.

I guess what i am saying is that for people that want a black and white situation, look elsewhere, audio isn't it.
Best,
Tom
RE: DBT's in 2007 Prove People Can Detect Differences In Audio Components, posted on October 17, 2009 at 05:01:14
Pat D
Audiophile

Posts: 7558
Joined: June 20, 2000
 I am not convinced these were double blind tests. They describe the test in Denmark as "a blind A/B comparison" and in Cleveland "Bishop repeated the blind A/B test." So the description is compatible with single blind comparisons or tests.

As well, we have no indication of what controls were used or how many trials were done.

I would like to see if the results could be replicated in controlled double blind tests.

"Probability is the very guide to life."---Cicero
That's OK... I guess we'll have to live with reality of you not being convinced., posted on October 18, 2009 at 18:37:41
carcass93
Audiophile

Posts: 2634
Location: NJ
Joined: September 20, 2006
 That wasn't the goal anyway.

Ahh, but I supplied reasons, and you ignore those. (nt), posted on October 22, 2009 at 18:21:32
Pat D
Audiophile

Posts: 7558
Joined: June 20, 2000
 .

"Probability is the very guide to life."---Cicero
I am not convinced these were double blind tests., posted on October 17, 2009 at 11:40:58
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 Maybe they were maybe they were not. But we do know enough that one could varify the results if they care to do so. We don't know that bias effects had any influence on the results and we know that the tests were repeated with the same results.

Analog It Also Proves LFOs Will Find A Problem With Every Blind A/B Test, posted on October 17, 2009 at 14:55:40
thetubeguy1954
Audiophile

Posts: 3149
Location: Orlando, Fla
Joined: January 7, 2001
 When those tests don't confirm their mantra of: 3 decades of tests have proven not one audiophile can detect differences in components when brand names are hidden and spls are matched. Strangely details like statistical significance, which controls are in place or the possibily of biases coming into play only become a concern for objectivists when the blind A/B tests provides proof that people can detect differences!

There are certain people I can no longer respond to here as per "The Bored." I don't believe I can even mention their name. If I was you I'd ask objectivists why they never worry about details like statistical significance, which controls are in place or the possibily of biases coming into play when the blind A/B tests provides "proof" that people cannot detect differences! Isn't it equally a concern for them then?

Thetubeguy1954

~Rational Subjectivism. It's An Acquired Taste!~

I've called them on cherry picking numerous times, posted on October 17, 2009 at 15:50:58
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 It's easy to wave the science flag. Real science is hard work.

Falsely., posted on October 17, 2009 at 17:38:50
Pat D
Audiophile

Posts: 7558
Joined: June 20, 2000
 You're good at making up allegations, poor at supporting them.

For example, my questions about this one are perfectly objective and your defence is worthless.

"Probability is the very guide to life."---Cicero
Wrong, posted on October 17, 2009 at 18:53:21
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 everything you just said is bullshit. nuff said

Are ten sets of ears statistically significant?, posted on October 16, 2009 at 16:27:56
Ugly
Audiophile

Posts: 1420
Location: Des Moines, WA
Joined: August 22, 2006
 I'm no statistician but the data seems a bit light imo.

Number of ears is irrelevant, posted on October 17, 2009 at 11:43:50
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 it's the number of trials that matter.

I suspect that they are . . ., posted on October 16, 2009 at 19:31:52
caspian@peak.org
Audiophile

Posts: 386
Location: Oregon
Joined: January 12, 2008
 . . . when they are the ears of highly trained audio professionals, thoroughly familiar with the sound of live symphonic music, and they agree unanimously. These are people whose careers depend upon their ability to hear and define subtle differences in sound, and who have refined their ear/brain systems to precision instruments. I strongly suspect that their judgments correlate rather well with the acoustical and electronic measurements of the equipment in question.

RE: I suspect that they are . . ., posted on October 16, 2009 at 21:58:38
Ugly
Audiophile

Posts: 1420
Location: Des Moines, WA
Joined: August 22, 2006
 I would expect that for real "proof" objective analysis of the data disregards any potential biases such as appeal to authority even if it does seem very convincing. When you start talking about experimental data sizes I expect that the chances of luck potentially off-coloring the results gets bigger. What is required for proof?

95% confidence level. nt, posted on October 17, 2009 at 11:45:44
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 nt

95% confidence is required for proof, or a single trial with 10 sets of ears gives 95% confidence?, posted on October 18, 2009 at 00:11:45
Ugly
Audiophile

Posts: 1420
Location: Des Moines, WA
Joined: August 22, 2006
 Sorry Scott, I have spent too much time with Beavis and Butthead in my days and so often see ambiguity even when I shouldn't. BTW I truly am confused which you meant. I never did take any statistics while still in school and have regretted it since.

who said anything about single trials?, posted on October 18, 2009 at 09:11:52
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 The original article sure didn't. 10 trials is a bit low. Although 10 out of 10 should meet 95% confidence I think. And that was the result was it not?

I don't know. me?, posted on October 18, 2009 at 12:52:06
Ugly
Audiophile

Posts: 1420
Location: Des Moines, WA
Joined: August 22, 2006
 I just through that out there since I was confused by your response.

OK, posted on October 18, 2009 at 19:21:39
Analog Scott
Audiophile

Posts: 4001
Joined: January 8, 2002
 sorry if my response was confusing.

Valid question, posted on October 17, 2009 at 07:43:00
caspian@peak.org
Audiophile

Posts: 386
Location: Oregon
Joined: January 12, 2008
 Perhaps sampling a larger pool of audio industry professionals (and maybe musicians) would yield a more statistically meaningful result.

Since the test concerns the detection of extremely subtle audible differences, it is unlikely that members of the general public would fare as well as people who have spent years training themselves to discern precisely such differences.

RE: Valid question, posted on October 18, 2009 at 00:36:12
Ugly
Audiophile

Posts: 1420
Location: Des Moines, WA
Joined: August 22, 2006
 I honestly don't feel the need to ask for scientific proof of this particualr type of claims. My own experience has left me with strong enough impressions in similar situations to not be surprised when I hear of others drawing similar conclusions as I have, even as non objectively as I have gotten my conclusions in some cases. For me often times even though I can see the glaring logical fallacies in my thinking plain as day it still doesn't bother me. I'm Free. whoohoo! but that isn't my point...

It's just that when I start seeing terms like proof thrown around I almost can't help but gawking. Maybe I was just being a little ornery but IMO most of what likely continuously attracts Tubeguys1954's regular troupe of hecklers is what comes accross to me as somewhat overly zealous language in his posts even when it isn't necessary IMO due to an already apparently strong rational argument. "Proof" (or the words proper conjegation in the context it was used) is some pretty strong language. The question of whether the test in question truly is is really all that I wondered about. You might call it a hunch since I already admitted my ignorance about these things.

Page processed in 0.102 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2009, Audio Asylum, all rights reserved ] [ Go Green ] [ Take5 ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]