Propeller Head Plaza Technical and scientific discussion of amps, cables and other topics.

---

[ Follow Ups ] Thread: [ Display  All  Email ] [ Propeller Head Plaza ]

---

[ Alert Moderator ]

---

Probability is the guide to life.

Posted by Pat D (A) on December 22, 2004 at 08:14:59

In Reply to: It's the essence of the DBT that you don't understand posted by Commuteman on December 21, 2004 at 21:29:16:

You have tried to attribute to me the position that cables make no difference. Indeed, I could point out DBTs which show that some of them do--about 30 foot lengths of 16 and 24 gauge cables, I think--but this apparently makes no difference to you.

I should also comment that whether I or anyone else believes that, this does nothing to establish when and under what conditions differences are audible. So, your argument is irrelevant.

Whether you actually spend more than a few dollars for interconnects or more than $.30 a foot or so for speaker cables is also irrelevant. It's your choice.

"The mechanical activity of using an ABX test on interconnects is indeed easy; designing a real DBT that actually mimics the normal use of the interconnect, and introduces NO other variables, is not."

You have not shown that any relevant variables are changed. A person using an ABX machine can do so at home using their own system, at their own leisure. They can switch between the known A and B as often as they want (the user controls the switching), they can compare them to the unknown X as often as they want, and they can decide when to choose whether X is A or B. This is exactly what happens with sighted auditioning except the switching is probably much faster.

According to jj, sometimes the statistics will show that subjects can differentiate a difference close to the JNDs when they thought they were guessing. But they actually were not guessing. So, people can sometimes detect differences below conscious awareness in a DBT. That's how sensitive a DBT is. I mention this because I suspect you think detecting a difference is necessarily conscious.

"In every other use of blind testing (at least those of which I am aware), the thing being tested is used in its intended, normal application. The user is unaware whether they are using the test item or not, thus eliminating the possiblity of bias. A placebo is used to provide a reference (and provide two sample sets for comparison).

Oh. So you want the subject to differentiate between A and a placebo! Well, let's call the placebo B. GMAB In any case, the use of the ABX switcher can be as close to normal use as desired. And as for not being aware which item is being used, that's exactly what the ABX box does. You are not making sense here.

"In the case of ABX audio testing, the test DOES NOT mimic real world usage, and the outcomes are not defined the same way. The ABX test requires that the test subjects identify differences. "

If the person can determine better or worse, this implies they can differentiate. Again, you are not making sense.

"I contend that there is no proof that this activity is the same, or even uses the same brain functions, as listening to music."

Ay there's the rub! As I have said, if you think there is a relevant difference, then it is up to you to show it. One cannot prove two things are the same--just as one cannot prove two wires, two amps, two CDPs are audibly the same. For some reason, you forget this little bit of logic when you try to make your case.

"We can't measure the emotional response to audio equipment directly, so we have tried to use listening tests as a substitute. IMO, the burden on proof is on the advocates of such tests to PROVE that they adequately mimic the real thing. To the best of my knowledge, that has never been demonstrated."

No, the burden of proof is on those who maintain there is a difference. I don't know why people find that so difficult to grasp. You reference to emotional response is a red herring. If someone has a different emotional response to two things, that means they can differentiate them. They would be able to differentiate them in sighted listening, but that introduces those extra variables you worry about above.

I mean you say someone might react differently to the sound of two components but then you turn around and deny this is a difference! This is not coherent, man!

DBT is not itself a measurement.

"So, IMO, the ABX tests proposed so far fail on 2 counts: they measure the wrong thing (detectability vs. quality), and they measure it under the wrong conditions (listening for differences vs. listening to music)."

"Can anyone prove me wrong?"

I just have. DBT is not a measurement. As for detectability vs. quality, differentiating the quality implies detecting a difference. If you don't think one can do DBTs for listener preferences, I suggest you look a the White Papers on the Harman International Site.

As for listening for differences as opposed to listening to music, you suggest a difference but offer no evidence; the burden of proof is on those who assert a difference. Does it make any difference to the results of a DBT? Part of this I suspect is that you do not consider the way you react as showing a difference, which is not coherent.

"As to my second point: I can only reiterate that, since the current state of play is an absence of absolute proof either way, your requirement that any advice be offered from a position of proof REQUIRES that no advice be given."

There are not too many absolutes in everyday life. Statistics doesn't give them, just probabilities. Here's a quote from the ABX site, linked below. The second part is often quoted, but strangely, the first is not.

"When an individual ABX Comparision is completed, the responses are checked against the key in the ABX Comparator in ANSWER mode. The number of correct responses is compared to the number of response attempts to give a score such as the example of 20 correct out of 30 attempts, which we have written briefly as a fraction or a percentage: 20 / 30 = 67%. The score is then compared with the probability table from which the probability that it is a random score is determined. Thus the result is stated: 20 / 30 = 67% (p = 0.049). This literally means that the score (20/30) is probably not random except for a 0.049 chance that it is random. Thus the result of the experiment is that the null hypothesis is not true except for a 0.049 chance it is true, or in audiophile terms, A sound different from B except for a 4.9% chance that they are the same.

Note that no matter what score is achieved, A = B cannot be proven. That is the ABX Double Blind Comparison can never be used to prove two audio components sound the same. The notion that ABX can prove components sound the same is a common misconception about ABX."

I believe Jeremy Benthan said probability is the guide to life. Indeed, we often give advice on probabilities. To require absolutes is an unreasonable standard. To allocate resources based on probabilities and risk assessment is not.

____________________________________________________________
"Nature loves to hide."
---Heraclitus of Ephesus (trans. Wheelwright)

• http://www.pcavtech.com/abx/abx_p9.htm   (Open in New Window)

This post is made possible by the generous support of people like you and our sponsors:

Dynaco Doctor

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2013, Audio Asylum, all rights reserved ] [ Take5 ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---