|
|
|
|
Why DBTs in audio do not deliver (was: Finally ... The Furutech CD-do-something)
(KikeG) wrote in message ...
Since everything I have to say has been said by me and others at least
a dozen times already. Since, I myself, (let alone the readers), am
bored to tears with the whole thing I'll try a dodge to hold my and
their interest for the next 15 minutes. We'll ask you to paeticipate
in an easy multiple choice test.
Two such signs mean "I said it".. One stands for "Kike said it."
I have no knowledge, opinion, interest in the particulars of such
research because it does not research differences between real-life
audio components.
Why not? I'd say that some do. Probably not available to the general
public, but some of them do.
Explain: A) Does "general public" to whom it is "not available"
include a) you b) me and c) our readers?
1) If a) is not a fact, please give me and your public the benefit of
reference; author, date, page
2) If all 3 (a,b and c) are true is the alternative of staying
silent when one has nothing to say;
2a) prejudicial to your cause?? 2b) too cruel to bear?
What I do object to is the belief that to "prove" your opinions re
"difference" or "no difference"- a necessary preliminary to
preference- (more about difference/preference in my answer to
Audioguy.) one has to undergo a procedure known as ABX.
Please, don't one of you tell me that "nobody says that".
I believe many people say that, and I agree. Or at least the need of
some DBT controlled procedure that doesn't have to be just ABX, it
can be an alternative procedure such as ABC/HR or others.
A simple answer: In a WELL DESIGNED AND PERFORMED TEST, ( my
capitals L.M.) there will be
NO DIFFERENCE due to the test. The thing is to design a test that
duplicates the listening habits of the listener, if that is neccesary,
but is made double blind. Maybe the listener will feel uncomfortable
with the idea of being tested, or with the idea that he may do "bad".
That's a problem of the listener (unless he is not aware he is being
"tested"), because those can't be addressed with this or with any
other possible test or evaluation procedure.
A) Is a "well-designed" and "performed" test
1) available on this earth?
2) available only in paradise?
3) paired with the Holy Grail?
3) available only in the objectivist dry (this is a family forum!)
dreams?
If number one is the case , please quote joutrnal, author, date, page.
If you are only now setting out to design it, do tell;
Will the "well-designed" test follow the model of the medical drug
research DBT and list the number and the selection of test subjects
(aka "population") for statistical significance, relevance to the
test's objectives etc.? Or will it be
A) just any old number of the average Joes from the street ?
B) "trained" subjects only?
Ba) by whom? Bb) for how long? Bc) who grants the "pass mark"? Bd)
what's the tolerable % of untrainables before the "test" is no longer
a useful tool for the consumers at large?
C) selected populations only? ; 1) orchestra musicians 1a) before they
have been there for long enough to lose their hearing?
2) Rock musicians? 2b)- ditto?. 2c) only the rock musicians who have
ever heard a piano and a violin? or 2d) only those who have heard
nothing but electronically amplified instruments ?
3) Autosound lovers?
3a) Only the "My "bass" is louder than yours" category or 3b) those
who play chamber music in their car (if you can find them)?
4) Middle-aged chamber music lovers? 4a) with? or 4b) without a
hearing test?
4c) male? 4d) female? 4e) a mixture? What proportions?
(On the average middle aged women keep their hearing better than men)
Would you like to continue or should we take a rest... ?
An ABX/DBT test introduces no biases, just removes them. The only
other biases that can appear will be due to the listener but not due
to the procedure.
Exactly. We listed just a tiny portion of the biases introduced by the
lsteners . Do you want to continue?
For Nth time, I think nobody has said ABX results have to be the way you say.
But if nodoby in earth has been ABX to nominally competent cables,
that means a lot.
I assume you meant to say: But if nobody in the whole world has been
able to distinguish nominally competent cables when ABXing that means
a lot."
It would. But we don't have to ABX *everybody* on earth. The one and
only detailed, proctored and statistically valid available to me
cable ABX test is Greenhill's in The Stereophile, Aug. 83.( Know any
others?)
This is what Greenhill had to say: " Final significant conclusion one
can draw is that at least one "golden ear exists".
Why? Because this participant reached and/or surpassed Greenhill's
statistical criterion for a "hit"( minimum 12 out of 15 "corrects") in
5 out of six different cable comparisons. (SCORE 83%)
HE ALSO DISTINGUISHED CORRECTLY MONSTER FROM ZIPCORD OF SIMILAR GAUGE
WHEN PINK NOISE WAS PLAYED.
His one and only failure was when Monster was compared with zipcord
and MUSIC was played. Sort of interesting -see below-isn't it?
One other participant scored 66,7%.
Three (3) scored below random chance-33% hits.
Immaculately objectivist Geenhill reviewed the majority performance
and concluded : "We can only conclude therefore, that there is little
advantage besides pride of ownership in using these thick expensive
wires.
Who do you think was "right"?
A) The "golden ear'?
B0 The majority?
C) were they all "right" within their abilities and their
limitations.?
D) Was Greenhill right when he allowed his "golden ear"? or
E) Was he right when he said it didn't matter anyway?.
Not good taste and discrimination but majority rules. The decisions
which is the better instrument are decided by a public opinion poll.
(Lloyd "integrated" wins and ABX proves it by showing "no difference"
majority vote)
You ask:
Where is the evidence that difference in performance with music as
opposed to pink noise has anything to do with using an ABX-DBT
procedure? Nowhere.
Convincing, controlled
experiment with random control subjects etc. is missing.
That's an oxymoron, since there's no way to prove that without a DBT.
Oxymoron sounds nice. Let's see. You put on wide band choral music,
and sit down. A friend operates your stepped volume control in steps
of one db. up or down out of your line of sight.
No problem hearing the up and down 1db changes unless you're somewhat
deaf.and NONE hearing the 1.75 db changes as in Greenhill test that
73% failed when ABXed.
You now switch to ABX and can't tell the difference just like those
unfortunate ABX rabbits.
A) Should you have a hearing test? or
B) should you blame your friend or your equipment? Or
C) music for "making the test difficult"
D) or ask the ABXer nearest to you for HIS evidence that music doesn't
bother ABX and/or ABX doesn't bother music
If he tells you that it is up to YOU to provide such evidence,
because he doesn't like oxymorons, will you tell him to:
A) .......!!! or
B)!!!!.......??!!! or
C) or you will set about pronto to set up a "well-designed test" ,
finding participants, hiring the venue and the equipment and so on.
After all you can't keep the scientists waiting all these years ,
holding their breath.
If you agree that they would differ how do you justify YOUR
challenges to all and sundry to prove their perceptions by ABX.
ABX/DBT removes false positives, just that. It doesn't imply that
everybody should perform equally using ABX. Stop repeating that, please.
Sorry, being dense. Don't get too irritatated by me. If "everybody"
does not "perform equally using ABX" should your challenges be to
"everybody" or to the selected good performers only?
I'm not proopunding any "test". You are. I have
no faith to stick to. You BELIEVE in ABX. It is MY right to ask YOU
for evidence. And it is your job to give it.
It is proved that non-DBT tests are unreliable when it comes to
detecting differences and deciding preferences just from the actual
sound, and can't be trusted. It happens every day. So DBT is the only
way to go. Period. Show me any evidence against that.
I'm not in the "evidence" game because I don't have any "test" worth
bothering with. In secret: just to you: I don't believe you have
either. When I point out areas of doubt and flaws you say they can
only be dispelled by a DBT. And you call this nonsense " others'
oxymoron",
Quote references to "well-designed" component comparison tests
performed according to your "good test" criteria, and published in
the form readable on this earth. A test has to be shown to WORK for
its purpose. Not because Enrique sees no reason why it shouldn't, not
because it is like other tests which work in other fields with other
controls but HERE, NOW, comparing compoents.
In the meantime ponder what the piano virtuosi, violinists,
cello,viol, trumpet, clarinet players would say to you if you told
them that they must " prove" their personal instrument choices by a
DBT. I shall not list the choices this time. Why should just us , poor
audiophiles (that includes easily suggestible newbies) be so plagued?
Ludovic Mirabel
|