| Author |
Message |
Mark DeBellis
Guest
|
 Posted:
Sat Nov 12, 2005 5:42 am Post subject:
discrimination and perception (da capo, in the Italian sense |
|
|
Hi all,
I've been away for a while, but, looking back, I don't think my basic
question was ever answered. Obviously I didn't do a good job of
explaining it.
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
An example to illustrate the worry: You hear a 10-second passage of
music from either of two sources, ending with a big chord. The thought
is that the first nine seconds has an effect on the way that that final
chord will sound. If there are differences in the first nine seconds
between the two sources, then that may make a difference in how the
final chord sounds. But how do you measure that? If you listen to 10
seconds of source A followed by 10 seconds of source B, you are making
a comparison of events separated in time, and you are likely not to
reliably detect the difference. And if you simply compare the final
chords in immediate succession, you will eliminate the effect of
context, which is exactly what we are, or should be, trying to measure.
What, moreover, assures us that any difference in the first nine
seconds that can make a difference to how the final chord sounds will
itself be a difference that is detectable (using, say, a quick-switch
method)?
If you're going to say I have no evidence that a preceding context can
ever make a difference to how a final chord sounds -- that it's
physically or psychologically possible -- or that I'm speculating, I
agree, but I don't see how that answers anything.
Stewart, I looked back just now at your last response to me, which
reads, "you have *not* 'remarked on [the existing paradigm's]
limitations', you have merely *claimed* that such limitations exist,
without showing a shred of evidence or indeed reasoning to support such
a claim."[1] Really, it takes a lot of temerity to say that, given
that I said exactly what was wrong with your argument, and you didn't
respond, and I referred to it, and you insisted I had to *quote* it[2],
and I did, and again you didn't respond. You hardly come off then as
someone who is interested in finding out what's *true*, but rather in
dismissing and belittling others, and it's hard to see how you could
possibly think that that serves your credibility. Would it be too much
to suggest that a less confrontational tone, all round, might be a good
thing here?
Mark
[1] Subject: Re: A model of the brain, & quick-switch, Date: 4
Oct 2005 02:09:09 GMT, Message-ID: <dhso8501v4h@news2.newsguy.com>.
[2] Subject: Re: Validity of audio tests, Date: 6 Sep 2005 23:33:59
GMT, Message-ID: <dfl9170jfq@news2.newsguy.com>
|
|
| Back to top |
|
 |
Guest
|
| Posted:
Sun Nov 13, 2005 12:16 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
Mark DeBellis wrote:
| Quote: | Hi all,
I've been away for a while, but, looking back, I don't think my basic
question was ever answered.
Obviously I didn't do a good job of explaining it.
|
My irony meter just blew up. (You did a great job of explaining it,
but I think a few people around here didn't want to hear it.)
| Quote: |
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
An example to illustrate the worry: You hear a 10-second passage of
music from either of two sources, ending with a big chord. The thought
is that the first nine seconds has an effect on the way that that final
chord will sound. If there are differences in the first nine seconds
between the two sources, then that may make a difference in how the
final chord sounds. But how do you measure that? If you listen to 10
seconds of source A followed by 10 seconds of source B, you are making
a comparison of events separated in time, and you are likely not to
reliably detect the difference. And if you simply compare the final
chords in immediate succession, you will eliminate the effect of
context, which is exactly what we are, or should be, trying to measure.
What, moreover, assures us that any difference in the first nine
seconds that can make a difference to how the final chord sounds will
itself be a difference that is detectable (using, say, a quick-switch
method)?
If you're going to say I have no evidence that a preceding context can
ever make a difference to how a final chord sounds -- that it's
physically or psychologically possible -- or that I'm speculating, I
agree, but I don't see how that answers anything.
|
Right. Their problem is the lack of interest in distinguishing between
various possibilities. They've never done a test that would indicate
whether context affects perception or not.
On the other hand, it's common knowledge to musicians---you know, the
people who understand perception well enough to make music? Oh, I
forgot, music isn't measureable and therefore isn't real.
| Quote: |
Stewart, I looked back just now at your last response to me, which
reads, "you have *not* 'remarked on [the existing paradigm's]
limitations', you have merely *claimed* that such limitations exist,
without showing a shred of evidence or indeed reasoning to support such
a claim."[1] Really, it takes a lot of temerity to say that, given
that I said exactly what was wrong with your argument, and you didn't
respond, and I referred to it, and you insisted I had to *quote* it[2],
and I did, and again you didn't respond. You hardly come off then as
someone who is interested in finding out what's *true*, but rather in
dismissing and belittling others, and it's hard to see how you could
possibly think that that serves your credibility.
|
Anyone who refuses to examine the basis of their evidence has lost
credibility in my eyes.
Mike |
|
| Back to top |
|
|
BEAR
Guest
|
| Posted:
Sun Nov 13, 2005 5:42 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
michaelmossey@yahoo.com wrote:
| Quote: | Mark DeBellis wrote:
Hi all,
I've been away for a while, but, looking back, I don't think my basic
question was ever answered.
Obviously I didn't do a good job of explaining it.
My irony meter just blew up. (You did a great job of explaining it,
but I think a few people around here didn't want to hear it.)
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
An example to illustrate the worry: You hear a 10-second passage of
music from either of two sources, ending with a big chord. The thought
is that the first nine seconds has an effect on the way that that final
chord will sound. If there are differences in the first nine seconds
between the two sources, then that may make a difference in how the
final chord sounds. But how do you measure that? If you listen to 10
seconds of source A followed by 10 seconds of source B, you are making
a comparison of events separated in time, and you are likely not to
reliably detect the difference. And if you simply compare the final
chords in immediate succession, you will eliminate the effect of
context, which is exactly what we are, or should be, trying to measure.
What, moreover, assures us that any difference in the first nine
seconds that can make a difference to how the final chord sounds will
itself be a difference that is detectable (using, say, a quick-switch
method)?
If you're going to say I have no evidence that a preceding context can
ever make a difference to how a final chord sounds -- that it's
physically or psychologically possible -- or that I'm speculating, I
agree, but I don't see how that answers anything.
Right. Their problem is the lack of interest in distinguishing between
various possibilities. They've never done a test that would indicate
whether context affects perception or not.
On the other hand, it's common knowledge to musicians---you know, the
people who understand perception well enough to make music? Oh, I
forgot, music isn't measureable and therefore isn't real.
Stewart, I looked back just now at your last response to me, which
reads, "you have *not* 'remarked on [the existing paradigm's]
limitations', you have merely *claimed* that such limitations exist,
without showing a shred of evidence or indeed reasoning to support such
a claim."[1] Really, it takes a lot of temerity to say that, given
that I said exactly what was wrong with your argument, and you didn't
respond, and I referred to it, and you insisted I had to *quote* it[2],
and I did, and again you didn't respond. You hardly come off then as
someone who is interested in finding out what's *true*, but rather in
dismissing and belittling others, and it's hard to see how you could
possibly think that that serves your credibility.
Anyone who refuses to examine the basis of their evidence has lost
credibility in my eyes.
Mike
|
I can't speak to the certainty that this is/is not and issue with the
typical "testing paradigms commonly used for discerning differences in
audio gear".
However, one thing is quite certain to me (at least) and that is that
there is a wide range in the "depth" and "resolution & detail" of
perception both over time and in multiple presentations at different times.
A good example is the music stand that gets whacked during a recording
of Bolet & Dutoit doing Prokovieff (sp???) and Shy Covskee (heh) that I
have. Some days you hear it getting wacked, some days not. Most people
never notice it. IF I do the A/B repeat trick with the CD everyone hears
it!
What does that mean?
You decide.
_-_-bear |
|
| Back to top |
|
|
Guest
|
| Posted:
Mon Nov 14, 2005 12:06 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
BEAR wrote:
| Quote: | michaelmossey@yahoo.com wrote:
Mark DeBellis wrote:
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
On the other hand, it's common knowledge to musicians---you know, the
people who understand perception well enough to make music? Oh, I
forgot, music isn't measureable and therefore isn't real.
I can't speak to the certainty that this is/is not and issue with the
typical "testing paradigms commonly used for discerning differences in
audio gear".
However, one thing is quite certain to me (at least) and that is that
there is a wide range in the "depth" and "resolution & detail" of
perception both over time and in multiple presentations at different times.
|
Musicians who listen and analyze and play a piece over and over have
the experience of getting closer to the heart of it. Not only are new
details noticed, but all details are put into a balance of awareness.
One musician who recorded himself described the experience of listening
to the recordings and gradually coming to clarity about the strengths
and weaknesses of them.
It would certainly seem a good idea to check if listeners pick up
important things in contexts other than quick-switching. As far as I
know, this has never been tested in a controlled way, and hence remains
an unknown.
Mike |
|
| Back to top |
|
|
Chung
Guest
|
| Posted:
Mon Nov 14, 2005 12:10 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
Mark DeBellis wrote:
| Quote: | Hi all,
I've been away for a while, but, looking back, I don't think my basic
question was ever answered. Obviously I didn't do a good job of
explaining it.
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
|
Your statement above has a few problems:
1. Discrimination tests are intended to reveal any audible differences
between two presentations. They are not intended to measure "perception
over time", whatever that means.
2. For something to be able to measure "perception over time", you
have to define "percertion over time" in a way that is measureable. You
have failed to do that. You cannot say tool A does not do task B, when
you do not define task B in a way that is doable.
3. Qualitatively speaking, "percerption over time" is a function of
many things: our moods, the aesthetics of the gear, the reliablity of
the gear, and so on. How can a double-blind controlled discrimination
test ever be able to control those non-sonic effects?
| Quote: |
An example to illustrate the worry: You hear a 10-second passage of
music from either of two sources, ending with a big chord. The thought
is that the first nine seconds has an effect on the way that that final
chord will sound. If there are differences in the first nine seconds
between the two sources, then that may make a difference in how the
final chord sounds. But how do you measure that? If you listen to 10
seconds of source A followed by 10 seconds of source B, you are making
a comparison of events separated in time, and you are likely not to
reliably detect the difference. And if you simply compare the final
chords in immediate succession, you will eliminate the effect of
context, which is exactly what we are, or should be, trying to measure.
What, moreover, assures us that any difference in the first nine
seconds that can make a difference to how the final chord sounds will
itself be a difference that is detectable (using, say, a quick-switch
method)?
|
You do not have cause for worry. When you postulated that two pieces of
gear sound different, you must have a set of test conditions under which
those differences were revealed, correct? In your discrimination test
(double-blind I presume), simply use the same test conditions. In other
words, use the same music segments or pieces in your discrimination test
as in your sighted tests where you had detected differences.
If you need to listen to the full 10 seconds to detect any difference,
that's OK. Just run the discrimination test to include the full 10
seconds. When you stated that there were differences, you clearly were
inmplying that your memory was good enough to remember those two
presentations separated in time, by 10 seconds or whatever.
If your position is that your memory simply is not good enough to detect
differences in presentations separated by 10 seconds or whatever, then
how could you possibly state that you detected differences before? |
|
| Back to top |
|
|
Stewart Pinkerton
Guest
|
| Posted:
Mon Nov 14, 2005 12:17 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
On 12 Nov 2005 03:20:19 GMT, "Mark DeBellis" <mad1@columbia.edu>
wrote:
| Quote: | I've been away for a while, but, looking back, I don't think my basic
question was ever answered. Obviously I didn't do a good job of
explaining it.
|
It seemed more that it veered with the wind, whenever one of it's
basic precepts was knocked down.
| Quote: | Stewart, I looked back just now at your last response to me, which
reads, "you have *not* 'remarked on [the existing paradigm's]
limitations', you have merely *claimed* that such limitations exist,
without showing a shred of evidence or indeed reasoning to support such
a claim."[1] Really, it takes a lot of temerity to say that, given
that I said exactly what was wrong with your argument, and you didn't
respond, and I referred to it, and you insisted I had to *quote* it[2],
and I did, and again you didn't respond. You hardly come off then as
someone who is interested in finding out what's *true*, but rather in
dismissing and belittling others, and it's hard to see how you could
possibly think that that serves your credibility. Would it be too much
to suggest that a less confrontational tone, all round, might be a good
thing here?
|
All the armchair philosophising and sophistry in the world is not
worth one solid experiment. You make claims that there is something
wrong with level-matched quick-switch DBTs for determining subtle
differences between audio equipment. Fine - go and *do* whatever
experiment you like, and *prove* that whatever method you support
actually works better.
Until you've done that, and can provide some *evidence* to support
your assertions, you are merely wearing out your keyboard.
BTW, my confrontational tone was created by the myriads of armchair
philosophisers who magically vanish on the breeze when challenged to
stand behind their reams of purple prose. My 'credibility' isn't of
any importance in this respect, what matters is the *facts*, and they
remain unchanged after all these years.
--
Stewart Pinkerton | Music is Art - Audio is Engineering |
|
| Back to top |
|
|
Mark DeBellis
Guest
|
| Posted:
Mon Nov 14, 2005 12:19 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
And about science and pseudoscience ("you wave your arms and make
pseudoscientific arguments...").[1]
Bob, This is *exactly* how a scientist should think. A scientist asks,
"What do the data mean? How should they be interpreted?" If an
experimental method is unsuited, in principle, to reveal certain
differences or to give certain information, that is a limitation.
What you advocate is not scientific thinking but rather a complacent,
consumerist attitude toward scientific results. No one ever became a
scientist thinking like you advocate. What has become manifestly
apparent is that you are less concerned with substantive discussion of
the issues than with browbeating others who you perceive to be less
well informed than you; you're just a bully.
Mark
[1] bob, Subject: Re: A model of the brain, & quick-switch, Date: 30
Sep 2005 02:31:51 GMT, Message-ID: <dhi82n01qss@news3.newsguy.com>) |
|
| Back to top |
|
|
Mark DeBellis
Guest
|
| Posted:
Mon Nov 14, 2005 12:22 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
michaelmossey@yahoo.com wrote:
| Quote: | Mark DeBellis wrote:
Hi all,
I've been away for a while, but, looking back, I don't think my basic
question was ever answered.
Obviously I didn't do a good job of explaining it.
My irony meter just blew up. (You did a great job of explaining it,
but I think a few people around here didn't want to hear it.)
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
An example to illustrate the worry: You hear a 10-second passage of
music from either of two sources, ending with a big chord. The thought
is that the first nine seconds has an effect on the way that that final
chord will sound. If there are differences in the first nine seconds
between the two sources, then that may make a difference in how the
final chord sounds. But how do you measure that? If you listen to 10
seconds of source A followed by 10 seconds of source B, you are making
a comparison of events separated in time, and you are likely not to
reliably detect the difference. And if you simply compare the final
chords in immediate succession, you will eliminate the effect of
context, which is exactly what we are, or should be, trying to measure.
What, moreover, assures us that any difference in the first nine
seconds that can make a difference to how the final chord sounds will
itself be a difference that is detectable (using, say, a quick-switch
method)?
If you're going to say I have no evidence that a preceding context can
ever make a difference to how a final chord sounds -- that it's
physically or psychologically possible -- or that I'm speculating, I
agree, but I don't see how that answers anything.
Right. Their problem is the lack of interest in distinguishing between
various possibilities. They've never done a test that would indicate
whether context affects perception or not.
On the other hand, it's common knowledge to musicians---you know, the
people who understand perception well enough to make music? Oh, I
forgot, music isn't measureable and therefore isn't real.
Stewart, I looked back just now at your last response to me, which
reads, "you have *not* 'remarked on [the existing paradigm's]
limitations', you have merely *claimed* that such limitations exist,
without showing a shred of evidence or indeed reasoning to support such
a claim."[1] Really, it takes a lot of temerity to say that, given
that I said exactly what was wrong with your argument, and you didn't
respond, and I referred to it, and you insisted I had to *quote* it[2],
and I did, and again you didn't respond. You hardly come off then as
someone who is interested in finding out what's *true*, but rather in
dismissing and belittling others, and it's hard to see how you could
possibly think that that serves your credibility.
Anyone who refuses to examine the basis of their evidence has lost
credibility in my eyes.
Mike
|
Exactly. Whether the context of a musical event affects the way it
sounds (or how it does) is a factual question which, along with
everything else, I submit for discussion. If someone thinks I have to
prove the matter one way or the other before they consider the
discussion worth participating in, then of course they are free not to
participate, but they shouldn't think that that attitude is anything
but head-in-the-sand, one of intellectual compacency and laziness.
Obviously, to argue on the basis of discrimination tests that
discrimination tests are not limited in the way I have suggested, as
some have done here, is circular.
Coincidences are amazing, by the way. I had occasion to read "On
Liberty" by J. S. Mill this weekend. He writes, "There is the greatest
difference between presuming an opinion to be true because, with every
opportunity for contesting it, it has not been refuted, and assuming
its truth for the purpose of not permitting its refutation." His
argument explains *exactly* why discussion groups, in our day, are
valuable.
Mark |
|
| Back to top |
|
|
bob
Guest
|
| Posted:
Mon Nov 14, 2005 2:54 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
Mark DeBellis wrote:
| Quote: | And about science and pseudoscience ("you wave your arms and make
pseudoscientific arguments...").[1]
Bob, This is *exactly* how a scientist should think. A scientist asks,
"What do the data mean? How should they be interpreted?"
|
Not to beat a dead horse, but you *have* no data.
| Quote: | If an
experimental method is unsuited, in principle, to reveal certain
differences or to give certain information, that is a limitation.
|
"In principle"? That's a concept alien to empiricism.
| Quote: | What you advocate is not scientific thinking but rather a complacent,
consumerist attitude toward scientific results. No one ever became a
scientist thinking like you advocate. What has become manifestly
apparent is that you are less concerned with substantive discussion of
the issues than with browbeating others who you perceive to be less
well informed than you; you're just a bully.
|
Rest assured, you will never have to worry about my bullying you--or
replying to you--again. You've just been added to my list of people not
worth engaging with.
bob |
|
| Back to top |
|
|
BEAR
Guest
|
| Posted:
Mon Nov 14, 2005 5:42 am Post subject:
Re: discrimination and perception (da capo, in the Italian s |
|
|
michaelmossey@yahoo.com wrote:
| Quote: | BEAR wrote:
michaelmossey@yahoo.com wrote:
Mark DeBellis wrote:
What seems to me to be the case is that discrimination tests are not a
good way to measure *perception over time*; or, to put it differently,
I am asking why we should think that they are.
On the other hand, it's common knowledge to musicians---you know, the
people who understand perception well enough to make music? Oh, I
forgot, music isn't measureable and therefore isn't real.
I can't speak to the certainty that this is/is not and issue with the
typical "testing paradigms commonly used for discerning differences in
audio gear".
However, one thing is quite certain to me (at least) and that is that
there is a wide range in the "depth" and "resolution & detail" of
perception both over time and in multiple presentations at different times.
Musicians who listen and analyze and play a piece over and over have
the experience of getting closer to the heart of it. Not only are new
details noticed, but all details are put into a balance of awareness.
One musician who recorded himself described the experience of listening
to the recordings and gradually coming to clarity about the strengths
and weaknesses of them.
It would certainly seem a good idea to check if listeners pick up
important things in contexts other than quick-switching. As far as I
know, this has never been tested in a controlled way, and hence remains
an unknown.
Mike
|
I am unaware of this being tested per se.
But I am confident that the ability to percieve *anything* is a variable
within one person, ergo within a group of people.
On the other hand, the *mean* for a group is what appears to be taken as
being the standard by which to make determinations... ergo, using my
example, the music stand being hit is statistically inaudible -
therefore it is *not in the recording*! :- )
Just a thought...
_-_-bear |
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in