Again: why do amps sound different?
Phil Birnbaum
Given that I started this thread with my question, I should probably
add my own two cents worth. I've definitely found that low- and
mid-price amps can sound different ... I bought a NAD 304 and hated
it, and I've compared a mid-price Onkyo A-B with a cheap JVC and
there were definitely differences.
I am therefore leaning towards the "of course there are differences"
school. Although I find Gene's "double-blind tests show they're
pretty much the same" hypothesis appealing, my own experience
doesn't support it at my price point.
Anyway, my original question remains unanswered, I think. If we
assume that amps *do* sound different, WHY? Is it that
(a) all manufacturers are trying to do a perfect reproduction, but
fail in different ways, or
(b) manufacturers deliberately engineer for a certain flavor of
sound?
I'll ask this question again about CD players in another post, in
case some of the contributors are really sick of this thread. :)
Phil
Sebastien P. McIntyre
In article <5a69tf$4...@news.istar.ca>
birn...@magi.com (Phil Birnbaum) writes:
>
>Given that I started this thread with my question, I should probably
>add my own two cents worth. I've definitely found that low- and
>mid-price amps can sound different ... I bought a NAD 304 and hated
>it, and I've compared a mid-price Onkyo A-B with a cheap JVC and
>there were definitely differences.
>
>I am therefore leaning towards the "of course there are differences"
>school. Although I find Gene's "double-blind tests show they're
>pretty much the same" hypothesis appealing, my own experience
>doesn't support it at my price point.
>
>Anyway, my original question remains unanswered, I think. If we
>assume that amps *do* sound different, WHY? Is it that
>
>(a) all manufacturers are trying to do a perfect reproduction, but
>fail in different ways, or
>
>(b) manufacturers deliberately engineer for a certain flavor of
>sound?
"definitely differences"???
"Gene's hypothesis"???
"price point"???
"if amps *do* sound different, WHY?"???
IMO, you're ripe for a little reading...
"Do All Amplifiers Sound the Same", Ian G. Masters and D. L. Clark,
Stereo Review, January 1987, pp.78-84.
"The Amp-Speaker Interface (Tube vs. SS)", E. Brad Meyer, Stereo
Review, June 1991, pp.53-56.
"Six Power Amplifiers: How Did They Sound?", Report prepared by
Ian G. Masters, AudioScene Canada, May 1977, pp.44-50.
"Amplifiers & Speaker Cables", Follow-up report to preceding
prepared by Ian G. Masters, AudioScene Canada, June 1981,
pp. 24-27.
The key is in _LEVEL MATCHING_ (at 1 kHz is probably OK but to
make sure I prefer weighted equalization over the spectrum... It
gives cleaner results :) ) and _BLIND CALLING_. To give you an idea
of the results, here's a summary of the 1987 Stereo Review controlled
test:
#From: "Sebastien P. McIntyre" <NA...@music.mus.polymtl.ca>
#Newsgroups: rec.audio.opinion
#Subject: Re: Stereo Review BLT's of amps (LONG)
#Date: 17 APR 96 15:09:28 EST
#Message-ID: <17APR96.16...@music.mus.polymtl.ca>
#
# Two kinds of listeners: Believers (believe that amplifiers sound
# significantly different) and Skeptics (are skeptical of that claim).
#
# 15 Believers and 10 Skeptics.
#
# Mr. Clark was hired as a consultant by Stereo Review to design and
# carry out the controlled listening tests.
#
# Five power amps and one receiver were compared (quoted prices are
# those of January 1987):
#
# - Julius Futterman OTL-1 (80-Watt mono tube amp, $12,000/pair)
# - Mark Levinson ML-11 (50W/ch, #2,000)
# - Counterpoint SA-12 (tube-SS hybrid, $995)
# - NAD 2200 (100W/ch, 6-dB dynamic power, high current/voltage, $548
# - Hafler DH-120 (62W/ch, $320)
# - Pioneer SX-1500 (45W/ch receiver, $220)
#
# The _opinions_ were that differences existed and could be heard.
# However, statistical interpretation of the _results_ of these
# controlled listening tests led to the conclusion that correct
# answers to "Which source, A or B, is identical to X?" were made
# by chance -- amplifier sound differences could not be heard to
# statistical significance, let alone psychoacoustical significance.
#
# | | | % Probability
# | No. of calls | | result
# Amplifiers | (Correct/Total) | % Correct | due to chance
#________________________|_________________|___________|_______________
# Counterpoint vs. NAD | 30/48 | 63 | 5.6
# Futterman vs. Hafler | 28/48 | 58 | 16.
# Futterman vs. Pioneer | 114/212 | 54 | 15.
# Futterman vs. Levinson | 106/204 | 52 | 31.
# Hafler vs. NAD | 63/128 | 49 | -
# Levinson vs. Hafler | 40/84 | 48 | -
# Futterman vs. NAD | 14/32 | 44 | -
# Pioneer vs. Levinson | 4/16 | 25 | -
#
# NOTE: The Counterpoint-vs.-NAD test results were the closest to
# statistical significance with a just a bit over the 5%
# probability that results were due to chance. But the
# Counterpoint power amp blew up during the test, so it can't
# be ascertained if its last breathe was abnormal and could
# have influenced the results.
#
#
# The results suggest that for these amps, listeners and setup,
# differences between amp sounds were very small or nonexistent.
# Because sound differences were small, test results based on
# small numbers of trials are less reliable than those with large
# numbers of trials (for the same comparison). Pure guessing gives
# a 50% score (similarity to coin flipping).
--
,
Sebastien
Phil Birnbaum
In article <29DEC96.23...@music.mus.polymtl.ca>,
NA...@music.mus.polymtl.ca says...
>
> [impressive summary of blind tests where listeners couldn't tell
> the difference between a $12,000 amp and a $220 amp]
>
Well, Sebastien, I must say I'm impressed ... and again, this is
what I thought would be the case before I started amp-shopping. My
own experience has been, so far, that I can't tell much difference
between mid-price amps except that the NAD 304/314 pair doesn't
sound right. However, I also thought the cheap amps I listened to
sounded poor ... but, hey, maybe they weren't level matched ...
should I check out that $169 JVC special at Future Shop?
One question: did the listeners here listen to music? I assume they
did, and that "level matching at 1kHz" refers to the level at 1kHz,
and not that the listeners listened to a pure 1kHz tone.
If it was real music the listeners heard, I must admit that I'm
surprised by these results. Given how the thread is going, perhaps
I should sit back for now and watch the shit continue to hit the
fan. :)
Phil
Gary Barger
years old. I've also noticed this characteristic in other bibliographies
you have listed recently. Do you believe the findings of such dated
research remain valid? Can you cite any more recent studies supporting your
position? Thanks.
Sebastien P. McIntyre
In article <32C80A...@res.mpls.frb.fed.us>
Gary Barger <g...@res.mpls.frb.fed.us> writes:
>
> With a single exception, the tests and articles you cite are more than 10
>years old. I've also noticed this characteristic in other bibliographies
>you have listed recently. Do you believe the findings of such dated
>research remain valid? Can you cite any more recent studies supporting your
>position? Thanks.
I think you can answer this by yourself. Are today's amp topologies
as good as (or better than) they were 10 years ago? Has human hearing
bettered over the last 10 years?
(In your favour, I'm an evolutionist. :) )
--
,
Sebastien
Stewart Pinkerton
birn...@magi.com (Phil Birnbaum) writes:
>> [impressive summary of blind tests where listeners couldn't tell
>> the difference between a $12,000 amp and a $220 amp]
>>
>Well, Sebastien, I must say I'm impressed ... and again, this is
>what I thought would be the case before I started amp-shopping. My
>own experience has been, so far, that I can't tell much difference
>between mid-price amps except that the NAD 304/314 pair doesn't
>sound right. However, I also thought the cheap amps I listened to
>sounded poor ... but, hey, maybe they weren't level matched ...
>should I check out that $169 JVC special at Future Shop?
There are a couple or three points worth mentioning to newcomers.
Firstly, there IS a minimum level below which cost is such a factor that
most product is pretty poor, often referred to as the 'knee' of the
cost/performance curve. This hits amps and CD players around $4-500
'street' price - below this price, look out!
Secondly, above the knee of the curve there are a few exceptional
'mid-price' products such as the Sony XA3ES CD player and Audiolab 8000S
amp which stand comparison with equipment at ten times the price
(seriously!), but essentially you are in an area where you can pay huge
ampounts of money for little or no audible improvement - indeed some of
the 'high-end' gear is WORSE than mid-fi, believe it or not!
Particularly in the area of amplifiers, there have been several
well-publicised examples of competent and highly experienced listeners
failing to differentiate between good mid-fi amps and audiophile units
with 5-figure price tags.
Finally, much 'audiophile' equipment is deliberately tailored away from
neutral in order to sound different from 'the pack'. For example, NAD
amps have a smooth and soft sound which is attractive against the edgy
sound of many of their older competitors, but after a while you realise
that they are in fact a little muffled and lacking in detail (hastily
dons Nomex underwear!).
Happy listening!
--
Stewart Pinkerton | If you can't measure what you're making,
A S P Consulting | how do you know when you've got it made?
(44) 1509 880112 |
"I canna change the laws o' physics" - the other Scotty
Alan Derrida
Phil Birnbaum (birn...@magi.com) wrote:
: Given that I started this thread with my question, I should probably
: add my own two cents worth. I've definitely found that low- and
: mid-price amps can sound different ... I bought a NAD 304 and hated
: it, and I've compared a mid-price Onkyo A-B with a cheap JVC and
: there were definitely differences.
: I am therefore leaning towards the "of course there are differences"
: school. Although I find Gene's "double-blind tests show they're
: pretty much the same" hypothesis appealing, my own experience
: doesn't support it at my price point.
If it doesn't support it at your price point, imagine how much those
differences come out when you're talking about equipment of high
resolution. And I'm curious to know why, if your own experiences don't
support it and you've discovered what most of us already have about audio
components, do you find the Steinbergian hypothesis appealing?
: (a) all manufacturers are trying to do a perfect reproduction, but
: fail in different ways, or
I don't think intelligent manufacturer's 'try' to reach perfection in
reproduction, they leave that to the mass market Japanese amateurs who
largely feel that perfection has been reached. They just have different
approaches to a similar ideal. See my response to your post on CD players
if you want to read further into it.
: (b) manufacturers deliberately engineer for a certain flavor of
: sound?
Some do, and that's not always a bad thing. Bob Carver was said to have
attempted this with an amplifier design a while back, in an attempt to
duplicate high end designs.
Phil Birnbaum
>And I'm curious to know why, if your own experiences don't
>support it and you've discovered what most of us already have about
>audio components, do you find the Steinbergian hypothesis
>appealing?
>
>And I'm curious to know why, if your own experiences don't
>>support it and you've discovered what most of us already have
about
>audio components, do you find the Steinbergian hypothesis
appealing?
Well, look at it this way: if the task is to reproduce an input
signal exactly, the difference between amps is (a) how close they
come, and (b) in which ways they come close. Now, if one amp comes
99.9% close, and another comes 99.8% close -- although, of course,
they might come close in quite different ways -- there really
shouldn't be much difference between them.
Now, consider speakers: I've seen frequency response graphs for
speakers (in Consumer Reports, for instance), and they're way off
flat compared to amps. If speakers can have such huge differences
and still sound reasonably good, shouldn't the smaller differences
between amps be discountable?
Anyway, that's why the Steinberg hypothesis appeals to me. However,
as you point out, I *have* been able to notice differences (which
Steinberg would explain by the Placebo Effect). Therefore, my
current hypothesis is that amps would sound very close if, in fact,
engineered to reproduce the sound exactly. However, perhaps
manufacturers deliberately tamper with the music to make them sound
different, and this tampering cuts out during frequency response
testing (perhaps because they test pure tones or something, and the
amp is "smart" enough to reproduce those exactly.
Anyway, my plans are to trust my ears; however, this is still a very
interesting question. Why? Because if manufacturers are trying to
produce a certain type of sound, and that's the sound that appeals
to me, it might make shopping easier in the future. Plus, it's hard
to A-B receivers, and knowing what's happening increases the chances
that I'll wind up with something I like.
Phil
Tom Morley
> >And I'm curious to know why, if your own experiences don't
> >support it and you've discovered what most of us already have about
> >audio components, do you find the Steinbergian hypothesis
> >appealing?
> >
> >And I'm curious to know why, if your own experiences don't
> >>support it and you've discovered what most of us already have
> about
Some amps sound distinctly different. An example is the
Mesa Baron. With a damping factor on the order of 1
(or less) this amp will sound distinctly different (with
many speakers) than a high dampind factor SS amp.
Play David Helfgott's Rachmaninov #3 (see the movie!),
Tom Morley | I shall now therefore humbly
mor...@math.gatech.edu | propose my own thoughts, which
tmo...@bmtc.mindspring.com | I hope will not be liable to
| the least objection -- J. Swift
KERADWC
In article <5aeav3$9...@news.istar.ca>, Phil Birnbaum <birn...@magi.com> wrote:
>>And I'm curious to know why, if your own experiences don't
>>support it and you've discovered what most of us already have about
>>audio components, do you find the Steinbergian hypothesis
>>appealing?
>Anyway, that's why the Steinberg hypothesis appeals to me. However,
>as you point out, I *have* been able to notice differences (which
>Steinberg would explain by the Placebo Effect).
While the placebo effect will explain 'a difference', it doesn't explain
why multiple listeners report *the same* difference.
I've been auditioning amplifiers for [far too long] a while, and had an
opportunity to listen to two different, quite good, amplifiers plus my
original system (a not-so-good amp). All were solid-state, with "excellent
specifications". Two other audiophiles also were present. We listened to
the same pieces of music through each amplifier, changing nothing but the
amps. Volumes were *not* exactly matched, though after one listening
round (all units), we replayed most of the pieces at different volumes
on all units. (i.e., we played track 1 at low, medium, and high volumes,
though low, medium, and high weren't level-matched.)
Nobody said what they thought about any of the amps until we were finished;
we wrote down out opinions.
The results? All three of us agreed that amp M had more dynamics, "punch",
and was more "forward" (those quoted terms being in all three notes) than
either of the other. It was also noted that amp W had a wider soundstage;
I thought it was a slight difference, while the others thought it was
a big difference. Image specificity and instrument placement was also
noted, with all of us showing M, W, then K as the order in which the
placement was most clear.
Other observations didn't necessarily match 1:1, as each of us consider
different things to be The Most Important Compromise for equipment, but
there were no *disagreements* about which amp did which things better/
worse/differently.
This might not be a 100% reproducable test (no level matching, etc.),
but it certainly matches my own experiences: the amps *don't* sound
alike, and the differences don't readily submit to being disregarded
as just placebo effect.
Other people's mileage may vary.
--kdc
Sebastien P. McIntyre
In article <5agnp1$c...@herald.concentric.net>
Ker...@cris.com (KERADWC) writes:
>In article <5aeav3$9...@news.istar.ca>, Phil Birnbaum <birn...@magi.com> wrote:
>
>
>
>>Anyway, that's why the Steinberg hypothesis appeals to me. However,
>>as you point out, I *have* been able to notice differences (which
>>Steinberg would explain by the Placebo Effect).
>
>
>While the placebo effect will explain 'a difference', it doesn't explain
>why multiple listeners report *the same* difference.
In the 1987 Stereo Review article, the 15 believers were quite
consistent identifying amp sound differences during the sighted part
of the listening test and also had pretty uniform descriptions for
them (read their comments in a sidebar of the article). In the
controlled (or blind) part, they were also very consistent in their
identification of the differences. The problem is that the results
from both parts of the test do not agree at all (sighted: plenty of
audible differences; blind: no audible differences -- statistical
sense).
How do you explain this?
--
,
Sebastien Best wishes for the New Year to "soldiers" of
both camps.
(Only in fake wars do soldiers observe what they
call the traditional New Year's cease-fire. >:-> )
Swanlee
On 3 Jan 1997 05:35:41 GMT, "Sebastien P. McIntyre"
<NA...@music.mus.polymtl.ca> wrote:
>In article <5agnp1$c...@herald.concentric.net>
> Ker...@cris.com (KERADWC) writes:
>>In article <5aeav3$9...@news.istar.ca>, Phil Birnbaum <birn...@magi.com> wrote:
>>
>>
>>
>>>Anyway, that's why the Steinberg hypothesis appeals to me. However,
>>>as you point out, I *have* been able to notice differences (which
>>>Steinberg would explain by the Placebo Effect).
>>
>>
>>While the placebo effect will explain 'a difference', it doesn't explain
>>why multiple listeners report *the same* difference.
>
>
> In the 1987 Stereo Review article, the 15 believers were quite
> consistent identifying amp sound differences during the sighted part
> of the listening test and also had pretty uniform descriptions for
> them (read their comments in a sidebar of the article). In the
> controlled (or blind) part, they were also very consistent in their
> identification of the differences. The problem is that the results
> from both parts of the test do not agree at all (sighted: plenty of
^^^^^^^^^^^^^^
> audible differences; blind: no audible differences -- statistical
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
sense)
>
> How do you explain this?
>--
> ,
>Sebastien Best wishes for the New Year to "soldiers" of
> both camps.
>
> (Only in fake wars do soldiers observe what they
> call the traditional New Year's cease-fire. >:-> )
>
This is a good example of how blind testing screws up your sences to
the point it is difficult to determine differences or actually changes
what you hear when compared to reality. Actually this shows alot.
When each person had full control over the way they listened and
their sences were not handicapped and were not under the pressures of
an abx test they could give a certain estimate of what they thought it
sounded like but then under the abx condititions the estimate was
completey different.This right here tells you abx tests change your
perception of sound because of what the abx tests do to your human
perception of hearing which is not a perfect measuring device.
Right here a number different of people heard things differently
under abx conditions then sighted not just one but 15 peoples
perceptions changed while under the abx enviroment. Is this proof
enough that abx test changed how your hearing works under abx
conditions. And since you can't take this same way of hearing with you
after the test what good does it do?
So are you saying that 15 people were biased in the same way as to
report the same type of differences in sighted listening.These amps
produced a bias in each person which correlated with each other when
describing differences but under abx conditions no differences were
found.What conclusion can we gather from this------ 1. these amps are
putting out subliminal messages or somehow biasing each person the
same way and are saying we sound like this you will hear this sound
as better that that other amp ooooh buy me please.
or 2. that abx conditions distort human percpetions to the point it
gets very difficult to determine differences in audio even when they
were aparent in reality.
Hmm which one would you choose
ae...@flight.els
>Is this proof enough that abx test changed how your hearing works under abx
>conditions. And since you can't take this same way of hearing with you
>after the test what good does it do?
No, physiologically the human ears sensitivity will not be changed,
regardless of other senses being muted. In a blind test, the subject
is challenged to "hear" the differences as opposed to "seeing/hearing"
the differences.
Ever notice that your car runs better after it's been washed?
Blind tests are necessary because they remove the visual cues that
have a profound effect on perception and convolute the decision making
process.
Sebastien P. McIntyre
In article <32ccb58e...@nntp.ix.netcom.com>
Swa...@ix.netcom.com (Swanlee) writes:
>On 3 Jan 1997 05:35:41 GMT, "Sebastien P. McIntyre"
><NA...@music.mus.polymtl.ca> wrote:
>
>> In the 1987 Stereo Review article, the 15 believers were quite
>> consistent identifying amp sound differences during the sighted part
>> of the listening test and also had pretty uniform descriptions for
>> them (read their comments in a sidebar of the article). In the
>> controlled (or blind) part, they were also very consistent in their
>> identification of the differences. The problem is that the results
>> from both parts of the test do not agree at all (sighted: plenty of
>> audible differences; blind: no audible differences -- statistical
>> sense)
>>
>> How do you explain this?
>
>the point it is difficult to determine differences or actually changes
>what you hear when compared to reality. Actually this shows alot.
>
>
> (...) Is this proof
>enough that abx test changed how your hearing works under abx
>conditions. And since you can't take this same way of hearing with you
>after the test what good does it do?
>report the same type of differences in sighted listening.
Ever heard of many people believing in the same thing? After all,
they were "believers" (in audible differencesbetween amps). :)
> (...) These amps
>produced a bias in each person which correlated with each other when
>describing differences but under abx conditions no differences were
>found.What conclusion can we gather from this------ 1. these amps are
>putting out subliminal messages or somehow biasing each person the
>same way and are saying we sound like this you will hear this sound
>as better that that other amp ooooh buy me please.
>or 2. that abx conditions distort human percpetions to the point it
>gets very difficult to determine differences in audio even when they
>were aparent in reality.
>Hmm which one would you choose?
I choose the one you conveniently left out:
3. The "differences heard" in the sighted part of the test were
products of the listeners' imagination (also termed "flight of
fancy") and/or the non-matching of levels.
In the sighted part, listeners could __force__ an answer based on
bias (preconceptions and/or unequalized levels), something readily
randomized and/or controlled when using an appropriate listening
paradigm. Enrol for "Psychology 101" one of these days, as my
girlfriend would suggest. Mine is: Enrol for "ABX 101". :)
--
,
Sebastien
F. Blaine Dickson
You assume that in blind testing listeners do not have any biases.
Listeners have biases before they even come to the test that could
affect the outcome of the test results. In a properly carried out
examination, these would be controlled for so that there would be no
possibility that they could confound the test. IMO, this is not
accomplished in most audio ABX studies.
Gene Steinberg
In article <1997010511...@pme005.awinc.com>, bdic...@awinc.com (F.
Blaine Dickson) wrote:
>You assume that in blind testing listeners do not have any biases.
>Listeners have biases before they even come to the test that could
>affect the outcome of the test results. In a properly carried out
>examination, these would be controlled for so that there would be no
>possibility that they could confound the test. IMO, this is not
>accomplished in most audio ABX studies.
Sorry you are not really understanding the point of the test. It is double
blind, meaning that the person who administers the tests (regardless of
that person's point of view) cannot influence the outcome. We know everyone
has a bias of one sort or another. The purpose of the ABX test is to allow
you to judge a component on its sound alone. You are not influenced by its
appearance, purchase price, the review you read in Stereophile, a slight
level mismatch, etc. The ABX test is best suited to help you determine if
an audible difference exists. If after determining that differences are not
audible, you may still prefer one component over another on other grounds
(brand reliability, amplifier power output, appearance, snob appeal, price,
whatever). I don't think anyone has a problem with that.
--
Peace,
Gene
Norm Strong
04JAN97.19...@music.mus.polymtl.
Organization: Seattle Community Network
In a previous article, bdic...@awinc.com (F. Blaine Dickson) says:
>
>You assume that in blind testing listeners do not have any biases.
>Listeners have biases before they even come to the test that could
>affect the outcome of the test results. In a properly carried out
>examination, these would be controlled for so that there would be no
>possibility that they could confound the test. IMO, this is not
>accomplished in most audio ABX studies.
Sounds interesting. Tell me more. What biases might a listener have that
would affect the outcome of a blind test?
--
Norm Strong (bg...@scn.org)
2528 31st S. Seattle WA 98l44
F. Blaine Dickson
Norm Strong <bg...@scn.org> wrote:
Expectations, mainly. The subjects still go into the tests with biases
about audio in general regardless of whether they see the components or
not. They have expectations that there may be differences or not, or
that CD players or amps should sound a certain way etc. For example,
someone may actually think that shrillness in some poor CD players is
pleasurable or even accurate representation of real music. These
expectations can still confound the tests. Furthermore, subjects may
have expectations about how they *should* perform in the tests,
especially if any pre-test discussion has given the purpose of the test
in the first place. They may expect to find or not find a difference
between components, and this may affect their decision making,
especially if the components sounded almost the same or almost
different. This has a possibility of confounding the experiment as
well.
Any good researcher tries to hide the test's purpose until after the
subject is finished so that it limits the potential biases or
expectations the subjects may acquire and use during the tests. That is
why I do not like the sole use of audio specific people in ABX tests
because they have more biases going into the tests than people who have
no interest in audio. If you are going to use the results as a basis for
saying "most people..." you better have a test population that actually
represents "most people" rather than a group of subjects limited by
their characteristics.
Alan Derrida
Sebastien P. McIntyre (NA...@music.mus.polymtl.ca) wrote:
: In the 1987 Stereo Review article, the 15 believers were quite
: consistent identifying amp sound differences during the sighted part
: of the listening test and also had pretty uniform descriptions for
: them (read their comments in a sidebar of the article). In the
: controlled (or blind) part, they were also very consistent in their
: identification of the differences. The problem is that the results
: from both parts of the test do not agree at all (sighted: plenty of
: audible differences; blind: no audible differences -- statistical
: sense).
: How do you explain this?
Hmmm... that's a tough one. Did SR happen to use an ABX box in the signal
path in the tests in that article?
Swanlee
Its easy to explain.Blind tests dampen the sence of hearing to the
point where they can't discern differences. All these people found
similar differences in sghted test without talking to one another or
revealing what they thoughtr but in blind tests they couldn't tell a
difference. Why isn't anyone a little skeptical of this type of test?
What the hell is the relevence anway im not blind?
Why is everyone so blind to maybe flaws in blind testing?
Swanlee
>In article <1997010511...@pme005.awinc.com>, bdic...@awinc.com (F.
>Blaine Dickson) wrote:
>
>>You assume that in blind testing listeners do not have any biases.
>>Listeners have biases before they even come to the test that could
>>affect the outcome of the test results. In a properly carried out
>>examination, these would be controlled for so that there would be no
>>possibility that they could confound the test. IMO, this is not
>>accomplished in most audio ABX studies.
>
>
>Sorry you are not really understanding the point of the test. It is double
>blind, meaning that the person who administers the tests (regardless of
>that person's point of view) cannot influence the outcome. We know everyone
>has a bias of one sort or another. The purpose of the ABX test is to allow
>you to judge a component on its sound alone. You are not influenced by its
>appearance, purchase price, the review you read in Stereophile, a slight
>level mismatch, etc. The ABX test is best suited to help you determine if
>an audible difference exists. If after determining that differences are not
>audible, you may still prefer one component over another on other grounds
>(brand reliability, amplifier power output, appearance, snob appeal, price,
>whatever). I don't think anyone has a problem with that.
>
>--
>Peace,
>Gene
Please produce scientific studies of how abx test effects a person
percpetion.When doing an abx test we may very well be trading on evil
for another.please produce evidence that these test do not obscure
percpetion to the point it is impossible to tell differences even if
they were their.Real evidence by real scientists not audio engineers.
And people trained in human percpetion not just audio.Untill you do
this the abx test is just a new toy for audio skpetics to play with.
You have to porve the test is scientificallt valid for the results to
be taken seriously.and just because the results are repatable does not
make it a worth test.If the same falws keep apearing in the test of
course it will be repeatable.But until more is learned about human
perception in the abx enviroment i will not take the results seriously
Instead i will listen ot prodcut in my own home and ones that [prform
better on a consistant basis in my home with my ears under my
condititons I will keep.
Please answer me this Gene why were you trying to get Sanders to give
out and confirm all of his personal info while you have yet to reveal
your audio syatem or even what music you like in of all places an
audio newsgroup? Are you just here to argue? If not how come you do
not ask questions about systems or ask advice or even tell us what
music you listen to?
Per Strömgren
In article <32d22ab5...@nntp.ix.netcom.com>,
Swa...@ix.netcom.com (Swanlee) wrote:
>Its easy to explain.Blind tests dampen the sence of hearing to the
>point where they can't discern differences.
That could very well be the case. How would you prove os disprove this?
The problem is to remove the bias from seeing and knowing about the
difference between the objects under test.
>All these people found
>similar differences in sghted test without talking to one another or
>revealing what they thoughtr but in blind tests they couldn't tell a
>difference. Why isn't anyone a little skeptical of this type of test?
Because there is a much simpler reason: there is no difference. Until
we come up with a better reason, we are stuck with this one.
>What the hell is the relevence anway im not blind?
>Why is everyone so blind to maybe flaws in blind testing?
There may be flaws, but noone has found them yet. If you think
there are any, you must use SOME method of testing this. How would
you do this?
May a suggest you read Daniel Normolle's references to lay
a solid foundation for your research?
Per Stromgren,
Karlstad,
Sweden.
From: mo...@sunm4515.sph.umich.edu (Dan Normolle)
Subject: Re: Double Blind Testing
Date: 30 May 1995 14:34:38 GMT
In article <3q8mom$q...@ethereal.magic.ca>, mea...@astral.magic.ca (Richard Dal
Farra) writes:
|> A question for the rec.audio.SDuraybito group. For all the debate over
|> objectivism vs. subjectivism, I don't believe I've ever seen this asked.
|>
|> What tests were used to establish the supremcy of double-blind testing?
|>
|> It's not a facetious question. If, in science, all assertions must be
|> tested and verified, then somewhere someone must have shown that, among
|> all the other possible testing methodologies, double-blind is superior.
|> How was this established?
This is a good question. An honest discussion (as opposed to a foul-mouthed
flame war) of the use of double-blind testing in audio actually points to
some core issues in the history and philosophy of science. For an
*introduction* to these issues, I suggest
Karl Popper, The Logic of Scientific Discovery
Thomas Kuhn, The Structure of Scientific Revolutions
And, so I am not labelled as some sort of falsifiablist Luddite
Paul Feyerabend, Against Method
Or, as an alternative, a good textbook introduction to the philosophy of
scientific method is,
David Oldroyd, The Arch of Knowledge
Popper and Kuhn lay down what most 20th Century scientists would agree
delineate science from non-science, although many philosophers of science
(e.g., Feyerabend) might argue that the practicing scientists are behind the
curve. A good-faith discussion of what science can claim to know and
what it cannot looks a lot different from the twaddle that's been passed
off in this neighborhood recently.
dpn
-----------------------------------------------------------------------------
Daniel Normolle, PhD
mo...@umich.edu
-----------------------------------------------------------------------------
Tom Albertz
Swanlee wrote:
> Please produce scientific studies of how abx test effects a person
> percpetion.When doing an abx test we may very well be trading on evil
> for another.please produce evidence that these test do not obscure
> percpetion to the point it is impossible to tell differences even if
> they were their.
You have a problem with ABX methodology. OK.
What about blind listening comparisons in YOUR listening environment?
Would you buy $100.00 cables because you thought they improved
the sound of your stereo?
Do you believe that SIGHTED listening is completely flawed and
unreliable?
F. Blaine Dickson
Gene Steinberg <Ge...@worldnet.att.net> wrote:
> Sorry you are not really understanding the point of the test. It is double
> blind, meaning that the person who administers the tests (regardless of
> that person's point of view) cannot influence the outcome. We know everyone
> has a bias of one sort or another. The purpose of the ABX test is to allow
> you to judge a component on its sound alone. You are not influenced by its
> appearance, purchase price, the review you read in Stereophile, a slight
> level mismatch, etc. The ABX test is best suited to help you determine if
> an audible difference exists. If after determining that differences are not
> audible, you may still prefer one component over another on other grounds
> (brand reliability, amplifier power output, appearance, snob appeal, price,
> whatever). I don't think anyone has a problem with that.
Absolutely, but the subjects still go into the tests with biases about
audio in general regardless of whether they see the components or not.
They have expectations that there may be differences or not, or that CD
players or amps should sound a certain way etc. These can confound the
tests. Furthermore, they may have expectations about how they *should*
perform in the tests, especially if the pre-test discussion has given
the purpose of the tests in the first place. This has a possibility of
confounding the experiment as well. Any good researcher tries to hide
the test's purpose until after the subject is finished so that it limits
the potential biases or expectations the subjects may acquire and use
during the tests. That is why I do not like the sole use of audio
specific people in ABX tests because they have more biases going into
the tests than people who have no interest in audio. Moreover, if you
Swanlee
I bleive in buying something taking it home trying it out for however
long they allow me to while I can still bring it back and then
determining whether the product is worthy of the money I spent.
I do not beleive in blind level matched etc. etc. test of any kind.
If a product performs better consistantly in my home with my equipment
listened how I will normally listen to the product on a consistant
basis then whatever differences are their are real enough for me to
justify keeping it.If you have a porlbme with that oh well.Im to bust
actually listening and enjoying my msuic to run around do abx tests to
death. But if I feel the products performace justifies the cost when
used how I will use it then I will keep it.
Steve Maki
bdic...@awinc.com wrote...
>Absolutely, but the subjects still go into the tests with biases about
>audio in general regardless of whether they see the components or not.
>They have expectations that there may be differences or not, or that CD
>players or amps should sound a certain way etc. These can confound the
>tests.
Often, if not usually, the subject's bias (at first) is that everything
sounds different. Usually, the test results indicate otherwise.
Under this scenario, how has the subject's bias confounded the test?
--
Steve Maki K8LX
st...@oakcom.com
Walter Tice USG
In article <29DEC96.23...@music.mus.polymtl.ca> "Sebastien P. McIntyre" <NA...@music.mus.polymtl.ca> writes:
> # - Julius Futterman OTL-1 (80-Watt mono tube amp, $12,000/pair)
> # - Mark Levinson ML-11 (50W/ch, #2,000)
> # - Counterpoint SA-12 (tube-SS hybrid, $995)
> # - NAD 2200 (100W/ch, 6-dB dynamic power, high current/voltage, $548
> # - Hafler DH-120 (62W/ch, $320)
> # - Pioneer SX-1500 (45W/ch receiver, $220)
> #
What was the rest of the equipment?
Bose 901's
Adcom pre-amp (the budget one from the late 80's) - never gone off on this 'piece' in this
group yet, but, what an overrated piece.
Magnavox 502 CD
Anybody who can't hear the differance between an OTL-1 and an ML-11 on a
system of serious resolution should be buy a rack system and forget about
anything else.
W
Sebastien P. McIntyre
On Sun, 5 Jan 1997, F. Blaine Dickson wrote:
> Sebastien P. McIntyre <NA...@music.mus.polymtl.ca> wrote:
>
> > In the sighted part, listeners could __force__ an answer based on
> > bias (preconceptions and/or unequalized levels), something readily
> > randomized and/or controlled when using an appropriate listening
> > paradigm. Enrol for "Psychology 101" one of these days, as my
> > girlfriend would suggest. Mine is: Enrol for "ABX 101". :)
>
> You assume that in blind testing listeners do not have any biases.
Absolutely not. This is the very reason for conducting it blind.
> Listeners have biases before they even come to the test that could
> affect the outcome of the test results.
Yes listeners come with biases. No this won't affect the outcome
of a properly conducted test that eliminates not the biases in the
mind of the listener but their effect on the results (i.e. you are
free to keep these biases after testing even if you fail to give
evidence supporting them...).
> In a properly carried out
> examination, these would be controlled for so that there would be no
> possibility that they could confound the test.
You mean listeners would assuredly discern differences between the
DUTs? Not necessarily if differences are below threshold *__and__*
the test carried out blind (re: my weekend post about a "positive"
listening test with _identical_ DUTs). I assume here you are fully
aware that perfectly analytical hearing is only seen in "The Outer
Limits". :)
> IMO, this is not
> accomplished in most audio ABX studies.
How and why? Suggestion(s)? I have one to demonstrate that it is.
Take an ABX test of, say, two SS amps (A and B in full view), come
with your mind full of all the biases you can think of and try to
force them into your calls (answers). Then post the results of your
scoresheet for us to appreciate if you succeeded. May differences
be large enough or luck with you. :)
--
,
Sebastien Update your kill-file. This e-mail account is new.
Sebastien P. McIntyre
On Tue, 7 Jan 1997, Swanlee wrote:
> I do not beleive in blind level matched etc. etc. test of any kind.
Finally one clear statement. But blind listening isn't something to
believe in. Blind listening is for facts, and facts are facts and
hold without beliefs of any kind.
Swanlee
On 9 Jan 1997 04:22:10 GMT, "Sebastien P. McIntyre"
<s...@meca.polymtl.ca> wrote:
Yea you can qoute facts all day long. Giving the results of an abx
test is a fact but if I dont like the way the test is or feel its
flawed I could care less for those facts.The only facts in abx tests
is saying these are the results. It will never prove their are no
differences between two amps cause you can't prove a negative.The
results of the test only apply to those amps using those people at
that particualr time.It has no application to anything outside of that
test.And since I don't listen to musici in that test I could care
less.
DrifterUSA
ae...@flight.els wrote:
>>Blind tests are necessary because they remove the visual cues that have
a profound effect on perception and convolute the decision making
process.<<
Or perhaps blindness in itself (to a sighted person) convolutes the
decision making process. I.e., visual disorientation may in fact have an
effect on our other senses' ability to perceive. Have you ever tried to
figure out what something is by touch only? It's not always easy.
Which brings up a thought: surely there are blind audiophiles. Do they
hear a difference?
John Bauer
Swanlee
Try doing a blind taste test while holding your nose shut.
Chuck Ross
> ae...@flight.els wrote:
>
> >>Blind tests are necessary because they remove the visual cues that have
> a profound effect on perception and convolute the decision making
> process.<<
>
> Or perhaps blindness in itself (to a sighted person) convolutes the
> decision making process. I.e., visual disorientation may in fact have an
> effect on our other senses' ability to perceive. Have you ever tried to
> figure out what something is by touch only? It's not always easy.
>
> Which brings up a thought: surely there are blind audiophiles. Do they
> hear a difference?
>
> John Bauer
This threadlet is getting really silly. Firstly, "blind" testing does not
by any means mean that you are blindfolded, or in any way have any vision
impairment or obstruction.
Both (or all) of the DUTs (Devices Under Test) are always VISIBLE; it's
just that you don't really know which one you are listening to at the moment.
Listening for differences with a blindfold in place is an entirely different
matter. Also, people who have been truly blind for a long time generally
have amazingly developed sensory perception in other areas to compensate.
I had a blind uncle who could recognize me from 50 feet away by the sound
of my footsteps on the sidewalk. I have always wondered how he would hear
two different amplifiers compared.....
If there are any blind audiophiles reading this, I would be very interested
to hear from them. No humor intended here...there are quite a few blind
invididuals on the internet who have others read for them.
--
Sebastien P. McIntyre
On 7 Jan 1997, Alan Derrida wrote:
> Sebastien P. McIntyre (NA...@music.mus.polymtl.ca) wrote:
>
> : In the 1987 Stereo Review article, the 15 believers were quite
> : consistent identifying amp sound differences during the sighted part
> : of the listening test and also had pretty uniform descriptions for
> : them (read their comments in a sidebar of the article). In the
> : controlled (or blind) part, they were also very consistent in their
> : identification of the differences. The problem is that the results
> : from both parts of the test do not agree at all (sighted: plenty of
> : audible differences; blind: no audible differences -- statistical
> : sense).
>
> : How do you explain this?
>
> Hmmm... that's a tough one. Did SR happen to use an ABX box in the signal
> path in the tests in that article?
Hmmm... Another article le Sieur de Derrida hasn't read?... Otherwise
he would already know that listeners were allowed to chose manual cable
swap (still with equalized levels and blind calling, naturellement) in
any given listening session if they felt this would increase the
accuracy of their identifications. Indeed, 4 of the 15 believers
chose this method of switching over ABX-box switching. I suspect
that, like Messrs. Derrida and Dickson, they didn't understand what
quasi-instantaneous switching is all about/for or, like Mr. Zipser,
they were very distrustful, publicly at least, of the switcher for
having heard of its indemonstrable "veiling effects"... :) (Five
other listeners tried cable swap but didn't retain it as their
preferred method of switching)
As for the results of those listeners who chose not "to use an ABX
box in the signal path", well... they show even "better" randomness
(statistical sense) than those from ABX-box switching (but this is
mean talking). IMO, it's the equalized-blind nature of the listening
tests, coupled with lowered resolution by not using a good switcher,
that is the cause of their bad showing... or is it the inaudibility
of the sonic differences? :) If only Stereo Review had used green
ink... Wait, were the "benefits" of green ink known back in 1986?
If not, Stereo Review owe these believers a fairer test... conducted
in the dark to see all these fluo-green painted ears in action! :-D
--
,
Sebastien
KERADWC
In article <5atlqp$c...@newstoo.ericsson.se>,
Per Strömgren <Per.St...@ein.ericsson.se> wrote:
>In article <32d22ab5...@nntp.ix.netcom.com>,
> Swa...@ix.netcom.com (Swanlee) wrote:
>>All these people found
>>similar differences in sghted test without talking to one another or
>>revealing what they thoughtr but in blind tests they couldn't tell a
>>difference. Why isn't anyone a little skeptical of this type of test?
>
>Because there is a much simpler reason: there is no difference. Until
>we come up with a better reason, we are stuck with this one.
FALSE REASONING!
If you accept the premise that there is no difference when ABX shows no
difference, this conclusion is acceptable. Since that premise, however,
is the one being disputed, using it as a given is fatuous, er, invalid.
Read the phrase to which you responded: "All these people found *SIMILAR*
differnces ... without talking to one another or revealing what they thought".
Does this sound like something where there is no difference? If there is no
real difference, why were the declared differences the same from one listener
to another?
I've participated in a limited number of auditions like that, with between
3 and 5 listeners, listening to different equipment (amplifiers and speakers
only). If the differences reported were merely different, I'd be inclined
to say the no-difference claim wasn't completely wrong--but the differences
were *consistently* reported; if A said it sounded 'bright', B, C, D, and E
usually did also. Same for imaging, forwardness, soundstaging, etc.
To claim, without any evidence, that this difference is due to everyone's
built-in perceptions/anticipation is pretty weak. And it certainly wouldn't
work for the time all of the listeners gave a nearly identical *trash*
review of a particular, very well reputed speaker. It turns out that while
that model is usually a bit forward, with a rolled-off treble, this unit
had the wrong cross-over, and nobody reported what it *should* have sounded
like. Similarly, why would top brand units be given poor scores (in the same
categories for each listener) if that brand is generally known for being
good in that area? (Recent amplifier tests did this.)
You'll need to *prove* your claim that no difference via ABX really means
no difference in teh real world.
"All models are flawed. Some models are useful." Is ABX more flawed, or more
useful?
--kdc
Sebastien P. McIntyre
On Fri, 10 Jan 1997, Chuck Ross wrote:
> This threadlet is getting really silly. Firstly, "blind" testing does not
> by any means mean that you are blindfolded, or in any way have any vision
> impairment or obstruction.
>
> Both (or all) of the DUTs (Devices Under Test) are always VISIBLE; it's
> just that you don't really know which one you are listening to at the moment.
In the ABX paradigm, you know exactly when DUT A, DUT B or DUT X plays.
The A <---> B directions of the ABX paradigm thus constitute a fully
SIGHTED evaluation of the components. Only the A <---> X and B <---> X
directions are "blind" (you don't know in advance whether X is A or B).
Imagine the following three-position joystick:
X
/ \
blind / \ blind
/ \
/ \
A --------- B
sighted
> Listening for differences with a blindfold in place is an entirely different
> matter. Also, people who have been truly blind for a long time generally
> have amazingly developed sensory perception in other areas to compensate.
> I had a blind uncle who could recognize me from 50 feet away by the sound
> of my footsteps on the sidewalk. I have always wondered how he would hear
> two different amplifiers compared.....
I'm sure he would have, given a "foot-dragging" amp... ;-)
> If there are any blind audiophiles reading this, I would be very interested
> to hear from them. No humor intended here...there are quite a few blind
> invididuals on the internet who have others read for them.
Some have their NGs (artificially) speech-processed, assuming spelling
mistakes to be nearly absent (I would have designed these processors
to work only on spelling mistake free posts... so the ignorance of the
ABX paradigm of some couldn't contaminate them... >:-> ).
--
,
Sebastien
Sebastien P. McIntyre
On Thu, 9 Jan 1997, Swanlee wrote:
> On 9 Jan 1997 04:22:10 GMT, "Sebastien P. McIntyre" wrote:
>
> >On Tue, 7 Jan 1997, Swanlee wrote:
> >
> >> I do not beleive in blind level matched etc. etc. test of any kind.
> >
> > Finally one clear statement. But blind listening isn't something to
> > believe in. Blind listening is for facts, and facts are facts and
> > hold without beliefs of any kind.
>
> Yea you can qoute facts all day long. Giving the results of an abx
> test is a fact but if I dont like the way the test is or feel its
> flawed I could care less for those facts.The only facts in abx tests
> is saying these are the results. It will never prove their are no
> differences between two amps cause you can't prove a negative.
Always uttering the same straw-man argument. Which means that although
you read my/others' earlier posts on this, you still don't understand them.
Or is it that you _refuse_ to understand them? You seem not to realize
how bad you hurt your position by allowing us to erect the highest
platform for this ABX "dogma". Thanks, Mr. Davenport, for your invaluable
help to push back the frontiers of ignorance. :-D
> The
> results of the test only apply to those amps using those people at
> that particualr time.It has no application to anything outside of that
> test.And since I don't listen to musici in that test I could care
> less.
Test after test.
--
,
Sebastien
Bruce Armstrong
another in topology in many cases, the power supplies are inevitably
different, and so on. The same question can be asked of anything made
by man. Ain't nothing the same out there. But that's to me. I cannot
help it if someone else is unable to tell the diference between a BMW
and a Hyundai when blindfolded and driven as a passenger over the same
30 mile course. Of course, as a driver, you can, and in the privacy of
your own home, you should be able to tell the difference between
amplifiers. If you can't, learn to love your Radio Shack salesman by
buying the cheapest, and keep off opinion newsgroups! Instead, devote
your time to haranguing Analog Devices or Linear Technology for wasting
their time developing new op-amps, when for obvious reasons the ones we
have are perfectly wonderful anyway. An opinion of no difference is
valid one time only. After that, the rest of us don't care about you.
Attacking people who do hear a difference is intellectual arrogance, and
worse, boring.
I have designed many different amplifiers, and they all sounded
different, even those that merely doubled up on the output transistors
from the original, where the quiescent power for class AB operation is
actually less for the same effect of reducing crossover distortion --
that's how Krell get away with those huge Class A ratings (or did,
before they went off into "anticipatory" output biasing).
All these *scientific* types who cannot hear a difference operate under
a different paradigm of reality from the average Joe, and quite honestly
cannot hear a difference. So, do I care? And neither should you.
Forget them. They're earnest, and probably have some aspect of their
life where they will argue until blue in the face that differences
exist. Audio just doesn't happen to be it.
Yawn...
Bruce Armstrong
)
In article <32d6016c....@nntp.ix.netcom.com>,
No problem. Salt and sugar in an ABX test, with water between
identifications. I'm willing to bet you up to a thousand dollars
that I can identify the X in the ABX test. Would you like to
take me up on this?
Young-Ho
Swanlee
On 11 Jan 1997 21:43:37 GMT, "Sebastien P. McIntyre"
<s...@meca.polymtl.ca> wrote:
>
>
>On Thu, 9 Jan 1997, Swanlee wrote:
>
>> On 9 Jan 1997 04:22:10 GMT, "Sebastien P. McIntyre" wrote:
>>
>> >On Tue, 7 Jan 1997, Swanlee wrote:
>> >
>> >> I do not beleive in blind level matched etc. etc. test of any kind.
>> >
>> > Finally one clear statement. But blind listening isn't something to
>> > believe in. Blind listening is for facts, and facts are facts and
>> > hold without beliefs of any kind.
>>
>> Yea you can qoute facts all day long. Giving the results of an abx
>> test is a fact but if I dont like the way the test is or feel its
>> flawed I could care less for those facts.The only facts in abx tests
>> is saying these are the results. It will never prove their are no
>> differences between two amps cause you can't prove a negative.
>
>
> Always uttering the same straw-man argument. Which means that although
> you read my/others' earlier posts on this, you still don't understand them.
> Or is it that you _refuse_ to understand them? You seem not to realize
> how bad you hurt your position by allowing us to erect the highest
> platform for this ABX "dogma". Thanks, Mr. Davenport, for your invaluable
> help to push back the frontiers of ignorance. :-D
The only ignorance is thinking abx actually proves anything. What do
the results of an abx test prove? They prove that these particlular
people at this particular time using these particular amps at these
particular levels in this partilcuar room using this particular music
with these paicular speakers,cables,cd players,interconnectsspaker
wires and how these particular people felt at the time could not tell
differences in said equipment. It has no and I repeat no application
outside of all of the factors.You cannot take those results and apply
them to anything outside of all of the factors mentioned here cause a
change in one of those factors could completely change the result of
the test.
And you cannot even apply the results to get a benifical results in
your home which you still do not seem to grasp.The results still does
not make the 200$ sound as good or better than the 2000$ no matter how
many tests you do. What ever bias the abx test supposingly takes away
its still their in your home after the test so you still have to deal
with it no matter what the abx test say.The only ignorance is
excepting a flawed limited test as the holy grail behind your entire
belif system.
>
>
>> The
>> results of the test only apply to those amps using those people at
>> that particualr time.It has no application to anything outside of that
>> test.And since I don't listen to musici in that test I could care
>> less.
>
>
> Test after test.
Yea Im not an abx tester im a music listener.
>--
> ,
>Sebastien
Swanlee
On 11 Jan 1997 23:52:34 -0500, you...@larmor.nmr.mgh.harvard.edu (\)
wrote:
I bet with your nose closed you could not tell coke form pepsi.
It is also science fact that the sence of smell has a direct effect on
the sence of taste any 6th grade science book will tell you that.
Comapring salt and sugar would be like comapring test tones that are
miss matched in abx tests and not music wih matched levels so again
your analogy and challange is not valid.
But try coke and pepsie in a blind taste test with your nose closed
and see what happens.
Swanlee
On 7 Jan 1997 14:54:44 GMT, pav...@niktow.canisius.edu (Greg Pavlov)
wrote:
>
>
>Per Strvmgren (Per.St...@ein.ericsson.se) wrote:
>: In article <32d22ab5...@nntp.ix.netcom.com>,
>: Swa...@ix.netcom.com (Swanlee) wrote:
>: >Its easy to explain.Blind tests dampen the sence of hearing to the
>: >point where they can't discern differences.
>:
>: That could very well be the case. How would you prove os disprove this?
>: The problem is to remove the bias from seeing and knowing about the
>: difference between the objects under test.
>:
>
> The basic problem is that the bias may produce most of the differences :-)
Yes the MAY and I repeat MAY cause them but you cannot and never
scientificall statistically prove that as fact. But thinking of stats
is it statistically possible that all differences in audio heard all
over the world over 50 years have had just 1 cause.That millions if
not billions of perceived audio differences caused by just one thing
and not the many billions of other valid factors.Bad anology but I
think you get my point.
Your using science against itself by trying to factually prove
something that cannot and never will be factually proved outsiode of
the test conditions it just can't be done. ABX tests are only
relavanet to the people used in the test the
amps,cable,speakers,abxbox,music listened in the test the exact
conditions of the room and the hearing of each individual person and
how they flet that day. You cannot apply the test results to anything
outside of these factors it simply doesn't work.You can prove at that
time that person using that amp with that cable with those speakers
listening to that music on that day when that person felt exactly the
way they did at the precise moment of the test. But those conditions
will never be reaptable again and I mean those conditions any
variuation could change the results.You can't apply it to real world
home listening so why the hell bother. It doesn't prove anything it
doesn't improve anything and violates the scienitifc range rule to
boot. And some base their whole audio belif syatem on them.
Hey read them take what they say into account but the bottom line is
how it sound in your hom eto you no mater what the tests say.
But aside from that, one can create visual cues to substitute for the
brand
> names and model numbers. For instance, one could put different-colored
> boxes over the units being tested. Or, instead of colors, a different
> cue could be used, such as a unique picture on the front of each box. How
> about portraits of Bill Clinton, Newt Gingrich, and Jim Wright ? Somehow
> these seem particularly appropriate, tho Sam Tellig, Jonathan Scull, and
> Robert Harley would probably do just as well...
>
> It would be interesting to run tests in this manner with *two* panels, with
> the new-and-improved visual cues randomly reallocated from one panel to the
> next. Would the results follow the cues ? For example, would more people
> perceive commonality `tween Gingrich Amp and Wright Amp or `twixt Gingrich
> Amp and Clinton Amp ? Enquiring minds wanna to know.
>
>
> [sorry for all of the U.S.-centric "cultural" references ...]
What I want is for 20 people go into a room individually with two amps
give them an hour to play with the amps all they want listen to what
ever music they want then individually write down what they though the
differences were. Take what they wrote by themsleves while listening
by themsleves and if they compare pretty well as in they heard the
same differences have them do the abx test. If they can't tell any
differences in abx then Id say this is pretty good proof the abx test
is flawed when humans are involved. I read about a test like this that
had about the same result but i do not know if the procedure was
exactually the same.
>
>
>
> greg pavlov
> [not affiliated with Canisius College]
>
> *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
> The truth of a proposition has nothing to do with its
> credibility. And vice versa.
> *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
>
)
In article <32d88384...@nntp.ix.netcom.com>,
Swanlee <Swa...@ix.netcom.com> wrote:
>>>
>>>Try doing a blind taste test while holding your nose shut.
>>
>>No problem. Salt and sugar in an ABX test, with water between
>>identifications. I'm willing to bet you up to a thousand dollars
>>that I can identify the X in the ABX test. Would you like to
>>take me up on this?
>>
Hmmm. I've never tried this. I don't drink cola.
>It is also science fact that the sence of smell has a direct effect on
>the sence of taste any 6th grade science book will tell you that.
Gee, really? It is also "science fact" that the sense of smell is not
the sole determinant of the sense of taste. I'm glad that you're reading
6th grade science books, though. It's a good start.
>Comapring salt and sugar would be like comapring test tones that are
>miss matched in abx tests and not music wih matched levels so again
>your analogy and challange is not valid.
I was not making an analogy. You simply said "Try doing a blind taste
test while holding your nose shut." Your tongue has chemical receptors
that can identify salt and sugar (this forms the basis for our sense of
taste, and the olfactory receptors significantly enhances it), and so
I offered a clear example of a blind taste test where tastes could be
identified with an impaired sense of smell. You did not say "Try doing
a blind taste test that is analogous to listening to music with two
level-matched components. This test should be done while holding your
nose shut."
>But try coke and pepsie in a blind taste test with your nose closed
>and see what happens.
I think Armand wrote about the significantly higher sugar level of Pepsi
as a key to distinguishing between and positively identifying Coke and
Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
I'm willing to try this myself, and I'll report on the results later this
week when I get the other guys from work to try it as well, if people
are really interested. I'd be happier trying Bass and Guinness in a blind
taste test with my nose closed, and I'm willing to bet money that I'm
able to do so. Easy money for me!
Young-Ho
Swanlee
On 12 Jan 1997 22:30:12 -0500, you...@larmor.nmr.mgh.harvard.edu (\)
wrote:
>In article <32d88384...@nntp.ix.netcom.com>,
>Swanlee <Swa...@ix.netcom.com> wrote:
>>>>
>>>>Try doing a blind taste test while holding your nose shut.
>>>
>>>No problem. Salt and sugar in an ABX test, with water between
>>>identifications. I'm willing to bet you up to a thousand dollars
>>>that I can identify the X in the ABX test. Would you like to
>>>take me up on this?
>>>
>>I bet with your nose closed you could not tell coke form pepsi.
>
>Hmmm. I've never tried this. I don't drink cola.
>
>>It is also science fact that the sence of smell has a direct effect on
>>the sence of taste any 6th grade science book will tell you that.
>
>Gee, really? It is also "science fact" that the sense of smell is not
>the sole determinant of the sense of taste. I'm glad that you're reading
>6th grade science books, though. It's a good start.
Yes not the sole determinant did I say it was but it does have a
dramatic impact on your sence of taste.Where in this post did I ever
say it was the sole determinant but it does have a pretty dramatic
effect. Hell eat anything while holding your nose then release it and
you will get a rush of flavor do it yourself. Unless you are blind of
taste as well as blind of hearing.
>
>>Comapring salt and sugar would be like comapring test tones that are
>>miss matched in abx tests and not music wih matched levels so again
>>your analogy and challange is not valid.
>
>I was not making an analogy. You simply said "Try doing a blind taste
>test while holding your nose shut." Your tongue has chemical receptors
>that can identify salt and sugar (this forms the basis for our sense of
>taste, and the olfactory receptors significantly enhances it), and so
>I offered a clear example of a blind taste test where tastes could be
>identified with an impaired sense of smell. You did not say "Try doing
>a blind taste test that is analogous to listening to music with two
>level-matched components. This test should be done while holding your
>nose shut."
It was in context with an abx disscussion if you are to linear to
realize that then to bad.
>
>>But try coke and pepsie in a blind taste test with your nose closed
>>and see what happens.
>
>I think Armand wrote about the significantly higher sugar level of Pepsi
>as a key to distinguishing between and positively identifying Coke and
>Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
>I'm willing to try this myself, and I'll report on the results later this
>week when I get the other guys from work to try it as well, if people
>are really interested. I'd be happier trying Bass and Guinness in a blind
>taste test with my nose closed, and I'm willing to bet money that I'm
>able to do so. Easy money for me!
>
>Young-Ho
How stupid are you you do not think that the sence of smell has
anything to do with the sence of taste my god this is ridiculous.
DrifterUSA
Chuck Ross wrote:
>>This threadlet is getting really silly. Firstly, "blind" testing does
not
by any means mean that you are blindfolded, or in any way have any vision
impairment or obstruction.
>>Both (or all) of the DUTs (Devices Under Test) are always VISIBLE; it's
just that you don't really know which one you are listening to at the
moment.<<
Excuse my ignorance and I'll excuse your condescension.
>>Listening for differences with a blindfold in place is an entirely
different matter. Also, people who have been truly blind for a long time
generally have amazingly developed sensory perception in other areas to
compensate.<<
Gee, maybe that's why I brought it up.
[Further elaboration of my idea snipped.]
John Bauer (cyberwounded, but not beaten)
Alan Derrida
Sebastien P. McIntyre (s...@meca.polymtl.ca) wrote:
On 7 Jan 1997, Alan Derrida wrote:
> Sebastien P. McIntyre (NA...@music.mus.polymtl.ca) wrote:
>
> : In the 1987 Stereo Review article, the 15 believers were quite
> : consistent identifying amp sound differences during the sighted part
> : of the listening test and also had pretty uniform descriptions for
> : them (read their comments in a sidebar of the article). In the
> : controlled (or blind) part, they were also very consistent in their
> : identification of the differences. The problem is that the results
> : from both parts of the test do not agree at all (sighted: plenty of
> : audible differences; blind: no audible differences -- statistical
> : sense).
> : How do you explain this?
>
> Hmmm... that's a tough one. Did SR happen to use an ABX box in the
signal
> path in the tests in that article?
: Hmmm... Another article le Sieur de Derrida hasn't read?...
: Otherwise he would already know that listeners were allowed to chose
: manual cable swap (still with equalized levels and blind calling,
: naturellement) in any given listening session if they felt this would
: increase the accuracy of their identifications.
Why do you think I ASKED you about the test McIntyre?? I take it that
somehow, you didn't understand the question since you're answer
presumed I should have already read it. What do you think I do - sit
at home and clip ABX articles out of magazines, line them up on a
corkboard and compare each for research in a debate somewhere? Unlike
you, I would rather use that time to actually listen to music. You know...
ecouter la musique??
: Indeed, 4 of the 15 believers chose this method of switching over
: ABX-box switching. I suspect that, like Messrs. Derrida and Dickson,
: they didn't understand what quasi-instantaneous switching is all
: about/for or, like Mr. Zipser, they were very distrustful, publicly at
: least, of the switcher for having heard of its indemonstrable "veiling
: effects"... :) (Five other listeners tried cable swap but didn't retain
: it as their preferred method of switching)
I suspect that you have no knowledge whatsoever about what I
do or do not know about "quasi-instantaneous switching" in ABX trials,
especially from the one line reply I gave, but I also suspect that
the prejudices you are showing here would eliminate you from being a
good scientist and test subject.
: As for the results of those listeners who chose not "to use an ABX box
: in the signal path", well... they show even "better" randomness
: (statistical sense) than those from ABX-box switching (but this is
: mean talking). IMO, it's the equalized-blind nature of the listening
: tests, coupled with lowered resolution by not using a good switcher,
: that is the cause of their bad showing... or is it the inaudibility of
: the sonic differences? :)
Since you like these ABX tests so much, why don't you reprint this article
on RAO for us since you're trying to convince everyone of its validity? I
can't very well criticize it if I don't have the full article in front of
me, I can't even discuss it if I don't know what you're leaving out, and
I'm not going to be caught dead trying to run around and buy an issue of
Stereo Review to argue a silly ABX test with you!
Sebastien P. McIntyre
On Mon, 13 Jan 1997, Swanlee wrote:
> How stupid are you you do not think that the sence of smell has
> anything to do with the sence of taste my god this is ridiculous.
How stupid are you you do think that the sence of sight has
anything to do with the sence of hearing my god this is ridiculous.
---> ;-) <--- (Sorry I couldn't resist)
--
,
Sebastien
)
In article <32d9e864....@nntp.ix.netcom.com>,
Swanlee <Swa...@ix.netcom.com> wrote:
>Yes not the sole determinant did I say it was but it does have a
>dramatic impact on your sence of taste.Where in this post did I ever
>say it was the sole determinant but it does have a pretty dramatic
>effect. Hell eat anything while holding your nose then release it and
>you will get a rush of flavor do it yourself. Unless you are blind of
>taste as well as blind of hearing.
Did someone say that Mark Davenport was a college student? That cannot
be correct. I DID say that the sense of smell does affect your perception
of taste, so you're beating a dead horse here. I also said that you
can taste things even if you hold your nose. I did NOT say that you said
that smell was the sole determinant of taste. Is that clear?
"Blind of taste as well as blind of hearing?" People who can't hear
at all are usually referred to as "deaf." "Deafness" refers to the
condition of not being able to hear anything. "Deaf" and "deafness"
are definitely useful words to know, Mark.
>>I was not making an analogy. You simply said "Try doing a blind taste
>>test while holding your nose shut." Your tongue has chemical receptors
>>that can identify salt and sugar (this forms the basis for our sense of
>>taste, and the olfactory receptors significantly enhances it), and so
>>I offered a clear example of a blind taste test where tastes could be
>>identified with an impaired sense of smell. You did not say "Try doing
>>a blind taste test that is analogous to listening to music with two
>>level-matched components. This test should be done while holding your
>>nose shut."
>
>It was in context with an abx disscussion if you are to linear to
>realize that then to bad.
The comparison is simply not valid. You yourself pointed out the problem
of level-matching in taste tests. How do you level-match two colas? Do
you use the sugar concentration? How do you level-match two beers?
Alcohol concentration? Obviously, you can't use taste tests as a valid
analogy to ABX testing.
Furthermore, while there is a great deal of empirical evidence that the
sense of smell enhances the sense of taste (when you have a cold, for
example), there is no such evidence that vision enhances hearing.
Instead, the evidence actually suggests that visual stimuli can
*distract* the brain.
>>I think Armand wrote about the significantly higher sugar level of Pepsi
>>as a key to distinguishing between and positively identifying Coke and
>>Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
>>I'm willing to try this myself, and I'll report on the results later this
>>week when I get the other guys from work to try it as well, if people
>>are really interested. I'd be happier trying Bass and Guinness in a blind
>>taste test with my nose closed, and I'm willing to bet money that I'm
>>able to do so. Easy money for me!
>>
>How stupid are you you do not think that the sence of smell has
>anything to do with the sence of taste my god this is ridiculous.
You obviously have no reading comprehension skills. I said that the sense
of smell significantly enhances our sense of taste, but we can definitely
taste things even if we hold our noses shut! In the simplest terms, here
it is: SMELL AFFECTS TASTE, BUT YOU CAN TASTE LOTS OF THINGS EVEN WHEN
YOU HOLD YOUR NOSE SHUT. IF I TOOK A BLIND TASTE TEST AND HELD MY NOSE
SHUT, I WOULD STILL BE ABLE TO ACCURATELY DISTINGUISH BETWEEN SALT AND
SUGAR OR EVEN BETWEEN TWO BEERS LIKE BASS AND GUINNESS. I WOULD BE WILLING
TO BET MONEY ON MY ABILITY TO DO SO.
Is that clear enough now? Perhaps if I said it again for you, you might
understand my position: You can taste things even if you hold your
nose shut but not as well. However, even if I close my eyes and hold
my nose shut, I can tell whether I'm drinking Bass or Guinness beer.
If you don't believe me, then I'm willing to bet you money that I can
do it.
I want a retraction of your statement "you do not think that the sence (sic)
of smell has anything to do with the sence (sic) of taste" since I never
said anything to that effect. And after that, I hope we can stop talking
about smell and taste in an audio newsgroup. This should be my last post
on smell and taste, unless you continue to distort reality or misrepresent
my position.
*sigh*
Young-Ho
)
>>the sole determinant of the sense of taste. I'm glad that you're reading
>>6th grade science books, though. It's a good start.
>
>dramatic impact on your sence of taste.Where in this post did I ever
>say it was the sole determinant but it does have a pretty dramatic
>effect. Hell eat anything while holding your nose then release it and
>you will get a rush of flavor do it yourself. Unless you are blind of
>taste as well as blind of hearing.
In my previous post, I said that the sense of smell significantly
enhances the sense of taste. I think we're in perfect agreement there,
Mark, and I wish you would stop misrepresenting my position and beating
this dead horse. I also said that you can taste some things even if
you can't smell them. I NEVER said that you said that smell was the sole
determinant of tate. Is that clear?
The word for "blind of hearing" is "deaf." "Deafness" is the condition
of severely impaired hearing. You probably knew that already, but I'm
just making sure.
>>identified with an impaired sense of smell. You did not say "Try doing
>>a blind taste test that is analogous to listening to music with two
>>level-matched components. This test should be done while holding your
>>nose shut."
>
>It was in context with an abx disscussion if you are to linear to
>realize that then to bad.
Too linear? Taste tests are simply not valid analogies to ABX testing.
How does you level-match flavors? When comparing two colas, do you
dilute one cola until they have the same sugar concentration? When
comparing two beers, do you dilute one beer until they have the same
alcohol concentration?
Furthermore, there is a great deal of empirical evidence that the
sense of smell enhances the sense of taste (just think of the last
time you were congested), but there is no such evidence that seeing
enhances hearing. In fact, the evidence suggests that visual stimuli
*distracts* the brain.
I don't think that you can make a valid comparison between ABX testing
and blind taste tests where your nose is held shut for the above reasons.
Also, opponents of ABX testing often claim that everything always sounds
the same in ABX tests (despite much evidence to the contrary), but
everything does not taste the same in blind taste tests where your nose
is held shut. I hope my position is absolutely clear now, so that we
can stop arguing about nothing.
>How stupid are you you do not think that the sence of smell has
>anything to do with the sence of taste my god this is ridiculous.
I never wrote anything to indicate that I "do not think that the sence (sic)
of smell has anything to do with the sence (sic) of taste," and I want
you to retract that statement since it is clearly false. In case I
haven't made myself clear by now, I'll state my position again:
Smells affects your perception of taste, but you can still taste things
even if you can't smell them. For example, I am willing to bet money that
I can accurately identify Guinness and Bass beer in a blind taste test
with my nose held shut. I think that blind taste tests are a bad analogy
for ABX testing because of the problems with level-matching. Blind taste
tests with your nose held shut are an even worse analogy because smell
directly affects taste but vision does not affect hearing in the same way
(if at all).
I hope we can stop talking about smell and taste now. I don't know what
kind of educational background or learning disabilities you have, but
I think that I've explained myself as clearly as I can. "This *is"
ridiculous" (my emphasis on "is") but not for the reasons you think.
Young-Ho
Steve Collier
Swanlee wrote:
>
> On 11 Jan 1997 21:43:37 GMT, "Sebastien P. McIntyre"
> <s...@meca.polymtl.ca> wrote:
> The only ignorance is thinking abx actually proves anything. What do
> precise test conditions the people ...] could not tell
> differences in said equipment. It has no and I repeat no application
> outside of all of the factors.
Oh - similar to the buyer in a shop who says he likes the sound of that?
He gets home and the purchase is invalid?
> You cannot take those results and apply
> them to anything outside of all of the factors mentioned here cause a
> change in one of those factors could completely change the result of
> the test.
Well that applies equally to a non-abx preference too. Abx is not the
sole sufferer from general problem of induction or generalisability. Put
it another way - if one sent two identical twins out to research CD
players - The first to do listening tests under specified and controlled
conditions and the second to make judgements under unknown conditions,
and they came back with their recommendations, whose answer would you
trust more if you had to bet on it?
Ans: the second because I didn't try to specify the exact conditions for
his judgement, so with a bit of luck his findings will apply everywhere.
Yeah, sure. Roll over Bacon.
--
OECD Halden reactor project ^ fax: +47 6918 7109
Postboks 173, @ /|\ tel: +47 6918 3100
1751 Halden, _< \, '/|\` http://www.ife.no/
Norway_________________________(*)>(*)_______' | `_____I pre-fir cycling
Alan Derrida
\ (you...@larmor.nmr.mgh.harvard.edu) wrote:
>I think Armand wrote about the significantly higher sugar level of Pepsi
>as a key to distinguishing between and positively identifying Coke and
>Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
>I'm willing to try this myself, and I'll report on the results later this
>week when I get the other guys from work to try it as well, if people
>are really interested. I'd be happier trying Bass and Guinness in a blind
>taste test with my nose closed, and I'm willing to bet money that I'm
>able to do so. Easy money for me!
Swanlee at (Swa...@ix.netcom.com) wrote:
: How stupid are you you do not think that the sence of smell has
: anything to do with the sence of taste my god this is ridiculous.
Listen Young-Ho, I hate to see these things go on to ridiculous lengths
because the discussion breaks down as someone goes off on a tangent to
pursue on irrelevant line of logic. I understood Swanlee's point fine and
I'm sure you're not "too linear" to understand it as well, so allow me to
explain it to you before you needlessly get everyone at your work drunk on
imported beer.
Swanlee gave an example where sense of taste can be affected when your
sense of smell is impaired. This is an old trick he described and a well
known fact. Regardless of whether this will always occur no matter how you
conduct the test is not relevant to the point being made by this analogy.
The analogy is that when you eliminate one of your senses in an ABX test,
namely your sense of visual reference to what you are listening to, it may
impair one of your other senses, namely your sense of hearing. Compris?
Ok, now let's take it from here, go out there and have a good debate!
Swanlee
On 17 Jan 1997 03:51:18 GMT, "Gregory T Freitag"
<gfre...@northnet.org> wrote:
>
>
>Alan Derrida <derr...@CAM.ORG> wrote in article
><5bl5be$r...@ocean.CAM.ORG>...
>> \ (you...@larmor.nmr.mgh.harvard.edu) wrote:
>> >I think Armand wrote about the significantly higher sugar level of Pepsi
>> >as a key to distinguishing between and positively identifying Coke and
>> >Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
>> >I'm willing to try this myself, and I'll report on the results later
>this
>> >week when I get the other guys from work to try it as well, if people
>> >are really interested. I'd be happier trying Bass and Guinness in a
>blind
>> >taste test with my nose closed, and I'm willing to bet money that I'm
>> >able to do so. Easy money for me!
>> >Young-Ho
>>
>> Swanlee at (Swa...@ix.netcom.com) wrote:
>> : How stupid are you you do not think that the sence of smell has
>> : anything to do with the sence of taste my god this is ridiculous.
>
>I knew a guy who lost his sense of smell in a camping misfortune. His sense
>of taste was almost nonexistent.
>
>> The analogy is that when you eliminate one of your senses in an ABX test,
>> namely your sense of visual reference to what you are listening to, it
>may
>> impair one of your other senses, namely your sense of hearing.
>
>I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
>better find a better analogy.
>
> Greg
After a persons brain has time to adjust and rewire the connestions
when he is blind his hearing will actually get better. This is becuase
of the unused space that still works which was previously used in
seeing but now has been rewired and refocused on the other sences thus
enhancing them. But in a short period of time before the person can
get use to it it impairs the sences. We all agree that human sences
have many flaws why would it work any better in short term conditions
it is not use to working in.
Rick Yerke
In article <32dfa9c9...@nntp.ix.netcom.com>, Swa...@ix.netcom.com
says...
distractions and you will concentrate more on sounds rather than having to
concentrate on both sight and sound.It`s like closing your eyes when you
want to hear more of the music.
--
Rick Yerke
42 Williams St.
P.O. Box 392
Moscow Pa. 18444 Phone (717)842-4857 ye...@csrlink.net
USA
)
In article <01bc0426$9a434720$255ee8cd@gfreitag>,
Gregory T Freitag <gfre...@northnet.org> wrote:
>I knew a guy who lost his sense of smell in a camping misfortune. His sense
>of taste was almost nonexistent.
>I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
>better find a better analogy.
Anecdotal evidence is problematic, but your anecdote does not contradict
the fact that the sense of smell significantly enhances but not solely
constitute the basis for the sense of taste. What sort of misfortune
befell your friend?
The problem with Mark Davenport's assertions are that they are fundamentally
unscientific, in the classic Popperian sense. They cannot be tested.
It is impossible for us to verify that a listener can identify or even
distinguish between two components when s/he already knows the identity
of the components, but when listeners can distinguish between two components
under double-blind conditions with statistical significance, we can
say with reasonable confidence (say, 95%) that the listener is able
to do so (actually, this is a simplification for brevity's sake). We
cannot make such a statement with any confidence whatsover under sighted
conditions.
In short, it is impossible to verify the existence of a phenomena whose
existence cannot be verified.
I might also point out that Mark Davenport seemed to indicate that
he believes in a Lamarckian view of evolution, a view that was shown to
be inconsistent with empirical evidence a century ago.
Lastly, science is not about making analogies. Direct current
can be compared to water flowing through a hose, but a knowledge of
hydrodynamics does not translate through analogy into a knowledge
of electricity and magnetism.
Young-Ho
Gregory T Freitag
Alan Derrida <derr...@CAM.ORG> wrote in article
<5bl5be$r...@ocean.CAM.ORG>...
> \ (you...@larmor.nmr.mgh.harvard.edu) wrote:
> >I think Armand wrote about the significantly higher sugar level of Pepsi
> >as a key to distinguishing between and positively identifying Coke and
> >Pepsi. That was the reason why Pepsi "won out" in blind taste tests.
> >I'm willing to try this myself, and I'll report on the results later
this
> >week when I get the other guys from work to try it as well, if people
> >are really interested. I'd be happier trying Bass and Guinness in a
blind
> >taste test with my nose closed, and I'm willing to bet money that I'm
> >able to do so. Easy money for me!
> >Young-Ho
>
> Swanlee at (Swa...@ix.netcom.com) wrote:
> : How stupid are you you do not think that the sence of smell has
> : anything to do with the sence of taste my god this is ridiculous.
I knew a guy who lost his sense of smell in a camping misfortune. His sense
of taste was almost nonexistent.
> The analogy is that when you eliminate one of your senses in an ABX test,
> namely your sense of visual reference to what you are listening to, it
may
> impair one of your other senses, namely your sense of hearing.
I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
better find a better analogy.
Greg
Stewart Pinkerton
Swa...@ix.netcom.com (Swanlee) writes:
>After a persons brain has time to adjust and rewire the connestions
>when he is blind his hearing will actually get better. This is becuase
>of the unused space that still works which was previously used in
>seeing but now has been rewired and refocused on the other sences thus
>enhancing them. But in a short period of time before the person can
>get use to it it impairs the sences. We all agree that human sences
>have many flaws why would it work any better in short term conditions
>it is not use to working in.
Pay more attention to the neurophysiology lectures Mark! Taste and smell
are linked senses, sight and hearing are not. In any event, the whole
thing is ridiculous and irrelevant because the listeners in a blind test
have full use of their visual sense, it's just that they don't know
which unit is actually playing of those in view.
--
Stewart Pinkerton | If you can't measure what you're making,
A S P Consulting | how do you know when you've got it made?
(44) 1509 880112 |
"I canna change the laws o' physics" - the other Scotty
Ray Knuth
Swanlee wrote:
>
> On 17 Jan 1997 18:33:22 -0500, you...@larmor.nmr.mgh.harvard.edu (\)
> wrote:
>
> >In article <01bc0426$9a434720$255ee8cd@gfreitag>,
> >Gregory T Freitag <gfre...@northnet.org> wrote:
> >>of taste was almost nonexistent.
> >>I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
> >>better find a better analogy.
> >
> >the fact that the sense of smell significantly enhances but not solely
> >constitute the basis for the sense of taste. What sort of misfortune
> >befell your friend?
> >
> >The problem with Mark Davenport's assertions are that they are fundamentally
> >unscientific, in the classic Popperian sense. They cannot be tested.
> >It is impossible for us to verify that a listener can identify or even
> >distinguish between two components when s/he already knows the identity
> >of the components, but when listeners can distinguish between two components
> >under double-blind conditions with statistical significance, we can
> >say with reasonable confidence (say, 95%) that the listener is able
> >to do so (actually, this is a simplification for brevity's sake). We
> >cannot make such a statement with any confidence whatsover under sighted
> >conditions.
And a conclusion to this could be
> that not being able to fix the point of sound and other abx condtions
> impairs the senses to the point it is nearly impossible to deteremin
> edifferences even if they are really their under nomral listening
> conditions.
You need to list your documentation and refs, not just stories and
anticdotes.
Ray
Chuck Ross
Swa...@ix.netcom wrote:
> Actaully its pretty simple. Do an abx test on equipment that
> dramatically measures differently see what happens. If he can't guess
> which is which the test is invalid because the test impairs the
> ability to discern differences in the abx condition when they were
> emasured to have had real audible differences.Actaully it was quoted
> here that some people individually lsitened to some amp wortoe donw
> their differences it turned out the people without talking to each
> other described basica;lly the same differences but under abx
> conditions could not discern any differences. yuo can't say that the
> apearance of each amp caused hese people to describe the same
> differences and charactoristics can you?That some how these amps
> produced a subliminal singla that triggered a brain reaction which was
> the same in each person.Seeing these people heard the same differecnes
> alone sighted but could not tell the differences in the abx test then
> something is wrong with the test.And a conclusion to this could be
> that not being able to fix the point of sound and other abx condtions
> impairs the senses to the point it is nearly impossible to deteremin
> edifferences even if they are really their under nomral listening
> amps have no differences which is what is stated quite often because
> oif abx results. Just because they did not prove positive in the test
> does not mean that audible differences do not exsist in the enviroment
> they were designed for and under long term circumstances as is shown
> in this example. But it has gottren to the pouint who really gives a
> damn you can do all the number crunching all the test you want what
> matters to most is long term enjoyment of audio whihc the abx deos not
> take into account.You can get off the subject by this and that but
> bottom line the test does nothing to improve the average consumers
> system.
Well, I'm certainly glad we got that problem solved.
Next problem, please!
--
Swanlee
>In article <01bc0426$9a434720$255ee8cd@gfreitag>,
>Gregory T Freitag <gfre...@northnet.org> wrote:
>>I knew a guy who lost his sense of smell in a camping misfortune. His sense
>>of taste was almost nonexistent.
>...
>>I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
>>better find a better analogy.
>
>Anecdotal evidence is problematic, but your anecdote does not contradict
>the fact that the sense of smell significantly enhances but not solely
>constitute the basis for the sense of taste. What sort of misfortune
>befell your friend?
>
>The problem with Mark Davenport's assertions are that they are fundamentally
>unscientific, in the classic Popperian sense. They cannot be tested.
>It is impossible for us to verify that a listener can identify or even
>distinguish between two components when s/he already knows the identity
>of the components, but when listeners can distinguish between two components
>under double-blind conditions with statistical significance, we can
>say with reasonable confidence (say, 95%) that the listener is able
>to do so (actually, this is a simplification for brevity's sake). We
>cannot make such a statement with any confidence whatsover under sighted
>conditions.
>
Actaully its pretty simple. Do an abx test on equipment that
Gregory T Freitag
Swanlee <Swa...@ix.netcom.com> wrote in article
<32e0201c...@nntp.ix.netcom.com>...
>Actaully it was quoted
> here that some people individually lsitened to some amp wortoe donw
> their differences it turned out the people without talking to each
> other described basica;lly the same differences
Way too many variables.
>but under abx
> conditions could not discern any differences. yuo can't say that the
> apearance of each amp caused hese people to describe the same
> differences and charactoristics can you?
No, there is a lot more to it than that, do you think David Copperfield
really made the Statue of liberty disappear? Mr. Copperfield makes it quite
apparent to me that even when we don't want to be fooled, we still are.
James Randi (a magician) does a type of blind test for people who claim to
be psychic. He knows how to not give out information, he understands the
type of communication that can occur under non-blind conditions. The
problem with your argument is you assume there was no communication. Why do
you think the tests have to be "double blind"?
>...could not tell the differences in the abx test then
> something is wrong with the test.
When I was in college James Randi did a show. After the show he held a
small gathering to show us some of the tricks of the so called psychics.
They were fairly simple once they were explained. The key is in the
illusion, the sighted tests are wide open to this kind of influence, there
is a lot more wrong with sighted tests
> But it has gottren to the pouint who really gives a
> damn you can do all the number crunching all the test you want what
> matters to most is long term enjoyment of audio whihc the abx deos not
> take into account.You can get off the subject by this and that but
> bottom line the test does nothing to improve the average consumers
> system.
There is information to be gathered from these tests, If your going to
design something you have to know what is important and what is not. The
bottom line is that these tests are research which is used by engineers to
improve the products. Of course this translates to improved systems for the
consumer.
Greg
Swanlee
On Fri, 17 Jan 1997 20:56:43 -0800, Ray Knuth
<Rayk...@sprintmail.com> wrote:
>Swanlee wrote:
>>
>> On 17 Jan 1997 18:33:22 -0500, you...@larmor.nmr.mgh.harvard.edu (\)
>> wrote:
>>
>> >In article <01bc0426$9a434720$255ee8cd@gfreitag>,
>> >Gregory T Freitag <gfre...@northnet.org> wrote:
>> >>I knew a guy who lost his sense of smell in a camping misfortune. His sense
>> >>of taste was almost nonexistent.
>> >...
>> >>I also knew a blind drummer, it didn't effect his hearing at all, Swanlee
>> >>better find a better analogy.
>> >
>> >Anecdotal evidence is problematic, but your anecdote does not contradict
>> >the fact that the sense of smell significantly enhances but not solely
>> >constitute the basis for the sense of taste. What sort of misfortune
>> >befell your friend?
>> >
>> >The problem with Mark Davenport's assertions are that they are fundamentally
>> >unscientific, in the classic Popperian sense. They cannot be tested.
>> >It is impossible for us to verify that a listener can identify or even
>> >distinguish between two components when s/he already knows the identity
>> >of the components, but when listeners can distinguish between two components
>> >under double-blind conditions with statistical significance, we can
>> >say with reasonable confidence (say, 95%) that the listener is able
>> >to do so (actually, this is a simplification for brevity's sake). We
>> >cannot make such a statement with any confidence whatsover under sighted
>> >conditions.
>> >snip snip
>And a conclusion to this could be
>> that not being able to fix the point of sound and other abx condtions
>> impairs the senses to the point it is nearly impossible to deteremin
>> edifferences even if they are really their under nomral listening
>> conditions.
>
>You need to list your documentation and refs, not just stories and
>anticdotes.
>
>Ray
>
Why should I those will just be attacked and deamed unworthy no matter
what is said or who said it.Just like those abx tests that show
siginifcant differences are pulled part for errors but ones tha show
no difference are let by and put in the abx bible. Your type wopn't
believe anything other than that which suits your needs and purpose in
the abx realn instead of reality.Fine but I find the tests invalid and
will not join in on them.Waste your time all you want trying to
factually prove what sounds like what. While ill actually be at home
listening to my sytem instead of a lab with abx boxes around.Thier
comes a time whenyour done trying things out or testing and actually
need to sit down and enjoy what you have bought or otherwise it is
worthless.
Why don't you people start an abx group to talk about this crap?
Arnold B. Krueger
Swanlee <Swa...@ix.netcom.com> wrote in article
> Your using science against itself by trying to factually prove
something that cannot and never will be factually proved outsiode of the
test conditions it just can't be done. ABX tests are only
> relavanet to the people used in the test the amps, cable, speakers,
abxbox, music listened in the test the exact conditions of the room and
the hearing of each individual person and how they flet that day. You
cannot apply the test results to anything outside of these factors it
simply doesn't work.
If I apply this same reasoning a tiny bit more globally, I conclude that
there is no purpose to testing because a test cannot occur under conditions
exactly identical to intended use later on. Therefore, its unreasonable for
anyone to reply to questions on rec.audio.opinion about what something
sounds like, because any who respond cannot duplicate the exact operational
conditions of the person asking the question.
>You can prove at that time that person using that amp with that cable
with those speakers
listening to that music on that day when that person felt exactly the way
they did at the precise moment of the test. But those conditions will
never be reaptable again and I mean those conditions any variuation could
change the results.
Then why bother asking anybody else what they think of a component? Its
obvious that their opinion will not be formed under exactly identical
conditions to the ones you intend to use the component under.
> You can't apply it to real world home listening so why the hell bother.
Then why do the high end magazines bother to publish their opinions? Its
obvious that their tests don't involve the same persons, program material
and associated components as the ones you use at home.
> It doesn't prove anything
Depends on what degree of proof you require, eh?
>it doesn't improve anything and violates the scienitifc range rule to
boot.
Therefore, according to your logic, Rec.audio.opinion should cease
operation since all the information it publishes about how audio equipment
sounds is at least as invalid as ABX tests.
> What I want is for 20 people go into a room individually with two amps
give them an hour to play with the amps all they want listen to what ever
music they want then individually write down what they thought the
differences were. Take what they wrote by themsleves while listening
by themsleves and if they compare pretty well as in they heard the same
differences have them do the abx test. If they can't tell any differences
in abx then Id say this is pretty good proof the abx test is flawed when
humans are involved.
In other words, you are saying that if people think they can hear audible
differences when they know the identity of the equipment, but can't hear
audible differences when the identity of the equipment is concealed, then
the problem is that the identity of the equipment was concealed from them;
not that they were imagining that they heard differences when they knew the
identity of the equipment.
Next thing you are going to tell me that math tests in school are invalid
because they are administered blind.
Enjoy that dream world!
Kevin Connery
In article <01bc04fd$55b39240$125ee8cd@gfreitag>,
Gregory T Freitag <gfre...@northnet.org> wrote:
>Swanlee <Swa...@ix.netcom.com> wrote in article
><32e0201c...@nntp.ix.netcom.com>...
>
>
>>Actually it was quoted here that some people individually listened
>>to some amp wrote down their differences it turned out the people
>>without talking to each other described basically the same differences
>
>Way too many variables.
Please expand on this. As one of the posters who described such a listening
session (over New Years, as part of my <sigh> ongoing amplifier auditions.
I'm curious as to how you would improve the tests--without, if possible,
the addition of equipment I neither have nor have access to.
There were three listeners. We played the same pieces of music through
each amp once, then gave each listener the remote and let them adjust
volume to taste, and played the cuts as often as they liked through
each amp, switching cables from one amp to the next as requested. NONE
of us said anything about the way any given amp sounded.
Now, while I agree that body language can give away preferences, I
am at a complete loss to understand how body language could convey
the item that one of the amps had a wider soundstage than the others,
while having less bass. Or that another had no soundstaging or imaging
to speak of. (Each one is 'less good', and were that the only criteria
stated, it could be so conveyed. But the DETAILS were written down, and
the bloody things matched!)
>The key is in the
>illusion, the sighted tests are wide open to this kind of influence, there
>is a lot more wrong with sighted tests
I see. Each amplifier projected the 'illusion' that they sounded the way
they did, and all the listeners picked up that illusion. As long as that
amp continued to project that 'illusion' after purchase I wouldn't care
in the slightest that it wasn't a 'real' difference, and only an
'illusion'. But, you know, the only way the amps had to communicate at
all (save for their appearance, and I do *not* believe that is sufficient
to cast such a 'spell') was through their *sound* -- and as such, any
illusion so cast was not an illusion, but part of the way the amp
sounded--no hands to have up their sleeves, no presto, and no magic rabbit
in a magic hat.
--kdc
Art Sackman
Arnold B. Krueger <ar...@pop3.concentric.net> wrote in article
<01bc0673$ee0bfce0$f176...@crc3.concentric.net>...
>
>
> Swanlee <Swa...@ix.netcom.com> wrote in article
> > Your using science against itself by trying to factually prove
> something that cannot and never will be factually proved outsiode of the
> test conditions it just can't be done. ABX tests are only
> > relavanet to the people used in the test the amps, cable, speakers,
> abxbox, music listened in the test the exact conditions of the room and
> the hearing of each individual person and how they flet that day. You
> cannot apply the test results to anything outside of these factors it
> simply doesn't work.
>
> If I apply this same reasoning a tiny bit more globally, I conclude that
> there is no purpose to testing because a test cannot occur under
conditions
> exactly identical to intended use later on. Therefore, its unreasonable
for
> anyone to reply to questions on rec.audio.opinion about what something
> sounds like, because any who respond cannot duplicate the exact
operational
> conditions of the person asking the question.
>
You seem to have a conceptual problem based upon a stubborn paradigm of
of thought. Not everything is cut and dried, A or B. It seems you perceive
knowledge to be only
that which has been proven. There ARE shades of gray.
The answer is that this hobby relies on part art and part science. I have a
healthy respect for science,
but I do not worship it as religion and proclaim it the one and only true
way. Its
limitations must be acknowledged, so that its results can be correctly
interpretated, so that
we do not state some grand generalization is proven, when the only purpose
was
to test for differences between two particular DUT's
> >You can prove at that time that person using that amp with that cable
> with those speakers
> listening to that music on that day when that person felt exactly the
way
> they did at the precise moment of the test. But those conditions will
> never be reaptable again and I mean those conditions any variuation could
> change the results.
>
> Then why bother asking anybody else what they think of a component? Its
> obvious that their opinion will not be formed under exactly identical
> conditions to the ones you intend to use the component under.
There is no cut and dried answer. Subjective evaluation is opinion, not
fact. Opinions
can change.
>
> > You can't apply it to real world home listening so why the hell
bother.
>
> Then why do the high end magazines bother to publish their opinions? Its
> obvious that their tests don't involve the same persons, program material
> and associated components as the ones you use at home.
>
Can't you tell the difference between an opinion and a fact.
Don't you read the part of the review where they list or describe
the associated equipment and listeneing environment?
>
> > It doesn't prove anything
>
> Depends on what degree of proof you require, eh?
Evidently tou accept nothing less than 100%
Good luck
>
> >it doesn't improve anything and violates the scienitifc range rule to
> boot.
>
> Therefore, according to your logic, Rec.audio.opinion should cease
> operation since all the information it publishes about how audio
equipment
> sounds is at least as invalid as ABX tests.
You really do have that paradigm problem.
You are stuck on absolute validity.
Everything with you must be absolute.
Relax, its alright.
There is nothing wrong with expressing or having an opinion.
>
> > What I want is for 20 people go into a room individually with two amps
> give them an hour to play with the amps all they want listen to what
ever
> music they want then individually write down what they thought the
> differences were. Take what they wrote by themsleves while listening
> by themsleves and if they compare pretty well as in they heard the same
> differences have them do the abx test. If they can't tell any
differences
> in abx then Id say this is pretty good proof the abx test is flawed when
> humans are involved.
>
> In other words, you are saying that if people think they can hear audible
> differences when they know the identity of the equipment, but can't hear
> audible differences when the identity of the equipment is concealed, then
> the problem is that the identity of the equipment was concealed from
them;
> not that they were imagining that they heard differences when they knew
the
> identity of the equipment.
The abx test, as usually practised, does more than conceal the identity of
the equipment.
It introduces an additional piece of equipment.
It introduces the authority figure of the tester.
It introduces a lack of control over the switching of the equipment.
It introduces lack of control over the selection of music.
It is performed outside of the listeners usual listening environment
It is conducted under time constraints not usually associated with typical
music listening
It is ironic that in the attempt to eliminate the one variable of listener
knowledge of the devices,
it introduces six times as many variables that are NOT conistent with usual
listening practices.
>
> Next thing you are going to tell me that math tests in school are invalid
> because they are administered blind.
The problem is , you have no coneption of the proper role of science in
this endeavor,
and no knowledge of the limits of scientific inquiry.
Math tests in school have nothing to do with abx testing.
>
> Enjoy that dream world!
>
Chuck Ross
As most readers on this forum know, I'm anything but an AB/X proponent, but
this post has to be answered;
" Art Sackman" <idk...@smart.net> wrote:
> The abx test, as usually practised, does more than conceal the identity of
> the equipment.
> It introduces an additional piece of equipment.
Most testees believe that the equipment does not color the sound in the
slightest and also believe that the AB/X box is much more convenient than
cable swaps, after listening THRU the AB/X box.
> It introduces the authority figure of the tester.
I think I'll pass on this one; if you fear the tester, that's your problem.
> It introduces a lack of control over the switching of the equipment.
Not at all. The testee has complete control over switching A or B; it's
X that is unknown.
> It introduces lack of control over the selection of music.
Bourjois! You can play whatever music you want to, or repeat any portion
of it that turns you on.
> It is performed outside of the listeners usual listening environment
Ridiculous! You can do an AB/X test in your own living room, or wherever
you feel most comfortable.
> It is conducted under time constraints not usually associated with typical
> music listening
Nope...wrong again. You can take as long as you want to listen to one "X"
then another. Months, if you want to.
My problem with AB/X testing involves none of the above. I feel that,
after a given number of DUT swaps (generally a small number, like 5)
it's impossible to tell the difference between one DUT and another,
unless there's a really gross difference. I'm not sure what the reason
for this phenomenon is, but I've experienced it a number of times. I
think the same applies to any type of comparison test, coke-tasting,
coffee, wine, etc. After a few trials, everything tastes the same.
--
Arnold B. Krueger
" Art Sackman" <idk...@smart.net> writes:
>The abx test, as usually practised, does more than conceal the identity of
the equipment.
This is simply not true and has been disproven on this forum many times.
>It introduces an additional piece of equipment.
The sonic transparancy of the ABX switchbox can be established in any test
environment. It has been established in the test environments for which
most test results have been published.
>It introduces the authority figure of the tester.
When the testers are the listeners (the most common case) this is simply
not true and has been disproven on this forum many times.
>It introduces a lack of control over the switching of the equipment.
This is simply not true and has been disproven on this forum many times.
>It introduces lack of control over the selection of music.
This is simply not true and has been disproven on this forum many times.
>It is performed outside of the listeners usual listening environment
This is simply not true and has been disproven on this forum many times.
>It is conducted under time constraints not usually associated with typical
music listening
This is simply not true and has been disproven on this forum many times.
Stop wasting our time with demonstrations of your poor reading
comprehension and/or inattention to this and related forums, as well as
ignorance of scientific literature that has been cited many times..
Arnold B. Krueger
" Art Sackman" <idk...@smart.net> writes:
>You seem to have a conceptual problem based upon a stubborn paradigm of of
thought. Not everything is cut and dried, A or B.
Just another tired, straw man argument.
>It seems you perceive knowledge to be only that which has been proven.
Hardly. I've written here and on RAHE that the study of perception
recognizes two kinds of perception: perceptions that are reliable and those
that are not. The latter are known as illusions. Either kind of perception
is knowlege. Therfore what is not proven is knowlege, but its not yet
reliable knowlege.
Your claim is false - I've said no such thing and I've even contradicted it
publically.
>There ARE shades of gray.
You are preaching that line to a member of the choir. Lets not waste each
other's time.
>The answer is that this hobby relies on part art and part science.
Hardly. I've written here and on RAHE that Audio is generally recognized as
being both an art and a science and therfore my tacit agreement with that
is public record..
Your claim is false - I've said no such thing and I've even contradicted it
publically.
>I have a healthy respect for science, but I do not worship it as religion
and proclaim it the one and only true way.
You are preaching that line to a member of the choir. Lets not waste each
other's time.
>Its limitations must be acknowledged, so that its results can be correctly
interpretated, so that we do not state some grand generalization is proven,
when the only purpose was to test for differences between two particular
DUT's
That the two DUT's are different is a trivial observation. Of course they
are different - they occupy two different spaces,. The interesting question
is whether they sound different.
>There is no cut and dried answer. Subjective evaluation is opinion, not
fact. Opinions can change.
Opinions about sound quality are qualified by establishing whether or not
the person with the opinion can really tell that there is an audible
difference between the DUT's. If you can't hear a difference, what is the
validity of your opinion?
>Can't you tell the difference between an opinion and a fact.
Actually, all opinions are facts. That you have a certain opinion is a
fact. The real question is whether your opinion is valuable or worthless.
If you have an opinion about a sound quality difference that you can't
actually prove you hear, then I think your opinion is pretty worthless.
Art Sackman
Chuck Ross <ckr...@enteract.com> wrote in article
<ckross-2101...@belial-42.d.enteract.com>...
> As most readers on this forum know, I'm anything but an AB/X proponent,
but
> this post has to be answered;
>
> " Art Sackman" <idk...@smart.net> wrote:
> > It introduces an additional piece of equipment.
>
> Most testees believe that the equipment does not color the sound in the
> slightest and also believe that the AB/X box is much more convenient than
> cable swaps, after listening THRU the AB/X box.
But it nevertheless adds another piece of equipment and associated cables,
which may
make a difference to some listeners.
>
> > It introduces the authority figure of the tester.
>
> I think I'll pass on this one; if you fear the tester, that's your
problem.
It may affect others.
Do you have someone looking over your shoulder when you listen to music at
home?
>
> > It introduces a lack of control over the switching of the equipment.
>
> Not at all. The testee has complete control over switching A or B; it's
> X that is unknown.
But this was not the case in all previous tests.
It may be the case in a test YOU design, but it may not have be the case
in tests done by others
>
> > It introduces lack of control over the selection of music.
>
> Bourjois! You can play whatever music you want to, or repeat any portion
> of it that turns you on.
Again, you miss the point.
Yes, of course YOU can, but that is not to say
that was the case in all tests that have been or will be conducted.
>
> > It is performed outside of the listeners usual listening environment
>
> Ridiculous! You can do an AB/X test in your own living room, or wherever
> you feel most comfortable.
But the ones conducted in the past that I have read about WERE NOT. They
involved panels of listeners
who all listened to the SAME music on the SAME equipment in the SAME
environment. Well, in the real world, outside of the artificial
tst environment, all the testees did not live in the same house, or usually
listen
to the same music, or all listen to exactly the same equipment.
You are talking about how a theoretical test condition can be set up
in your home, and you are correct, within those limitations.
But if the purpose is to test a GROUP of people, so that a body of
verifiable and
statistically significant data can be derived, then the test parameters
have to be
the same for each person and the "designer" abx test you refer to is not
possible.
The problem is that the objectivists have a need for proof. That involves
testing
a number of people under controlled and identical situations.
That is where my comments as to the artificiality and dissimilarity of
testing
environment vs usual environment apply.
> > It is conducted under time constraints not usually associated with
typical
> > music listening
>
> Nope...wrong again. You can take as long as you want to listen to one "X"
> then another. Months, if you want to.
Again, you misinterepreted.
A test with one testee CAN be conducted as you say.
A test with a panel of testees CANNOT.
It is such tests involving panels that the objectivists base their "all
sounds the same"
mantra upon.
>
> My problem with AB/X testing involves none of the above. I feel that,
> after a given number of DUT swaps (generally a small number, like 5)
> it's impossible to tell the difference between one DUT and another,
> unless there's a really gross difference. I'm not sure what the reason
> for this phenomenon is, but I've experienced it a number of times. I
> think the same applies to any type of comparison test, coke-tasting,
> coffee, wine, etc. After a few trials, everything tastes the same.
>
We are agreed here.
I hope you have come to understand that we are really talking about
two different types of abx tests.
One type is the one you, or I, would
take individually in the home.
The other is one that a researcher such as Mr.Noussaine would conduct upon
a
panel, or separately (but identically upon a vriety of individuals) in one
common setting.
Art
Sebastien P. McIntyre
On 14 Jan 1997, Alan Derrida wrote:
> Sebastien P. McIntyre (s...@meca.polymtl.ca) wrote:
>
> On 7 Jan 1997, Alan Derrida wrote:
> > Sebastien P. McIntyre (NA...@music.mus.polymtl.ca) wrote:
> >
> > : (...) The problem is that the results
> > : from both parts of the test do not agree at all (sighted: plenty of
> > : audible differences; blind: no audible differences -- statistical
> > : sense).
> > : How do you explain this?
> >
> > Hmmm... that's a tough one. Did SR happen to use an ABX box in the
> > signal path in the tests in that article?
>
> : Hmmm... Another article le Sieur de Derrida hasn't read?...
> : Otherwise he would already know that listeners were allowed to chose
> : manual cable swap (still with equalized levels and blind calling,
> : naturellement) in any given listening session if they felt this would
> : increase the accuracy of their identifications.
>
> Why do you think I ASKED you about the test McIntyre?? I take it that
> somehow, you didn't understand the question since you're answer
> presumed I should have already read it.
I mistook your question for an answer... Sorry. Let me answer your
question this way: How the presence of an element (The Box) that has
been found audibly transparent in the most stringent listening tests
could make the results of sighted and blind tests so much different?
Please tell us.
> What do you think I do - sit
> at home and clip ABX articles out of magazines, line them up on a
> corkboard and compare each for research in a debate somewhere?
I obviously don't think so.
> Unlike
> you, I would rather use that time to actually listen to music. You know...
> ecouter la musique??
What, you do neither of the two???
I would like to listen to la musique more often than the actual
once-a-week but my research work/TA commitments (as well as my
girlfriend's odd schedule of internship) and other activities
prevent this to materialize. Not that Montreal's music scene is
boring, au contraire! So I'm left with a lot of false music at
home and in the lab... you know, shiny discs and two weird sound
boxes?
> I suspect that you have no knowledge whatsoever about what I
> do or do not know about "quasi-instantaneous switching" in ABX trials,
> especially from the one line reply I gave,
From your one-line reply, no, but from the archives (DejaNews), yes.
If your understanding of the matter has changed since last June, I
invite you to post a revised statement superseding those in the
archives, as they _might_ no longer reflect your present viewpoint
about sound juxtaposition in comparative listening... :)
> but I also suspect that
> the prejudices you are showing here would eliminate you from being a
> good scientist and test subject.
"Bad scientist", I concede but "bad test subject", I object fiercely.
The best way to test a protocol is to employ the most prejudiced
subjects you can find. If these subjects' preconceived opinions can be
filtered out of the results in standard control sequences, the protocol
used is definitely worth considering for serious listening comparisons.
> Since you like these ABX tests so much, why don't you reprint this article
> on RAO for us since you're trying to convince everyone of its validity?
This test seems quite valid, judging by the very weak opposition, at
worst, it received over the past 10 years (from The New York Times
to RAO).
> I can't very well criticize it if I don't have the full article in front
> of me, I can't even discuss it if I don't know what you're leaving out,
Should I also read it for you?
> and I'm not going to be caught dead trying to run around and buy an
> issue of Stereo Review to argue a silly ABX test with you!
Well... S**t or get off the pot! :)
--
,
Sebastien Playing: A VHS HiFi recording of The Naked Voice
(a Radio McGill programme). A cappella
music by a San Francisco-based band
called Chanteclair <sp?>. Exxxxquisite!
Wolph
In article <32f5be42....@nntp.ix.netcom.com>,
lucif...@ix.netcom.com wrote:
>Pretty funny
> to, you often make reference to high end scam artists furthering
> falsehoods in audio to get money, but you would not have a job or the
> name recognition you do if it were not for your furthing of your abx
> tests and their results. I seem to think you also have a vested
> interest in abx tests and their everything is the same results for
> personal gain.
No, we'ld still get a big laugh over his editorials in Car Audio
and Electronics. It's really sad some of the uninformed comments,
as well as passing off personal opinion as "facts", he gets away
with.
--
Please remove the "*" infront of my account name, for any coorespondance.
Nousaine
re:
<<The abx test, as usually practised, does more than conceal the identity
of
the equipment.
It introduces an additional piece of equipment.
It introduces the authority figure of the tester.
It introduces a lack of control over the switching of the equipment.
It introduces lack of control over the selection of music.
It is performed outside of the listeners usual listening environment
It is conducted under time constraints not usually associated with typical
music listening
It is ironic that in the attempt to eliminate the one variable of listener
knowledge of the devices,
it introduces six times as many variables that are NOT conistent with
usual
listening practices.>>
This is simply not true. None of the conditions implied and stated are
part of ABX or alternaive controlled listening tests excpet the "authority
figure" part. However, the "administrator" role is also present in every
open listening presentation as well.
Steve Zipser (Sunshine Stereo)
Nousaine wrote:
>
> re:
>
> <<The abx test, as usually practised, does more than conceal the identity
> of
> the equipment.
> It introduces an additional piece of equipment.
> It introduces the authority figure of the tester.
> It introduces a lack of control over the switching of the equipment.
> It introduces lack of control over the selection of music.
> It is performed outside of the listeners usual listening environment
> It is conducted under time constraints not usually associated with typical
> music listening
>
> It is ironic that in the attempt to eliminate the one variable of listener
> knowledge of the devices,
> it introduces six times as many variables that are NOT conistent with
> usual
> listening practices.>>
>
> part of ABX or alternaive controlled listening tests excpet the "authority
> figure" part. However, the "administrator" role is also present in every
> open listening presentation as well.
Hey Tom, it introduces you! Don't complain!
lucifersam
On 3 Feb 1997 08:04:15 GMT, nous...@aol.com (Nousaine) wrote:
>re:
>
><<The abx test, as usually practised, does more than conceal the identity
>of
>the equipment.
>It introduces an additional piece of equipment.
>It introduces the authority figure of the tester.
>It introduces a lack of control over the switching of the equipment.
>It introduces lack of control over the selection of music.
>It is performed outside of the listeners usual listening environment
>It is conducted under time constraints not usually associated with typical
>music listening
>
>It is ironic that in the attempt to eliminate the one variable of listener
>knowledge of the devices,
>it introduces six times as many variables that are NOT conistent with
>usual
>listening practices.>>
>
>This is simply not true. None of the conditions implied and stated are
>part of ABX or alternaive controlled listening tests excpet the "authority
>figure" part. However, the "administrator" role is also present in every
>open listening presentation as well.
So in an abx test when an abx switch box is used you do not consider
that another piece of equipment being brought into the equation. He
also said abx tests as usually practiced most of the quoted tests have
a number of the charatoristics described. Seems kind of funny you are
now slumming with us in RAO. Seems you caught wind your paid shill is
getting beat up and came here to try and make the save. Pretty funny
Chuck Ross
lucif...@ix.netcom.com wrote:
> On 3 Feb 1997 08:04:15 GMT, nous...@aol.com (Nousaine) wrote:
>
> >re:
> >
> ><<The abx test, as usually practised, does more than conceal the identity
> >of
> >the equipment.
> >It introduces an additional piece of equipment.
> >It introduces the authority figure of the tester.
> >It introduces a lack of control over the switching of the equipment.
> >It introduces lack of control over the selection of music.
> >It is performed outside of the listeners usual listening environment
> >It is conducted under time constraints not usually associated with typical
> >music listening
> >
> >It is ironic that in the attempt to eliminate the one variable of listener
> >knowledge of the devices,
> >it introduces six times as many variables that are NOT conistent with
> >usual
> >listening practices.>>
> >
> >This is simply not true. None of the conditions implied and stated are
> >part of ABX or alternaive controlled listening tests excpet the "authority
> >figure" part. However, the "administrator" role is also present in every
> >open listening presentation as well.
>
> So in an abx test when an abx switch box is used you do not consider
> that another piece of equipment being brought into the equation. He
> also said abx tests as usually practiced most of the quoted tests have
> a number of the charatoristics described. Seems kind of funny you are
> now slumming with us in RAO. Seems you caught wind your paid shill is
> getting beat up and came here to try and make the save. Pretty funny
> to, you often make reference to high end scam artists furthering
> falsehoods in audio to get money, but you would not have a job or the
> name recognition you do if it were not for your furthing of your abx
> tests and their results. I seem to think you also have a vested
> interest in abx tests and their everything is the same results for
> personal gain.
I don't believe this is true at all. I do think that Mr. Noussaine did,
at some point, have the amazing revelation that amplifiers were, indeed,
pretty similar-sounding with double-blind testing, and instantly decided
that this was a really good thing to talk and write about. It of course,
removes a lot of the enjoyment of audiophilia and makes people who spend
large amounts of money on equipment whose sound can be duplicated at
1/10th the price feel like idiots, but if that's what he believes,
then that's what he believes. Consumer's Union feels pretty much the
same way, completely ignoring such things as pride of ownership, etc.,
in the quest for saving money for the consumer who can't hear the
differences anyways.
I certainly cannot conceive of the idea that Mr. Noussaine really
doesn't believe in what he says and does what he does for personal
gain, sneaking back home to listen to his multi-thousand-dollar
system playing with really expensive cables. He probably does indeed
use that Yamaha amp and whatever speakers are the best buy, and enjoys
the music.
--
For three days after death, hair and fingernails continue to grow
but phone calls taper off. -- Johnny Carson
email replies to ckr...@enteract.com
Art Sackman
The abx industry is alive and well !!!!
lucifersam <lucif...@ix.netcom.com> wrote in article
<32f5be42....@nntp.ix.netcom.com>...