The Connecticut Cable Caper...
SJMARCY
During one of the wire-war threads at AR, Jon Risch offered to send me a set of
his DIY high end Belden 89259 one meter interconnects with the provision that I
post my findings at AR and over at AOL And at RAHE, why not? . I am doing this
with this post. For the most part I use Radio Shack gold tips of short lengths
for interconnects. I accepted his offer and received the cables
a week or two later after some emails. I also received some of his AES
papers, which I read. (ahem) One of them (AES 3178) involved subjective
listening tests. I was asked to use the cables starting out with a few
manual swap comparisons and then I was to listen to them for a week or
so and then compare some more along the way using his suggested
guidelines. After this I would do various other comparisons.
Here are my results and impressions. I am not drawing a conclusion, but
leave that up to you.. I am happy to do follow up tests using
suggestions which you may offer as well as to answer any questions which
may arise. This is not an AES paper or whatever. Still, I spent quite a
bit of time on this effort. Jon and I emailed each other about the
overview results recently and I have tried to expand upon areas which
may have been unclear and recognize that more testing can and should be
done.
Equipment-wise (Jack G just perked up) I will offer the following. Two
channel, music only material was played on a two channel system. CD
music, no HT or multichannel stuff, mostly jazz, female voice and other
selections. Lots of familiar material. I tried various components out
and settled on a few variations. I thought that one of the more
sensitive setups involved using a CD player with a well regarded
internal DAC connected directly to a pair of respected biamped monitors.
In that case the interconnects were plugged into the back of the active
speakers which were properly set up in the room, although not very far
apart due to the wire length limitations. The other end of the
interconnects were attached directly to the CD player. Volume control
was via the level adustment on each speaker, which was a pain but not a
functional problem as I would set and forget them at various listening
levels. I listened near field and far field. I also used my preamp's DAC
and other interconnects from the preamp to the active speakers OR power
amps and passive speakers. Purist-wise it seemed to me like the first
very direct set up I mentioned had fewer variables since I only had
access to one set of Belden cables.. I avoided reading comments by Risch
and others specifically about the Belden wire so as to be able to form
opinions on my own as much as possible.
Here goes. The wires were: one meter (or three feet) Belden 89259 versus
Radio Shack Gold Tips versus 5-10 year old old blister pack cables which
I grabbed from a box. The first two were individual wires per channel,
the latter was siamesed. Jon supplied nicely terminated, well soldered
Belden. The Belden wire is pretty stiff and has a fairly thin and shiny
look which I considered to be exotic and elegant.
1) Measured differences. I found no significant, measured, functional
differences between them. The Belden and RS gold may have been a hair
better than the blisterpack wire but I thought that none of them had any
glaring problems of any kind. They measured a bit different of course
but I found super small variations not necessarily favoring one or the
other. No surprises. Bear in mind that these are pretty short wires too
and that the input /output impedance characteristics of associated
equipment is a variable too. I used stuff with low output and high input
impedance, which is typical these days. The measurements were done day
one and if I were to repeat them I would do tests which would focus upon
the characteristics noted below. (future followup potential)
2) Manual swap tests, early on, sighted (CD player phrase repeat & music
just playing on) Preference for the Belden. Impression of no bass range
differences. but differences in the mids / highs. More detailed yet
nonharsh. Sharper. Crisp. Focussed Other differences noted but those are
the ones I perceived as most obvious to me on nearly all program
material. Heard them immediately and consistently but not quite 100 % of
the time. Near field / far field. I started out including some old
blister pack wires of 3 foot length and they seemed to sound like the
r-s golds, until one channel's connector became intermittent after lots
and lots of swapping so the three way comparison evolved into a two way
test. Durability during lotsa swaps is not a strong point of very cheap
interconnects. I liked the terminations on the Belden wire better than
the r-s ones, the r-s cable was much less stiff though. One important
future follup up: blister pack wire, since I didn't get very far with
it.
3) A/B switched swaps using a manual switchbox and 6 inches of r-s gold
, sighted, phrase repeat & music played on See above but less obvious.
Seemed more obvious if you just left it in place. At first easy to tell,
after blind tests, harder to tell sometimes impossible. Very cool
subjective experience. Near field / far field. Sometimes I thought wow
sounds great, there's that "snap" and then noticed that the box was set
for the R-S wire. Hmm. (I wear glasses and can't reliably tell what the
across the room box is set for if I take them off.) Some obvious future
follow ups include more box testing and listening to see if it hurts
anything and short bypasses, perhaps 2 inches, perhaps made of 89259.
4) Just listen for a week or so. I gave 'em a few weeks and enjoyed lots
of music. They sound fine. Returning my old wire, sighted.. Let down?
YUP! What happende to the transients? Later, wait a sec. No difference.
After many swaps, come back the next day, forgot which is in, oops
thought the Belden was in place. Near field / far field. Several swaps.
Mixed results though tending to favor the Belden. Hmm. See 2 above for
the blister pack wire too.
5) Blind tests. Manual swaps Coin flip random number generator for
isolated assistant (behind screen, hand signals only) Phrase repeat only
10 to 90 seconds using familiar sure fire material. 7 and 9 correct out
of 16 Coin flips, the same approx totals I think - wasn't 8 heads for
example. (I can list raw data) The blind tests were done after the above
and I tried to use music that would tend to highlight the differences I
had perceived earlier. Felt kinda stressful. Actually, I hated it. Too
slow, hard to remember things. Frustrating. Who turned out the lights,
I'm banging into things! A fascinating experience which I hope some of
you will try out to see what I mean. (See addenda.)
6) Blind tests switched swaps as above. Results: random, 7 and 9 correct
out of two runs of 16 trials. I greatly preferred this type of test
compared to the manual swaps and a number of times felt strongly that I
totally NAILED the selections and did not need much time at all to make
a selection. I was able to hear the Belden and Radio shack before making
a decision as to what I thought the randomly assigned x was in this and
the manual swap cases.. The box makes a quiet mechanical click (no click
thru the speaker) when a button was pushed, so the procedure was to flip
a coin determining what x would be and then push an unused switch
position and then the A or B choice so that I couldn't tell if say A was
the X if the number of clicks would give it away. During many of my
selections I felt very decisive and confident. I had listened to these
wires for countless hours and of course was already very familiar with
the RS cable. Some of the blind test raw data showed mini "streaks"..
3-4 in a row correct (or inversely wrong which also indicates
differences potentially) I don't know if the mini streaks were
significant or not but they did tend to cancel themselves out. They
happen with coin flips too of course. Maybe more testing should be done
to look into the ministreaks.. I'd like to do say a 1-10 forced choice
where I would make a choice and then when I thought that I nailed it
that would be a 9 or 10 for example. Degrees of choice rather than black
and white. Then we could look into the correlations, if any. I recorded
my reasons for choosing but not the DEGREE of my conviction. My reasons
for selecting one or the other as my choice for what X was involved the
same sonic characteristics mentioned above. Perhaps meaningless
technically BUT an important human factor nonetheless.
8) More blind tests, TBD. Any suggestions? More listeners for one, some
sensitivity testing too. Training.
9) After the blind tests, it seemed like there were no longer any
differences in sound during sighted tests, or at least no longer any big
ones and not consistently. Disconcerting. Expectations? Let down?
Reality? Sometimes I felt "there's the improvement or flaw" and found
that the opposite wire was in place in the days following the blind
testing.. A large contrast with the earlier perceptions. No image diffs
found if i R-S on just one side, although I tried this only briefly
after the blind test and on a sighted basis. That might also be a cool
thing to blind test and play around with.
10) It seemed like the more rapid the swap, and the less I knew about
it, the worse my discrimination became, whether a switchbox was used or
not. Very pronounced effect and very, very interesting and really
entertaining to experience. A real doubletake "say what?" type of
sensation. Did the switch box and jumper wire inclusion just gobble up
all the good sound? It sure measured great. And I couldn't tell r-s or
belden apart in sighted tests when using switched one channel
comparisons of say r-s A switch versus r-s B switch or bypass tests and
so forth when I tried to test the "sound" of the switchbox. Again, the
same thing seemed to be happening with manual swapping.
11) I'd like to do some sensitivity tests to determine say blind and
sighted test thresholds of tiny level differences, distortion etc
discrimination. Various JND kinda things to get some positive blind test
results. Time consuming.
Offhand, I think that the "ABXers" would like to consider 12 out of 16
as showing staistically significant differences, which was not achieved
in this case. I am going to suggest that the reader draw their own
conclusions about wire and will try to retest as posters may suggest and
I'll try to answer their questions as best I can. Also, I am having some
minor construction done soon (windows, finally) and when complete I
would like to have a get together at some point if there is any interest
out there and if a balance of subjectivist/objectivist folks are
interested. I'd be happy to test or be tested by any interested parties.
My gut feel is that the Belden/ R-S conclusion depends upon the validity
of blind testing. If its valid then there is no difference, under these
circumstances, in this case, with this equipment and this listener at
the very least. Or perhaps when the sonic differences are nonexistent or
very very subtle, open listening is not very reliable. I used very, very
good equipment in my library and in my dedicated sound proof listening
room for these tests. If blind testing has some as yet unknown fatal
flaw that noone has been able to identify despite strenuous efforts over
many many years, then I think that there may be a difference between
interconnects but its not huge, at least for short lengths. Most
magazines and stores don't seem to be very helpful on a subject such as
wire...see the end of the addenda..I may be able to get permission to
share some or all of the emails I mention.. Also, I am very disturbed by
the pronounced differences in my sighted perceptions before and after
the blind test. The perception of things like speaker sound diffences
seem to remain unaffected by blind tests. Also, maybe the r-s wire is
pretty good stuff. I understand that it is a durable, cheap, low
capacitance wire that has been described by the Audio Critic as all ya
need.. The r-s seemed to be much more durable than the blister pack wire
at the very least and looks much nicer too. I use a lot of RS cable, one
thing I dislike about it is that it can be a bit tight fitting which
makes me nervous when I am plugging and unplugging it into pricey
components. I think that there are more significant differences in
speaker wire than interconnect but that its easy to keep them inaudible
using normal equipment.. Maybe audible interconnect differences (if any)
would be much more audible if the lengths were much, much greater. If so
then I would tend to regard that more as a set up issue rather than a
wire issue. Don't use a 25 foot turntable to preamp connection for
example!
I'd like to see more tests along these lines, particularly ones which
give test results for a wide variety of comparisons rather than just
"falled to distinguish Belden from R-S under controlled listening
conditions to a satisfactory statistical degree" or "try it you'll like
it." Human factors are important too..I experienced a variety of
emotions during this test and it made me think and wonder what-if quite
a bit. I hope that this was an interesting "read" for you (see addenda
below).
Stan
Addenda: an email I sent to Jon with some of his thoughts and my
reactions to an earlier email of mine..
------------------------------------------------------------------------
--------------
In a message dated 12/2/98 6:59:43 PM EST, jri...@cybertron.com writes:
<< I find it distressing that you did not experience a more positive
result
after the long term listening, as this is usually the place where the
reinsertion of the old (RS) cable is quite noticable.>>
:I know. Don't be distressed! This process dragged on (my
responsibility) for a longer period than I expected. My email to you is
pretty much a non-proofread overview. I DID get that "let-down" effect.
And then it vanished or diminished. Poof. Inconsistent. And seemed to
change depending upon other factors such as the blind test. I listend to
the 89259 exclusively for quite awhile and then put in other wire
several times. But the immediate perception especially early on WAS as
you described. Perhaps my description needs to be improved there. There
were a number of shifts if you will as things progressed.
<< [[ 1) Measured differences. I agree with your conclusion that there
are no significant measured differences between them. ]]
Not quite my opinion, as I still feel that thre is some way to
differentiate in a valid measurement good sounding cables from mediocre
sounding cables. What I said was that the FR didn't vary significantly,
and that the capacitance and inductance roll offs are well above the
audio range for either, so any differences in bulk capacitance or
inuctance between them would be insignificant.>>
Okay. From your email to me that mentioned some measurements you seemed
to imply that you did not expect much in the way of any measured diffs.
I can measure some more but I thought that the r-s and belden both
seemed to do just great and I didn't find ANY glaring problems of any
kind with either. If you are aware of any let me know. The measurements
were done the day I received the wire. Since I later perceived or
thought I perceived mid and treble differences I would focus on mid and
high related things were I to redo some objective tests. You know,
iterative, subjective and objective evaluation until you get to the
answer.
<<< [[ 2) Manual swap tests, ]]
Now be honest with yourself: did you have expectations or desires to
hear the Belden as sounding better? If anything, I would think that
your internal biases would tend to favor the RS. Interestingly enough,
the qualities you relate are generally what others hear when switching
from OEM cables to the 89259. In addition, they usually hear a
tightening up of the bass, slightly better pitch definition, but this is
lower in absolute terms than the mid and HF effects noticed.>>
I deliberately avoided reading your words and those of others about the
"sound" of this specific wire until after I had formed an opinion. So
what I wrote is my impression. I'm sure I had and have biases. Biases
work both ways ..either wire could be unconsciously favored...either way
a difference is involved so its moot in terms of can you tell a diff or
not..in this case the perception seemed to favor the Belden. Do you
realize that you are making a case FOR blind testing? I noticed or
thought I noticed other things besides the mid and high items but
thought that they were the most obvious and consistent characteristic.
Sometimes changes in the bass are really changes in the lower midrange.
And bass guitar has significant midrange energy, even lower treble
activity.
<<[[ 3) A/B switched swaps using a manual switchbox and 6 inches of r-s
gold ]]
Now think about this for just a moment: if the cables _do_ indeed have
differences that escape the current traditional measurements, how can
the switchbox be measured and given a clean bill of health?
Food for thought number two: if the RS cable is losing some
information, or blurring it in the mids and highs, by running the signal
through some RS cable, even a short piece, isn't it possible that it
would now be harder to hear those qualities on the 89259 now that they
had been compromised already?>>
Well we can't say for sure if there IS some magic going on now can we?
And boxes can be tested AND listened to too. And as for the bypass,
that's easy, try again with say 2 inch 89259 interconnects or whatever.
I am happy to try this if you send me some tiny short cables or else if
you can live with whatever I can come up with.
<<[[ 5) Blind tests. .... Felt kinda stressful. ]]
I hope that this personal experience will allow you to understand the
claims that get made for stress during listening tests. What stresses
people out will be different, and everyone will react differently to the
test conditions.>>
Oh I felt stress and confusion or disorientation all right. But did the
stress affect the listening results? Its tough to say. Was the stress
caused by...not hearing a difference and trying and trying to hear one?
I can try all I want but I can't lift a car for example, and if I kept
trying I would feel lotsa stress.. I can't say what the exact reason for
the stress was BUT remember that I mentioned that I'd like to do various
sensitivity tests. If I can feel that stress and still perform... This
would shed some light on the subject I think. And in comparison I LOVED
the switched testing. I felt POSITIVE that I had nailed certain
decisions but the grand totals worked out to be comparable to coin flips
The 1-10 "sureness" info would help there I think. In other words were
my occasional BINGO feelings indicative of better resulta were I to
review only those selections. If they were, maybe the audibility of wire
is not a constant steady thing but an occasional flicker which is real
but not under our control. In that case I would want the fancy wire even
if its benefits are transient and come and go as they please. Still I
have my doubts here since if a subset of my selections were much better
guesses than others then there should be an effect on the overal results
UNLESS it was also accompanied by an exactly offsetting opposite effect.
Get it? The Nousaine effect? LOL.
<< You mention the dichotomy between the sighted test before and after
the
blind tests. I think that there is ineed some psychology at work here,
but perhaps not quite the kind you may have thought it was.
What if you really were hearing differences at first, and then when
exposed to the blind tests which tended to hide the minor differences,
they destroyed your confidence in your abilty to hear the small changes.
This could render you psychoacoustically impotent!>>
Possibly. Like I said a lot of this depends on blind test validity. And
I can always test some more! I hope that when people respond, that some
interesting and worhwhile ideas and comments are forwarded. Some may
consist of ideas such as the bypass wire issue - EZ to minimize that and
test the box too to make sure that it is satisfactory.
<<As agreed, if you post you results, you can keep the Belden
interconnect.>>
Thanks. I appreciate it. It will certainly serve as a reminder of a very
interesting experience for me.
<< Now, one final question: which cable would you use between your CD
player and your preamp - the Belden, or the RS? :-) >>
RS. I pulled the belden to use for some frankenstein speaker cable tests
where I wanted to measure them as crossconnected speaker wire and made
little adaptors with wire pin pig tails to try this out without cutting
anything apart. No listening just measuring. That's the first clue I had
that those double coax suckers were definitely not 10 gage!!! And they
are really stiff even without the heat shrink tubing. For some of the
interconnect tests I used the cd players DAC and plugged the wires into
the cd player and then directly into a pair of well regarded biamped
monitors. Volume control (seldom changed) via the speaker's controls.
And next to no speaker wire. Nice and simple.
Oh, BTW I have received some rather interesting emails from well known
folks about the dollars behind the wire industry. Really sobering. The
margins, ad revenues and so forth are amazing and there apparently is
some degree of direct or indirect strong arming of the magazines going
on. I may mention that in the Caper piece. Then again, perhaps I
shouldn't .
Stan
jj, curmudgeon and tiring philalethist
>I don't know if the mini streaks were
>significant or not but they did tend to cancel themselves out.
They are expected, given the 50% chance that pure random guessing
would result in, i.e. they are not significant.
>Offhand, I think that the "ABXers" would like to consider 12 out of 16
>as showing staistically significant differences, which was not achieved
>in this case.
N p(n) sum p(n)
0, 1.52588e-05, 1
1, 0.000244141, 0.999985
2, 0.00183105, 0.999741
3, 0.00854492, 0.99791
4, 0.027771, 0.989365
5, 0.0666504, 0.961594
6, 0.122192, 0.894943
7, 0.174561, 0.772751
8, 0.196381, 0.59819
9, 0.174561, 0.40181
10, 0.122192, 0.227249
11, 0.0666504, 0.105057
12, 0.027771, 0.0384064
13, 0.00854492, 0.0106354
14, 0.00183105, 0.00209045
15, 0.000244141, 0.000259399
16, 1.52588e-05, 1.52588e-05
The three columns are
N, the number right.
p(n) the chance of randomly getting exactly that number right
sum p(n) The chance of getting N or more right randomly.
Decide what significance means, for yourself.
I notice that there is some very insidious attacking going on in some
of the replies to the test, where someone suggests that the switchbox
could affect the test, etc.
Simply put, tests HAVE BEEN RUN with putting switchboxes in and out,
and they come out NEUTRAL. Ergo, any complains along that line
require extraordinary proof, blind, well controlled, and from
multiple subjects.
--
Copyright j...@research.att.com 1998, all rights reserved, except transmission
by USENET and like facilities granted. This notice must be included. Any
use by a provider charging in any way for the IP represented in and by this
article and any inclusion in print or other media are specifically prohibited.
Jay P. Kapur
"Oh, BTW I have received some rather interesting emails from well known
folks about the dollars behind the wire industry. Really sobering. The
margins, ad revenues and so forth are amazing and there apparently is
some degree of direct or indirect strong arming of the magazines going
on. I may mention that in the Caper piece. Then again, perhaps I
shouldn't "
I think that this comment isn't entirely shocking. Most of us that have
glanced through numerous hi-fi mags have seen the large number of glossy ads
for cables. Magazines don't tend to like to bite the hand that feeds them.
The same thing that goes for auto enthusiast mags also goes for audio
enthusiast mags. I can't remember the last time I read about a new car that
drove like a truck (even though many do). I can't remember that last time I
read about a new cable that sounded like zip cord (even though many do).
What isn't yet obvious is whether the cable phenomena is a case of the
emperor's new clothes or a legitimate piece of equipment. I'm inclined to
think it is a bit of both.
--
Jay P. Kapur
jka...@andrew.cmu.edu
SJMARCY
emperor's new clothes or a legitimate piece of equipment. I'm inclined to
think it is a bit of both.>>
I'd love to see more wire tests using more than just one test method. Sighted
AND controlled listening tests for example.
I have done some detailed objective measurements of speaker wire which showed
the main problem with 12 feet of12 gage zip fed by a typical SS amp and loaded
by a minimonitor was a few tenths of a dB los at about 20 khz which appears to
be well below the error that has been detectable by humans in DBTs. With a
lower inductance wire we can get some of those few tenths back. Is it worth it?
Also setting up your system to use short wire runs helps the measured
performance a bit as well as your wallet.
Stan
TEAShea
prinicple involved is congitive dissonance. This is well established and much
explored issue among experimental psychologists.
Essentially, a person can hear what he wants to hear or expects to hear in
sighted tests. A person investing hundreds of dollars in interconnects want to
hear a difference - and usually does.
This is why double blind testing is the only effective and reliable way to
determine whether there is a true difference. Of course there is very little
of this done because it is time consuming. More importantly, it destroys
myths.
You did a nice job of single blind testing and I credit your most excellent
testing. However, you must expect to receive a lot of criticism because the
results you found are threatening to many audiophiles. They have a great
vested interest (emotional, financial) in believing that there are differences,
even thought the differences do not exist.
It is the same as the story of the Emperor's new clothes. You have revealed
that the Emperor is not wearing any clothes. Expect threats and disbelief.
Many will write you off and ignore you. Others will deride you. Ignore them.
You have done very well. Stand up and stand by what you have done.
Tom Shea
>During one of the wire-war
at AR, Jon Risch offered to send me a set
>of
>his DIY high end Belden 89259 one meter interconnects with the provision that
>I
>post my findings ------
- TEA (Tom)
Rob Pannoni
intermittently is an interesting one. Certainly there are plenty of
subtle variables in the process that are difficult to control for,
particularly if tests are run across multiple days. Who's to say
that what a subject ate for breakfast, or allergens in the air, or
mood, or fatigue, or relative humidity, or subtle changes in the
power grid, or any one of a million other things might be significant
in some odd combination. Some of these things can be controlled for,
but the number of permutations of these subtle variables is
mindboggling. I know that my system sounds better to me sometimes
than others. I have no idea why.
Has anyone done an ABX test where only responses where the subject
was relatively sure were counted and the rest discarded? Is there a
theoretical reason to not accept such a test as valid? It's probably
been done. But if not, it would be interesting.
--Rob Pannoni
SJMARCY
a flame backlash. It hasn't happened, at least not yet.
If anything I have experienced the reverse, particularly at Audio Review and
Stereo Review's board on AOL. Lotsa positive comments from subjectivists and
objectivists. Also I have received lots of email on the topic from some well
known folks, including details about the difficulties in getting wire articles
published in major mags, and specific examples of ad revenue loss tied to wire
blind test pieces. I attribute the favorable response to the wide range of
tests performed at the same time and the fairly open, leave it up to the
individual tone of the posting. I think that the concept of a test performed
a few different ways combined with anecdotes can make some progress where more
limited yet still technically rigorous tests have not. Less
confrontational/lecturing/debating and more focussed on the dichotomy between
sighted and unsighted tests.
In my opinion DB testing seems to have a bit of an image or PR problem amongst
lay people, which hugely affects its credibility . Combined with the ads in
nearly all magazines and "reviews" in others and its an uphill battle sorta
thing. Sooner or later perhaps that factor should be a larger part of the
discussion.
Also, I think that there are a few easy things to do when blind testing which
would increase test credibility. Human factors. Along the lines of the old
spoonful of sugar helps the medicine go down...
Stan
Robert Lowery
>The conjecture that cable differences may only manifest themselves
>intermittently is an interesting one. Certainly there are plenty of
>subtle variables in the process that are difficult to control for,
>particularly if tests are run across multiple days.
<snip for brevity>
This is the exact problem that fairly drove me mad for a year.
I have a Theta Miles balanced player running through Music Metre Silver
XLR cables to a BAT VK-500 and then to some new Chapman T-7's.
Pretty basic setup. Initially, the speaker cables were MIT T2.
I was never satisfied with the sound of the system and continually moved the
speakers around. It would sound great one night and then terrible the next.
The volume level seemed to be variable from one day to the next. Weird.
After about six months of this crap, I decided it must be some sort of hardware
problem. The Miles was sent into Theta for service and they could not find
anything wrong with it (but updated the firmware anyway). Theta even talked to
BAT about some sort of impedance mismatch, but could find no culprit. They
suggested I look at the speaker cables or speakers.
When the Theta came back, I paid close attention to the sound from power up.
Aha.
The MIT cables would change their character, starting at the warm end of the
scale and very gradually slide to tinny, cold, and nasty over about a four-day
cycle. Power down and up and the cycle would repeat. Damn.
I would usually run the system continuously for several days prior to a
critical audio visitor. No wonder it sounded terrible every time.
I replaced the speaker cables with Music Metre Silver and it all stabilized.
Bob
Thor...@tnt-audio.com
SJMARCY wrote:
> During one of the wire-war threads at AR,.....
<snip, snip, snip>
Just a few thoughts from me here, thorwn in from the side.
First of all, my own experiments using Belden 89259 where quite
disappointing. It did sound somewhat different from Rat-Shack
"Gold-Patch" (which I still keep around as "real world" Cable).
I'm not sure it sounded significantly better, though fitted with
better Phono- Plugs.... Test where both sighted and blind (but not
double blind).
> Equipment-wise (Jack G just perked up) I will offer the following. Two
> channel, music only material was played on a two channel system. CD
> music, no HT or multichannel stuff, mostly jazz, female voice and other
> selections. Lots of familiar material.
Now, again this is only an observation made by me during my testing
of various DIY and commercial Interconnects.
I can hear a darn flipping nothing with Seedee.... My own confidence
Level about hearing Cable Difference with CD-Playing is near Zero....
Only the most gross differences (between say XLO and Cardas) are
perceptible with any confidence to myself under SIGHTED
conditions.... Under blind conditions it's even worse....
At the same time playing good old "so much worse than the perfect CD"
Vinyl the differences are easily acertained and even blind a almost
100% accuracy of identifying which Cable is in circuit is
possible.... Go figure.
> My gut feel is that the Belden/ R-S conclusion depends upon the validity
> of blind testing. If its valid then there is no difference, under these
> circumstances, in this case, with this equipment and this listener at
> the very least.
Indeed. I think it is often easy to imagine differences where they
are not. I also feel that the setup used is of crucial importance....
> Oh, BTW I have received some rather interesting emails from well known
> folks about the dollars behind the wire industry. Really sobering. The
> margins, ad revenues and so forth are amazing and there apparently is
> some degree of direct or indirect strong arming of the magazines going
> on. I may mention that in the Caper piece. Then again, perhaps I
> shouldn't .
I do agree. I have found specifically in the Mid-Budget Sector very
few reliable differences between a variety of Cables and must note
that my above experience with using CD-Standard Digital Sources as
primary signal source as revealing only extremely gross differences,
that much of the testing makes little sense....
Incidentally similar findings can be noted between the lines in Cable
Group- Tests carried out by the UK Mag "HiFi-Choice" (which uses
listening Panels and blind tests). Indeed in at least the larger
majority of cases where dramatic differences are noted they seem
dominatly the result of "unusual" RLCG Values.
In some cases they are not and that is when I start taking note....
And most interestingly, Teflon Insulated Stranded Copper is not often
in the "unusual" Category....
Ultimatly I think Cables of any form are the last bit of fine-tuning.
And simply using Cat-5 UTP Network Cable with Solid Core copper in a
3 X 4 pair of 4 X 4 pair as Speaker Cable combined with Solid Copper
Core and Solid Copper Foil Shield Coax (50pf/m) get's you most of the
way dirt cheap....
More important to sort out room accoustics and speaker positioning
first....
Kind Regards Thorsten
Visit TNT-Audio on the Web - the only advertising
free audio web-zine. http://www.tnt-audio.com
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Rob Pannoni
>
> Stan, your post is very intersting and is not surprising to me. The main
> prinicple involved is congitive dissonance. This is well established and much
> explored issue among experimental psychologists.
>
> Essentially, a person can hear what he wants to hear or expects to hear in
> sighted tests. A person investing hundreds of dollars in interconnects want to
> hear a difference - and usually does.
I think a more correct psychological term would be "expectation bias."
We're likely to hear what we expect to hear. It certainly plays a part
in audio. But that's too simplistic an explanation to account for the
variety of audio experiences reported through subjective listening
tests. What if a subjective test goes against expectations? I know of
no psychological theory that would account for an unexpected result.
I suspect there are simply more variables and permutations of variables
than we can realistically control for. As a methodology, controlled
experiments are very useful, but not perfect--particularly when dealing
with complex systems with many potential interactions.
--Rob Pannoni
Rob Pannoni
> I think a more correct psychological term would be "expectation bias."
Oops. Brain cramp. I meant "confirmation bias." But the principle
is the same.
--Rob Pannoni
jj, curmudgeon and tiring philalethist
Rob Pannoni <pan...@sera.com> wrote:
>I think a more correct psychological term would be "expectation bias."
>in audio. But that's too simplistic an explanation to account for the
>variety of audio experiences reported through subjective listening
>tests.
Oh?
>What if a subjective test goes against expectations? I know of
>no psychological theory that would account for an unexpected result.
You certainly do, we've discussed it. The expectation of a
difference, in any direction, can and will cause the listener to
focus (pay attention) to different things on the different stimulii,
even when they are the same, in fact.
That's all it takes.
Trouble
possible that simply changing the cables, then replacing them, then
returning the originals back to the system for comparison testing
can make any system sound better IF terminal oxidation is at
issue...a point which jj and others have belabored to death in
recent RAHE discussions <that's a joke, jj, not a jab>
S. Churchill,
Pullman, WA