Digital Cable Audition (Long)
Arthur LeGrand Shapiro
[ Even cable aetheists mind want to read Art's digital interconnect
review, as he cites more than a few recordings of interest. -- jwd ]
At the Stereophile Show in New York City last Spring, a drawing was
held in which some RAHE attendees won merchandise generously donated
by a number of retailers and manufacturers. Due to other personal
commitments I was unable to be present at the drawing, held the last
afternoon of the Show; I figured my odds of winning anything
significant were rather remote.
Subsequently, upon my return to California, it was my pleasure to
receive a phone call from His Moderatorship, James Durkin, informing
me that I had been among the winners. I had won a digital
interconnect, a category of audio that has occupied quite a bit of my
time over the last several years in doing auditioning and product
comparisons for improving my own system. As expected, I was quite
pleased at this unanticipated happening. I hoped for the best and
inquired as to whether this was an ultra-high-end digital product such
as the NBS. No, I was told, this was a Canare cable.
Well, Canare certainly is a reasonable acquisition for the entry-level
audiophile trying to improve his lot on limited funds; for a $45 or
whatever cable it certainly represents an intelligent expenditure and
good value. I've heard it in familiar systems in the past, and while
a great value for the money, it was readily bettered by some of the
more esoteric cables that were available within my circle of
audiophile friends. My own system is some number of parsecs beyond
the entry level point, and I had to confess to James that the Canare
was unlikely to be of the slightest use to me, suggesting that it
might find a more appreciative home elsewhere amongst the RAHE
readership. I was told that things weren't as bleak as I might
expect; this was a "statement" cable from Sound & Video, the pro arm
of HAVE, Inc., makers of Canare cabling, and allegedly quite a step up
from the more pedestrian Canare product with which I was somewhat
familiar. At the least, it was suggested that I could give the cable
an honest audition and report to RAHE what I heard. So, ever anxious
to advance the Subjectivist Art against the heathen engineering hordes
in our midst, I agreed to receive the cable.
Readers not intimately familiar with the mysterious, dark, underbelly
of the Internet may be unaware that a small portion of the shadowy
"modem tax", currently levied only on the largest ISPs and huge
corporate entities with direct high-speed connections to the Internet
backbone, is funneled back into the Internet newsgroups. The intent
is to allow "controlling interests" the opportunity to "facilitate and
enhance their mastery and understanding of, and expertise in," each
newsgroup's subject matter so as to promote the goals of the group.
(Perhaps I should volunteer to moderate alt.fan.traci-lords!) In the
case of RAHE, it means the Moderators have funding to attend relevant
audio shows and visit dealers, manufacturers, distributors, and even
significant individual systems, as long as not for personal gain. If
my understanding of this is slightly wrong, I'm sure the Moderator
will correct me. In any event, James hoped to deliver the cable to me
in person, as he had previously expressed some interest in auditioning
my system; I likewise would be honored to have such an illustrious
visitor for a day or two.
[ Art, whatever am I going to do with you. First you clue the readers
in to our dirty little slush fund. What's next, revealing the
details of the Stereophile headquarters break-in last year in Santa
Fe? Ever heard the phrase "sleeping with the fishes"? Get a pillow
Art ... -- jwd ]
Well, things just didn't work out, despite our best efforts. A change
in business plans here, an injured kitty there, etc., etc., and he was
never able to find his way to southern California. In fact, he has
yet to appear, although we have good hopes for the early Spring.
Finally, in November, the postman delivered a package to my door,
which I opened to find the by-then-mythical Canare "DigiFlex +Plus"
cable.
I was immediately impressed by the packaging. Those of us into the
phono cartridge scene are well aware of the ludicrous excesses of
packaging that are de rigour in that industry. One company's faceted
Lucite box is bested by the next company's box "hand carved from rare
endangered hardwoods pillaged from the Amazon rain forest." All this
for something that's used only once: until the cartridge is removed
and installed. While cable manufacturers have not reached this level
of conspicuous overconsumption, the Canare immediately rubbed me the
right way with its practical container. The base was a square of
corrugated cardboard. Over top was a congruent square of glossy
product blurb, turned at 45 degrees. The four corners of the
cardboard and the offset four corners of the blurb sheet made an
attractive star pattern. The cable sat on top of these two pieces,
and on the bottom side of the cardboard was a small brochure
describing and pricing the company's products. The whole works was
simply covered with transparent heatshrink wrap. Bravo to Sound &
Video for practical, attractive, informative, and environmentally
appropriate packaging.
The Digi Flex +Plus itself is a rather plain black cable whose body
shows a braided exterior of about a third of an inch in diameter.
Heavy black heatshrink connects the body to sturdy gold-plated RCA
plugs of unknown manufacture, while functioning to provide strain
relief. The cable is quite flexible, a far cry from the
massively-unwieldy Marigo Reference I am currently using in The
System. These RCA plugs grip the jacks firmly, although not to the
almost preposterous degree of the Marigo's plugs. A piece of
heatshrink in the middle is inscribed with the manufacturer's
name/logo and telephone number. Just as with the wrapping, this all
seemed eminently reasonable and practical. I was hoping to appreciate
the cable's sonics as much as I respected the manufacturing efforts
that I had seen thus far.
The literature suggested that at a dozen or so hours would be needed
to enjoy full cable break in. As this was a Friday evening, and I had
some boxing and business shows to enjoy on TV, I simply replaced the
Marigo with the Canare, turned on the system aside from the
amplifiers, and put a sampler CD from "Classic CD" magazine on repeat
mode. I let it run continuously for fifteen hours before ever
listening to the new cable. For reference, the relevant parts of my
system consist of a Wadia 3200 transport and EAD 7000 III dac, fed by
Kimber KCTG (silver) interconnects into a CAT SL1 Signature preamp.
Tara RSC Master Gen II interconnects feed the Clayton M-70 monobloc
amplifiers, which power Legacy Signature III speakers by way of
bi-wired Cardas Golden Hex V cables. The Claytons are plugged into
one dedicated line, while the CAT and a Power Wedge 116 occupy the
other dedicated line. All other equipment, in this case the dac and
transport, go through the Power Wedge. The system errors on the side
of being excessively detailed, which seems to facilitate its
impressive reproduction of "classical" piano, that forming the vast
majority of my listening. For orchestral, chamber, vocal, etc.
music, the system, while very fine, could clearly benefit from
additional warmth. I mention the sonic characteristics of the system
in order to give the reader a better feel for the environment in which
the Canare was auditioned.
Over the course of the time that I've had the cable, I was assisted by
two other audiophiles: Mark Katz, an occasional contributor in this
forum and a frequent visitor, and John Nye, one of the luminaries
behind Clayton Audio. John, who I've known electronically for quite a
few years, finally managed to wend his way here for an overnight visit
accented by many hours of music listening. Both of these guys have
pretty formidable ears, and I'd say all our opinions of the Canare
cable were similar. I'll note that John actually preferred the one
other digital cable in my collection: the "old" Illuminati Datastream,
to the megabuck Marigo that Mark, especially, finds so spectacular
(and uses himself). I've waffled more than usual between the two,
although I tend to use the Marigo and considered it the primary
comparison vehicle to the Canare.
My initial impressions of the cable were not extremely positive. Over
the several detailed comparative listening sessions that resulted in
this report, I came to enjoy what I was hearing more and more.
Whether the cable was subtly breaking in, or whether I was just
growing accustomed to the sound of what I was hearing, I can't say.
But had I written this summary back in November, it would have been
somewhat lackluster. But let's look at some specific CDs out of the
many I used to compare the cables.
I start with my all-time auditioning favorite: the Bach "Coffee
Cantata" on a disc of Bach secular cantatas on Dorian 90199. This
was, deservedly, Stereophile's Record of the Month a couple years ago.
Both the performance and the sonics are spectacular. This work
features a small orchestra, beautiful solo flute passages, basso
continuo (background harpsichord/double bass accompaniment) and male
and female solo singers. With the Marigo, the flute soli are just
gorgeous- open and airy, with wonderful decay. There is a trace of
sibilance to the voices. The Canare gave slight bass emphasis which
made the orchestra somewhat more prominent with respect to the
singers. I found the violins a little more brittle, and the voices
more sibilant. The flute seemed quite a bit flatter, with less
audible decay to the notes and somewhat less open sound. Voices
decayed too fast. Male voices were somewhat richer with the Marigo.
All in all, the Marigo was the winner here.
I then selected a well-received recent piano disc: a recording of the
complete sonatas by Erich Korngold on Chandos 9389, with Geoffrey
Tozer performing. I chose the graceful minuet of the 3rd Sonata.
Starting with the Canare on this round, I enjoyed what I was hearing,
with nice piano presence and appropriate tonal balance. Moving to the
Marigo, I was more conscious of the micro-dynamics in this
performance, and found the sound slightly more open with nicer, more
apparent note decay. It made the Canare seem a smidgen flat by
comparison. Interestingly, on some of the crisper, staccato passages,
the Canare seemed to present the sound quite well. And the slight
bass emphasis noted with the Bach carried through to this work as
well, giving a very pleasant foundation to the music. Both cables
were nice in different ways.
I turned to a CD that has always seemed a bit aggressive, being
recorded a little too close. It's a disk of the Suites for Two Pianos
by Anton Arensky, on Hyperion 66755; Stephen Coombs and Ian Munro
perform these works that deserve a little more publicity than they
receive. I chose the Polonaise from the Suite #1. The Canare had
very strong bass, indeed too overwhelming. The Canare seemed to tame
the aggression of the CD, perhaps due to being a little less revealing
in general. Once again, the Marigo proved a little more subtle and
gave slightly more spacious sound. The Canare really worked well here
and tamed the harshness of this CD.
To get another feel for human voice, I selected another previous
Stereophile Record of the Month: the Fairfield Four doing "Standing in
the Safety Zone" on a CD of that name on Warner 26945. (Why are there
five men on the cover?) This is an a cappella male group performing
some nice toe-tapping gospel music. Once again, the Marigo seemed to
reveal a little more subtlety in the voices, and I sensed that the
spread of the voices from side to side was slightly more apparent.
Although the recording is in general excessively sibilant, this
negative quality did not seem overemphasized with the Canare, unlike
the aforementioned Bach CD.
I turned to the Overture to Barber's "The School For Scandal" on a
performance by Yoel Levi and the Atlanta Symphony Orchestra on Telarc
80250. I enjoyed both cables here, and wasn't conscious of the
Marigo's superior resolution when the orchestra was playing en masse.
Indeed, the nice bass heft of the Canare proved beneficial to the
music. There's a beautiful solo oboe passage introducing the
secondary theme of this overture, and here, true to form, the Marigo
forged ahead with its more delicate, subtle presentation.
It seemed useful at this point to pull out a rather curious CD on
Crystal 801: a most fascinating piece by Alan Hovaness entitled the
"Prayer of Saint Gregory", with trumpet soloist Thomas Stevens and the
Crystal Chamber Orchestra (any relation to the label?) conducted by
Ernest Gould. This features a distant, poorly-recorded, indistinct
orchestra with a beautiful, soaring trumpet solo throughout the work.
On the right system - warmer than mine - this trumpet can bring tears
to your eyes. It turned out that the orchestra was a little more rich
through the Marigo, and conversely the increased resolution made the
violins seem excessively steely on this poor recording. In this sense
the Canare's attributes were not a detriment. As usual, on the solo
instrument, the Canare was bested; I was more conscious of the trumpet
overtones and the resonances of Mr. Stevens' instrument.
Finally, I auditioned a harp recital by Isabelle Moretti on Harmonia
Mundi 905184. I chose the Perpetuum Mobile, the third movement of the
Harp Sonata by Germaine Tailleferre. If that name doesn't sound
familiar, Tailleferre was the only female member of the famous
avant-garde group of early 20th Century composers known as "Les Six",
of which Poulenc is best known. Nothing new to report; the Canare was
slightly less spacious, although interestingly there did not seem to
be a noticeable difference between the two cables in the decay of the
torrent of rapid notes. The Canare acquainted itself quite well on
the harp. There's a point with a little under a minute left in this
movement where there's some quite obvious sound from Ms. Moretti's
foot hitting the stage, probably in the process of quickly diddling
the harp's pedals. I set the Wadia to endlessly repeat these few
seconds, while switching back and forth between the Canare and the
Marigo. I found this stomping to be a little clearer through the
Marigo.
There were several problems that were consistently shown by the Canare
in my system. Most significantly, I was bothered by problems with the
decay of notes. Individual notes seemed to just stop, whereas with
the Marigo one could sense them fading gradually into silence. I
found this characteristic to be somewhat obvious, both for my beloved
piano and for other solo instruments such as the flute in the Dorian
Bach CD. The beautiful open transparency on that flute was clearly
diminished with the Canare.
John Nye, especially, felt that the Canare moved one well back in the
hall, and more so that he could accept in a high-end product. Perhaps
this is just another way of saying that the sound was less open. He
(John) thought that the sound was somewhat laid back or rolled off,
and that this actually might be a good choice for folks whose digital
setups are excessively bright. It's an interesting hypothesis. Mark
Katz appreciated the Canare, but observed that it just couldn't
compete with the beautiful, spacious, smooth openness of the Marigo.
In a word, the $110 Canare DigiFlex +Plus was somewhat bettered by the
MUCH pricier Marigo product. This is hardly a surprise; while Marigo
products are anything but inexpensive, they generally give a lot of
performance in return. As I previously noted, in the several
auditioning sessions I spent going back and forth between cables, I
found myself gradually warming up to the sound of the Canare, meaning
I appreciated it more as the weeks ran on. Maybe comparing the
eminently-practical Canare to the high-spirited Marigo was grossly
unfair, but that's what I had to work with. I think it would prove an
excellent performer in its price range, and an audiophile looking to
spend in the low three figures for his digital cable should well
consider the product. I feel almost guilty putting the Canare cable
into my sporadically-used video system, where it connects my laser
disc player to the VCR that functions as a control center. But I'll
never need to worry about that video link with this solid a performer
installed.
--
Arthur L. Shapiro Art...@mpa15c.mv.unisys.com
Rarely an official Unisys corporate spokesman, and certainly not now.
Michael Ford
In article <5dnh2e$6...@eyrie.graphics.cornell.edu>, Arthur LeGrand
Shapiro <ART...@mpa15c.mv.unisys.com> wrote:
> [ Even cable aetheists mind want to read Art's digital interconnect
> review, as he cites more than a few recordings of interest. -- jwd ]
>
> Subsequently, upon my return to California, it was my pleasure to
> receive a phone call from His Moderatorship, James Durkin, informing
> me that I had been among the winners. I had won a digital
> interconnect, a category of audio that has occupied quite a bit of my
> time over the last several years in doing auditioning and product
> comparisons for improving my own system. As expected, I was quite
> pleased at this unanticipated happening. I hoped for the best and
> inquired as to whether this was an ultra-high-end digital product such
> as the NBS. No, I was told, this was a Canare cable.
>
> Well, Canare certainly is a reasonable acquisition for the entry-level
> audiophile trying to improve his lot on limited funds; for a $45 or
> whatever cable it certainly represents an intelligent expenditure and
> good value. I've heard it in familiar systems in the past, and while
> a great value for the money, it was readily bettered by some of the
> more esoteric cables that were available within my circle of
> audiophile friends. My own system is some number of parsecs beyond
I have a VERY, VERY, hard time believing digital cables have any
audible effects whatsoever. I have my doubts that they even have any
meaningfull measured effects. I bought the very first consumer digital
separates, the Sony 650 player and 702 DAC, which being the first, one
might suspect would be most suseptible to problems with the digital
linkage. Nothing I have ever tried out of the entire contents on my
large and varied "box of wires" that years of AV ownership have
accumulated made a bit of difference. I admit to a couple of Radio
Shack ferrite clamps on the current piece of coax, but again no effect
I have ever noticed.
I find it VERY easy to believe OTOH that many people do "think" they
hear all sorts of differences when any type of blind testing is
avoided. I find it especially amusing when the "differences" are so
tangible, notes chopped off instead of smooth decay, etc., but almost
like magic when nonaudible cues are removed so are the differences. Of
course no one could be expected to hear even gross differences under
the pressure of a real test. It is just luck where people "seem" to
hear tiny real variations in sound best under the pressure of such
tests.
Doug Weinfield
In article <5dt52j$k...@canyon.sr.hp.com>, mike...@kaiwan.com (Michael
Ford) wrote:
(snip)
>
> I have a VERY, VERY, hard time believing digital cables have any
> audible effects whatsoever. I have my doubts that they even have any
> meaningfull measured effects. I bought the very first consumer digital
> separates, the Sony 650 player and 702 DAC, which being the first, one
> might suspect would be most suseptible to problems with the digital
> linkage. Nothing I have ever tried out of the entire contents on my
> large and varied "box of wires" that years of AV ownership have
> accumulated made a bit of difference. I admit to a couple of Radio
> Shack ferrite clamps on the current piece of coax, but again no effect
> I have ever noticed.
>(snip)
(1) I've read that those ferrit clamps on digital cables can cause
some kind of problem, perhaps bandwidth limiting.
(2) I've read that if the digital cable (and plugs) don't provide a
true 75 ohm connection, that this could cause problems. Audible? I
don't know either way.
(3) Some might have opinions about the quality of sound from the 650
and 702, if they're that old.
--
You can learn to fly, but that cocoon has got to go.
Bob Olhsson
>I have a VERY, VERY, hard time believing digital cables have any
>audible effects whatsoever.
I would suggest that you arrange to hear the particular systems and
combinations of equipment being described. Most of the effect is, in my
experience, found in the focus of the stereo image and the clarity of the
5kHz. & higher signal. If you are not listening on a system with 1% matched
channels and a very flat frequency response, you will indeed not hear it.
Otherwise, it is surprisingly clear.
Certainly these effects must reflect design engineering problems but they
can be audible with certain components.
>I find it VERY easy to believe OTOH that many people do "think" they
>hear all sorts of differences when any type of blind testing is
>avoided.
It would be very difficult to arrange a double-blind test of digital
connections that does not introduce its own impedance and rfi problems.
Having a single piece of wire between two devices with no connecters in
between is one thing. The performance at high frequencies is controlled by
the length and construction of the wires and by the termination on both
pieces of hardware.
Introducing extra connecters and termination is another issue as it
disrupts the signal with reflections unless everything involved has a
precision impedance design which most consumer gear, common connecters and
wire does not. Transmission-line issues are very much a part of digital
interconnection while they are of questionable importance at analog audio
frequencies.
--
Bob Olhsson Audio | O tongue, thou art a treasure without end.
Box 555,Novato CA 94948 | And, O tongue, thou art also a disease
415.457.2620 | without remedy. == Jelal'uddin Rumi ==
415.456.1496 FAX |
Doug Weinfield
> I have a VERY, VERY, hard time believing digital cables have any
> audible effects whatsoever. I have my doubts that they even have any
> meaningfull measured effects. I bought the very first consumer
> digital separates, the Sony 650 player and 702 DAC, which being the
> first, one might suspect would be most suseptible to problems with
> the digital linkage. Nothing I have ever tried out of the entire
> contents on my large and varied "box of wires" that years of AV
> ownership have accumulated made a bit of difference. I admit to a
> couple of Radio Shack ferrite clamps on the current piece of coax,
> but again no effect I have ever noticed.
(1) I've read that those ferrit clamps on digital cables can cause
Andre T. Yew
doug...@his.com (Doug Weinfield) writes:
>(1) I've read that those ferrit clamps on digital cables can cause
>some kind of problem, perhaps bandwidth limiting.
I remember propagating this misinformation myself, so please allow me
to correct it. It doesn't limit bandwidth because the ferrite bead
sees a net zero current running through it and therefore doesn't do
anything. The ferrite helps a bit with rejecting common-mode noise,
which I believe can wreak more havoc than just plain old jitter, since
screwing up the internal clock and data lines is just one of the many
things it could do. Transformer-coupled inputs help a lot with
common- mode noise (Bob Fitzgerald measured a 40 dB reduction in his
DAC after installing a good pulse xformer), but not too many people
implement this in their DACs.
>(2) I've read that if the digital cable (and plugs) don't provide a
>true 75 ohm connection, that this could cause problems. Audible? I
>don't know either way.
This is true. I've seen 75 Ohm RCAs, but not 75 Ohm RCA jacks. Some
would argue it's inaudible because the length of the connector is
small compared to the wavelength of the signal. My opinion is that if
you're going to take the time and trouble to fix something else in the
DAC, you might as well do the connectors right because it's very
little effort for, at the very least, the peace of one's mind. It's
amazing how good a stereo sounds to a happy, relaxed mind. Seriously.
--Andre
--
PGP public key available
Andre T. Yew
jim howat
>I have a VERY, VERY, hard time believing digital cables have any
>audible effects whatsoever. I have my doubts that they even have any
>meaningfull measured effects.
I had the same opinion until I substituted a cheap'n chearfull
interconnect because I thought it wouldn't make the slightest bit of
difference - but sitting through edgy, smeared images all the time
thinking this can't be happening... I'm 100pc positive this has a
physical explanation - Probably terminations and impededences etc, But
I know when what what I hear is wrong, and by substituting another
cable it becomes better I'm at least one ahead of the "head in the
sand" school of so called objectivists who can't back up their
argument objectivley or subjectively.
--
jim howat
Steven R. Rochlin
Dudettes/Dudes,
Well, on Sunday night i was subjected to an ABX test by none
other then E. Brad Meyer of the Boston Audiophile Society. Mr. Meyer
is quite good with ABX and is VERY experienced with properly carrying
out and ABX test. We used his system (for which i'm not too familiar
with audibly speaking). The outcome. After 7 trials i scored 6
correct answers in the digital cable test.
The digital cables ABX was part two of the test. Part one was
with analogue interconnects where i only scored 5 of 7 correct
answers. Still best "the odds" in a system that i was not familiar
with audibly. So in conclusion, can you ABXers who feel ALL cables
sounds the same please give it a rest. And here's Mr E. Brad Meyer
e-mail address for those who doubt his qualifications or who have
questions for him:
What REALLY matters to me is that YOU...
Enjoy the music (The Lemon Pipers on vinyl right now),
Steven
"Now listen while i play
My green tambourine"
Music lover, Head toilet bowl scrubber, workaholic, drivin' maniac,
Baron of Grey Matter, and just one kooky dude who's havin' a
goooood time as he travels through this time/space continuum.
Bob Olhsson
In article <reg1ywk...@eyrie.graphics.cornell.edu>,
and...@alumnae.caltech.edu (Andre T. Yew) wrote:
>doug...@his.com (Doug Weinfield) writes:
>> (2) I've read that if the digital cable (and plugs) don't provide a
>> true 75 ohm connection, that this could cause problems. Audible? I
>> don't know either way.
>This is true. I've seen 75 Ohm RCAs, but not 75 Ohm RCA jacks. Some
>would argue it's inaudible because the length of the connector is
>small compared to the wavelength of the signal.
When I queried a pro digital manufacturer about their use of 50 Ohm
BNC jacks for 75 Ohm connections, I was told that it wasn't critical
because they HAD to trim the actual impedences out to compensate for
the circuit board, etc. and the connecter impedance comes out in the
wash.
They said what IS critical is the precision of the cable's impedence
and any intermediate (and obviously un-trimable) connectors.
Obviously this also leads to the question of how many consumer
manufacturers trim out THEIR impedances...
Craig Stark
"Steven R. Rochlin" <enjoyt...@top.monad.net> writes:
[cutting -- rgd]
> After 7 trials i scored 6
> correct answers in the digital cable test.
Not shabby, assuming valid methodology. According to Splus:
Exact binomial test data: 6 out of 7 number of successes = 6,
n = 7, p-value = 0.0703 alternative hypothesis: true p is not equal to
0.5
> with analogue interconnects where i only scored 5 of 7 correct
> answers. Still best "the odds" in a system that i was not familiar
Well, did you really best "the odds"? The odds are:
Exact binomial test
data: 5 out of 7
number of successes = 5, n = 7, p-value = 0.2891
alternative hypothesis: true p is not equal to 0.5
This says, in effect, that if it were the case that you had 0
sensitivity to the measure and were purely guessing (e.g. if the
cables were identical in every way, shape, and form), if we ran the
test 100 times, we would expect to get 5/7 correct about 29 of those
times. Not all that convincing. Now, your digital cable results seem
better at only 7 times, but keep in mind that the N here is small and
the only difference is one trial.
To humor me (and I know you exist only to humor little ol' me), do the
following test: Grab a coin and flip it 7 times. Before each flip,
call the Psychic Friends (or just do it yourself) to come up with a
prediction for the outcome (or just say it will be heads each time).
We'd all agree that you should be correct 50% of the time. Now, try
the experiment a few times. Were you ever right 5/7 times? 6/7? Do
you think you're really psychic?
Now then, given Mr. Meyer's experience with ABX testing and the
subject pool he has access to (the Boston Audiophile Society), we
should have the possibility of some more data. More data = higher N =
better ability to rule out the chance we're looking at normal noise.
Please don't mistake me for belittling your experiment or data. I
applaud you for doing the test and your data (again, assuming valid
testing methodology) provides an important piece. It's not wholly
conclusive on its own, but it still is valuable. Run the test a total
of 10 times for 70 comparisions, and keep the same rate in either (or
even 5 times for 35 sessions) and you're a lot more convincing.
- Craig
Tony F
>> (2) I've read that if the digital cable (and plugs) don't provide a
>> true 75 ohm connection, that this could cause problems. Audible? I
>> don't know either way.
>
>This is true. I've seen 75 Ohm RCAs, but not 75 Ohm RCA jacks. Some
>would argue it's inaudible because the length of the connector is
>you're going to take the time and trouble to fix something else in the
>DAC, you might as well do the connectors right because it's very
>little effort for, at the very least, the peace of one's mind. It's
>amazing how good a stereo sounds to a happy, relaxed mind. Seriously.
>
>--Andre
>
Could this be why I'm experiencing very brief intermittent drop-outs?
I'm using a Van Den Hul 75 ohm digital cable with RCA ends but on the
jitter reducer (DTI Pro) side I have to use one of the BNC adapters
that came with it because it doesn't have RCA. On the DAC end I have
both inputs so I've tried it both ways but this didn't make any
difference.
Should I get a good 75 ohm cable with BNC ends?
I also tried a Radio Shack video cable (still using the adapters) but
this made no difference, still got the drop-outs.
The lock lights do not go off during the drop-outs. BTW, I'm using
toslink on the input side of the DTI Pro. I realize this is not
optimal, but I have to live with it for now
Tony F
Arnold B. Krueger
Craig Stark <sta...@CMU.EDU> wrote in article
<5el7pj$f...@tolstoy.lerc.nasa.gov>...
> "Steven R. Rochlin" <enjoyt...@top.monad.net> writes:
> > After 7 trials i scored 6
> > correct answers in the digital cable test.
>
> Not shabby, assuming valid methodology. According to Splus:
>
> Exact binomial test data: 6 out of 7 number of successes = 6,
> n = 7, p-value = 0.0703 alternative hypothesis: true p is not equal to
> 0.5
My table says 0.062, but no matter how you split it, this is not even
95% confidence. One would like to see 99% confidence before making any
public claims.
>
> > with analogue interconnects where i only scored 5 of 7 correct
> > answers. Still best "the odds" in a system that i was not familiar
>
> Well, did you really best "the odds"? The odds are:
>
> Exact binomial test
> data: 5 out of 7
> number of successes = 5, n = 7, p-value = 0.2891
> alternative hypothesis: true p is not equal to 0.5
Again, I get 0.227 but now we are at 77% confidence which the bank
won't buy at all.
I'm seeing a pattern here. You must be using a table for Chi Squared
which is an approximation of binomial. Binomial has a lot of factorals
in it and it was hard to compute until we had big computers. The
numbers I cite are for true binomial, but took a large computer (in
1981) about an hour to compute out to 500 trials. That would be a lot
of folks cranking a Comptometers for a long time in 1940.
> This says, in effect, that if it were the case that you had 0
> sensitivity to the measure and were purely guessing (e.g. if the
> cables were identical in every way, shape, and form), if we ran the
> test 100 times, we would expect to get 5/7 correct about 29 of those
> times. Not all that convincing. Now, your digital cable results seem
> better at only 7 times, but keep in mind that the N here is small and
> the only difference is one trial.
From doing a lot of ABX tests I can tell you that 20:1 shots do happen
about one time in 20 (!). And, surprise, surprise, when you do more
trials to improve your results things have this irritating tendency to
go towards random guessing, more times than not.
> To humor me (and I know you exist only to humor little ol' me), do the
> following test: Grab a coin and flip it 7 times. Before each flip,
> call the Psychic Friends (or just do it yourself) to come up with a
> prediction for the outcome (or just say it will be heads each time).
> We'd all agree that you should be correct 50% of the time. Now, try
> the experiment a few times. Were you ever right 5/7 times? 6/7? Do
> you think you're really psychic?
This is, of course, the classic test of the binomial theorem. The
other thing to note is that the mean convergence theorum says that
experimental results for a small number of trials converges to the
mean for a large number of trials with the square root of the number
of trials. This means that if your results have low confidence, you
have to do a lot of trials to improve them, if doing more trials will
improve them at all. We ran the binomial theorum out to
> Now then, given Mr. Meyer's experience with ABX testing and the
> subject pool he has access to (the Boston Audiophile Society), we
> should have the possibility of some more data. More data = higher N =
> better ability to rule out the chance we're looking at normal noise.
The classic ABX test had 16 trials. This meant 12 correct answers for
95% confidence and 14 for 99%. As I said before, we would occasally
have someone who got 12 out of 16. Maybe we had a golden ear?
Regrettably, further trials always showed nothing of the sort. Just
lucky guessing.
Check out http://mars.acs.oakland.edu/~djcarlst/abx_wire.htm for the
results of a fair number of double blind tests of wires.
Bob Myers
Bob Olhsson (o...@hyperback.com) wrote:
> When I queried a pro digital manufacturer about their use of 50 Ohm
> BNC jacks for 75 Ohm connections, I was told that it wasn't critical
> because they HAD to trim the actual impedences out to compensate for
> the circuit board, etc. and the connecter impedance comes out in the
> wash.
This is correct; connector impedance can generally be ignored in all
but EXTREMELY high-frequency applications (UHF and above), as the
connector is electrically very short (the distance travelled by the
signal in the connector is very small compared to the signal
wavelength). In "normal" applications, the effects of the connector
will be swamped out (and then some) by the cable itself, the load,
etc..
There are quite a few examples of applications where the standard
connector isn't even particularly well-controlled, impedance-wise. A
really good one is probably very close to most of you right now - the
15-pin D connector commonly used for PC video outputs.
> They said what IS critical is the precision of the cable's impedence
> and any intermediate (and obviously un-trimable) connectors.
"Intermediate" connectors - presumably, those within a line, as
opposed to being at either end - CAN be more of a concern, as they put
a "bump" (and therefore a possible source of reflections) in the
impedance profile of the line. That's why it's almost always better
to use a single long cable than several strung together.
Bob Myers KC0EW Hewlett-Packard Co. |Opinions expressed here are not
O- Workstations Systems Div.|those of my employer or any other
my...@fc.hp.com Fort Collins, Colorado |sentient life-form on this planet.
Andre T. Yew
ro...@novagate.com (Tony F) writes:
>Could this be why I'm experiencing very brief intermittent drop-outs?
>I'm using a Van Den Hul 75 ohm digital cable with RCA ends but on the
>jitter reducer (DTI Pro) side I have to use one of the BNC adapters
>that came with it because it doesn't have RCA. On the DAC end I have
>both inputs so I've tried it both ways but this didn't make any
>difference.
Actually, in my modified transport, I'm using van den Hul's 75 Ohm
cable (the light blue one), and I haven't noticed any dropouts. I've
also used the DTI-Pro with the BNC-RCA adaptors and haven't noticed
any drop outs there. This was with Toslink feeding into the DTI-Pro.
>I also tried a Radio Shack video cable (still using the adapters) but
>this made no difference, still got the drop-outs.
This probably indicates that the problems lies not in the cabling.
Have you tried connecting your transport directly to your DAC? If you
still get dropouts, then, at least you know it's not the DTI-Pro. At
this point you'll have to find another DAC or transport to eliminate
either one. Perhaps an equipment list would be helpful in fixing this
problem. What do the dropouts sound like?
Tony Fordham
In article <5ev3k2$h...@eyrie.graphics.cornell.edu>,
and...@alumnae.caltech.edu says...
> ro...@novagate.com (Tony F) writes:
> >Could this be why I'm experiencing very brief intermittent drop-outs?
> >I'm using a Van Den Hul 75 ohm digital cable with RCA ends but on the
> >jitter reducer (DTI Pro) side I have to use one of the BNC adapters
> >that came with it because it doesn't have RCA. On the DAC end I have
> >both inputs so I've tried it both ways but this didn't make any
> >difference.
> Actually, in my modified transport, I'm using van den Hul's 75 Ohm
> cable (the light blue one), and I haven't noticed any dropouts. I've
> also used the DTI-Pro with the BNC-RCA adaptors and haven't noticed
> any drop outs there. This was with Toslink feeding into the DTI-Pro.
I was also using the light blue Van den Hul 75 ohm cable.
> >I also tried a Radio Shack video cable (still using the adapters) but
> >this made no difference, still got the drop-outs.
>
> This probably indicates that the problems lies not in the cabling.
> Have you tried connecting your transport directly to your DAC? If you
> still get dropouts, then, at least you know it's not the DTI-Pro. At
> this point you'll have to find another DAC or transport to eliminate
> either one. Perhaps an equipment list would be helpful in fixing this
> problem. What do the dropouts sound like?
When I connect the transport directly to the DAC I have no problems.
But the thing about it is this, before I got the DTI-Pro I had a SF
Ultrajitterbug. With it the drop-outs were even worse than they are
now, so I returned it (SF tech support had some suggestions but they
did not solve the problem). Since my original post I have noticed
that the drop- outs become fewer in frequency as the unit is on longer
(maybe as it gets more warmed up) to the point where there is only
maybe one per every hour or two of listening. My CD player is a Sony
CDP-CX250 (which is the main reason I got the DAC and the jitterbug in
the first place, it has quite a grainy sound with it's internal DAC).
My DAC is an Assemblage DAC-2. The drop-outs sound like nothing - for
just a very brief split second there is no sound. With the DTI-Pro
the drop-outs seem to be briefer in duration and less abrupt sounding
than they were with the Ultrajitterbug. I'm starting to wonder if
maybe the Toslink sending unit in the Sony is the problem (maybe the
jitter reducers are more sensitive to whatever it's doing than the DAC
is).
I've been looking into adding a coax out to the Sony and I think I may
pursue this.
Anyway, I would hate to give up the DTI-Pro because I love what it does
for the sound of my system.
Thanks,
Tony F
Arnold B. Krueger
We have a bit of a controversy...
Steven R. Rochlin <enjoyt...@top.monad.net> wrote in article
<5edb54$h...@tolstoy.lerc.nasa.gov>...
> ...on Sunday night I was subjected to an ABX test by none
> other then E. Brad Meyer of the Boston Audiophile Society.
> We used his system (for which I'm not too familiar with audibly
speaking).
>After 7 trials I scored 6 correct answers in the digital cable test.
(Writer's note - IME Brad Meyer is a fine gentleman whose hospitality
I have enjoyed in the past when traveling on business in Boston. He
provided this reply by private Email with permission to post.)
Brad writes:
"Here's a message to post on that part of the Net; I don't have access
to it, and I'd be grateful if you would forward it.
"To all who have seen my name mentioned in connection with alleged ABX
testing conducted by Steve Rochlin: Mr. Rochlin has, I understand,
been claiming to have established the audibility to him of differences
in digital cables at my studio, in a test conducted by me. Please be
advised that Mr. Rochlin's score on the digital cable test was not
6/7 but 1/7, and that the test was not an ABX test but a single-blind
test in which I switched the cables manually behind an equipment rack
next to the subject.
"Rochlin had a few minutes to familiarize himself with the "sound" of
the two cables and then got 6/7 wrong. He then decided that he just
was confused because he thought all he had to do was identify when the
cables changed, and that he had made this mistake because the cheaper
cable sounded more "integrated" than the expensive one with my
particular hardware -- an NAD 5000 CD player feeding a Lexicon 300
acting as a D/A converter -- rather than the other way around. This
meant to him that his real score was 6/7, because if the answers were
reversed this is what his score would have been. I proposed that as a
working hypothesis for further experiments but do not allow it as a
valid interpretation of what took place.
"One reason to doubt Rochlin's interpretation was that if he was
listening only for changes in the cable, between trials two and three
he announced that he was very sure that there had been a change, and
those two trials used the same cable.
"When we tallied the results he pronounced himself satisfied that he
now knew the differences in the cable sounds. Good, I said, then lets
do three or four more trials and prove it. "Oh, no," he said, "I'm
going to quit while I'm ahead." I have offered him further trials so
he can conclusively prove that he hears the difference. At this point
I don't expect him to risk being proved wrong.
"FWIW, a score of 5/7 happens just over 1/5 of the time by accident,
giving that score a significance of about 78%. In fact, on that very
trial, I flipped a coin to determine which cable to use and it came up
heads 5/7 times! The usual criterion for a valid positive result is
95%, which for seven trials would require 7/7 correct. A score of 6/7
is close, at around 94%. This makes the test result interesting, and
I would very much like to do further tests with someone who seems on
other evidence to be a very keen listener. -- E. Brad
Craig Stark
> which is an approximation of binomial. Binomial has a lot of factorals
Nope, I'm using an exact Binomial. SPlus is about as hard-core a
package as you're going to find. The test I reported is a two-tailed
test and you're reporting a one-tailed test. I've had a discussion
with Tom about whether we should be doing one or two-tailed tests on
these data and we've agreed to disagree on this fairly minor detail.
> From doing a lot of ABX tests I can tell you that 20:1 shots do happen
> about one time in 20 (!). And, surprise, surprise, when you do more
> trials to improve your results things have this irritating tendency to
> go towards random guessing, more times than not.
Well, unless, of course there is something there...
> This is, of course, the classic test of the binomial theorem. The
Yup, hence the demo. A few weeks ago I gave it in my class and it
really does drive home a point. People tend to think that if they do
well on a test (such as a blind cable audition) over a few trials that
they are really accurate. They may be, but there is a chance that it
is merely chance. Get 20 people in a room doing an ABX test and
you're bound to get one who looks good over a few trials-- just do the
math. (Did anybody see the gerbil that predicted football wins? Case
and point.)
- Craig
Steven R. Rochlin
Arnold,
> My table says 0.062, but no matter how you split it, this is not even
> 95% confidence. One would like to see 99% confidence before making any
> public claims.
First, my humble apologies for chiming in late. Had to go to NYC
for a week. My usual fun and bagel/pizza addiction :-{)'''' .
As for the tests, well, i'll uphold my promise to not say anymore
about them per se. It satisfied me and my curiosity. If you 95%
confidence then i'm ok with that. Personally i could really care
less. i heard differences before the test. i hear differences after
the test. 'Nuff said
> Now, try
> > the experiment a few times. Were you ever right 5/7 times? 6/7? Do
> > you think you're really psychic?
(Said with humor) Psychotic, maybe. Psychic, hmmmm, gotta call
my friends on the phone for that :-{)+ .
Bottom line, no more tests for me. Since i personally proved
something to myself then i'm happy. My curiosity has been filled. i
hear differences and my time is getting too short for things i already
know. The definition of insanity: Trying things over and over
expecting different results. Besides, i'd rather be...
Enjoy(ing) the music (ELO "Brain Salad Surgery" right now),
Steven
"Welcome my friends to the show that never ends..."