Tweaks, test results that prove...
Peter Tomaszewski
OK, here is the test of tweaks on demand.
This test was published in a Swedish magazine. They used cd:s from
two different factories and tried four tweak products on them.
The reflection of both of the CD:s was around 75%. Philips specify the
lower limit to 70%. Gold has the best reflection, BUT beacuse gold is
so expensive they usually put on a very thin layer on the CD:s so
that the reflection still is around the same as for aluminum (75%).
The test was performed at Audio Development in Sweden. The company is
one of the world leading in CD test/measuring equpiment and their
products are used by many CD-plants around the world.
The cd.s were:
- Roxettes "joyride" EMI 7960482-59:11
- British and American band classics MERCURY 432009-2-61:30
The Mercury cd:s were manufcatured at Polygram in USA. Their machines
are older and use an older process. The Roxette cd:s were made in
Holland at EMI in a fully modern and up to date factory.
BLER = Block error rate, tells you how many blocks per second that are
faulty. The digital signal is divided into blocks, 7350 every second.
According to the specification only 220 of those are allowed to be
faulty. The errors are grouped into three groups: one symbol-error,
two symbol-errors and more than two symbol-errors. This is done both
in circuit A and circuit B. When a large error occurs in circuit B3
(more than two symbol-errors), THEN the cd-player can be forced to
interpolate, ELSE the digital signal is always correctly retreived
(no errors). As you can see the Cd-player never has to interpolate
and you can also see that the cd from Mercury (made with older
machinery) has more and "worse" errors than the Roxette cd made in
Holland.
The tweaks used were:
-A green pen from AudioPrism, CD-stoplight.
-Damping-rings from Mod Squad, you put it round the center-hole on the disc.
-A CD washer/polisher from Marigo Aurio Lab.
-A Cd washer from Audio Technica.
The untreated Cd was as it came in the box.
Errors/Second BLER A1 A2 A3 B1 B2 B3
Mercury untreated 37 23 5 8 54 0 0
Mercury green pen 24 21 7 8 41 0 0
Mercury damping-rings 31 20 6 18 142 0 0
Mercury AT-wash 44 27 11 19 139 5 0
Mercury MARIGO-wash 33 29 10 11 148 3 0
Roxette untreated 13 12 5 0 5 0 0
Roxette green pen 11 11 4 2 17 0 0
Roxette damping-rings 17 15 5 8 34 0 0
Roxette AT-wash 17 15 4 2 4 0 0
Roxette MARIGO-wash 23 23 7 6 23 0 0
As you also can see that the green-pen (AudioPrism) has so little effect that
you can't tell if it actually helps at all.
Save your money.
Peter
--
Internet: d8...@efd.lth.se |
inca...@solace.hsh.se | It is easier to curse the darkness,
inca...@df.lth.se | than to light a candle...
Paul Guy
>
>
>OK, here is the test of tweaks on demand.
>
>This test was published in a Swedish magazine. They used cd:s from
>two different factories and tried four tweak products on them.
>
>
>As you also can see that the green-pen (AudioPrism) has so little effect that
>you can't tell if it actually helps at all.
>
>
in the original article are a bit suspicious, since they seem to be just
the results of ONE test on ONE CD player use just TWO CD's.
I deal on a regular basis trying to prove or disprove a number of
research hypotheses, and it strikes me that the test is interesting but
doesn't have enough 'data points' to really make a point.
If the results are the averages of many trials, it would add a
tremendous weight to the argument if the averages, and standard
deviations, T and F tests, etc, blah,blah were applied under many
realistic conditions. From my own personal experience, I found that the
error rates on my NAD5100 vary considerably from day to day using even
the same CD, and can vary quite a bit as the player warms up, or if I
clean the suface of the CD.
It would be an interesting test if repeated using many CD's,CD
players, diffferent brands, scratched CD's, etc. The trouble is of
course, it would cost MONEY to do a significant set of tests, and would
need to be set up and performed by a reliable agency. Hell is going to
freeze over before that day comes, despite all the hot air the tweakers
can emit.
-Paul
--
-----------------------------------------------------------------------------
Paul J Guy work phone:519-885-1211 ext 6371
pa...@gaitlab1.waterloo.edu home/FAX/message:519-576-3090
pg...@healthy.waterloo.edu ..remember...bullshit baffles brains...
Kerry Shetline
wrote:
> As you also can see that the green-pen (AudioPrism) has so little effect that
> you can't tell if it actually helps at all.
>
>
> Save your money.
You fail to understand tweakophiles. The tweakophiles **KNOW** that the
green pens make their CDs sound better. So all your test can prove is this:
1) The test was badly run.
2) The test was faked by the evil objectivist conspiracy.
3) Test systems used did not have "high enough resolution." (Ever notice
how some audiophiles won't believe in acoustic masking when you say that
any possible effect of a tweak will be swamped the noise of a refrigerator
two houses down the street, but their explanation of why you can't hear the
tweak is that the distortion of your obviously inferior equipment masks the
benefit of the tweak? Here's to having it both ways!)
4) The supposed audible improvements are due to something apart from error
rate and data recovery. After all "We don't know everything there is to
know about digital audio equipment." We don't know everything there is to
know about food either, so if I claim that eating an apple, another apple,
a banana, and a papaya, with intervals of 4, 3.5, and 7 minutes in between,
will turn a human into a llama, then this, of course, stands as a valid and
sensible assertion until someone performs the experiment. (And then I can
always claim that the timing wasn't precise enough or the fruit fresh
enough.)
So you see, normal (i.e. reasonable) standards of burden of proof must be
set aside. The tweakophiles *KNOW* that their tweaks work. They *don't need
explanations* to enjoy them. (In fact, understanding would probably
interfere: Ignorance is bliss, you know.) It is thus our duty if we care
about such silly things as explanations to further develop the science of
audio until the explanations (which *MUST* exist) are found.
-Kerry
Peter Tomaszewski
Hmmm,
I don't see what you mean with "ONE test on ONE CD player..". You see this
was not an ordinary CD player. A special test equipment was used which
included a special lab quality CD-drive with a lot of different sensors.
You seem to think that it would be a good idea to use ordinary
CD-players for the test, but can't you see that the reason for using
that test equipment was to avoid the fact that different CD players are
DIFFERENT. We only want to measure the effect of the tweak NOT the combined
effect of a not so good cd player and the tweak. You say yourself that your
NAD behaves differently every day and needs warm-up. That is exactly what they
were trying to avoid, those factors.
And the tweak should have an effect even if the CD is not scratched.
Kong Kritayakirana
For ya objectivists: Read my .sig
Best Regards to both camps.
--
Kong Kritayakirana (ko...@leland.stanford.edu OR ko...@nova.stanford.edu)
--
"The foolish reject what they hear, not what they think.
The wise reject what they think, not what they hear." - Apologies to Huang Po
Paul S. Winalski
In article <28kfaa$1...@nic.lth.se>,
d8...@efd.lth.se (Peter Tomaszewski) writes:
|>In article <CE83s...@watserv2.uwaterloo.ca>
|>pa...@gaitlab1.uwaterloo.ca (Paul Guy) writes:
|>> I'm not much of a green-pen believer, but the tests that are quoted
|>>in the original article are a bit suspicious, since they seem to be just
|>>the results of ONE test on ONE CD player use just TWO CD's.
|>> I deal on a regular basis trying to prove or disprove a number of
|>>research hypotheses, and it strikes me that the test is interesting but
|>>doesn't have enough 'data points' to really make a point.
|>
|>
|>I don't see what you mean with "ONE test on ONE CD player..". You see this
|>was not an ordinary CD player. A special test equipment was used which
|>included a special lab quality CD-drive with a lot of different sensors.
|>You seem to think that it would be a good idea to use ordinary
|>CD-players for the test, but can't you see that the reason for using
|>that test equipment was to avoid the fact that different CD players are
|>DIFFERENT. We only want to measure the effect of the tweak NOT the combined
|>effect of a not so good cd player and the tweak. You say yourself that your
|>NAD behaves differently every day and needs warm-up. That is exactly what they
|>were trying to avoid, those factors.
It is a fact of life that there are going to be (hopefully) small differences
in the behavior of test equipment. Furthermore, there are manufacturing
variances between discs of the same music. Also, repeated measurements taken
using the same sample and the same equipment are going to vary slightly. A
properly-designed experiment tries to minimize, or at least quantify, the
effects of these variances by performing the experiment many times.
In the case of this particular experiment, I would want to run the experiment
using:
a) more than one of the special lab-quality CD drives (all of the same
model, and calibrated as closely as possible, but different units). This will
allow one to measure how much variance in the test equipment affects the
results.
b) more than one copy of each CD. This will allow one to measure the variance
between different copies of the CD.
c) more than one (at least 10; the more the merrier) run of each experiment.
So, for example, you would use 3 different recordings, and buy 10 copies of
each. Using three instrumented CD players, play each copy of each of the
recordings 10 times on each player and record the results. Now apply the tweak
to all of the CDs and repeat. Compile all the results, computing the mean
and standard deviation for each measurement. You now know how statistically
accurate your measurements are. The results should be reported as the
mean of the observed numbers of errors, plus or minus some interval, depending
on what statistical level of confidence you wish to achieve. For measurements
in physics and chemistry, an interval of plus or minus 3 times the standard
deviation, or better, should be employed. If I remember my statistics
correctly, that amounts to 99.99% confidence that the true value lies within
the stated interval.
If you do only one measurement, as appears to be done in this experiment,
then all you can say is that things happened to come out that way that one
time you tried it. The results could be a fluke. Without repeating the
measurements many times, you cannot say how accurate they really were.
--PSW
Bob Myers
Unfortunately, Kong, hearing occurs in the brain, not in the ears. You
cannot reject what you think or what you hear SEPARATELY; they both come
from the same place, and are influenced by the same things.
Bob Myers | "Man will occasionally stumble over the truth, but
my...@fc.hp.com | most of the time he will pick himself up and continue."
| - Winston Churchill
Kong Kritayakirana
>> "The foolish reject what they hear, not what they think.
>> The wise reject what they think, not what they hear." - Apologies to Huang Po
>
>Unfortunately, Kong, hearing occurs in the brain, not in the ears. You
>cannot reject what you think or what you hear SEPARATELY; they both come
>from the same place, and are influenced by the same things.
Unfortunately, Bob, Huang Po (or I in "requoting" him my way) does not care
whether hearing or thinking has anything to do with brains or whether the
hearing or thinking can or cannot be separated.
Fortunately, Bob, in a Zen way of life, I don't either agree or disagree with
your statement.
Best wishes
Kong
Karl Hansell
>
>
>OK, here is the test of tweaks on demand.
>
>This test was published in a Swedish magazine. They used cd:s from
>two different factories and tried four tweak products on them.
>
UNTRUE, this is HALF of a test performed by the Swedish magazine
HiFi & Musik (no. 12/91).
I certainly hope that Peter Tomaszewski omited mentioning the other
part of the test because he missed it, because otherwise he has a very
strange view on objectivety.
>
>The tweaks used were:
>
>-A green pen from AudioPrism, CD-stoplight.
>-Damping-rings from Mod Squad, you put it round the center-hole on the disc.
>-A CD washer/polisher from Marigo Aurio Lab.
>-A Cd washer from Audio Technica.
>
>The untreated Cd was as it came in the box.
>
First they measured the BLER, and the amount of correction made by the CD-player,
on all the records, and made the table that was included in Peter's original post.
THEN one the journalists made recordings of each disc (digitaly) to a DAT,
in a A-B-A fashion with another disc. Then he gives the DAT to one of his
colleagues, who has no knowledge of which recording is of which disc, so it
is a good blind test.
He then had to write down which on he did think sounded the best.
He picked the "green" disc over the untreated disc six times out of six !!!!
They also made notes, and afterwards compared these notes, and they are very
much the same !
best regards
-Karl Hansell
*Any flames about spelling or grammatics is kindly accepted in Swedish only*