On Mon, 2 May 2011 04:09:19 -0700, Gary Eickmeier wrote
(in article ):
"Audio Empire" wrote in message
...
On Sun, 1 May 2011 16:00:25 -0700, Gary Eickmeier wrote
(in article ):
"Audio Empire" wrote in message
...
A lot more work needs to be done, but it looks like the British
measurement
company, Acuity Products (http://www.acuityproducts.co.uk/) has
developed
a
waveform analysis test for the effects of cables on sound(!).
Listen fellers - this stuff is really incredibly simple.
The only way to tell if a SOUND is AUDIBLE is with a LISTENING test. All
the
king's meters and all the king's scopes do not mean anything unless
correlated with LISTENING tests.
All together now: How can you tell if two cables sound different from
each
other?
______________ ___________
Gary Eickmeier
Easy. They don't because they can't. speaking of effects that are below
the
threshold of hearing!
Probably correct! But they are going on and on about all of these technical
details without correlating anything to a listening test. And a lot of
responders are getting sucked in by it.
I'm just sayin'.
Gary Eickmeier
That's why I brought it up. OK, they have developed software based on sonar
correlation algorithms. Sounds reasonable that some navy has developed
sophisticated programs that can "listen" through the background clutter of
undersea spunds and find intelligence such as prop noise, engine signatures,
and can even tell hull reflections from temperature inversion layers or
whales. It is also reasonable that these programs can be modified to look for
other things within an audio signal. Things such as minute differences in the
way that two different cables conduct an audio signal, or how two different
amplifiers do their job.
I might even buy that such software could tell the difference between a
component powered through a mains conditioner, and one that wasn't (although
that's really stretching my willing suspension of disbelief),
But, given all that, there remain two points about this test that I cannot
get around.
1) If such military auto-correlation software were designed to pick up sounds
in the water that trained human sonar operators cannot hear, and thus miss
when doing strategic listening, what does that say about the audibility of
the differences picked-up by the modified software? Nobody has ever said, for
instance, that cables and interconnects have NO effect on the signals they
pass. We all realize that any conductor will have resistance, capacitance,
and inductance. What is important is that over the audio spectrum (and quite
a bit beyond) and in the lengths commonly used for a home audio system (a
couple of meters maximum for interconnects, and probably ten meters maximum
for speaker cable) that the effects of these three parameters are so far
below the threshold of human hearing as to be inconsequential. The fact that
a computer algorithm can detect these minute differences between these
characteristics in different cables (if indeed it can), is completely
irrelevant to music reproduction except in, perhaps, the most academic
manner.
2) If the people who developed this test methodology are technically
competent enough to conduct these tests and to compile the data shown in the
paper, then why is it that they don't seem to understand that results with as
many variables as these tests seem to exhibit don't MEAN ANYTHING?
Instead of saying that their tests reveal that the Vertex mains cables BY
ITSELF showed a significant difference in the cleanliness of the component's
sound, for instance, or that the Nordost mains conditioner BY ITSELF effected
a marked improvement in noise, these people lumped them both together. This
is ludicrous! Was the "improvement" in noise a result of the mains cable or
the mains conditioner, or the sum of both? And is the difference with and
without these two components in the tests setup something that one is likely
to hear? Nowhere in the entire paper are we told the scale of the shown
oscillographs except to say that the noise difference scale was 10X that of
the actual mains noise seen on the raw AC. The entire paper is riddled with
fundamental testing "errors" of this sort.
Frankly, I have looked at the raw AC feeding lots of amplifiers, pre-amps and
CD players on a dual-trace oscilloscope and then looked, simultaneously, at
the DC on the other side of that same power supply, and even at as much as
100X the gain of the AC trace, I have NEVER seen the AC line noise, even
severely attenuated, show up on a component's power supply DC. It just isn't
there. The reason? Power transformers designed for 50-60 Hz simply haven't
the bandwidth to pass the high-frequency grunge that can be riding on our
house current. Add to that the time constant of the filters on the output of
the rectifiers (and the frequency response of the rectifier diodes
themselves) and again, they act as low-pass filters. High-frequency noise and
transient spikes haven't a chance of getting through. So, while using power
line conditioners sounds, on the surface, like a good idea, the reality is
that they are totally redundant. That's why I say that if your components
NEED such power conditioning, then you bought very poorly designed
components. Expensive IEC line cords? Don't be ridiculous! That the last 2
meters of WIRE from the wall plate to your component could do ANYTHING to
improve the quality of an AC current that has probably travelled hundreds of
miles over all types of terrain and through countless transformers and
switching yards, is at the very least, the height of electrical naivete.