Sorry if this has already been posted in this thread: http://www.passlabs.com/downloads/articles/spkrcabl.pdf
Johan Potgieter said:
Yes, the numbers are really staggering.
Unfortunately, a discussion of audibility requires a definition of what is audible. Monophonic hearing capability does not support cable audibility, yet mono criteria has been historically used to refute cable audibility.
I posted analysis graphs at diy a year or two ago, I'll hunt the link down.
The 2 uSec number actually cannot equate to frequency, but I am reminded of early days when I had to buy scope probes...ya didn't buy a 1ghz probe, you bought one with a settling time sufficient for the application. 2uSec would equate to half a meghz, so I mention it just to provide the flavor of attempting to measure temporal shift audibility.
Cheers, John
thoriated said:
Exactly! test results of what Fourier series predicts.
The question then always becomes "so what if those harmonics aren't reproduced? We can't hear them anyway."
But others have suggested that our hearing may actually be quite sensitive to phase relationships well into the 100s of Khz. And I was suggesting that perhaps there is some non-linear mixing of those high frequencies that produce products within the audible range.
What we REALLY need for perfect sound reproduction is a Star Trek "holodeck"
Sorry, I'm WAY off topic.
panomaniac said:
I'm a tech at a cable company, so I'm familiar with the ringing you're talking about. I thought that perhaps you were running HF signals through a waveform monitor or oscilloscope and seeing "ringing" there.
The ringing you're talking about is most certainly reflections, also known as "multipath" due to the impedance mismatches in the switch, which are probably quite severe if it's not a proper video switcher.
The CRT is simply drawing a secondary image which is out of phase with the original signal, but high enough in amplitude to "see".
As a video projectionist, I'd have to ask if any of the video you were projecting had been digitally encoded? This "ringing" effect is quite common on lower quality Mpeg encoded video.
Interesting article by Mr. Pass. Comments?
And even more than a secondary image. There are several reflections that trail off in amplitude.
It is, but I know an Mpeg artifact when I see one! Same for Jpeg artifacts and "over sharpened" video. They do tend to look similar at 1st glance, though.
But we seem to be saying that these effects will not occur at audio frequencies because the cables are too short. Though one might suspect reading the Nelson Pass article that HF reflections and oscillations might affect the audio band by way of intermodulation.
macgyver10 said:
And even more than a secondary image. There are several reflections that trail off in amplitude.
It is, but I know an Mpeg artifact when I see one!
Same for Jpeg artifacts and "over sharpened" video. They do tend to look similar at 1st glance, though.But we seem to be saying that these effects will not occur at audio frequencies because the cables are too short. Though one might suspect reading the Nelson Pass article that HF reflections and oscillations might affect the audio band by way of intermodulation.
Back to the issue of speaker cable design. One article that I quite enjoyed was this:
http://www.avahifi.com/root/audio_basics/1990-02_wire_challenge.htm
It's purhaps a quite practical, James Randiesque, approach to ending the debate (although with a somewhat soft ending...). I like the idea of throwing down the gauntlet and expecting the manufacturers to just "prove it". Not surprisingly, almost none of them take him up on it.
So here's my challenge:
I've tried, and according to many, failed to measure speaker cable and Cat5 impedance in a manner that would suggest it would make any difference in the audio frequency range. I accept that I made some errors, and those errors may have produced erroneous results.
However, in the interest of "science", rather than trotting out the math and theory to debunk it. Why doesn't someone devise a better test, and falsify my results through experiment? That's the scientist way, isn't it?
http://www.avahifi.com/root/audio_basics/1990-02_wire_challenge.htm
It's purhaps a quite practical, James Randiesque, approach to ending the debate (although with a somewhat soft ending...). I like the idea of throwing down the gauntlet and expecting the manufacturers to just "prove it". Not surprisingly, almost none of them take him up on it.
So here's my challenge:
I've tried, and according to many, failed to measure speaker cable and Cat5 impedance in a manner that would suggest it would make any difference in the audio frequency range. I accept that I made some errors, and those errors may have produced erroneous results.
However, in the interest of "science", rather than trotting out the math and theory to debunk it. Why doesn't someone devise a better test, and falsify my results through experiment? That's the scientist way, isn't it?
An externally hosted image should be here but it was not working when we last tested it.
[/IMG]
This is the suggestion for a differential listening test from the german hififorum.de: http://www.hifi-forum.de/viewthread-42-87.html
Any audible differences in cable - it should work also with loudspeaker cables - should show up in some kind of signal in one of the channels.
Criticism of this test involved threshhold (if the signal is so small not to be perceived - what does this say about audibility of differences?) and not being able to perceive any differences in spatial presentation of cables.
This is beside the point here - if there exists a difference between cables that have any audible significance - a signal should be audible from the speakers. If the cables transmit the signal without differences - no signal is present.
A clever idea. But I see can some problems with it. If you do find a difference, then how do you know which cable is "best"?
You would need to start with 2 identical cables, just to be sure that nothing else in the system is causing an imbalance.
Then how do you load the cables? Are the cables post amp? If so how would you wire that?
I had been thinkng about a Wein bridge or null type of test that could show very slight differences. This is along the same lines.
You would need to start with 2 identical cables, just to be sure that nothing else in the system is causing an imbalance.
Then how do you load the cables? Are the cables post amp? If so how would you wire that?
I had been thinkng about a Wein bridge or null type of test that could show very slight differences. This is along the same lines.
What in the world could be discussed about spatial dimensions on a forum???macgyver10 said:More than likely, however I'm jumping topics from another forum I'm on right now where we're arguing spatial dimensions. That seems more appropriate for a holodeck reference
I did not expect anyone to take the challenge. Why should a manufacturer do so? Is it in his best interest?macgyver10 said:
Excerpts from the article:
""This cable measured very low capacitance — 113 pF for 2 meters, but we could induct more hum into this cable than we could in either the blue Kimber Kables or Radio Shack cables by holding an AC power cord close to them — even with the cable ends shorted together. The increase in noise was not terrible, but it was observable. We would suggest you be very careful in using this cable in long runs near AC power lines. You may get lower noise with better shielded cables.""
The reason the cable was more sensitive to hum was because the capacitance was lower...this means that the inductance was higher. Inductance is, by definition, the relationship between the current and the energy that is stored within a magnetic field..Higher inductance, more external field... The consequence of that, is the higher inductance means external influences will be more coupled to the wires, hence more hum sensitivity.
""We compared one channel to the other attempting to hear any difference between the speakers. Nobody could detect any differences.""
Why in the world would one use MONOPHONIC tests??? This is only testing the frequency response sensitivity of humans, but has nothing to do with localization..
While interesting reading, the article presents no actual methodology consistent with the problem.
macgyver10 said:
You have not failed. What you have done is repeated a test which has been done now for 2 or 3 decades, by hundreds of people. The fact that you have duplicated previous results is good.
macgyver10 said:
To devise a better test, you first need to know what it is you are looking for. If you wish to spend the next 3 decades repeating the same tests, only to get the same results, then I will not be one to stop you.
I have pointed out what you will see at the extreme, using conventional analysis, and confirm that nothing further is to be seen using conventional techniques and conventional wisdoms.
How does one devise a "new" test, if one does not have new knowledge? That is why theory and math is so very important. One must understand exactly what one is testing for.
Cheers, John
Hi macgyver10,
I agree with John. One very important thing that you did find out was that the differences would be swamped by amplifier output impedances and speaker impedances.
I don't think anyone is criticizing your test, just suggesting how to make it more relevant to it's intended task.
As I said before, you took the time and trouble to plan out and execute a test. You didn't do such a bad job either.
-Chris
I agree with John. One very important thing that you did find out was that the differences would be swamped by amplifier output impedances and speaker impedances.
I don't think anyone is criticizing your test, just suggesting how to make it more relevant to it's intended task.
As I said before, you took the time and trouble to plan out and execute a test. You didn't do such a bad job either.
-Chris
FWIW, when I once compared Kimber speaker wires to some older Tara Labs cables, I felt the Kimber speaker wires gave a tonally more neutral presentation overall by comparison but lacked slightly some of the inner detail of the Tara labs cables, while the TARA cables sounded a bit rolled at the highs and a little colored. The construction of the two types of cables was quite different though, being that the Tara cable appeared not much different than two twisted pairs of conventional solid core 12AWG THHN, although with better materials, while the Kimber cables consisted of several pairs of braided conductor with teflon insulation for each multistranded copper conductor.
From this and other listening experiences with wire, both at speaker and line level, I drew the empirical conclusion that individually insulated braided solid conductors (with low dielectric constant insulation) offered the highest neutrality at least choosing from these configurations.
I then constructed my own speaker cables for a bit over a hundred dollars for the pair that IMO outperformed both these cables (that were loaners), using 4 braided pairs of 14 AWG solid conductor copper each (equivalent to 8 AWG overall per conductor) with 1/8" thick foamed polyethylene insulation (well bonded to the conductor so there is no fear of surface oxidation over time, unlike with much commercially available wire). I formed the individual conductor ends into ring lugs soldered with 4% lead free silver solder, and have happily used this cable since. If I were to build more speaker cables, I would do the same, but use 18 AWG, since while the inner detail was as good as I've heard the highs were just a tad deemphasized subjectively with this cable, and I would imagine it's probable that a little DCR could be traded to improve this to advantage as well as making the cable a bit more manageable physically (the cables I made are about 1" diameter).
From this and other listening experiences with wire, both at speaker and line level, I drew the empirical conclusion that individually insulated braided solid conductors (with low dielectric constant insulation) offered the highest neutrality at least choosing from these configurations.
I then constructed my own speaker cables for a bit over a hundred dollars for the pair that IMO outperformed both these cables (that were loaners), using 4 braided pairs of 14 AWG solid conductor copper each (equivalent to 8 AWG overall per conductor) with 1/8" thick foamed polyethylene insulation (well bonded to the conductor so there is no fear of surface oxidation over time, unlike with much commercially available wire). I formed the individual conductor ends into ring lugs soldered with 4% lead free silver solder, and have happily used this cable since. If I were to build more speaker cables, I would do the same, but use 18 AWG, since while the inner detail was as good as I've heard the highs were just a tad deemphasized subjectively with this cable, and I would imagine it's probable that a little DCR could be traded to improve this to advantage as well as making the cable a bit more manageable physically (the cables I made are about 1" diameter).
jneutron said:
How objects like a torus can be described using only 2 dimenions, and that curvature is intrinsic to the manifold and doesn't require higher dimensions to explain....that sort of thing....
jneutron said:
I meant to say that I have failed to provide evidence of anything new. I took a look at Nelson Pass's tests, and they were much like my own (high impedance source, voltage divider with low terminating impedance....0.1 ohm in fact!) I don't feel too bad that my attempt can be lumped in with someone like that.
jneutron said:
This is why I ask that somebody do these tests, because I'm not the one with the knowledge to do them. I can't say that I truly understand what you're looking for. My understanding is completely "conventional".
Of course you have to begin with a hypothesis -- perhaps we could start there? What would you present as your hypothesis?
I'm asking this completely from a humble position, I feel that you're discussing theories that I don't quite understand, but would like to.
- Status
- Not open for further replies.