Should be easy enough to do the test? It does mean I'll have to buy another speaker though 😉 and some zip wire, I use twisted mains conduit wire at the moment, I suppose I could use some loose wire I have knocking around to do the initial test
No. There has been no logical scientifically supported argument that the cable should matter. No measurement, no experiment, no ABX, nothing. Just someone's unsupported hypothesis. One link leading to technically flawed article. But the ignorant crowd is happy. I will not continue in useless turn in circle debates. Only results count to me. I have shown mine, waiting for those of others.
Last edited:
Ultimately I think if there is any image shift to be heard it will just show how important it is to have symmetry between channels as far as is possible
Yes, of course they can. However, we do not necessarily choose to measure everything that might be audible for at least some listeners. For instance, there are now some dacs that measure quite well, Gustard offering one example. However, their great measuring dac is quickly earning a reputation for sounding worse than some other dacs in the same price range that do not measure as well. It means that designers are starting to get good at designing to a standard regimen of tests, but they still have not learned to listen. Or, it could mean that the standard regimen of dac tests needs to be augmented with additional tests that better correlate with how humans hear.
That is what I asked ScottJ about his fired off post (as if it was the end of the matter) in answer to Johnego statement "What is your criteria of accurate to the source?" "Preservation of all the cues that aid production of a believable image in the auditory cortex of the brain. Next question"
If we don't know all the "cues", we can't assure that we are measuring audio signals fully or even accurately (as far as auditory cortex is concerned). Accuracy defined simply as the set of measurements we choose to use (because they were historically used) is not very rigorous
This is the concept that many measurists need to grasp
Last edited:
Awww, it's so nice of you to think of me, why thank you.
My cables: lots of neutrik and 12/3 in lengths from 6 to 90 feet. Sams club had a sale on 90 foot #12 extension cords in neon colors..filled a shopping cart..90 footers neon green, 45 footers neon purple, 22.5 footers neon orange, small various neon yellow. Nobody wants to steal my cords...go figure..
1/4 plugs with 12 zip in 25 to 150 foot lengths.
simple stripped 14-2 zip in 50 to 100 foot lengths
lepai 20/20 w t amps
LM3886 single chippers
Dynaco 400
Tigersaurus pair
coupla QSC rms1450
I combine then in any way I need, and the needs vary tremendously.
Never ever had an issue, ever.
The neatest thing was when I had to make a 25 foot long 4 ohm cable to feed a 4 ohm cryogenic stripline to a 4 ohm bifilar 1 mil stainless steel heating element I designed and planted into a 2000 ampere superconducting quadrupole I built that ran in a 3 tesla background field. The purpose of the heater was to instantly dump energy into the magnet. Pulses of 40 volts, rise time 10 uSec, duration variable almost to 10 milliseconds.
The problem was, there was a very large electron beam that would pulse through the magnet, and the radiation spray would put energy into the niobium titanium. Because metals have almost no heat capacity at 1.8 kelvin, we needed to test how much instant energy the coil could take without quenching.
The e beam pulse was 192 nanoseconds long, but there was no way I could get a 40 volt ten ampere pulse at that speed. So, I made a non inductive resistor capable of single digit nanosecond response, connected 6 cat5e cables in massive parallel for 4 ohms, used 1.5 mils of kapton between two 3 mil thick copper ribbons .5 inches wide, connected a tigersaurus and pulse generator, and bingo, worked like a charm. 10 uSec, 40 volts, and a pulse width variable from about 10 uSec out to 10 milliseconds. Also made a 50 ohm copper/kapton stripline to monitor the voltage at the heater. Flat top, no ringing, 10 uSec to full current...
Physicist did his thing, wrote a paper, another happy customer.
What did you do today Pavel?
jn
Last edited:
I have never considered envelope or timbre, so I would have to defer to someone more expertise in those matters as to listening tests, as I have no content to add in that regard.
The cables should not be capable of altering envelope, and I suspect not timbre either.
jn
For a short line at most audio frequencies if R is irrelevant then L is even more irrelevant, as R exceeds the impedance of L.scott wurcer said:
The energy in a short line which is terminated in its RF impedance is stored equally in L and C. That is simply a consequence of the coincidence that the RF impedance approximation is sqrt(L/C) (which only applies for RF) and the value of terminating resistance which happens by Ohm's Law to give the right current and voltage to give equal energy storage is also sqrt(L/C) (which is valid down to DC). The fact that two different concepts with different domains of applicability happen to be numerically equal does not mean that they are the same concept.
Given an inductance and a capacitance the rules of dimensional analysis mean that about the only way to get a resistance from them is to form sqrt(L/C). Thus the answer to both (quite different) questions has to be a dimensionless number times sqrt(L/C). It just happens to be the case that the number is 1 for both questions.
You persist in taking a short delay and inverting it to get a frequency. Why? Let us be clear: I am talking about audio frequency signals in a cable which is dominated by R and C. The cable causes phase shifts. It cannot cause reflections of any time period, because such a cable does not support wave propagation; it is simply a distributed low pass filter.jneutron said:
Not a bad idea. I remember Nordmark had ungodly numbers like 1.2 uSec, but recent researchers are showing 6 plus. I am wondering if the use of headphones (Nordmark, lateralization) is a more discerning test that the use of speakers in a room for localization.
Equal lengths, well I would guess within a foot or two wont make a difference.
jn
I believe that it follows logically from the idea that timing shifts in individual frequencies can happen at the speaker load possibly causing perception of ITD that therefore envelopes (which are comprised of these same frequencies) can be affected - I suggest in shape.
A change in envelope shape is a change in timbre, AFAIK.
You crack me up. A step test that is over in 10 uSec has no audio frequencies. Why you continue to talk about an entity that cannot exist in a 10 uSec timeframe is beyond me. Unless, your intent is obfuscation. (you can tell, I learned a new word..)
Equal energy.
E = 1/2 L I^2
C = 1/2 CV^2
equate both...
1/2 LI^2 = 1/2 CV^2
LI^2=CV^2
L/C=V^2/I^2
Sqr(L/C)=V/I
jn
Last edited:
The series inductance (Ls) of 6 feet of 12 awg zip cord would, alone, cause a 1dB drop at 10KHz for typical dome tweeters (Le).
If Zo of amp rises with freq, (VFB) then that is added to the zip cord for greater atten. at the tweeter.
THx-RNMarsh
If Zo of amp rises with freq, (VFB) then that is added to the zip cord for greater atten. at the tweeter.
THx-RNMarsh
Last edited:
Yes, of course. However, few people here (or anywhere else) have superconducting audio cables. Hence over much of the audio range R dominates over L, and C dominates over G. Only right at the bottom do we have an RG cable, and only around the top (or a bit higher) do we have an LC cable. In the middle we have an RC cable, for which diffusion is a better conceptual model than wave propagation.scott wurcer said:
Take a sheet of metal. Heat it at one point (or a small area). The heat will flow out to the rest of the metal. When this heat flow reaches the edge does it reflect back to the hot point? No, because there is no wave propagation. Of course, the edge heating up will reduce the heat flow to the edge but this is not reflection. Over much of the audio frequency range this diffusion model is what we should think of when considering a cable; not a wave propagation model.
So you instantly recommend parallel connections of zip to lower the inductance and center the HF image.
No fair, we need tests!!!😉
jn
I am not putting steps into cables. I am taking about putting audio into cables. Why do you talk about something different?jneutron said:
Thanks for showing the derivation of the equal energy condition. It may help those who have never seen it before. Nothing to do with calculating characteristic impedance, but this just happens to give the same result if you use the high frequency approximation.
- Status
- Not open for further replies.