[Joong]

Quote:

+1
 
No science involved in that one.  No cable twist changes reactive parameters enough to affect anything audible.
 
However, reducing resistance very definitely can be audible, something like changing from 18ga to 12ga over 50' or so.  Depending on the speaker load, that change could be audible.  But without much wire resistance you have to have a TON of L and C to audibly change much around an 8 ohm load with a near zero ohm source.

There is science in twist due to generate magnetic dipole or multi-pole of opposite winding neighboring pairs.
However I do not know the reduction due to more than magnetic multi-pole is enough audible amount or not.
 
I think reactance of cable can be combined with that of driver ( especially dynamic phone) in such that the combination makes a filter to affect sound.
Here I am very interesting in relating to the 2nd or higher order amount of electrical quantity to SQ.
The fundamental order quantity can be easily charted and understood in terms of SQ like THD, IMD, or resistance etc.
However even harmonics impact to SQ is questionable because it is an octave higher.
 
I want this thread to keep going.

 

[jaddie]

Quote:

There is science in twist due to generate magnetic dipole or multi-pole of opposite winding neighboring pairs.
However I do not know the reduction due to more than magnetic multi-pole is enough audible amount or not.
 
I think reactance of cable can be combined with that of driver ( especially dynamic phone) in such that the combination makes a filter to affect sound.
Here I am very interesting in relating to the 2nd or higher order amount of electrical quantity to SQ.
The fundamental order quantity can be easily charted and understood in terms of SQ like THD, IMD, or resistance etc.
However even harmonics impact to SQ is questionable because it is an octave higher.
 
I want this thread to keep going.

Yes, twisting conductors reduces inductance...but, if the cable inductance is already down quite low, making it a tiny bit lower wouldn't result in an audible change.
 
Here are some numbers.  
 
16ga flat wire has about .18uH per foot inductance.  
16ga twisted wire has about .15uH per foot inductance.
 
In order to detect an audible change we need to affect at least a .25dB change over a 1/3 octave in the mid band.  Since inductance has more effect as frequency goes up, lets look at what it might take to change 20KHz by .25dB.  Every speaker is different, but it's not unusual for impedance to be rising above 10KHz.  Many speakers end up at around 10 or 11 ohms at 20KHz.  So, just working with 10 ohms as an example, for there to be a .25dB loss we would need .29 ohms of inductive reactance at 20KHz.  That would be 2.32uH, or 12ft of flat wire, or 15ft of twisted wire.  To put it another way, 12ft of flat wire could possibly produce a .25dB loss at 20KHz while 12ft of twisted wire could possible product a .19 dB loss.  Thats a difference of .07dB.  
 
The problem is, we can't hear a .25dB change at 20KHz!  We only have that kind of sensitivity in the mid frequencies.  So, we need a MUCH larger difference in inductance for it to produce an audible effect.  And, as you can see, just twisting the same gauge wire won't make nearly enough difference. 
 
You can always say that a speaker with a much lower impedance at 20KHz will produce a more dramatic change.  Well, it would have to.  We can't hear .25dB difference, so we sure can't hear a .07dB difference.  If you had 10 times the wire length, we could talk about it.  Otherwise, wire twist, while reducing inductance, doesn't reduce it enough to be audible.  
 
As a footnote, the two wire types used in the examples are manufactures data, and while they are the same conductor gauge, and differ in the fact that one is flat and the other is twisted, conductor spacing has a large effect on inductance as well.  The greater the space between conductors, the more inductance.  I found it pretty much impossible to find two wires with identical conductor spacing, but differing only in twist.  But that fact can practically be ignored, because the figure you need to concern yourself with is the inductance per foot of any wire, twisted, flat, ribbon, or coat-hanger.  That, of course, being completely secondary to DC resistance.  

 

[mat-t]

truly a science. Thanks for sharing guys! loads to learn

 

[jaddie]

Quote:

Here I am very interesting in relating to the 2nd or higher order amount of electrical quantity to SQ.
The fundamental order quantity can be easily charted and understood in terms of SQ like THD, IMD, or resistance etc.
However even harmonics impact to SQ is questionable because it is an octave higher.

While it's true that harmonics add character to the fundamental,  it is also true that the range of human hearing is limited to 20KHz for good hearing, as high as 24KHz for exceptional young ears before their first concert, and below 15KHz for older adults.  Hearing response is not a hard limit, but a slow roll-off in sensitivity.  When we say that a young person can hear 24KHz, that means that is the highest detectable frequency, and the level at which that frequency must be presented is much louder than it would be it it were a harmonic in music.  Though music has measurable high frequency information, it's a simple fact that it isn't audible above our own individual maximum audible frequency, and in practice, quite a bit below that.  So, the second harmonic of 20KHz is 40KHz, way out of our hearing range.  The second harmonic of 10KHz is 20KHz, which is just in our hearing range.  This means that harmonics above 10KHz add little if anything to the SQ.
 
The argument in favor of higher bandwidth audio usually centers around some form of high bit rate audio, like 96KHz, which has a theoretical maximum usable frequency of 48KHz.  Unfortunately, the comparisons that are usually made are mostly invalid because when comparing an "original CD" and a "high rate file", we have no idea of the total audio path.  We are not simply comparing the result of higher bandwidth, we are comparing the entire audio chain that gets to the file, and the playback chain too.  Those who have had the rare opportunity to compare the same analog signal sampled simultaneously by two identical A/D converters, but run at two different rates, have shocked by how little, if any, difference there is.  
 
The second pro high-bandwidth argument usually includes a comparison between some digital format an an analog one.  The same problem exists, though, as again you aren't comparing just the difference in the final medium, but the difference in the entire path to the final medium. 
 
But in reality, if someone can't hear 20KHz, then he can't hear the second harmonic of 10KHz either.  Or the third harmonic, or any others.

 

[mikeaj]

Maybe I'm just misinterpreting things, but I'll just throw it out there that in some audio circles, people talk about 2nd-order and 3rd-order effects not in terms of harmonic frequencies but in a more general sense.
 
But anyway, if we're talking about 2nd harmonics, there really aren't a lot of tones out there in music with fundamentals anywhere around 10 kHz anyway.  Ear-piercing piccolo stuff is say 2-4 kHz fundamental.  It's the more broadband sounds, and mostly weak higher harmonics, that breach 10 kHz.

 

[jaddie]

Quote:

Maybe I'm just misinterpreting things, but I'll just throw it out there that in some audio circles, people talk about 2nd-order and 3rd-order effects not in terms of harmonic frequencies but in a more general sense.

Yes, I know.  A lot of people choose to deal with things only in the "general sense", which allows enough room to be vague and still be sort of right.
 
Quote:

But anyway, if we're talking about 2nd harmonics, there really aren't a lot of tones out there in music with fundamentals anywhere around 10 kHz anyway.  Ear-piercing piccolo stuff is say 2-4 kHz fundamental.  It's the more broadband sounds, and mostly weak higher harmonics, that breach 10 kHz.

Percussion gets up there pretty high.  A triangle hit goes will into utrasonics.  A lot of what's heard in the top octave is transients in support of lower sounds.  That's why if you take off the octave above 10KHz things sound sort of papery and dull, but you don't have a problem identifying instruments.

 

[stv014]

Quote:

Originally Posted by jaddie /img/forum/go_quote.gif
 
So, just working with 10 ohms as an example, for there to be a .25dB loss we would need .29 ohms of inductive reactance at 20KHz.

 
And even that is only true if the speaker impedance is purely inductive. If the 10 Ω impedance is 8 Ω resistance and 6 Ω serial inductance, for example, then 0.475 Ω of additional serial inductance would be needed for a 0.25 dB loss.

 

[doublea71]

You guys are speaking a language that I can't comprehend

, but I still appreciate it and try to glean anything I can from it. On another note, a funny post that I read recently was made in which the poster said that their new mini-to-mini connector was "the best sounding and increased the clarity, tightened the bass, etc..." and this was a very short piece of cable, perhaps 10-12 cm in length. It was connected to about $2,000 worth of portable high-end gear. It struck me as a whole new level of expectation bias...

 

[uchihaitachi]

I am surprised this thread isn't closed yet. Doesn't everything here go VERY against the sponsors lol? 

 

[anetode]

Sure there's an occasional raid on the ghetto, but the sponsors don't care so much about this place as compared to the headphone or cable subforums.

 

[doublea71]

hahaha raid on the ghetto lmao
 

 

[Joong]

Hi guys
 
do you know that toxic cable has some trouble?
That company is known to make a quality cable with Helix wound type to improve SQ?
This company has a strong belief that SQ is strongly depending on what they are doing.
 
I am not sure what they do, even though I ordered He-500 cable 3 months ago, but they still require me of great patience.
 
I need to rethink in order to return to the solid science based on the facts on this thread.

 

[xnor]

I heard they have problems in their unicorn horn and fairy dust supply chain.
 
On a more serious note, I wouldn't even consider buying from a site that doesn't show proper legal notice.

 

[Joong]

Alas! when I changed from the 1meter long Western electric RCA cable (40 USD) to cheap and shorter (30cm)  RCA cable, SQ noticeably improved.
It is very obvious that the cost of the cable does not matter but the length of the cable does matter.
 
For this experiment, the common science stands and speaks.
In terms of impedance change, it must be negligible!!!!!

 

[xnor]

Well I agree that cost doesn't matter, nor does how fancy the cable looks. All it needs to be is well built and not excessively long.
 
Rod Elliot measured some interconnects:
The resistance (0.12 to 0.38 Ohm) and inductance (0.6 to 1.2 uH) "may be considered negligible at audio frequencies".
The capacitance was the dominant influence. A single core cable had only 77 pF per meter, but even with an unrealistic output impedance of the source of 10 kOhms "this will be 3dB down at 207kHz".
 
Line outs usually have an output impedance of a few hundred Ohm max, so even higher capacitance shouldn't matter at audio frequencies.