Some problems just have too darn many variables to reach any general or "rule of thumb" conclusions. Cables are one of them. Not only do you have all the recognized electrical parameters of the cable itself, but you have to take into account the input and output characteristics of the whatever's on each end of the cable. Tubes? Solid state? Hi Z, low Z? C? There are also systematic issues like certain conductor materials being used with certain insulation types for practicality of production. You'll never find tinned copper wire with Teflon insulation for example. It's all silver because it holds up to the temperature of the insulating process.
I've heard definite and distinct differences between ICs, and had very little luck sorting them out by measuring only the IC and not everything else along with it.
Some previous time this came up I wanted to compare copper and silver. I couldn't quite do that, but I did have some 3/8" pure silver tubing, and some 3/8" brass tubing. No doubt, brass would sound brassy ;-) I made up some 3' rigid uninsulated ICs and did both listening tests and measurements. Couldn't tell a bit of difference between them by listening, though out at the MHz there were some measurement differences. My conclusion is that if conductor material makes any difference, I can't hear it. IMO, that test reduced the variables as far as was practical.
Dielectric materials probably have a better chance of making a difference, as they determine capacitance to things like the chassis and to other conductors.
It's also a little disturbing when you see copper conductors that have oxidized and turned green inside clear insulation. Some of that insulation isn't so inert.
I wire my projects with stranded silver plated Teflon insulated hookup wire because I know it won't break and the insulation won't get stiff, not for any sound quality reasons. I make up my own low level interconnects with Belden low capacitance wire, but I don't worry a bit if it has a steel core. My speaker cables are heavy zip cord- FWIW, with a tri-amped system I never hear any difference between speaker cables at all. YMMV.
CH
I've heard definite and distinct differences between ICs, and had very little luck sorting them out by measuring only the IC and not everything else along with it.
Some previous time this came up I wanted to compare copper and silver. I couldn't quite do that, but I did have some 3/8" pure silver tubing, and some 3/8" brass tubing. No doubt, brass would sound brassy ;-) I made up some 3' rigid uninsulated ICs and did both listening tests and measurements. Couldn't tell a bit of difference between them by listening, though out at the MHz there were some measurement differences. My conclusion is that if conductor material makes any difference, I can't hear it. IMO, that test reduced the variables as far as was practical.
Dielectric materials probably have a better chance of making a difference, as they determine capacitance to things like the chassis and to other conductors.
It's also a little disturbing when you see copper conductors that have oxidized and turned green inside clear insulation. Some of that insulation isn't so inert.
I wire my projects with stranded silver plated Teflon insulated hookup wire because I know it won't break and the insulation won't get stiff, not for any sound quality reasons. I make up my own low level interconnects with Belden low capacitance wire, but I don't worry a bit if it has a steel core. My speaker cables are heavy zip cord- FWIW, with a tri-amped system I never hear any difference between speaker cables at all. YMMV.
CH
Jacob2, if we are to debate a subject then no offence should be taken🙂 i air my views others air theirs.
On that note AD and DC convertion is changing an analogue signal into a bit pattern, be it parallel conversion or bitstream (delta sigma etc) and visa versa. It is the bits that define the signal, nothing else, anything else affecting the signal is noise. This method of signal transmission/control is not only used in audio, but is also common in other areas of electronics where an anologue control signal is required between different locations. Motor control is one such area (though numerous control and instrumentation scenarios use the same techniques). It is better signal wise to change an analogue signal to digital to transmit it from one point to another. Not only can error correction be applied but also you dont run into the problems of long cable runs for analogue signals, which do degrade the signals, this is not good when the electronics are controlling critical things such as life support, millitary/areospace etc. In these field everything including full systems are tested to the nth degree, also under adverse conditions full temerature range, EMC testing ...
I am not denying that cables can have no effect on any signal transmission, but generally in audio we are talking of 1m of cable for ICs, what I do keep asking for is some engineering based metrics on how a digital interconnect can effect the sound, and as I've siad if different cables can effect the sound quality noticably they are flawed, or the driving and recieving circtutry is flawed, cables should! be transparent, they are not an active device, and only passive due to their inherent electrical properties (LRC).
Now dielectrics...
On that note AD and DC convertion is changing an analogue signal into a bit pattern, be it parallel conversion or bitstream (delta sigma etc) and visa versa. It is the bits that define the signal, nothing else, anything else affecting the signal is noise. This method of signal transmission/control is not only used in audio, but is also common in other areas of electronics where an anologue control signal is required between different locations. Motor control is one such area (though numerous control and instrumentation scenarios use the same techniques). It is better signal wise to change an analogue signal to digital to transmit it from one point to another. Not only can error correction be applied but also you dont run into the problems of long cable runs for analogue signals, which do degrade the signals, this is not good when the electronics are controlling critical things such as life support, millitary/areospace etc. In these field everything including full systems are tested to the nth degree, also under adverse conditions full temerature range, EMC testing ...
I am not denying that cables can have no effect on any signal transmission, but generally in audio we are talking of 1m of cable for ICs, what I do keep asking for is some engineering based metrics on how a digital interconnect can effect the sound, and as I've siad if different cables can effect the sound quality noticably they are flawed, or the driving and recieving circtutry is flawed, cables should! be transparent, they are not an active device, and only passive due to their inherent electrical properties (LRC).
Now dielectrics...
Well with digital to digital interfaces I would assume jitter can come into play. But every one I have measured mercs it's analog counterpart in FR, PR, and THD. I really doubt people can hear a difference if there is low jitter and the clocks are running stable etc..
My experience, a good silver cable allow more HF detail to survive. If it makes a system sound bright or harsh, look for the problem in the rest of the system, it is just not hiding the flaws. Don't blame the silver cable.
@ marce,
that´s why it is important to keep the different purposes apart.
The transport to DAC over a S/P-Dif (or AES - EBU) connection is a special one as the clock recovery (or more precise the clock construction) inside the DAC depends on the signal.
If done properly it depends only on the preambles, if done a bit less properly it depends on the data (means coded audio signal) too.
The correctness of any analog signal after conversion depends on the data integritiy (thats the part you were referring too) and on the timing accuracy.
You mentioned "anything else is noise" but that depends on the distribution of the timing uncertainty (means error). If this timing error is random then you´re right, it would only lead to a raised noise floor, but if it isn´t random than you get distorsion (means newly generated signal content during the DA conversion).
Of course it remains questionable if this distorsion will be audible, but the original signal will be altered.
As a second thought, two RF sources (and sinks) were coupled over the digital interconnect and that could lead to sound differences as well. In this respect you were absolutely right- a optical connection should be favoured; unfortunately Toslink outputs for example tend to deliver more jittered signals than there nonoptical counterparts, so the DAC´s jitter immunity has to be even better for this.
Of course it shouldn´t but.....
As a bottom rule we surely could assume, that all human products were flawed; perfection is really, really rare if it comes to technical products. 🙂
Wishes
that´s why it is important to keep the different purposes apart.
The transport to DAC over a S/P-Dif (or AES - EBU) connection is a special one as the clock recovery (or more precise the clock construction) inside the DAC depends on the signal.
If done properly it depends only on the preambles, if done a bit less properly it depends on the data (means coded audio signal) too.
The correctness of any analog signal after conversion depends on the data integritiy (thats the part you were referring too) and on the timing accuracy.
You mentioned "anything else is noise" but that depends on the distribution of the timing uncertainty (means error). If this timing error is random then you´re right, it would only lead to a raised noise floor, but if it isn´t random than you get distorsion (means newly generated signal content during the DA conversion).
Of course it remains questionable if this distorsion will be audible, but the original signal will be altered.
As a second thought, two RF sources (and sinks) were coupled over the digital interconnect and that could lead to sound differences as well. In this respect you were absolutely right- a optical connection should be favoured; unfortunately Toslink outputs for example tend to deliver more jittered signals than there nonoptical counterparts, so the DAC´s jitter immunity has to be even better for this.
Of course it shouldn´t but.....
As a bottom rule we surely could assume, that all human products were flawed; perfection is really, really rare if it comes to technical products. 🙂
Wishes
Last edited:
When we talk about digital cables in audio, it is usually SPDIF signals. The problem is that the clock and data are 'mixed together' and send as one signal through the cable (according to me, already a flawed concept). On the other end, these two signals must be seperated, the problem is that any noise and reflections caused by the cable influence the clock signal and result in jitter. If this clock signal is used for the DAC also, it is easy to understand why cables have an influence on the sound.
Edit: Sorry Jacob2, just saw your post, my connection is veeery slooow tonight.
Last edited:
Conrad, I believe that amount of conductor material for an IC will cause enough problems to swamp the differences you might hear between silver and copper.
I expect you will hear a difference between a round conductor and a flat one
And between stranded variations, and a solid core
A various gauges
If you cant hear any difference, try use one type in left channel and a different one in right channel, and see(hear) what happens
And between stranded variations, and a solid core
A various gauges
If you cant hear any difference, try use one type in left channel and a different one in right channel, and see(hear) what happens
Jakob2😀
i agree, the interface is flawed, especialy when interconnects can cause problems and the way the signal is transmitted is prone to these errors, but it is not new technology, look how PC's have progressed in the years since digital audio became the standard over LPs and tape. As ive siad this is not an isolated problem, but music repreduction seems to provide the most emotional discussions over the minutiae of detail in thr signal path🙂
Regarding testing digital cables, eye diagrams should give some insight into different metals performance, if construction etc is the same.
i agree, the interface is flawed, especialy when interconnects can cause problems and the way the signal is transmitted is prone to these errors, but it is not new technology, look how PC's have progressed in the years since digital audio became the standard over LPs and tape. As ive siad this is not an isolated problem, but music repreduction seems to provide the most emotional discussions over the minutiae of detail in thr signal path🙂
Regarding testing digital cables, eye diagrams should give some insight into different metals performance, if construction etc is the same.
pot calling the kettle
Hmmnn.. who's telling who what they might hear, here???🙁
with no explanation of why, at that!!!😕
Guilty as stated!!😀😀
Hmmnn.. who's telling who what they might hear, here???🙁
with no explanation of why, at that!!!😕
Guilty as stated!!😀😀
I was doing some reading today when it hit me. I've been thinking all wrong, well maybe not all wrong but one could easily do the required mesurements for cables.
Me personally have up until now assumed everyone is measuring cables with the regular electronics test equipment and one can argue that one has to measure it in actual real world applications.
So, why not do some acoustiq measuremets? Why not measure the cables just like you measure a loudspeaker (with microphones)? Comparing two measurements should show any differences?
All the ear is doing is translating soundpressures anyway, right?
Me personally have up until now assumed everyone is measuring cables with the regular electronics test equipment and one can argue that one has to measure it in actual real world applications.
So, why not do some acoustiq measuremets? Why not measure the cables just like you measure a loudspeaker (with microphones)? Comparing two measurements should show any differences?
All the ear is doing is translating soundpressures anyway, right?
Yes, differences should be revealed, although remember that any changes are potentially system dependant (for e.g. a current source is rarely affected by the speaker wire). I doubt anything particularly profound would turn up that couldn't be easily predicted in advance via assessing the output impedance of the amp, speaker load, & wire spec. Still, worth doing IMO.
Reminds me of a 'test' (cough) on some Nordost wire in a certain high-end mag. a little while back. I can't remember the specific wire names now, but I do recall the circumstances. Priceless remark about how the bass suddenly seemed less compressed when the pricier wire was inserted. Nasty speaker load, big, hefty amp, fairly long run. Now, I wonder if that just might have something to do with the fact that the more expensive wire had substantially more conductor area / was of a lower effective gauge than the 'cheaper' one, with the result of less voltage drop at the speaker terminals, and reduced 'clipping' (for want of a better phrase) of LF dynamic peaks? Narry a mention of this rather basic fact in the mag., unsurprisingly.
Reminds me of a 'test' (cough) on some Nordost wire in a certain high-end mag. a little while back. I can't remember the specific wire names now, but I do recall the circumstances. Priceless remark about how the bass suddenly seemed less compressed when the pricier wire was inserted. Nasty speaker load, big, hefty amp, fairly long run. Now, I wonder if that just might have something to do with the fact that the more expensive wire had substantially more conductor area / was of a lower effective gauge than the 'cheaper' one, with the result of less voltage drop at the speaker terminals, and reduced 'clipping' (for want of a better phrase) of LF dynamic peaks? Narry a mention of this rather basic fact in the mag., unsurprisingly.
Last edited:
The ears (and the brains & mind attached) do much more than translating soundpressures. 😉 Or how do you explain sonic colors? No please don't resort to frequency response, which has nothing to do with them 🙂 Or, how to explain different perceptions of sonic refinement, even when measurable distortion is extremely low anyway? Or dynamics, again independent from freq. resp.? 🙂
Last edited:
😀 😛 Ouch. 😉
[insert Bagpuss style yawn here] Perhaps I'm loosing my way these days, but could you please point me to where Marc was suggesting basic frequency response measurements only Andrea?
[insert Bagpuss style yawn here] Perhaps I'm loosing my way these days, but could you please point me to where Marc was suggesting basic frequency response measurements only Andrea?
Last edited:
" All the ear is doing is translating soundpressures anyway, right? "
What I commented was this sentence, BTW.
What I commented was this sentence, BTW.
Right but remember we have two ears (microphones) and how are you going to translate these measurements?
Rather than that, how could you translate mechanically into scientific data the (so much more complex and unscientific) way the brain/mind interprets the stimulations it receives from the ears. 🙂
- Status
- Not open for further replies.