Don't get why "Audiophile" USB Cable would improve sound quality
Aug 3, 2011 at 7:15 PM Post #661 of 835
[size=10pt][size=10pt]Other than noise, conductivity and thus impedance, and simple durability I cannot think of any other reason why an expensive hi-fi cable would alter the sound any.  Noise, conductivity, and impedance are pretty important though.  I remember in my physics days that nasty reflections and attenuations can occur from non-impedance matching.  This means that the impedance of your gear, headphones, and cable should ideally be the same.  Perhaps a higher quality cable has better grounding and slightly better impedance matching to headphones and headphone amps.  The same would go for speaker wire, speakers, and speaker amps.[/size][/size]
[size=10pt] [/size]
[size=10pt][size=10pt]I guess the price does not really matter, but the afore mentioned quantitative qualities do matter and will affect the sound.[/size][/size]
 
[size=10pt][size=10pt]The following link shows the reflection coefficients for voltage down a transmission line relative to the impedance of two connected lines.  The only real way to preserve the amplitude and minimize the back reflection of other waves is to very closely match the impedance of each connection.[/size][/size]
 
[size=10pt][size=10pt]http://en.wikipedia.org/wiki/Impedance_matching[/size][/size]
 
[size=10pt][size=10pt]Here is another link specifically referring to audio equipment impedance matching:[/size][/size]
[size=10pt][size=10pt]http://hyperphysics.phy-astr.gsu.edu/hbase/audio/imped.html[/size][/size]
 
Aug 3, 2011 at 7:19 PM Post #662 of 835


Quote:
[size=10pt][size=10pt]Other than noise, conductivity and thus impedance, and simple durability I cannot think of any other reason why an expensive hi-fi cable would alter the sound any.  Noise, conductivity, and impedance are pretty important though.  I remember in my physics days that nasty reflections and attenuations can occur from non-impedance matching.  This means that the impedance of your gear, headphones, and cable should ideally be the same.  Perhaps a higher quality cable has better grounding and slightly better impedance matching to headphones and headphone amps.  The same would go for speaker wire, speakers, and speaker amps.[/size][/size]
[size=10pt] [/size]
[size=10pt][size=10pt]I guess the price does not really matter, but the afore mentioned quantitative qualities do matter and will affect the sound.[/size][/size]
 
[size=10pt][size=10pt]The following link shows the reflection coefficients for voltage down a transmission line relative to the impedance of two connected lines.  The only real way to preserve the amplitude and minimize the back reflection of other waves is to very closely match the impedance of each connection.[/size][/size]
 
[size=10pt][size=10pt]http://en.wikipedia.org/wiki/Impedance_matching[/size][/size]
 
[size=10pt][size=10pt]Here is another link specifically referring to audio equipment impedance matching:[/size][/size]
[size=10pt][size=10pt]http://hyperphysics.phy-astr.gsu.edu/hbase/audio/imped.html[/size][/size]



This is why we have a USB spec.
 
Aug 4, 2011 at 10:37 AM Post #663 of 835
 
DaBomb's Quote:
I would agree that we cannot measure a system's "imaging" in any real sense.

 
DaBomb's Quote:
As such, my view is ... our current equipment can measure everything that a cable change could possible affect.

 
Do you not think those two statements are contradictory?  For example, I would claim that one of the audible benefits I perceive of my Nordost USB cable (vs a stock USB 2.0 cable) is an improvement in the precision of the imaging.  Are we both agreed that imaging as a real and perceptible, yet non measurable or quantifiable, property exists?
 
ac500's Quote:
However if your DAC does run off a buffered source, and with its own local clock, then USB-dependent jitter simply should not be happening.

 
I would take issue with that.  Jitter is not something that is eliminated by a buffer, it is only reduced by a buffer.  All we can hope to do is reduce it to a level below the threshold at which it can have a theoretically audible effect.  In one of my earlier posts I gave my rationale for suggesting that 24/192 audio is susceptible to jitter at levels which fall deeply below the ability of instrumentation to measure it, let alone eliminate it.
 
ac500's Quote:
I wonder if anyone with a buffered DAC would comment.

I have direct experience with two high-quality buffered DACs.  One is the Calyx 24/192 DAC and the other is a Classé CP-800 preamplifier with a built-in USB DAC.  Both use buffers, but both go beyond the "first-order" effect of buffer vs no buffer.  The Calyx uses two totally independent clocks for buffering.  One clock is used for the 44.1/88.2/176.4 family of sample rates, and the other is used for the 24/48/192 family.  This approach has resulted in an extremely high quality sound.  The Classé goes one step further.  It uses totally independent clocks for each and every sample rate it supports.  The Classé sounds even better than the Calyx.  Indeed, it is the best sounding USB DAC I have ever heard (I am not claiming I have heard them all), so I bought it.  Neither of these gems can claim to eliminate jitter (I don't speak for either manufacturer, but I am sure they would concur), but I think they both do a great job of reducing it.
 
I also want to comment that jitter is a complex phenomenon.  To say that it can be "measured" is somewhat misleading.  Its presence can be measured, but its nature cannot.  How to explain this?  It is like the "decibel" app on my iPhone - it can measure the overall volume of the sound it hears, but it does not say anything descriptive about the sound, such as what tune is playing.  So, two different sources can have the same measurable amount of jitter, yet sound very different.
 
I would make one further comment on the effect of buffers on jitter, with no basis other than my own personal observations to support it:  The jitter present in the output of a buffer designed to reduce it, will still be dependent to some extent on the jitter present in the source data stream.  In other words, the sonic signature of a system comprising such a buffer will, to the extent that jitter has any audible effects, still be dependent on the jitter characteristics of the original incoming data stream.
 
 
 
Aug 4, 2011 at 11:02 AM Post #665 of 835
> All we can hope to do is reduce it to a level below the threshold at which it can have a theoretically audible effect. 
 
Yes, but a buffer will remove any clock dependence on the USB cable, and any "jitter" would be from a local clock, which would be absurdly low jitter - and in reality it would then become a DAC quality issue as opposed to a USB issue anyway.
 
> The jitter present in the output of a buffer designed to reduce it, will still be dependent to some extent on the jitter present in the source data stream.  In other words, the sonic signature of a system comprising such a buffer will, to the extent that jitter has any audible effects, still be dependent on the jitter characteristics of the original incoming data stream.
 
As you put it, "I take issue with this". A proper buffer will completely isolate the USB clock away from the DAC, and therefore completely eliminate any jitter that results from the USB source.
 
I'd like to hear your explanation on why you claim this. Please no voodoo or magic-based explanations.
 
 
Aug 4, 2011 at 12:47 PM Post #666 of 835
A 1974 BBC study on jitter and its audibility.
 
http://downloads.bbc.co.uk/rd/pubs/reports/1974-11.pdf
 
The conclusion is that 35ns rms sinusoidal jitter, in frequencies above 2kHz would be audible to 5% of the population, for random jitter it is 50ns rms from 30Hz to 16kHz. So for a few extremes of jitter are audible. But that is with measurements in one billionths of a second, yet here I see arguments about jitter measured at pico seconds or one trillioth of a second, or a billionth of a nanosecond.
 
It is absolutely astounding what audiophiles claim, with no evidence and significant evidence to the contrary, that they can hear.
 
Aug 4, 2011 at 12:57 PM Post #667 of 835
10101010101010101010   That's it! simple, on or off, 1 or 0, 0v or +5v. Only if it were so simple, we presume that digital signals are "clean" 1's and 0's sadly they are not. When things are happening so fast (a few megahertz) things can go crazy inside a cable. Inductance, interference, jitter, loss etc..etc..
 
and there's the best of them, what is 2.5v? 1 or 0? :p 
 
a cable's job is to deliver information from one place to another. Digital circuits have to be able to differentiate between, garbage and signal, the more garbage the cable adds the more it gets difficult for the circuit to differentiate between them. Soon, it may make mistakes of confusing noise as 1's or 0's which is avoided by other fancy stuff such as modulation and encoding.  Sadly, apart from providing the write electrical "connectivity" where fibers don't break in between and uniform in dimensions, there not quite much cables can actually do!
 
Simple rule, sturdy, tight and short :wink: , and a good sturdy connector always help , if you know hat i mean :wink:
 
 
And a 2400$ cable ! woof! I'd like to meet the guy who markets such stuff.
 
P.s vote up if you think there should be standards in audio equipment which is tested by an independent lab to provide true information and not fool people.
 
 
 
Aug 4, 2011 at 1:00 PM Post #668 of 835
Quote:
 
I also want to comment that jitter is a complex phenomenon.  To say that it can be "measured" is somewhat misleading.  Its presence can be measured, but its nature cannot.  How to explain this?  It is like the "decibel" app on my iPhone - it can measure the overall volume of the sound it hears, but it does not say anything descriptive about the sound, such as what tune is playing.  So, two different sources can have the same measurable amount of jitter, yet sound very different.


Jitter can be measured. And its effects on the analog signal can be mathematically and experimentally confirmed. Yes, yes they can. What that does to the "sound" is totally in your head, so you have a point there. A point that no one has any interest in arguing...
 
edit: I guess some people want to argue
rolleyes.gif

 
Aug 4, 2011 at 1:12 PM Post #669 of 835
>And a 2400$ cable ! woof! I'd like to meet the guy who markets such stuff.

in a dark alley. with no cameras... and a cleansing baseball bat (cricket bat as it may be in Australia :D ?)

>The conclusion is that 35ns rms sinusoidal jitter, in frequencies above 2kHz would be audible to 5% of the population, for random jitter it is 50ns rms from 30Hz to 16kHz. So for a few extremes of jitter are audible.

heh. considering most modern DACs have 10-50ps not ns ... :D and 35ns = 35 000ps :D... That's a whooping 4 orders of magnitude or 2-3 for the not so good devices.
 
Aug 4, 2011 at 1:31 PM Post #670 of 835
I think no matter how much scientific evidence we provide proving that the difference can't be heard, the believers will always claim it makes a difference.
 
For this reason I think the whole technical discussion is ultimately pointless. I think double blind testing is really the only undisputable way to prove whether there's any difference. People will always find a way to say "well your SCIENCE and MEASUREMENTS are faulty, all that matters is what you think you hear." To that, a double blind test is the perfect response, because it's the perfect measure of just that: what you hear.
 
Look, this thread has gone on for 45 pages. FORTY FIVE PAGES. Have we gotten anywhere arguing technicalities (with people who really don't understand digital processing to begin with)? I say anyone arguing that USB cables makes a difference should be required to refrain from posting unless they post actual double-blind data. This may seem one-sided, but:
 
a) This is the science forum.
b) We're not allowed to talk about double-blind outside of the science forum, so the reverse seems fair to me.
c) All peer-reviewed studies shows that there is no difference.
 
 
Aug 4, 2011 at 2:41 PM Post #671 of 835


Quote:
I think no matter how much scientific evidence we provide proving that the difference can't be heard, the believers will always claim it makes a difference.
 




That issue is redundant as there is no doubt that people do hear differences and there may be a absolute reality to those heard differences i.e. they are not fooling themselves. We are proving that the differences that are heard are not caused by the cable itself, instead they are down to the listener themselves.
 
Aug 4, 2011 at 2:52 PM Post #672 of 835

That issue is redundant as there is no doubt that people do hear differences and there may be a absolute reality to those heard differences i.e. they are not fooling themselves. We are proving that the differences that are heard are not caused by the cable itself, instead they are down to the listener themselves.


Yes, they're all mentally divergent. Your 37 year old scientific studies on jitter audibility nail it hard. really. I'm not. Kidding. Not Even. A bit.
 
Aug 4, 2011 at 3:10 PM Post #673 of 835
 
Quote:
I think no matter how much scientific evidence we provide proving that the difference can't be heard, the believers will always claim it makes a difference.

 
I don't know why everybody is getting so hot under the collar.  After all, as you point out, this is a science forum, and all I am trying to do is point out areas where science might have something useful to say on the topic, which is, in case we forget, "I don't get why audiophile USB cable would improve sound quality".  I don't think I have been dissing anybody.  It is not disrespectful to address somebody's viewpoint with a reasoned counter-argument.  I am open to that myself. 
 
I have pointed out a scientifically valid explanation for why the jitter threshold for 24/192 audio could be as low as 310 femtoseconds.
I have pointed out a scientifically valid explanation for why a USB cable's properties might affect the jitter in the range of hundreds of picoseconds and below.
I have tried to point out that some aspects of jitter are measurable, whereas some are not.
I have tried to point out that any theoretical analysis of the measurable effects of jitter on the audio signal is difficult, if not impossible.
I have tried to get some measure of acceptance of the notion that some things which are audible are not measurable.
 
I don't believe I am foisting any radical voodoo or scientific pseudo-babble on anybody.
 
Taken together, all these observations add up to a basis for postulating that there is a reasonable argument to be made that a USB cable can impact the sound quality of a high end audio system.  Further, I am pointing out that this may even be true despite the fact that nobody can tell you what those audible effects could or should be, not can they tell you what to go away and measure in order to verify or deny the existence of such effects.  That's all.
 
But instead, I am reduced to a "believer", a term used dismissively, if not pejoratively.  On the other hand "no matter how much scientific evidence we provide" attempts to stake out the rational high ground, based on ... what, exactly?.
 
Sorry, but if this is a Science Forum, I'm Obama's mullah.  I won't be back, and I'm sure I won't be missed.
 
 
Aug 4, 2011 at 3:15 PM Post #674 of 835
I don't think anyone -- well at least I was not calling you a "believer" in the negative sense, Talos. I was making an off topic observation of the fact that there was 45 pages of discussion on this anyway. The easiest way to conclude this thread is to do a double blind test on someone who claims they can hear a difference reliably with their cables.
 
In any case I'd still like to hear how you think a buffered DAC input stream running on a local clock can receive jitter effects from an external clock.
 
Aug 4, 2011 at 3:22 PM Post #675 of 835


Quote:
Yes, they're all mentally divergent. Your 37 year old scientific studies on jitter audibility nail it hard. really. I'm not. Kidding. Not Even. A bit.


confused_face_2.gif

 
 

Users who are viewing this thread

Back
Top