USB Cables and Digital Sound Quality

[DonH50]

Thanks Amir! That looks to be mostly bandwidth issues, same problem we have in our PCIe and SAS links. I'll have to read the article to see if he took the driver out of the picture.

 

[Lee]

Cool, we should try to hook up sometime!

Thank you for the link. The article (in my 2-second look) appears to define line-induced jitter as due to bandlimiting. I would certainly say that is by far the largest factor, but there are many others. However, I strongly suspect the other factors are in the mud for audio links. My guess for audio line-induced jitter causes would be (a) bandlimiting, (b) loss from long lines, (c) poor shielding leading to EMI issues, and (d) everything else at a much lower level.

On the coupling issue, I have a vague memory of an article discussing HDMI or USB links and discovering one of the benefits of a more expensive cable was the much better shield and heavier ground on the more expensive cables. However, for USB links, I still think galvanic isolation is the way to go.

Don, I will let you know when I am in CO next. We have an office in Broomfield.

 

[Lee]

Good info as usual Amir. Thanks!

 

[microstrip]

(...) In fairness, whether a cable makes a difference or not is source dependent. So Paul's tests can't be representative of all situations.

Although it is more than 10 years old, this document can help to understand why USB cables can make a difference and are source dependent - you will find that designers have to face compromises and may be they will not choose all the same options.

High Speed USB Platform Design Guidelines

www.usb.org/developers/docs/hs_usb_pdg_r1_0.pdf

 

[FrantzM]

Tim, the cable is $700 MSRP, the DAC1 Pre is $1,600. No doubt the cable is pricey nonetheless.

...

I understand the debate and am opening my mind to the idea of USB cables being sounding different. I am however extremely skeptical about such. I do not question whether the measurements could prove to be different. I am in the IT business and have seen different brand of Cat% cables measure quite differently .. but ... do these differences matter? Are they audible? Can we really ascertain that the differences between up-to-specs USB cables are such that we can perceive them? To me that is the crux of the matter.

 

[fas42]

Again, in electronics there is no such thing as a digital signal; at least at the level we worry about it. Everything, but everything, is analogue: it's just that another piece of electronics chooses to interpret what is input to it as representing digital information. But what is actually happening at the circuit parts level, the voltages and currents, is all that mucky, difficult analogue stuff that can get you into trouble.

And if you don't take that reality seriously, then you can have degraded sound ...

Frank

 

[Phelonious Ponk]

Again, in electronics there is no such thing as a digital signal; at least at the level we worry about it. Everything, but everything, is analogue: it's just that another piece of electronics chooses to interpret what is input to it as representing digital information. But what is actually happening at the circuit parts level, the voltages and currents, is all that mucky, difficult analogue stuff that can get you into trouble.

And if you don't take that reality seriously, then you can have degraded sound ...

Frank

Of course. And if someone really wants to talk themselves into an $800 cable (sorry about my earlier overstatement, I Googled and found a Diamond USB at much more...different length, no doubt), I wish them well. But it really is just data, noise and timing. All of them can be effectively addressed for much less than $800, though all of them cannot be effectively addressed by a USB cable, at any price. A good USB cable -- robustly built, overzealously sheilded, and meeting USB standards, is a good thing that's worth about 50 bucks. If you want to address noise (not just the noise in the cable) and jitter (not just the jitter created by the cable), put the $650 into isolation and re-clocking.

Tim

 

[DonH50]

+1.

The real problem with a USB cable is the USB part, not the cable...

 

[Orb]

I wouldn't be surprised these guys are shooting blind in changing cables and not knowing without measurements, whether they have moved forward or back. Orb mentioned the same regarding Paul's tests: http://www.whatsbestforum.com/showt...yond-say-10-feet&p=53850&viewfull=1#post53850

In fairness, whether a cable makes a difference or not is source dependent. So Paul's tests can't be representative of all situations.

Indeed and good point Amir.
Paul Miller mentions this in the article, however he did also note the same trend (but lesser numbers) with the said cables even with a more jitter resistant setup/isolation setup.
I could try to find the article from their digital online library (open to all with a Hifinews subscription) when I have time.
He did a few articles on this, and then also a couple more relating to Jim Lesurf's on the low frequency time related (more extended timeframe) jitter/isolation.

Cheers
Orb

 

[FrantzM]

Of course. And if someone really wants to talk themselves into an $800 cable (sorry about my earlier overstatement, I Googled and found a Diamond USB at much more...different length, no doubt), I wish them well. But it really is just data, noise and timing. All of them can be effectively addressed for much less than $800, though all of them cannot be effectively addressed by a USB cable, at any price. A good USB cable -- robustly built, overzealously sheilded, and meeting USB standards, is a good thing that's worth about 50 bucks. If you want to address noise (not just the noise in the cable) and jitter (not just the jitter created by the cable), put the $650 into isolation and re-clocking.

Tim

+2 ...

 

[microstrip]

Of course. And if someone really wants to talk themselves into an $800 cable (sorry about my earlier overstatement, I Googled and found a Diamond USB at much more...different length, no doubt), I wish them well. But it really is just data, noise and timing. All of them can be effectively addressed for much less than $800, though all of them cannot be effectively addressed by a USB cable, at any price. A good USB cable -- robustly built, overzealously sheilded, and meeting USB standards, is a good thing that's worth about 50 bucks. If you want to address noise (not just the noise in the cable) and jitter (not just the jitter created by the cable), put the $650 into isolation and re-clocking.

Tim

Although your proposition seems much more reasonable than the usd 800 USB cable , what happens if we buy the $650 more expensive USB DAC with better isolation and re-clocking and later we found that it sounds better with a $801 USB cable?

There is still a lot to be learned about digital audio, until them everyone who designs something sounding different is free to price it accordingly to the market rules - if users perceive it as better you can charge more, if it is overpriced someone will soon start selling it cheaper. Why should we expect digital to be different from anything else in audio?

Disclaimer - I never used an USB cable costing more than euros 20.00 (~ usd 30).

 

[Phelonious Ponk]

As long as someone will pay the price, that price is the value -- to the market consisting of someone who will pay the price. And perception is reality. But none of that changes the very likely fact that, in spite of what we don't know about digital audio, if the digital source is well-isolated from the DAC after the usb cable, and the digital signal is re-clocked after the usb cable, then beyond its basic functionality of carrying the digital data, the usb cable becomes a moot point. Of course we can "what if?" And "what we don't yet know" all day long, if the mystery makes us feel better, but in the end it's still about $50 worth of cable and $750 worth of hustle.

Tim

 

[microstrip]

(...) Of course we can "what if?" And "what we don't yet know" all day long, if the mystery makes us feel better, but in the end it's still about $50 worth of cable and $750 worth of hustle.
Tim

It is not mystery, now it is reality ...

We can be sure that in a short time the expensive USB cable would be obsolete, as the DAC designers improve their models. But for the guy who is buying today and wants to listen tonight, if he has the money and this particular USB cable sounds better why not buying it?

I have good friends who systematically postponed buying a videocamera recorder waiting for the next version, that would become better and cheaper, and one day found that the children they wanted to film were no more children anymore ...

 

[Phelonious Ponk]

We can be sure that in a short time the expensive USB cable would be obsolete, as the DAC designers improve their models. But for the guy who is buying today and wants to listen tonight, if he has the money and this particular USB cable sounds better why not buying it?

Because you can make any improvements the premium usb cable has any possibility of make, and more, by addressing the issues at hand instead of buying an absurdly expensive usb cable. What if you make all those appropriate, systemic changes and the usb cable still sounds better? It won't, even though you may very well hear it that way.

Tim

 

[Vincent Kars]

I expect the opposite
A couple of years ago you simply couldn’t buy an audiophile grade USB cable.
Likewise no audiophile cat5 cables
But as the popularity of file based audio increased overtime all what already has been proven very successful in the analog world, charging a premium for a wire, is now applied to the digital part.
With equal success.
I have seen $25 for 1 m Cat5.
$25 is way too low to be “high end”.
If people are willing to pay a couple of 1000 for a interlink, they sure are willing to pay the same price for a cat5 or USB.
So I expect the more popular file based audio will become, more “high end” digital cables will be sold at an increasing price level.

As you so aptly stated in this forum “a product doesn’t need to justify its price tag, it need somebody willing to buy it”.
This will work regardless of the design of a DAC

 

[amirm]

Well said Vincent.

 

[microstrip]

As you so aptly stated in this forum “a product doesn’t need to justify its price tag, it need somebody willing to buy it”.
This will work regardless of the design of a DAC

This type of argument is drifting away from the main point - why people buy the usd 800 cable and not the usd150 cable and the usd 650 more expensive DAC.

The more popular computer audio base systems will become the more exigent users will become, and some people will take profit of it in a objectionable way, but not all of them. IMHO focusing our attention in the charlatans and using them as arguments will only move us away from the key points.

 

[fas42]

This type of argument is drifting away from the main point - why people buy the usd 800 cable and not the usd150 cable and the usd 650 more expensive DAC.

And the simple answer of course is that people want "good" sound, not necessarily more accurate sound - see my posts on such in another thread - and the industry won't provide the information in an intelligent way so people can achieve such goals easily. There's a lot more money to be made, and it's much, much easier to do, in supplying "raw" components, rather than engineering overall systems to work properly. How many slick ads have you seen lately by companies aiming to do the latter ...??

Frank

 

[garylkoh]

I expect the opposite
A couple of years ago you simply couldn’t buy an audiophile grade USB cable.
Likewise no audiophile cat5 cables
But as the popularity of file based audio increased overtime all what already has been proven very successful in the analog world, charging a premium for a wire, is now applied to the digital part.
With equal success.
I have seen $25 for 1 m Cat5.
$25 is way too low to be “high end”.
If people are willing to pay a couple of 1000 for a interlink, they sure are willing to pay the same price for a cat5 or USB.
So I expect the more popular file based audio will become, more “high end” digital cables will be sold at an increasing price level.

As you so aptly stated in this forum “a product doesn’t need to justify its price tag, it need somebody willing to buy it”.
This will work regardless of the design of a DAC

Well put, Vincent. I have customers who won't spend the afternoon to build themselves a $400 music server because they are convinced that a $50,000 one will sound better. Unfortunately, these are the guys that keep some charlatans in business.

I have tried some expensive USB cables, but found that there is absolutely no correlation between price and SQ. If you believe that USB cables can have a different sound, take your armful of USB printer cables out of your computer cabinet and listen to all of them. You may just find one that sounds better than the others - and no, I can't figure out why. It was a off-white cheaply constructed freebie, but it had a ferrite bead at one end.

 

[fas42]

You may just find one that sounds better than the others - and no, I can't figure out why. It was a off-white cheaply constructed freebie, but it had a ferrite bead at one end.

And that's what it's all about: stopping interference or unwanted interactions occurring: not whether a certain quality of copper was used!

Frank