[watchnerd]

A new article from Dave Clark (if you don't know him, he's the reviewer at Positive Feedback who gushed about how the Clever Little Clock improved his sound).
 
In this article, spokespeople from various cable makers theorize about how the USB cables they sell make digital data sound better:
 
http://positive-feedback.com/audio-discourse/why-usb-cables-can-make-a-difference/

 

[sonitus mirus]

It was essentially just an advertisement for each of the cable makers that participated.  I'd suggest that all is required to silence the doubters would be to test the results of different USB cables, but anyone with a lick of sense and credibility would most likely consider this to be a fool's errand.  Is there any evidence out there to suggest that there may be a difference other than "just listen and make up your own mind"?  If anything, this should be a red flag to everyone that the industry is lost and confused.

 

[watchnerd]

  It was essentially just an advertisement for each of the cable makers that participated.  I'd suggest that all is required to silence the doubters would be to test the results of different USB cables, but anyone with a lick of sense and credibility would most likely consider this to be a fool's errand.  Is there any evidence out there to suggest that there may be a difference other than "just listen and make up your own mind"?  If anything, this should be a red flag to everyone that the industry is lost and confused.

 
What I thought was particularly interesting was:
 
1. They all had different 'theories', some of them contradictory with the next guy.  Yet the 'journalist' didn't point this out.
 
2. If manufacturer A's theory is true, then can manufacturer B's theory also be true?  They can't all be right. So I'm not sure how believers choose which church to worship at.
 
3. Not a single mention of actual tests being peformed

 

[spruce music]

   
What I thought was particularly interesting was:
 
1. They all had different 'theories', some of them contradictory with the next guy.  Yet the 'journalist' didn't point this out.
 
2. If manufacturer A's theory is true, then can manufacturer B's theory also be true?  They can't all be right. So I'm not sure how believers choose which church to worship at.
 
3. Not a single mention of actual tests being peformed

Believers are the same in every endeavour.  In this case, they listen to a cable or three, take a liking to them, and follow the correct guidelines from the particular guru they first come across.  In time maybe the evangelists ahem, I mean audio reviewers, convince the flock something better has come along and worth their trying it as it is better.  Lo and behold, the new things is better, scripture is slightly revamped and all the hosannas are heard anew.
 
Notable to me is nearly all mentioned errors in transmission, that USB has no error checking or no shielding from the voltage leads etc etc all leading to errors.  When in fact errors are so rare as to be non-existent effectively.  Certainly not common enough to be a reason for two cables to sound slightly different.

 

[watchnerd]

 
Believers are the same in every endeavour.  In this case, they listen to a cable or three, take a liking to them, and follow the correct guidelines from the particular guru they first come across.  In time maybe the evangelists ahem, I mean audio reviewers, convince the flock something better has come along and worth their trying it as it is better.  Lo and behold, the new things is better, scripture is slightly revamped and all the hosannas are heard anew.
 
Notable to me is nearly all mentioned errors in transmission, that USB has no error checking or no shielding from the voltage leads etc etc all leading to errors.  When in fact errors are so rare as to be non-existent effectively.  Certainly not common enough to be a reason for two cables to sound slightly different.

 
What I also found amusing was the idea that errors in transmission would lead to changes in tonal balance, timbre, soundstaging or other sonic subtleties.
 
When, in reality, transmission errors happen it's obvious because you get dropouts or buzzsaw digital distortion.

 

[Ruben123]

Want to chime in on the fact that usb cables can make a difference... many heard argument is that a cable works - or not. When it is a bad cable you wont hear any sound. In my experience that is wrong, since I was playing some music through my usb soundcard, bumped the cable and the sound turned mono and sounded from afar. Correctly inserting the cable in the pc again and the sound was OK again. Although this hasnt anything to do with the cable itself (and I am all but a believer), it means that a bad connection of the cable in the usb port could make a difference and not causing the sound card to not transfer any music.

 

[swannie007]

The HiFi world has always been awash with "snake oil" salesmen so this lot don't surprise me in the least. Let your own ears be your guide. If it sounds good TO YOU, then it is good, and damn the rest. After all, you have to listen to it, not anyone else.

 

[spruce music]

  Want to chime in on the fact that usb cables can make a difference... many heard argument is that a cable works - or not. When it is a bad cable you wont hear any sound. In my experience that is wrong, since I was playing some music through my usb soundcard, bumped the cable and the sound turned mono and sounded from afar. Correctly inserting the cable in the pc again and the sound was OK again. Although this hasnt anything to do with the cable itself (and I am all but a believer), it means that a bad connection of the cable in the usb port could make a difference and not causing the sound card to not transfer any music.

Do we really need to amend things to take into account cables not fully plugged in and connected on all wires?  Hardly a reason to think the cable salesmen have a point.

 

[Ruben123]

No not at all! Cables are some o of the biggest snake oil products. Only wanted to add that a not correctly inserted USB cable does not give no sound, but altered sound. Which means if a USB cable is very bad (very bad) the sound could degrade.

 

[SoundAndMotion]

Audio is not alone when it comes to marketing claims being borderline lies, often even well over the border to outright lies. But in our efforts to debunk them, let’s not make ourselves easy targets. When our debunking is itself easily debunked, we just contribute to the confusion, when we could be helping enlighten.
 
Quote:

  Is there any evidence out there to suggest that there may be a difference

 
Yes, there is evidence of a difference. Archimago’s blog on optical vs. USB, and a WBF thread on the Jitterbug and Uptone Regen, show measurements demonstating that the 8kHz USB microframe packet noise (and higher harmonics) can go through a DAC to the analog output. Interestingly, in the WBF thread the noise spikes appeared WITH the Regen and not without. BUT, this shows a difference is possible, although in this case the spikes were very low level (probably inaudible). We’ll blame the DAC for letting the analog noise on the digital line go through. We often see the disclaimer that “any properly designed item [DAC, amp, cable] will perform as I say it should”. The problem is that design teams, with at least competent if not brilliant, engineers/designers created the Takata air bag, the GM ignition switch, the Challenger’s O-rings, and the Fukushima power plant, which we hear about because lives are at stake. An improperly designed DAC (according to us) probably goes unnoticed. There can be a difference, so let's not base our debunking on this.

  Notable to me is nearly all mentioned errors in transmission, that USB has no error checking or no shielding from the voltage leads etc etc all leading to errors.  When in fact errors are so rare as to be non-existent effectively.

  When, in reality, transmission errors happen it's obvious because you get dropouts or buzzsaw digital distortion.

 
  No, only 2 mentioned no error correction and the cable guys on positive feedback are wrong here. Asynchronous USB (as used by asynchronous USB DACs) does have error detection and correction, and a hopelessly mangled packet will get resent. There’s plenty of time with the relatively low rates of audio. Only when there are multiple failures do you get dropouts or weird buzzsaw noise. Bit errors are more common than you think (not on all cables, but more than "effectively non-existent); they are simply masked by successful error correction. 
 

Only wanted to add that a not correctly inserted USB cable does not give no sound, but altered sound. Which means if a USB cable is very bad (very bad) the sound could degrade.

This truth does not justify crazy-expensive cables, but the answer IMHO is to request tests of audibility and state a priori the conditions needed to accept the results. You know the whole ABX argument...once again... redundantly.... repeated.... again.

 

[watchnerd]

   
  No, only 2 mentioned no error correction and the cable guys on positive feedback are wrong here. Asynchronous USB (as used by asynchronous USB DACs) does have error detection and correction, and a hopelessly mangled packet will get resent. There’s plenty of time with the relatively low rates of audio. Only when there are multiple failures do you get dropouts or weird buzzsaw noise. Bit errors are more common than you think (not on all cables, but more than "effectively non-existent); they are simply masked by successful error correction. 
 

 
Sure, but the point being that bit errors don't lead to changes in things like soundstaging.  I can manipulate a soundstage to make it move forward or backward by using a PEQ filter with a center at 1600 Hz, Q of 1.00, and -2 / +2 dB depending upon whether I want to move the image back or forward.  This persistent, broad, and very specific manipulation of the signal is well beyond what should happen in a digital error scenario.

 

[sonitus mirus]

   
Yes, there is evidence of a difference. Archimago’s blog on optical vs. USB, and a WBF thread on the Jitterbug and Uptone Regen, show measurements demonstating that the 8kHz USB microframe packet noise (and higher harmonics) can go through a DAC to the analog output. Interestingly, in the WBF thread the noise spikes appeared WITH the Regen and not without. BUT, this shows a difference is possible, although in this case the spikes were very low level (probably inaudible). We’ll blame the DAC for letting the analog noise on the digital line go through. We often see the disclaimer that “any properly designed item [DAC, amp, cable] will perform as I say it should”. The problem is that design teams, with at least competent if not brilliant, engineers/designers created the Takata air bag, the GM ignition switch, the Challenger’s O-rings, and the Fukushima power plant, which we hear about because lives are at stake. An improperly designed DAC (according to us) probably goes unnoticed. There can be a difference, so let's not base our debunking on this.
 
 

 
Maybe I wasn''t reading the same blog?
 
http://archimago.blogspot.com/2015/05/measurements-usb-hubs-and-8khz-phy.html
 
It appears that any differences were between optical and USB, and also between various powered USB hubs.  I'm not certain if a USB cable by itself would be responsible for any audible differences.  I'm assuming the cable is properly designed and adheres to standard specifications.  
 
Mass produced cables might also have a much smaller failure rate than more expensive handmade cables.  A 1% failure rate would be significant for an assembly line, though 1 of 100 boutique cables may not seem too terrible.  I just don't know, but I am curious.

 

[Roseval]

   Asynchronous USB (as used by asynchronous USB DACs) does have error detection and correction, and a hopelessly mangled packet will get resent. 

I'm afraid this not true.
Most DACs use USB Audio class 1 or 2.
This is done by using  Isochronous Transfers: time sensitive information, such as an audio or video stream

1. Guaranteed access to USB bandwidth.
2. Bounded latency.
3. Stream Pipe - Unidirectional
4. Error detection via CRC, but no retry or guarantee of delivery.
5. Full & high speed modes only

 
As it is a quasi real time stream, there is no time for a retry.
Bulk transfers do use a retry in case of mangled bits but this mode is hardly used in case of DACs as it requires developing a proprietary driver (expensive).
A bit more detail on my website: http://www.thewelltemperedcomputer.com/KB/USB.html

 

[SoundAndMotion]

Yes, I need a mea culpa post:
 
Quote:

  I'm afraid this not true.
Most DACs use USB Audio class 1 or 2.
This is done by using  Isochronous Transfers: time sensitive information, such as an audio or video stream

1. Guaranteed access to USB bandwidth.
2. Bounded latency.
3. Stream Pipe - Unidirectional
4. Error detection via CRC, but no retry or guarantee of delivery.
5. Full & high speed modes only

 
As it is a quasi real time stream, there is no time for a retry.
Bulk transfers do use a retry in case of mangled bits but this mode is hardly used in case of DACs as it requires developing a proprietary driver (expensive).
A bit more detail on my website: http://www.thewelltemperedcomputer.com/KB/USB.html

 
You have a nice website! I seem to have conflated "asynchronous mode" with "bulk transfers", which are asynchronous. You are right that "asynchronous mode" does not allow a resend, but bulk transfers do (and you point out they are much less common; rare probably works). But here you state most DACs use USB Audio class 1 or 2, which is done by isosynchronous mode, but on your website you state: "Almost all recent offerings of quality DACs have asynchronous USB input." Either way, thanks for the correction. I messed that up. It doesn't change my point that USB is not perfect in transmission of data(digital), and can also transmit analog noise (by this I would include the analog characteristics of the bits, e.g. rise time, etc.) and that can appear at the analog output of the DAC. Thanks for correcting my carelessness. 

   
Maybe I wasn''t reading the same blog?
 
http://archimago.blogspot.com/2015/05/measurements-usb-hubs-and-8khz-phy.html
 
It appears that any differences were between optical and USB, and also between various powered USB hubs.  I'm not certain if a USB cable by itself would be responsible for any audible differences.  I'm assuming the cable is properly designed and adheres to standard specifications.  
 
Mass produced cables might also have a much smaller failure rate than more expensive handmade cables.  A 1% failure rate would be significant for an assembly line, though 1 of 100 boutique cables may not seem too terrible.  I just don't know, but I am curious.

 
I meant:http://archimago.blogspot.de/2015/05/measurements-corning-usb-3-optical.html , but again I messed up by putting words in your mouth you didn't say. You said cables and I interpreted your meaning as USB transmission in general. With my interpretation, cables could contribute excessive 8kHz packet noise, or could have characteristics that modify the transmission, but you are right that the examples I gave cannot be attributed to bad vs. good cable. 

   
Sure, but the point being that bit errors don't lead to changes in things like soundstaging.  I can manipulate a soundstage to make it move forward or backward by using a PEQ filter with a center at 1600 Hz, Q of 1.00, and -2 / +2 dB depending upon whether I want to move the image back or forward.  This persistent, broad, and very specific manipulation of the signal is well beyond what should happen in a digital error scenario.

 
Well here I completely agree with your last statement, but only stick with a devil's advocate argument (no proof, just what if...): Although you offer a way to modify soundstaging, that is hard to imagine can occur by digital errors, it may not be the only way soundstaging can be modified. And I assume you mention soundstaging because the marketing people and subjectivists do, but my point is more general that analog modifications due to "poor" digital transmission is not excluded, and the objectivist argument "it can't happen" unnecessarily weakens their point.. 

 

[sonitus mirus]

  I'm afraid this not true.
Most DACs use USB Audio class 1 or 2.
This is done by using  Isochronous Transfers: time sensitive information, such as an audio or video stream

1. Guaranteed access to USB bandwidth.
2. Bounded latency.
3. Stream Pipe - Unidirectional
4. Error detection via CRC, but no retry or guarantee of delivery.
5. Full & high speed modes only

 
As it is a quasi real time stream, there is no time for a retry.
Bulk transfers do use a retry in case of mangled bits but this mode is hardly used in case of DACs as it requires developing a proprietary driver (expensive).
A bit more detail on my website: http://www.thewelltemperedcomputer.com/KB/USB.html

 
The question remains, how would an expensive USB cable offer any solution to these potential issues over a standard USB cable?  What physical/electrical properties would benefit that are not already well within the design specifications for proper operation of a typical 1-2 meter USB cable?