High End USB Cables - Real or Fantasy?

As to be expected, there is a small but apparently growing number of so called high end USB cables appearing in the market for connecting devices such as DACs to your computer and to some stereo components.  Has anyone taken the time to really determine if there is any significant verifiable difference in the sound quality over using a standard "high quality" computer USB cable?

Geofiz

Original Post

Not sure - I've been running a dealer-supplied Audioquest usb cable from the start, never tried anything else. I have a sneaking suspicion that differences are subtle if there at all, however, BNC/SPDIF cables clearly have their own signature so wouldn't discount the same for USB altogether. 

 

EJ

Interesting.. The so called sound quality is really determined by how susceptible your audio equipment is to interference, Powerline intermodulation, RF and EMI. If your equipment is immune from this (in the real world almost impossible) then different USB cable characteristics should give no variation in SQ. But since USB (1 and 2) is a reasonably loose standard in terms of physical electrical characteristics, USB leads will most likely therefore give the appearance of 'sounding' different, albeit it's their physical electrical properties affecting the performance of connected audio equipment.

.

So a well designed lead with a quality twisted differential signal pair, perhaps shielded and suitably decoupled powerlines into a quality USB connector could perform electrically better than a cable that wasn't as carefully, but still compliantly constructed.

 

However, the price of this should be minimal, and of course in this market anything slightly different has exponential margins and great scope for rip off.. Especially if it looks pretty.

 

BTW the same principle applies to USB memory sticks.

 

Simon

Isn't it also the case that many DACs take a 5V supply from the computer to power the USB interface, and that 5V supply is unlikely to be of the highest quality. In which case, mightn't USB cables that come with their own higher quality 5V power-supply, like say the ifi-audio iUSB, offer some advantages? 

My experience if testing cables from QED, Chord, Wireworld an Audioquest indeed showed that there are differences.  These tests were in effect A/B comparisons after 80 to 100 hours of burning in each cable.

In the end I choose the cable that I liked the sound of, which as it happens, was towards he high price range of the cables tested but not the most expensive. 

I could have tried more expensive cables but I am am more than delighted with the combination or effort on testing vs price vs SQ.

So, yes they can make a difference but time and effort is needed to verify this.

I bought a Wireworld USB A-B around £45 on recommendation of friends.

I think there was a slight increase in SQ but the main difference was that the plugs fitted a lot better.

So maybe the performance is in the plugs not so much the cable?


Poor socket fit had included "Belkin" too.

Originally Posted by Brilliant:

If the USB performance was so susceptible to cable construction - how come data loss is not common when working with computer files? Is it the interface software used in audio that is inadequate then, requiring extra dressing just for audio data?

+1

Originally Posted by cvrle:
Originally Posted by Brilliant:

If the USB performance was so susceptible to cable construction - how come data loss is not common when working with computer files? Is it the interface software used in audio that is inadequate then, requiring extra dressing just for audio data?

+1

Does it matter in the end....

Originally Posted by Jude2012:
Originally Posted by cvrle:
Originally Posted by Brilliant:

If the USB performance was so susceptible to cable construction - how come data loss is not common when working with computer files? Is it the interface software used in audio that is inadequate then, requiring extra dressing just for audio data?

+1

Does it matter in the end....

I am asking the same question. Obviously, some people find that there is difference in cables, so I believe it, but I'm having hard time to understand why, and how. 

Originally Posted by cvrle:
Originally Posted by Jude2012:
Originally Posted by cvrle:
Originally Posted by Brilliant:

If the USB performance was so susceptible to cable construction - how come data loss is not common when working with computer files? Is it the interface software used in audio that is inadequate then, requiring extra dressing just for audio data?

+1

Does it matter in the end....

I am asking the same question. Obviously, some people find that there is difference in cables, so I believe it, but I'm having hard time to understand why, and how. 

I wrestled with the 'how' question as I like to know how things work.  The research led me an understanding that audio data is a stream, but I have not heard an explanation as to how this translates to asynchronous audio data.  

 

One thing for certain is that jitter and EMI reduction is what makes the difference in audio and that what the cables are designed to minimise.

 

BTW, I end up choosing a Wireworld cable (Starlight 7). It took a long time to run in compared with the Audioquests.  I guess there is much more shielding that has 'burn in', which is said to be part of the electric circuit.

 

More food for thought.

 

Almost anyone who has used a USB device at sometime has encountered a "bad" cable for at least one of a multitude of reasons ranging from:

  1. Not supporting the 5V power standard (if the device on one end is to be powered in this fashion). A lot of USB2 devices still are shipped with an external power supply (usually a "wall wart") precisely for this reason.
  2. Poor connectors on the cable (or misshappen from improper storage; even more of a problem with the new USB 3 connectors and these cables are so terribly short, must be due to the wires being used being extremely thin even compared to the USB 2 standards)
  3. poor connectors on the device itself (tend to break off or break the circuit boards if too much force is used or from muliple connection/reconnection or if someone transports the device with the USB cable in place)
  4. poor adherence to standards (ie. too cheaply made)
  5. other problems in the manufacturering process.

In a tightly bound multi-conductor wire package like the USB cable in general (there are not surprisingly flat USB cables now available on the market from some of the mainstream computer cable manufacturers, often labelled as high performance data cables) crosstalk can be a major issue.  I can see some benefit from cables where the design decreases or eliminates to vanishingly small levels the potential for crosstalk, and also improves on the "fit" of the connectors,etc.  All of these should be very measureable improvements (and also probably audible), but alas, the manufacturers of these cables are probably not going to publish or list these specs or design benefits on their packaging. 

 

There is little doubt that improved connections, particularly if the material making the connection is the same throughout (same on the cable and on the device being connected), will improve electrical throughput, how audible this is will definitely be a subjective battleground. I suspect only the most "golden eared" of audiophiles will actually claim/hear a difference that can be described as more than very subtle, if subtle at all. 

 

These same issues will also extend into the realm of internet cables (cat 5 vs 5e vs 6, etc.) I would suspect, but given the longer runs and other issues with their use and installation, probably more measureable and audible in certain instances.

 

What I am also curious about is those who have reported "dropouts" etc with USB cables, what material have they been playing back at the time (low res MP3, hi-res 192KHz or what?)? As an earlier post has  mentioned, are the software drivers for the interfaces up to the challenge of "pushing" this data stream through the cable without overloading their own data buffers or the physical limit of the cable itself? (are we pushing the technology with some data transmission and hence "loosing" or changing the data).

 

Jitter will be a potential problem for digital transmission along a USB cable (computer to DAC, etc.) but not for analog audio.

 

 

In the interests of scientific research, you understand, I have recently set up a new office music system with a 2012 mac mini playing music via iTunes & the Bitperfact app through a new Naim DAC-V1& NAP140 into a pair of wall mounted PMC DB1i speakers.

 

I have used a USB cable to connect the mac mini to the DAC-V1.  It is really, really cheap.

The DAC-V1 and NAP140 are both powered from a SMPS infested ring main (at least 15 SMPS on this ring main) via a cheap DIY store 4 way mains distribution block and Naim mains cables.

 

Given such care with set up, the sound quality is, as you might image, simply brilliant

 

Having met Touraj, liked him and as I really enjoy my Roksan turntable, I've ordered a Vertere Pulse D-Fi USB cable (yes, it's a posh one!)

I'm also having a mains radial circuit installed for the new music system.

 

It will be interesting to hear the differences that these 2 changes make.

 

Best regards, FT

Couldn't quite work out issue about data loss and compliant cable construction. The specification has quite a loose standard as I said.. I was talking about EMI and coupled RFI which is quite separate... .??????

But any if there is a bu error through signal error the receiver flags it and asks the sender to resend, so you USB flow control and retry. It's quite a neat standard.

Simon

Originally Posted by Simon-in-Suffolk:

 

But any if there is a bu error through signal error the receiver flags it and asks the sender to resend, so you USB flow control and retry. It's quite a neat standard.

Simon

Interesting, so there theoretically shouldn't be any difference between a so called "crappy" cable and a "hi-tech" one except if the cable (or circuit) is physically broken at some point impeding the transmission.  Very neat.

 

So Foot Tapper, when the reno is completed, you should be able to switch USBs and there should be no audible difference other than those associated with the USB standard for flow control and retry.  I suspect the dedicated mains will make the usual documented improvement in sound quality.

Hi, not quite my meaning... A less optimum USB cable can be within the USB specification and work as intended, but can be more prone toi causing and be affected by interference and physical electrical transmission issues such as coupling, imbalances and reflections. These parameters can effect the connected equipment.

The USB standard has a means of error detection and  data retransmission. Therefore with a less perfected cable (but still compliant), or a cable working in a hostile electrical environment or near the edge of the USB specification the probability of signal degradation causing a bus error, and data retransmit is higher. 

 

Interesting point to me, I don't know without looking it up how data retransmit protocol works in asynchronous mode if at all, unless anyone wishes to share on here...

 

Simon

 

Edit. My curiosity got the better of me.. In asynchronous mode, which uses the USB isochronous transfer mode, there is no handshaking or retransmit protocol. There is simply a CRC check to discard the frame on signal corruption... Not dissimilar to SPDIF or Ethernet layer 2 frames for that matter.

Simon, good point re what happens to data in the asynchronous mode. This is curious to me, too.  Also in asynchronous mode what aspects of a computer are being used, (the V1 mimicks a keyboard on a mac, so which elements of the operating system and hardware control software is at play?)

Answers to these questions may shed some light on the point made in am earlier post about software drivers.

Hope someone has some insight...

J

There's a group  test ( blind listening auditions) of USB cables in the July edition of Hi-Fi-News.

Prices range from £18 to a gob smacking £6500.

In the same edition Paul Miller, the editor,  has publish a one page article on how he produced and conducted the necessary technical tests.

Originally Posted by sheffieldgraham:

There's a group  test ( blind listening auditions) of USB cables in the July edition of Hi-Fi-News.

Prices range from £18 to a gob smacking £6500.

In the same edition Paul Miller, the editor,  has publish a one page article on how he produced and conducted the necessary technical tests.

Just the booksellers here and they only have the May issue, guess will have to wait about 2 months for the July issue to show up on this side of the Atlantic.

I have just come across a review of high end usb cables ( just type 'audiostream usb cable shootout' in Google).

The review, comments and links in the comments may help.

The comments on the review also show the same perennial argument about whether digital cables making a difference.  One of the commenters provides links to explain jitter and the nature of asynchronous USB.

Interestingly, one of these links is to Audiphilleo, who supply the technology within the V1, which says that asynchronous USB reduces the affect the cable has, I.e. doesn't eliminate it.  So if one's system is capable of showing the differences, the cable choice matters.

Another fundamental point from one of the commentors -"Also keep in mind that a digital signal transmission is effectively the same as an analog transmission. The electrons in the conductor do not see any difference. And the 0 is flagged as a 1 when it approximately crosses the 0V on a pseudo-square wave with a specific rise-time on the nS time scale."

For the OP, it may be worth talking a audiophile cable specialist dealer who is able to allow a lengthy time scale in their returns policy to allow you time to run in and audition the cables at home in your system. I have used Futureshop in the UK and can recommend them.

J

+1 on a serial bus signal in a USB is an analogue differential signal that when decoded provides the logical bits of a data frame..

However can't agree about asynchronous mode is less dependent on the USB cable. In fact it's the complete opposite. In asynchronous mode which uses isochronous USB transfer, there is no flow control or frame retry. A poor cable or a cable working on the edge of its spec will more likely corrupt a frame and the data is gone forever.. There is no retry.. This is the same as SPDIF or using UDP on Ethernet. 

Simon

Originally Posted by Simon-in-Suffolk:

+1 on a serial bus signal in a USB is an analogue differential signal that when decoded provides the logical bits of a data frame..

However can't agree about asynchronous mode is less dependent on the USB cable. In fact it's the complete opposite. In asynchronous mode which uses isochronous USB transfer, there is no flow control or frame retry. A poor cable or a cable working on the edge of its spec will more likely corrupt a frame and the data is gone forever.. There is no retry.. This is the same as SPDIF or using UDP on Ethernet. 

Simon

 

 

Makes logical sense to me (i.e. if the V1 calls the data but there is no way of checking whether all of the data is sent then the it can only work with what it receives/has already).

 

I guess there is also  the issue of jitter and whether the cable introduces this and if so the effect this has on data loss....

 

So, my understanding is that the perfect cable is one that does not alter the data stream content caused by EMI.

 

Of course, in reality the decision is about choosing which cable minimises this and accepting the affect it has on the data stream in order to achieve this.

 

Of course, the system has to be capable of showing the difference.  I guess most Naim kit would.

 

I do love a good puzzle

 

J

 

 

Jude, I think you more or less sum it up.. However i suggest you don't get wrapped up about jitter.. This is transport jitter.. The same as for SPDIF jitter.. It's not sampleclock jitter. The data is sent in frame bursts anyway.. It's not a smooth bit stream of PCM. In the old days the DAC clock was often derived from the transport clock directly and so transport jitter could cause all sorts of upsets. These days DACs use there own precision clocks or at worse a phased lock loop clock, and the data is buffered as the transport frames are read and decoded. Therefore these days transport clock jitter, like SPDIF transport jitter , is more likely to let itself be known via EMI and indirect side effects rather than directly impacting the DAC clock.

 

Simon

 

PS Dr Mark.. Let your inner geek surface 

 

Well, the Vertere Acoustics Pulse D-Fi USB A/B cable arrived this morning and has been duly installed to replace the unbranded, shielded cable that I have been using (it's a spare cable that originally came with a wifi router).

 

I can confirm that the system definitely sounds different when the unbranded cable is replaced with the posh one.  Rhythm, timing, bass quality and musical involvement are significantly different.  It was obvious within 2-3 seconds of hearing the first song (an Eric Bibb track from his album Painting Signs) and wasn't subtle.

 

Why is it different?  What is the technical explanation?  Despite Simon's best, informed and technically eloquent efforts to explain such things, I can confidently assure you that I have no idea.  Absolutely none.

 

But one cable allows the system to play wonderfully engaging music, while with the other I hear a slightly ponderous, plodding, slow hi-fi that tries but fails to play music.

 

More to come as I dig into this a little further.

 

Best regards, FT 

Well, replaced my $10 Nexxtech 1m High-Speed data USB cable with a $75 1m Straight wire USB cable and the difference is akin to adding a NAIM Power Supply to any Naim component.  There is a difference and must be down to the quality of the wire used. 

Gotta love this snake oil - http://www.hifiart.se/index.ph...tID=4&prodID=224

 

"Cables can not actively improve a recording´s sound, of course, but they can adversely affect it to varying degrees. The key is to make a cable as neutral as possible when it comes to conveying information."


For people that don't understand the difference between a digital and analogue cable this must sound really scientific. LOL. In case people are wondering, when it comes to digital cables there is no such thing as 'neutral' - they either work or they don't - end of story!

Originally Posted by Conrad Winchester:

Gotta love this snake oil - http://www.hifiart.se/index.ph...tID=4&prodID=224

 

"Cables can not actively improve a recording´s sound, of course, but they can adversely affect it to varying degrees. The key is to make a cable as neutral as possible when it comes to conveying information."


For people that don't understand the difference between a digital and analogue cable this must sound really scientific. LOL. In case people are wondering, when it comes to digital cables there is no such thing as 'neutral' - they either work or they don't - end of story!

Oh dear another bits-are-bits adept. :-(

Aleg, exactly, when it comes to digital cables, the statement it either works or it doesn't is just sooo mis informed... It simply isn't the case LOL ... After all 'digital' cables carry an analogue electrical or light signal. Good old physical layer in the OSI 7 layer model if some one needs an introduction. If this wasn't the case we mostly wouldn't need many of the techniques and protocols of digital data transmission to ensure bits remain bits.... Even in our humble SPDIF framing protocol. And that's not even at looking intermodulation and EMI from the physical signal.

 

Simon

Likes (1)
Jude2012
×
×
×
×