Why Do Different USB Cables Sound Different

Steve Williams

Site Founder, Co-Owner, Administrator
#22

marty

Active Member
Apr 20, 2010
1,107
1
38
Far Hills, NJ
#23
I am in full agreement with my friend Steve that I don't quite understand why USB cables should have the sonic consequences they do. They're just transmitting 1's and 0's, right? I guess the old audiophile adage "everything matters" is no exception here. I also agree with Steve that until recently, I thought the Wireworld Platinum Series 7 was the cat's meow. But I've recently explored some alternative US cables that have exposed some liabilities of the Wireworld.

Actually, I've been evaluating 3 cables recently. The first was a new cable line that Vladimir Lamm is either importing or using at shows called "Tchernov". I evaluated the their reference USB cable- (not sure what its called exactly) Its mids and highs are neutral, natural, and musical. It is unfortunately, not the last word in definition at the bottom but its a damn good delightful cable. It was the cable that showed me for the first time that the Wireworld top end was in fact a bit bright. It isn't the last word in imaging perhaps, but overall, it is musically pleasing where most of the music lives. I can easily see why Vladimir likes it. I mean this as a compliment- one might even call it the "Lamm" of USB cables! I have no idea what the Tchernov price is but it could be quite a bargain if its really affordable. Bottom line, as good as it is, it isn't compelling enough to make me to give up the Wireworld.

The Wireworld platinum has been widely lauded universally and its easy to see why. It has a powerful, well defined a low end, a very good midrange and a very slightly bright top end that could be a great fit on some systems. It is certainly a very fine cable.

However it was the Stealth Select "T" that really knocked me out. The Stealth has an extraordinary midrange with spatial imaging that is not just the best of the three but is really in another league. It easy to love it for its midrange and its imaging alone which is drop dead impressive, but it is quite good elsewhere and is nicely balanced. If the purpose of improved performance of home audio hear is, by analogy, an exercise in trying to leave the room by moving half-way towards the door with every step, all the while knowing you can never actually leave the room that way (i.e. no matter what you do it's not going to sound like live music), then in the end, the Stealth USB is the one that provided me a real sense of taking a step half way towards the door. It's resolution and spatial imaging are exceptional, its timbres seem correct, and in the end, it puts a smile on my face as my new reference USB cable. (Oh, by the way, the "T"is a sliding adjustment filter of some sort that is pretty easy to set in the sweet spot, and does in fact make a difference in the sonic result that the cable delivers. For me it was about 2 inches past the 1/2 way mark closest to the DAC). One last point. My Meitner DA2 has a galvanically shielded input for the +5V supply. Meitner therefore says the choice of USB cable doesn't much matter. I can't say I'm in agreement with that sentiment. Of course, YMMV. I'm looking forward to hearing the MB USB shortly, which comes highly praised. Should a fun day on the game board.

Finally, a shout out to Anthony Perrotta in Bethel, CT who is one of those vanishing breed of high end dealers that lets you listen to stuff first without any obligations, is easy to work with, and offers good pricing.
Marty
 
Last edited:

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#24
....snip

. My Meitner DA2 has a galvanically shielded input for the +5V supply. Meitner therefore says the choice of USB cable doesn't much matter. I can't say I'm in agreement with that sentiment. Of course, YMMV. I'm looking forward to hearing the MB USB shortly, which comes highly praised. Should a fun day on the game board.
....snip
Marty
Anyone who has properly tested USB isolation realises that isolation on the USB 5V VBUS is simply not nearly a solution. I have found that galvanically isolating the USB VBus & ground wires would only address ground loops issues, if they existed. The majority of disturbance comes from the signal wires & these need isolation to mitigate this disturbance. The audible result of such mitigation is greater clarity, sound stage depth & a greater sense of realism.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#26
This article is the best and most technically comprehensive article on why USB cables can affect sound quality. Skip half way down to the part where he talks about contacting Gordon Rankin.

http://www.digitalaudioreview.net/2016/05/gordon-rankin-on-why-usb-audio-quality-varies/
I've seen this claim by Gordon Rankin before & I don't buy it. It's easy to check if asynchronous USB used in audio playback is dropping bits or not - play an audio file which is a HDCD file. Send this to a DAC with a HDCD indicator which turns on when it senses & plays a HDCD file.

How does this show the USB asynchronous transmission is bit perfect? HDCD is a one bit flag embedded in the audio file & if this bit isn't constantly sensed by the receiving DAC throughout the playback of the file, then HDCD is turned off. I have tested this & never found a dropped HDCD flag & I have never seen anyone report this.

Furthermore, there are USB audio devices which include a bitperfect test file for transmission via USB & never have I seen a report that there were bit errors found
 

marty

Active Member
Apr 20, 2010
1,107
1
38
Far Hills, NJ
#27
Apr 3, 2010
16,022
0
0
Seattle, WA
#28
I am in full agreement with my friend Steve that I don't quite understand why USB cables should have the sonic consequences they do. They're just transmitting 1's and 0's, right?
Hi Marty. They are explicitly transmitting ones and zeros. But implicitly, more is sent. See this article I wrote: http://audiosciencereview.com/forum...performance-pc-server-interfaces-async-usb.8/

Briefly, the process of converting a digital sample to audio involves both digital components and analog. Digital systems are quite perfect with respect to sending the digital data. What happens in analog world on the other hand, is subject to many obvious and non-obvious factors. Here is an example measurement from above article:



Here, both HDMI and S/PDIF transmit identical set of PCM bits. The analog output of the DAC however is drastically different. Why? Audio is slaved to video in HDMI so when using that for audio, you light up a lot of circuits in the receiver in addition to having to extract the audio clock from video. In less than optimal implementations like the above AVR, the measured performance over HDMI degrades substantially.

The "bits are bits" is an argument from extremist objectivists who shout louder than their level of knowledge :). It is not true and it is just an imagined way the system works, not how it really does.

Note that this doesn't necessarily explain the subjective reports from users. Other than extreme cases, I have not been able to find a measured difference in the analog output of the DAC when I change USB cables. I have not seen it published at such elsewhere either.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#29
Basically Rankin is saying that USB audio transmission (asynchronous) has no error correcting capability (which is correct) & he maintains that bit errors occur during transmission of audio (which is incorrect). If this happened we would likely hear this as obvious audible glitches plus it's wrong for all the reasons I already mentioned
 
Last edited:
Apr 3, 2010
16,022
0
0
Seattle, WA
#30
Thanks Caelin. I 'm not sure I understand it at the level I'd like to, but its helpful and I think I get it.
Marty
He is saying as speeds/sampling rates go up, you get USB data errors. I don't buy that argument. If audio samples are dropped, you will immediate and highly audible static, pops, glitches, etc. at that moment. You will not hear an overall change in tonality/fidelity of the music across all of the music you play. That would require data errors all the time which is not remotely supported by any data I have seen.

It is of course possible to cause errors by using poor cabling or instrumentation but I don't think any audiophile is hearing those artifacts per above.
 
Nov 27, 2013
217
0
16
#31
https://www.nexthardware.com/forum/pc-top-hardware/85947-abiura-cavi-usb-12.html#post966153
(with Chrome you can easily translate it)

Here you can read some measurements of various USB cables, made by a dear friend of mine (a real nuclear engineer), on how they radiate different electromagnetic fields that affect data flow. Obviously, the DAC is able to reconstruct the signal, but the downstream result is different.

For the record:
1.) Generic cable
2.) WireWorld UltraViolet 7
3.) WireWorld Starlight 5
4.) diyCAT7

You should also read the following messages because there are some clarifications about it.
 

marty

Active Member
Apr 20, 2010
1,107
1
38
Far Hills, NJ
#32
He is saying as speeds/sampling rates go up, you get USB data errors. I don't buy that argument. If audio samples are dropped, you will immediate and highly audible static, pops, glitches, etc. at that moment. You will not hear an overall change in tonality/fidelity of the music across all of the music you play. That would require data errors all the time which is not remotely supported by any data I have seen.

It is of course possible to cause errors by using poor cabling or instrumentation but I don't think any audiophile is hearing those artifacts per above.
Thanks Amir. I guess some of my take away from the article is that Rankin says, in essence, everything matters; the type of USB transmitter and receiver used, the ongoing computer operating system processing that occurs simultaneously with USB signal transmission, the quietness of the power supplies, the type of wire used in the digital cable, the type of insulation, etc. What seems somewhat surprising is that most of these things are constant in one's system and yet changing a single variable such as the USB cable itself is enough to alter the sound. Then again, I think the point of the article is that we shouldn't be surprised at all this can occur. Whether there are differences that can be measured to explain this phenomenon is certainly beyond my expertise and thus defer to others for discussion on this point.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#33
https://www.nexthardware.com/forum/pc-top-hardware/85947-abiura-cavi-usb-12.html#post966153
(with Chrome you can easily translate it)

Here you can read some measurements of various USB cables, made by a dear friend of mine (a real nuclear engineer), on how they radiate different electromagnetic fields that affect data flow. Obviously, the DAC is able to reconstruct the signal, but the downstream result is different.

For the record:
1.) Generic cable
2.) WireWorld UltraViolet 7
3.) WireWorld Starlight 5
4.) diyCAT7

You should also read the following messages because there are some clarifications about it.
Thanks bibo01 - JosephK is always interesting & his measurements well executed - He presented some very useful measurements on a thread of mine on DIYAudio "RF attenuators = jitter reducers?"
 

marty

Active Member
Apr 20, 2010
1,107
1
38
Far Hills, NJ
#34
Thanks bibo01 - JosephK is always interesting & his measurements well executed - He presented some very useful measurements on a thread of mine on DIYAudio "RF attenuators = jitter reducers?"
This is interesting as it's the first time I've seen something "measurable" to distinguish a USB cable that is just supposed to transmit "0"s and "1"s. Not that I have any idea as to what I'm looking at, but it's interesting. It would also be very presumptuous of me to try and correlate any of these measurements with sonic performance, but I sure would like to see the data of the cables I've reviewed lately.

I’ve listened to a bunch of USB cables over the past 2 months and for what it’s worth (YMMV) here are some thoughts on individual cables.

Wireworld Platinum Starlight 7- a good cable with lots of detail but a bit forward on the top end. I would not think it would be a great match with electrostats or some solid state amps

Tchernov – supposedly preferred by Vladimir Lamm. A very good cable, well balanced, very musical. The “Lamm” of USB cables as you might expect. Not the last word in bottom end definition but quite good.

Masterbuilt Ultra. This cable came with a lot of fanfare and I tried as hard as I could to cotton up to it but could not. It's got a very good midrange, a super top end, but it's quite bottom weighted, at least in my system. I even tried to adjust my subs by turning them down considerably, however I just couldn’t achieve the overall balance that I was seeking. For those of you that know JL subs, I had to turn down the “ELF” control by 6dB as well as the main gain. Without doing that, I thought an earthquake was occurring in my room. I plan on re-visiting this again after I receive some MB Ultra speaker cables since this could be a “loom” effect issue (i.e, things sound best with the same brand throughout) , though I’m not even sure why this might be the case. My gut tells me otherwise.

Canare. This is the cable John Atkinson supposedly uses on all his recordings. I figured, for $40 bucks, what’s not to try? My hope is that this cable would convince me that we are all nuts and irreparably damaged and that for all intent and purpose, this bargain basement special would perform as well as the high priced spread. Well, apparently, there is a high-end audio God. The cable is a lackluster performer at best. I would suggest that Mr. Atkinson avail himself of an alternative, although more expensive option.

Stealth Select “T”. I was essentially unaware of Stealth cables until I visited Philip O'Hanlon’s home a couple of years ago. When Philip played a tape source, I was knocked out and wandered by his equipment rack to see what he was using. That’s when I noticed these fat white interconnects on his front end gear and learned about Stealth for the first time. I don’t even think he was using the latest and greatest of their current line at the time. But it didn’t matter. It was clear that whatever the hell he was using, it sounded damn good and they have intrigued me ever since. I subsequently tried a pair of their Sakra 10 interconnects from my Meitner DA2 DAC to VTL 7.5 III preamp and confirmed they were superb. Somwhat surprisingly, they were essentially the equivalent of my MIT SHDs sonically, although I gave the slight nod to the MITs. Therefore I didn’t replace the MIT SHD’s but I have paid attention to Stealth ever since. When the opportunity came to try their USB, I gave it a whirl and am thankful I did. It replaced the Wireworld USB in short order. The more I listen, the more impressed I have become. This is by a considerable measure, the most impressive, best sounding USB cable of the bunch. It is balanced, detailed, musical and has spatial detail that is unlike any other USB cable I’ve heard, and is in fact, a very analog sounding USB cable in that regard. As mentioned previously, the placement of the sliding tunable thing-a-magig on the cable is critical, but it's fairly easy to set. Stealth cables have been widely reviewed but mostly outside the US. I really loved this review of their Select “T” USB cable in particular because it’s in Chinese and I can’t read a word of it!

http://www.stealthaudiocables.com/reviews/USB _review.pdf

However, if you look at the graphic ratings of the cables reviewed, it sure seems easy to conclude they loved the Stealth Select “T”. Actually, I did run this article in Google Translate and even though the English translation was at times comical, I got the gist- their one liner said it all: "True water without fragrance". I can’t say much more than that (even if I have no idea what the heck they meant by some of the things they said- you really need to read the translation for yourself.). It’s just been a wonderful performer in my system. And it's reasonably priced (1m=$1250 retail).

For now, my USB explorations have taken a much needed break. I’m happy to live with the Stealth Select “T” for quite some time.
Marty
 
Last edited:

Steve Williams

Site Founder, Co-Owner, Administrator
#35
Masterbuilt Ultra. This cable came with a lot of fanfare and I tried as hard as I could to cotton up to it but could not. It's got a very good midrange, a super top end, but it's quite bottom weighted, at least in my system. I even tried to adjust my subs by turning them down considerably, however I just couldn’t achieve the overall balance that was seeking. For those of you that know JL subs, I had to turn down the “ELF” control by 6dB as well as the main gain. Without doing that, I thought an earthquake was occurring in my room. I plan on re-visiting this again after I receive some MB Ultra speaker cables since this could be a “loom” effect issue (i.e, things sound best with the same brand throughout) , though I’m not even sure why this might be the case. My gut tells me otherwise.
Definitely system dependent Marty as fat bottom end is not anything I have ever heard with the MB USB cable. Definitely revisit it after you get your MB SC in and I bet it will be different. There is absolutely nothing that stands out with the USB Ultra cable in my system
 
Dec 12, 2013
935
0
16
NYC , USA
#36
USB cables matter no doubt on that. They matter less as the server is better they matter less . In mine I have many reg DC supplies one for the USB audio card alone. My msb needs the 5 volt handshake . My lampi does not I have cables for both . On my ref server the 5 volt ISO does not matter. Steve does that cable have a way to remove the 5 volt ? Also does this cable have anything more than the cable itself ? Steve Nugent I think it was him and also Paul pang audio supplies also said the length matters. Too short is not right I forget the distance. It has something to do with an echo. I make no claims in knowing only hearing. When I bought my msb stack I bought a audio quest USB cable its there diamond from a few years back. It's the one with a battery on it that is suppose to charge the shield. I never liked that cable lol. It's to hard sounding no matter what DAC or server I use with it.
What I do wonder about is AES and its lack of use in computer audio. While there are some cards made rarly does anyone use them. I don't. My msb stack has a server steamer the UMT plus . It has many outputs and the one they Recomend is gues what. AES , I had my lampi have one. While luk of lampi was not happy claiming it added another board inside . I can say for PCM it rivals my custom server USB . Msb claims balanced digital is better than USB. And consider i2s over cat6 is the choice from server to DAC.
They also now have a network renderer to keep us away from servers.
There claims seem to point get away from servers. I think it may have some merit but it also makes things simple too once you have some network storage with music on it.
I cannot say with certainty if my server is better than the msb streamer but they do sound a bit different . But my server does have win 2012 with AO and find pro so as they update my sound does change a bit.
Steve what type of server do you have ?
One model of arrender has a USB output I have heard it a few times on different systems it did sound good too.
Over time I think we all become to either like or hate our systems. If we love it changes need time like Steve did taking days to tune it in . Once we lock on an a b gives us some sanity in this hobby.
Now my take on why is jitter , in general it's noise be it voltage or timing errors in the digital domain. While a normal data stream is one and zeros and filling in the making bits by sampling over. In audio I think this matters . All of our servers and dacs have to take the stream that is recloked , stored in memory , converted and so on all before it's actually processed from digital to analog .
Even the hardrives in our servers move the data around causing fetch noise. I think the future of great digital is to get away from the many varibles . This means network attached storage or internal storage. Why we can spend 90 k on a msb sel 2 and it has no internal storage is just crazy and dcs with many boxes and no storage . On my setup the UMT plus plays tidal as good as my custom built server streamer does.
 

marty

Active Member
Apr 20, 2010
1,107
1
38
Far Hills, NJ
#37
I recently had an email conversation with Serguei Timachev from Stealth Audio who provided a cogent (at least for me) explanation of why USB cables can and do sound different. Thought I'd share it below. Interesting info.

"The common (incompetent) logic AGAINST ANY premium USB cables usually is that in digital signal transmission the data errors get corrected (and the incomplete or distorted data packets get re-sent). For some reason folks do not know (or do not realize) that music is being transmitted as a STREAM, without any correction, and thus gets distorted (more or less) in a similar fashion as analog signals.

On top of that, a lot of people do not know that S/PDIF protocol for digital audio is, in fact, an ANALOG protocol: digital signal, consisting of “ones” and “zeros” gets converted into a corresponding analog signal, then this ANALOG signal gets transmitted via a “digital” transmission line (75 Ohms characteristic impedance at BOTH ends), and then the receiver chip once again converts this analog signal into digital. Any distortion of this analog signal CHANGES “ones” and “zeros” in the received signal, after the second conversion at the receiver. I.E. the received “ones” and “zeros” digital signal after the S/PDIF transmission is NOT exactly the same as the signal that has been sent. And THIS IS why digital cables do make a sonic difference."
 
#38
I don't believe that any data gets corrupted in a 1-2 meter USB cable. It is noise and jitter that is introduced into the DAC USB interface.

I have seen several effects and I have my own theories:

1) improving the power supply for the USB output port on the computer makes a difference - This will generally only improve the risetime and signal integrity of the signal on the cable

2) introducing common-mode filters or isolation techniques makes a difference - This indicates that the common-mode rejection of the USB receiver is insufficient.

3) using a really low-loss cable with air or expanded Teflon dielectric makes a difference - This indicates that the risetime and signal integrity is important

4) cutting the 5V supply and replacing with a fast-reacting LPS makes a difference - This indicates that crosstalk in the cable or the quality of the 5V is important.

5) using a 1.25-2m cable length improves things - this indicates that impedance discontinuities and the resultant reflections on the cable matters.

As a result, I use a really fast Hynes-based LPS called the "Power Block" to provide 5V power to my USB interface. This inserts into the cable and provides a short stub cable to the DAC. I use a 1.5m USB cable that is well-matched to the spec impedance. I use a USB port on my Mac Mini that is superior to the other ports. I provide a fast LPS to power my Mac Mini.

Steve N.
Empirical Audio
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#39
Steve is right - the cause of different USB cables is not to be found in digital transmission errors.
It's easy to prove this by playing a HDCD audio file on computer via USB to a HDCD capable USB audio DAC or device - these usually have an indicator that shows when HDCD is being prcoessed
HDCD is just a single bit indicator in each byte transmitted. If even one of these bits are missing HDCD is turned off
Playing a HDCD audio file will always show HDCD is being processed - no bits missing i.e no errors

What this guy is talking about is that asynchronous USB audio (the common protocol used for audio transmission) has no error correction or retransmission of packets
 
May 30, 2010
13,925
14
38
Portugal
#40
I recently had an email conversation with Serguei Timachev from Stealth Audio who provided a cogent (at least for me) explanation of why USB cables can and do sound different. Thought I'd share it below. Interesting info.

"The common (incompetent) logic AGAINST ANY premium USB cables usually is that in digital signal transmission the data errors get corrected (and the incomplete or distorted data packets get re-sent). For some reason folks do not know (or do not realize) that music is being transmitted as a STREAM, without any correction, and thus gets distorted (more or less) in a similar fashion as analog signals.

On top of that, a lot of people do not know that S/PDIF protocol for digital audio is, in fact, an ANALOG protocol: digital signal, consisting of “ones” and “zeros” gets converted into a corresponding analog signal, then this ANALOG signal gets transmitted via a “digital” transmission line (75 Ohms characteristic impedance at BOTH ends), and then the receiver chip once again converts this analog signal into digital. Any distortion of this analog signal CHANGES “ones” and “zeros” in the received signal, after the second conversion at the receiver. I.E. the received “ones” and “zeros” digital signal after the S/PDIF transmission is NOT exactly the same as the signal that has been sent. And THIS IS why digital cables do make a sonic difference."
Marty,

I think this is not true - since long we know that digital audio of decent digital equipment is bit exact. This is the source of the headaches of those who want to understand the differences...