I have queried that myself on multiple occasions 
However, if I remember correctly, in some Z video ages ago, he mentioned that the owner of Geshelli was one of the people that developed the USB protocol and now refuses to include USB on any Geshelli product, so maybe it is voodooâŚ
Would be interesting to know the full story behind that!
At least thereâs a user guide with a troubleshooting section.
Back when movies had puerile plots, burn-in was free and resulted in a lot of trouble.
No need for shootingâŚ
Also being a computer guy, I had the same questions. I found a few articles on the subject, but this one was easiest for me to understand:
The quotes in that articles all come from people selling high priced USB cables, so I didnât necessarily take this as anything more than a marketing exercise, but some of the responses seemed aligned with more independently-sourced articles. Also, there were a few things that seemed to make sense to me (again, I donât have any understanding of electronics):
1. Music signals do not use the same error correction as files sent to a hard drive or printer
When sending a document to print, thereâs a code that tells the printer how big the packet was. The printer can check that code against the file received, and can ask the PC to resend the file until it confirms that the file is completely correct. Streamed music signals do not benefit from this resend function and therefore DAC error correction systems cannot replace missing or corrupted data.
2. USB cables do not pass binary data - they pass voltage signals which are interpreted as 1s and 0ss
Voltage below a certain level is interpreted as a 0 and voltage above a certain level is interpreted as a 1. Anything that interferes with the voltage signal can cause the data to be misinterpreted.
3. Most USB cables do not isolate the power conductor from the signal conductors
Noise from the power conductors can interfere with the signal unless suitable shielding is used.
That third point resonated with me, which was why I was looking for USB cables that isolate the power conductor. The Curious Cable seems to get the most accolades, and that physically separates the power conductor. I couldnât find a used Curious Cable, so I got the Black Cat instead, since that shielded the signal conductors from the power conductor.
This was an interesting experiment. I proved to myself that there can be audible differences between USB cables. But different doesnât necessarily mean better or worse, so I donât want to make any claims on which one is more musical. And those cables are very expensive, so I donât think they deliver value for money, unless youâre happy with what you have and donât mind spending money to get an incremental benefit (or worse case no benefit).
When the world is in better shape, if anyone is planning on driving through the Chicago Western suburbs, just PM me, and Iâd be happy for you to try the experiment for free. 
I think Iâm just going to FTP my music from the Mac to the Bifrost2 from here on out.
Thanks - Interesting read.
This is somewhat misleading.
Binary data is never passed as âbinaryâ. Wires and circuit board traces arenât capable of passing âbinaryâ. They can only carry analog signals upon which digital data is encoded.
Think of the old telephone modems where you put the phone in the cradle. The digital signal was encoded (modulated) into audio that could be sent over the phone line and then decoded (demodulated) at the other end.
The word modem came from sticking the terms modulate and demodulate together.
In this scenario you can see how interference on the phone line would prevent correct encoding and decoding. In the same way there could be interference in a usb cable strong enough to cause the voltage to be at a level it shouldnât be and thus interpreted incorrectly.
This seems to be in the interference territory again.
To me this reduces to minimizing interference as opposed to the wire itself having any special properties.
And how much emi/rf interference is required to do that?
My hunch is, far far more than a normal system experiences in normal conditions.
Upgraded, shielded, and enhanced USB cables may prove necessary with extreme use cases. These are the first thing that should be double-blind tested.
Maybe I didnât word it correctly. The point I was trying to make is that out of ignorance, I had always assumed that 1s and 0s are passed and it was surprising to me to find out that these are analog signals that need to be decoded. From other comments that Iâve read (on other forums), I think many others have made the same mistake as me. Your telephone example is very helpful.
I think you phrased it fine. The data is indeed passed as voltage changes, and mean nothing by themselves. Itâs just voltage in a wire.
The 1s and 0s are created when the usb controller on the other end interprets the voltage changes as digital data. And more than just the music is passed along in any USB connection, so if interference does harm the music it should harm this additional data equally. I donât see that happening on the systems Iâve used, so Iâm still very skeptical about the usb cable mumbo jumbo.
So, youâre both kinda right, as far as Iâm concerned.
Can you chase ultra clean signals in a USB cable? Sure. Have at it. Diminishing returns and marketing/greedy businesses wonât make that friendly on your wallet, but itâs your money so knock yourself out.
Iâd prefer to invest in items that improve on the USB controller inside the device where the real goings-on are happening vs spending obscene amounts of money to allow a crappy usb controller do itâs job.
I think Schiit would agree.
Alternatively, you can bypass all of the above and go AES or Coax with an Allo Digione Signature or Pie2AES device for less than one high end audiophile-approved, blessed-by-Tibetan-monks, cryo-treated USB cable. That solution is far more interesting and challenging to me than just getting better copper. Sometimes I have found that there are more effective and less costly solutions to problems that may not even exist but in our own minds.
(Can you tell Iâm not a big cable believer yet? LOL)
Oh, I didnât mean to imply that your statement was wrong.
I was trying to say that analog encoding is not unique to usb and wouldnât be a reason that usb is different than other data transmission.
Itâs important to have a quality receptacles. The key factor being plug retention. A basic hospital receptacle is good. Iâm using the lower end Furutechs in my dedicated 20A circuit. They sure grab the plug nicely and hold it firmly in place.
Dammit, Iâm curious again⌠Let me see if I can carry out another experiment thatâs a little cheaper than my USB one.
I donât have a dedicated 20A circuit, so Iâm âprotectedâ from spending a lot of money on audiophile outlets, which mostly appear to be 20A.
Furutech makes their own (and I respect Furutechâs engineering capability, as I have some cables and connectors made by them) but a lot of these audiophile outlets seem to based on customized Hubbell HBL5362 outlets. The 15A version of that is HBL5262, so I think I will order one of those for a few dollars off the shelf and see what happens.
I built a few Furutech power cables with Furutech connectors. DIY supples from VH Audio. FP-S032N wire.
I came across this product when looking for something else. Itâs an âaudio bufferâ that âprovides stable, high input impedance and low output impedance thus eliminating potential mismatch between a music source and an amplification stage.â The product literature doesnât specifically say what it does in order to achieve XYZ (or I missed it). Does it have an impedance converter (like the LTA ZOTL amps) or an impedance matching transformer (like the Hagerman Tuba) â which some say are no different from conventional output transformers?
Does anyone else use this or a similar product? In what types of chains and use cases can this type of product be favorable (presumably high impedance output amps with low impedance cans)?
I donât think I have EMI/RFI issues, but I came across ERS paper. My Schiit gear runs hot though!
Right, so putting a cloth under a unit that is exposed to RFI from above is supposed to help. Iâm not a believer.
Iâm a bit skeptical also. The first link goes through suggested uses, including placing on the underside of the top panel â which isnât really doable if your device has vents or holes or is transparent/clear.
Faraday cages are usually metal and fully documented as effective. Any conductive substance may be helpful with RFI.