Files are sended with TCP protocol that have error correction and retransmissions.
Audio Streaming is send by Streaming protocols similar to UDP that don't have integrity check and retransmission (like the une used in audio en video call/conferences).
This is because DACs don't have a "bufferer" to store the songs before play it, they are just processing the streaming "on the fly".
There are Jitters that generate errors and can be measured...
I hope with this explanation you can understand.

A guy in a forum made measurements, here you can see the difference between a standard usb cable and a WireWorld Starlight 5.
Standard USB cable:
https://uploads.disquscdn.com/images/6303d8df6a561cc08cdcaa24352ea61ed1301727851cf59c737de66c172771ff.jpg

WireWorld Starlight 5:
https://uploads.disquscdn.com/images/7b5e0b89607ac89d21c737411b119705a0ac97738954baaf5d359475f2e8d5b8.jpg

Quote from the article:
"Alex Sventitsky, WyWires

There is no such thing as a
purely digital signal. Digital "data" is represented by electrical
signals that meet certain conditions. In the case of USB cables, these
conditions are robustly defined in the USB Standards Document. The data
transfer within a USB cable is accomplished by using two electrical
conductors, D+ and D-. USB defines one's and zero's as a difference
between the two electrical signals transmitted by D+ and D-. A One is
when the D+ signal is 200 millivolts greater than the D-signal. A Zero
is when the D+ signal is 200 millivolts LESS than the D- signal."

+1.
From what I could read from the article, these people (cited, from each and every company) know what they are talking about. Much, if not all (I believe even all of it, I did read fast moments ago) of what they are citing can be found even in textbooks, technical documents and data-sheets from different manufacturers, pure EE stuff.

Proofs are already, from what I have read in the article, given (if the explanations given can be possibly attributed as the "indirect") already. So in a way, a textbook is given there, and details given in the article about digital signal being in effect an electrical signal, which is not strictly at the 0 or 1 level, but within the certain variation of the voltage levels (measured!) ) are telling clearly the story, even the explanation about jitter and possible internal interferences.

+1 on this. Not an employee, nor related to any cable company, having been formally educated within EE and CSE it is easy to recognize all the arguments and details raised by the people from the industry, technical details raised and explained in the article. Furhtermore, I disagree with one of the previous comments here that there is "so little or no useful end user info. here in the article". On the contrary, there is a wealth of the info provided from each and every participant in the article.

Furthermore, I have just opened the box and instead of the generic USB, thin cable (USB A-B) provided with my headpone amp/dac started using the dedicated USB A-B audio cable.
I was entirely skeptical about any possible effect on the sound quality, to me it was like, "well, yeah"...
To my surprise, there is actually very noticeable difference in the quality and clearness of the sound, which is the integrity of the signal in the end. Reading the article after testing might be the ground for suspicion that the article is providing the placebo or justification (to me personally) for what I am experiencing, but even without running across the article, prior to reading it I felt immediate difference in the sound. It is really that obvious.

Just as a reminder, I had a negative (totally) bias prior to opening the box, towards this "dedicated" cable, and I was ready and set to return it.

To my surprise, still, now I have no intention whatsoever of returning it.

It will not sound different. Especially with a digital signal it´s a hoax to claim otherwise!

By empirically you mean ABX testing?

It seems the writer of this articel has a specific agenda.

This is fantastic. It seems the very idea of digital data transfer seems to be missing here, no to mention concepts like "checksum" or "bitperfect". The digital data transfer was actually invented especially to assure that the transferred data is identical to the (digital) source data. That's the very idea. The cable manufacturers' standpoint seems cynical at best.

This whole post was a complete bunch of silliness. A gaggle of cable manufacturers post about how their simple data transfer cables can sound different from one another. What would you expect them to say? When they start to produce scientific evidence on how the bit stream is effected in a way that the DAC will recognize and produce a altered analog waveform from, you should then start paying attention. Or even a positive simple blind listening testing session might make it worth further investigation. If a USB cable does alter the data it is BROKEN. You'll never get anything in the way of scientific evidence because it doesn't and can't exist so what you get is statements that seed the mind with incorrect bias.
If any of this had validity you would have to worry about the digital data of anything else transferred via USB. But you don't, data transmission is a known technology and all this babble is just a maneuver to lighten your wallets.

The issue is it doesn't matter if file transfer takes nanoseconds longer due to retransmits; it's the streaming of audio and jitter in real time receipt of data that is a factor.

If the DAC were to wait until the entire file were transferred to the device as some memory-type players do, the cable would not matter as long as it transmitted the data intact.

Unfortunately most people would not find that solution acceptable as they would not want to wait 30 seconds for a song to transfer before they heard it, and since there are no de-facto stop points in audio streams, this would only work with data such as digital files you already own.

well said.

You know, I get so tired of hearing how only blind tests can be "trusted." We are not testing medications for efficacy here, we are listening to emotionally moving and sense stirring music. Two chicken breasts, from the same chicken, prepared differently, will taste different, whether you are blindfolded or not. The analogy is the same; two cables of identical length, even of identical purity and casting of copper (if that could be possible), but constructed with differing geometries, dialectic materials, shielding, casings, and terminations WILL SOUND DIFFERENT...regardless of whether you want to hear it or not, or whether you can see them or not. Which one is preferred MAY differ between listeners, but they WILL sound different. Give up on the hyperbole and learn to trust the empirical evidence presented by one of the most complex and discerning test instruments ever developed, the human ear/brain connection. So, essentially, what Kemper said! 😉

Sighted listening can supply no scientifically accepted evidence. All sort of bias will have control over what you believe to hear. If the differences in the cables were that "huge" you would be able to easily identify them under blinded conditions. I respect your listening impressions but to present them as anything more than opinions can not be justified. The gentleman sitting next to you could just a easily reported hearing something else or nothing at all. Genuinely "huge" differences in soundstaging should also be easily measured in such things as L-R balance, changes in phase, etc. Not something I see as possible in any correctly designed cable but I always remain open to looking at provided evidence.
Cheers, Sal

Not blind, but output levels carefully matched to 0.25 dB. It was very difficult, nearly impossible to tell 3 of the DACs apart, the cable differences were huge in comparison, mostly soundstage which changed from flat 2D, to expansive and a very developed sense of depth, again we were all surprised.

Where those differences heard and the evidence tabulated under tightly controlled blind listening circumstances? I didn't see the recorded results posted.
TIA

The incredible ignorance of audiophiles astounds me
of course digital signal transmission integrity can be measured lol

its a multibillion dollar business 🙂

http://www.onsemi.com/pub_link/Collateral/AND9075-D.PDF

please remember that the frequencies of transmission are around 6 MHZ, and therefore quite different properties from typical audio transmission frequencies

Q. What cable interface should I use?

A. With a properly designed DAC, as described above, none of these cables have a bearing, because all that the DAC expects from them is a truthful transmission of bits. Since all of them can do that I don’t expect any sonic differences. However, there are differences in length, cost, universal application. I personally use USB due to it’s bandwidth and ability to transfer the highest resolution files available. Since the cable is not transferring an analog signal, the differences you may hear between cables will most likely only be due to shielding.

When the owner of a Benchmark Audio DAC2 HGC DAC contacted Benchmark to get a recommendation for a long USB cable, they recommended to the owner buy the best $6.95 cable they could find.

So I would suggest as a followup article surveying DAC designers (who are not affiliated with cable companies), and get their input on the value of high end USB cables.