We Need To Talk About Ones And Zeroes

How many times have you entered $50 to deposit in your bank account and it became $5000 because the banks server was 3000 miles away?

We are just talking music here, not a lot of data and not hard to do no matter the miles away the disk is. When you watch (stream) an F1 race 10,000 miles away in 4K, how many times does Hamilton’s red Ferrari gets displayed in blue? 4K streaming is much more difficult than streaming a 16/44 song or a hires song. 
I setup very large enterprise databases and worked with Facebook/Apple/Yahoo/Comcast/banks and these companies have servers all over the world on AWS/Oracle/Azure/their own server farms and every transaction done is bit perfect, with thousands/millions of people concurrently accessing these servers and every transaction has to be bit  perfect or we would have some big issues. And this occurs over dirty WiFi or using a $1 Ethernet patch cable. Now you take an audiophile that uses a $1000 Ethernet cable going to a galvanized streamer going to a $10000 dac, I’m not worried that a 1500 byte packet will make it to my system perfectly.

@rbstehno - too logical, don’t you know audio is special?

"The digital data rides on an analog signal that can pick up noise"

I think you're trying to explain Wi-Fi, which is a Layer 2 Data Link Protocol, as is Ethernet.  Layer 3 Network Protocols, like TCP/IP, handle all the logical addressing, routing, forwarding, fragmentation and reassembly, error correction, and buffering, and diagnostics. Any 'noise' accumulated at Layer 2 simply isn't recognized at Layer 3, unless it corrupts a packet, in which case a checksum error is generated, the entire packet is discarded, and another packet retransmitted. Remember, thi is happening at 100 Mbit or Gigabit speeds, hundreds or thousands of time faster than even the highest audio signals. From a network challenge perspective, even 192KhzX24-bit audio is small beans.

I keep reading the same old stance from folks that either don’t stream or streaming for fun with their fancy analog or CD players. You know who you are :-)

It’s not about the data but about the context (emotional, physical, psychological) in which it’s delivered. Think about it before you eager to call out the cult-like belief in well designed streamers with premium parts without understanding why something works and works well over mass produced streamers. 

I have compared streamers ranging from $500 to $25K. You don’t have to spend $25K to get a great sound, but pick a streamer that is well engineered to deliver bit-perfect digital output by implementing low-noise design, stable clocking, robust power supply and isolation (ethernet noise or power rail interference). 

So why some of the streamers sound different? Because they prioritize  aforementioned underlined elements that ultimately impacts how bitstream is distributed to your DAC. 

The message should be; don’t fall for over-priced pseudo tech without due diligence be it a DAC, Streamer or Switches. 

A complex topic when allied to digital music reproduction, made more difficult by S/PDIF.

Assuming sound engineering the bits at the output of the receiver in your device at the end of the Cat 5 cable are precisely the same as the bits stored at the source.  If they were not the modern world would collapse as no transaction could be relied upon.

Setting aside the linearity of the DAC chip(s) there are then, however, two (digital) issues that arise; what data gets presented to the DAC processor (the component(s) inside the "DAC" box that generate the analog signal) and when is that data presented. 

Unlike the TCP/IP ethernet or WiFi connection the S/PDIF has error detection but not correction so a truly terrible digital interconnect could, I suppose, result in bit errors, resulting in data loss.

My suspicion is that any unwanted artifacts that are caused by changing between decent digital cables are caused by jitter being introduced and not eliminated by processing before the data is presented to the actual DAC chip(s).

I did an experiment with different cables between my Aurender N100SC and the Esoteric K-01XD.  My baseline is the Cardas Clear USB connection, and I do have an external Stanford Systems Research rubidium clock.

The three cables that I tested were an MIT from the 90's, a "Trubutaries" Video Cable - not high end, not even a specific digital interconnect, and a two meter audio junk interconnect from who knows when (or why).

I did not listen extensively but the MIT cable and the Video cable were indistinguishable.  The junk interconnect however was clearly not up to the task, the sonic quality sounded like hard clipping was occurring!  

The Esoteric K series DACs have buffering and clock synchronization before the actual DAC processor, that appears to have taken care of any jitter issues.

I suspect that the audio cable did not have the bandwidth to support the 1.2MHz 24 48 S/PDIF signal and so there were data errors. (S/PDIF has a parity bit, hence 25x48,000  = 1.2MHz).

BTW, I take exception to a total dismissal of switch mode power supplies.  OK regarding wall-warts, but, for example, Benchmark have transitioned to use SMS to reduce noise.  It all depends on the quality of the power supply.

Once again, my apologies if I am a bit off-topic, but an interesting question makes my mind diverge onto allied topics.