(...) While this is not my area of expertise (I don't design electronic circuits), it sounds like a false claim to me that the noise accompanying a digital bit stream over Ethernet would also be encoded onto a PC's hard drive, for example.
Over Ethernet we may see ones & zeros come across like this -
The ones & zeros are just transitions from one voltage level to another.
In a hard drive, a read-write head magnetizes microscopic spots on a spinning platter to represent those ones & zeros. There's no Ethernet-sourced noise also being recorded onto the platter. Just noise-free ones & zeros.
The last place that a PC will see Ethernet-sourced noise is on the I/O (Input/Output) of an Ethernet chip. The digital data stream is then rewritten without extraneous noise as it is copied into various locations (RAM, non-volatile memory) in the PC.
While that author is surely well informed about many technical issues, he also
makes claims that fall into the "audiophile vodoo" category IMO -
"
To match the aliveness and dynamic freedom of copper networking, you need the Startech 1550nm 80km modules I recommended. What you gain is more contrast, better colour saturation and darker backgrounds. The best analogy I can come up with is turning down the brightness and increasing contrast on a television set. Or have it ISF calibrated."
When someone claims that swapping in different networking modules will give your audio more "
aliveness and dynamic freedom" or "more contrast, better colour saturation and darker backgrounds", that just lowers the trustworthiness of their claims IMO.