On my tinkering list is connecting an endpoint to a specific Ethernet port. I have a 4 port Ethernet add-in card and I want to associate a specific port with a specific endpoint. I suspect switches add noise. What are the chances of this working?
The full plan is port->short audiophile Ethernet cable->Ethernet-fiber converter->long fiber run->fiber-Ethernet converter->short audiophile Ethernet cable->endpoint.
Yeah, I know this is overkill, but the OCD monster that shares my thoughts demands it.
The way ethernet works makes it impossible that a switch add noise to the tranfered data.
Otherwise it would not be possible to listen music from Tidal. The noise added by the thousands of switches the signal passes on it’s way around the world would make the music unhearable.
So don’t worry to use one or two switches.
Audiophile ethernet cables are pure snakeoil.
So port->endpoint is sufficient. Every additional component brings only the potential of problems and failure and no audible benefit.
Thanks for your response. These are partly my opinions also. Probably won’t tinker with this unless I find a bandwidth issue, not specifically with Roon and multiple endpoints, but with Roon and other inter/intranet traffic.
Really, with the possible exception of AC power cables, isn’t all audiophile grade cabling snake oil?
Yes, this is true for analog cables, science tells you that with the frequencies we use with audio each €10 cables is more than enough. And that is even more true for digital cables. It is not easy to add noise to 0s and 1s
And in my opinion audiophile power cables gives you only a benefit if you have a really bad build device. Wouldn’t Yamaha, Denon or another big manufactor equip their amps or streamers with these power chords if they have such an impact on the sound quality, only to get better reviews than their competitors?
In general I am skeptical of high-priced cables and suchlike. But you’re overlooking something: Although TCP/IP as a protocol does guarantee errorless deliver (or no delivery at all), that delivery may require retries/retransmissions (just attach a network sniffer to a poor connection). So retries may affect timing in a significant way, and in audio, timing is pretty important.
More expensive cables may have better shielding, and better shielding should decrease the likelihood of retries (all other things being equal; YMMV; etc ad nauseam). If a person had a long run of cable, a more expensive cable might be justified.
You would need to use the route command if using Windows to get this to work with a multi-port adapter since your scenario would mean the port to end point link represents it’s own collision domain.
You would also need a routing entry on the next upstream router, most likely your cable/dsl modem. That’s if it supports such a function.
