Full disclosure - I’m co-owner and developer of the Network Acoustics Ethernet Filters.
I wanted to add to this thread the result of my experience over many years of trying to achieve the best sound from streaming based systems.
I’ve listened to the following Ethernet delivery methods, which have been mentioned in this thread, all using the same Auralic Aries Femto streamer.
1. Direct wired physical connection via RJ45 from upstream network switch
2. Inbuilt Wireless connection.
3. Ethernet over the mains power-line extenders to the Auralic RJ45 port.
4. Ethernet from a WiFi extender to the Auralic RJ45 port.
5. Media converters, RJ45 to Optical and then Optical back to RJ45 and into the Aries RJ45 port.
To my ears, the order of sound quality was 1., 2., then 3. and 4. were pretty much the same, and the 5. Option 1. is still the way to go if at all possible.
The thing that changes everything though, when using all these options other than built in wireless, is to try to remove the ‘noise’ that Ethernet is infected with.
Option 1. has less noise, whereas the extra devices used in options 3., 4. and 5. Introduce additional noise.
you can reduce noise in all options (except 2.) by having an Ethernet switch (with decent PSU) between the ‘extending device RJ45 output’ and your streamer. The noise can be further reduced by using passive Ethernet filters too.
I’ve found with the correct noise reduction measures in place, options 3. and 4. Can sound very very close to 1. and better than 2.
I’m afraid that I’ve never managed to get option 5. to sound good, but still working on that.
I hope that helps.