Cat5e or Cat6

My only issue with wireless is its general stability. To be fair i have a pretty awesome wireless/router so it does not tend to be an issue but still, wires rarely go wrong.

From Wikipedia:

Error detection[edit]

Sequence numbers allow receivers to discard duplicate packets and properly sequence reordered packets. Acknowledgments allow senders to determine when to retransmit lost packets.

To assure correctness a checksum field is included (see TCP segment structure for details on checksumming). The TCP checksum is a weak check by modern standards. Data Link Layers with high bit error rates may require additional link error correction/detection capabilities. The weak checksum is partially compensated for by the common use of a CRC or better integrity check at layer 2, below both TCP and IP, such as is used in PPP or the Ethernet frame. However, this does not mean that the 16-bit TCP checksum is redundant: remarkably, introduction of errors in packets between CRC-protected hops is common, but the end-to-end 16-bit TCP checksum catches most of these simple errors.^[15] This is the end-to-end principle at work.

I figure if the networking code doesn't have to go down the error handling path, and the data is transmitted correctly the first time around, there's less load on the network card/the surrounding software. This will, in the case of NAIM, have a non-trivial effect on sound quality.

When folks talk about 0s and 1s, well that is a mental model that works on the software layer. When you're dealing with the network, all you have are voltages, with voltages above a certain threshold interpreted as 1 and below as 0.

Meerkat - as long as the cable is stated at least 5e - it should be fine. There may always be a bad dun, but if you buy from reputable retailer you should be ok.

MM, ok - I see what you are referring to -- quite honestly you are referring to TCP which is a transport mechanism designed to cater for unreliable networks or congested networks, typically routed across a WAN or the public internet where traffic (packets) may well be being dropped, both intentionally and unintentionally  at router boundaries. This certainly may be experienced with web radio, but very unlikely on the home LAN. If you have marginal  connectivity such as with edge of tolerance Wifi or PowerLines then the link layer will discard the frame if there is an error - and the higher level TCP windowing mechanism will try and recover the data through requesting a resend.

However in the typical home LAN there is far more network traffic from layer 2 devices chattering with each other for searching, discovering and confirming/updating purposes. The larger the number of devices on a subnet the larger this chatter becomes. Now a switch will limit to some extent the chatter from the unintended ports - but even so mac address discovery and broadcast/multicast  traffic will go to all devices and need processing in the application stack - you can see this from your switch LEDs blinking where there appears no network 'traffic'. This could have an impact on the load of the peer device, such as the Naim renderer , as this traffic typically needs processing outside the TCP state machine stack by the application.

So if you have a very busy network there may be benefit on creating a subnet just for the audio with very few devices on it. However I have failed to hear any real benefit in doing this - other than with wifi, where increased chatter consumes bandwidth and reduces effective throughput.

Simon

Simon:

what did help - and was relatively easy to implement - was to put all audio gear connected to just one switch. The switch is connected to the router.

Just get 2-3 of the new Apple Airport Extremes and place one near your streamer.  I can stream throughout my house at 192/24 with on 1-2 small dropouts / month.  Then buy an expensive Audioquest Ethernet cable from AAEx to streamer.