Originally Posted by Ian
Originally Posted by DaveyC
It's been a while since I posted here, but seeing the brave comments from Mwiebelhaus and Crion, based on their experience, I thought I'd share my own experience with Ethernet cables.
I think the use of different cables affecting audio presentation for SpeakerLink audio connections is largely accepted, it is the use of exotic cables over appropriately specified network cabling for network use that is debatable.

In the case of Crion, his network test was vs. a UTP cable, hardly a fair comparison as in some circumstances, omitting any form of shielding as per a UTP cable may not be appropriate for the job as it allows EMI ingress into the chain, again, an accepted issue. I don't know if he ever repeated using a shielded or screened cable.

Iirc that Mwiebelhaus comment was regarding SpeakerLink.

Hi Ian,

Agreed, intuition suggests reduced benefits or no benefits running audiophile Ethernet cables on a network link running TCPIP.

In spite of my own intuition, I've had surprising experiences in playing around with the link from MD600 to 808. The first was last year, was when I put a GISO in the last 30cm of the network cable run before the 808 (gains similar in nature to those from the AQ Vodkas). Just a couple of weeks ago, I borrowed a second GISO from a friend and put it in the Ethernet chain immediately up-stream of the MD600 (i.e. between switch & MD600). I had a hunch that there might be some gain, but the reality was a much bigger gain than putting in the first GISO. This was obviously an easy and completely cost-free test, but the scale of improvement really threw me.

How this might be true, against well-founded expectation with TCPIP in play, ties back to a thread we had going a year or so back. It talked about RF carried on the four twisted-pair signal conductors and on inner and outer screens in an Ethernet cable. Given what TCPIP does (explained in detail by Ronnie) I can only assume that the gains in an Ethernet network link are down to removing substantial amounts of RF noise that has made its way into the network at various points. While TCPIP is impregnable in preserving the integrity of the packet data stream, it doesn't operate at all in the analogue signal domain, and would therefore allow RF noise (on both signal and screen conductors) to propagate into audio devices and onwards down the food chain, at undiminished amplitude.

I'm no RF expert, audio engineer or high-frequency cable designer, but my gut feel is that RF noise inside audio devices does much more damage to music than we might expect, to the extent that merely achieving a CAT 6 or CAT 7 standard on a cable falls some way short of the optimum when it comes to limiting RF ingress along the full length of the chain. This would explain why two GISOs can have such a big impact. While this could explain what I heard, I'm still very surprised by the extent of the gain.

BTW the impact of the AQ Vodkas was very noticeable before adding the second GISO, but became much more dramatic after adding it. I ran another test with the second GISO in place, moving from Vodkas (between 808v5 & DSPs) back to CAT 7 and back to Vodka. My friend who'd brought over the second GISO was with me at the time and it was one of those instant-grin experiences when we put the Vodkas back in. So it looks like the gains from the Vodkas where much more apparent with the further reduction in noise and general grunge that the second GISO brought about.

The impact of the second GISO really gets me thinking about the extent to which RF impacts the music. I've spent several years putting a lot of thought and effort into getting rid of RF in the mains and other signal cables, all to good effect, but this experience suggests I may be less far down the road than I thought :-(

808v5, MD600 & DSP7200.1
Dedicated audio mains supply