Chord Ethernet cables

Then I think you potentially have a great deal to gain. You can never bet the farm on it, but for forty quid, most of which you could claw back...

A direct link to your ND5, even with cheap Cat 5 cable, will sound superior to wi-fi by a great margin.  

I got similar "breath & flow" experience when comparing Vodka to one cheaper but good CAT6. Also this CAT6 was clearly brighter. First I preferred CAT6, but later on when listening/swapping few times I found Vodka more natural, providing better dynamics & rythm, wider & punchier bass. Then I started to compare vocals: first the same thing, CAT6 sounded better/brighter but when comparing more tracks it was clear that Vodka is separating vocals better. When connecting later on CAT6 back sound is more flat. One example is heavy music which sounds much better with Vodka. So, this is very much about personal taste & also sepends from speakers a bit. I have very revieling Marten Getz 2, with some "darker" sound speakers Vodka might not be the best? Looking forward to see your longer time test results.

Just ordered a couple of C-Stream cables from the UK to compare to my current Audioquest Cinnamons. I do need one 3M length but all the others can be .75m. So I can send one Cinnamon farther back in the stream (main switch to modem), keep one Cinnamon (main switch to audio switch) and Chords from Vortexbox and UQ (to audio switch). That is if I prefer the Chords. Will let you know. 

Indigo from the switch to the streamer would be nice, but might be total overkill for a UQ from a price perspective. I think I'd rather spring for a powerline before that. 

I got C-Stream & compared to AQ Vodka - I will do more testing but so far AQ is still better: more detailed, more open. Differencies are not that big e.g. when comparing to cheap CAT5e/6 so C-Stream is also a good choise. 

Charlesphoto, what were your findings?

I never did get a C-stream as they can't be shipped to the US from the UK I found out, and here they are prohibitively expensive (almost twice the price), so just sticking with the cinnamons (which I got on sale). The V1 and a linear power supply for the switch (and modem) made a much bigger difference (esp the V1!). 

Ok, thanks, was just curious!

Hi interesting program on tv watchdog on cables it seems the more we pay the better the cable,I think some times the old idea having a bass and treble ajustment  on the amp works a treat are we not starting to take things a bit to far ,Try the old test close your eyes and listen to different ones Are we all spending our money rashly .regards

Yes and they come out with the usual catch all - its digital so it either works or doesn't and leave any other variables out. 

Interesting program on watchdog tv about cables it seems the more we pay ,We think the better the product we should not think with our pockets regards.

For those considering the Chord TA cables (Sarum and Indigo), although they should work fine; it's worth noting the information in this thread:

https://forums.naimaudio.com/to...und-different?page=1

starting with Simon's post 23 June 2014 22:57 and the following two posts.

I have experimented with the insertion of 2 Chord C-Stream cables into my network.

The heart of the network is a Cisco 3560 Series switch. Data comes from a Mac mini (early 2009) running Asset UPnP, with external firewire drives. I first had unshielded cables (flat Cat. 6 UTP) between the router and the switch, and between the switch and the streamers (SBT and NDS), that I replaced with shielded cables (flat Cat. 6 STP). Although this made an important sonic difference, the effect of the C-Stream cables remained the same. The cables were burnt in for 2 weeks 24/7 before assessment.

The first cable (0.75 m) replaced a Cat. 6 S/FTP between the Mac mini and the switch; the second cable (1.5 m) was added to the flat cable going from the switch to the NDS, using a female-female ethernet coupler, on the streamer end. The effect is somewhat additive, but the first one inserted has a greater impact, even if it's the one between the Mac mini and the switch (can't figure out why).

To my ears, in my system and network and environment, the Chord C-Stream adds an unsubtle veil over everything that conceals the details, flattens the microdynamics, kills the slam, tarnish the sparkle on the voices and the richness of the timbres (sorry for the strong words, I'm not good at that). But very bizarrely, it gives a sound that's soft and groovy, even beguiling somehow, very easy and unfatiguing, and still manages to be "musical". So I can understand why one might like it.

So, in opposition to what as been stated before about this cable, unfortunately for me the resulting sound is not only undesirable, it is also very wrong.

In my experience with Ethernet and audio, there has been two factors interacting:

1. the major source (NAS or switch)
2. the Ethernet cable itself

One major influence has been the source. I hear differences in the model of switch, powering  (all now on linear supplies) and grounding. Changing the switch closest to my audio, and in particular grounding it, brought the most significant musical gains. Grounding the switch removed the final sharp edge and smear, really opened up low level detail, and brought body/richness. I could meter 13mV AC between my switch and US grounds, which is a lot of noise; unfortunately I don't have a 'scope or spectrum analyser at home!

With cables I found successive and obvious gains moving from Cat5e, to Cat6, to c-stream on my local source/sink leads - reducing hash/and letting more of the music come through. I haven't yet been able to try anything more esoteric but I'm tempted.

I'very found combinations matter. The first time I introduced one local switch with the same CAT6 patch it sounded flat and grey. Reverted to a longer CAT6 and existing switch. Retried existing switch as local with c-stream vs CAT6 and c-stream obviously cleaner, but exposing the switch's noticeable brashness - a mixed bag but a step in the right direction. Swapped switches again and brashness largely goes. Grounded the switch and completely floored with elimination of grain and fundamental improvement in musicality - ah finally! All the time referencing local USB as a reference.

I can understand different people having different take-outs from similar tests, because in my experience effects are so context-dependent and interacting with the end-end configuration.  

In the old days Naim provided an end-end system that when installed correctly would assure a level of performance. But IMO digital mix'n match with commodity components is still a crap-shoot, has a way to go to mature just like the CD refinement life cycle which took a decade to reach CD555 peaks.

Interesting observations 40 below:

I have also been looking at discussions on switches & cables on the Linn forum & in particular Lejonklou the Swedish based linn-fan-forum.  They seem to have a preference for NetGear switches & not too fussed about cables other than Micro Connect Cable S/STP CAT6. 

They too have concluded the mix'n match crap-shoot does predominate,  “best switch” does always work out best for different systems ( I tend to believe it’s a bit subjective & maybe peeps trying to make a point or just be different on the unanimity of a forum)  but whatever, no matter.  

One question I have on your observations is grounding the switch:

Did the various switches all have metal cased (cross linked) RJ/E ports & was your CAT6 UTP (unscreened) or S/STP ?  (Chord C-Stream is CAT7 & therefore screened)

Without screened cables & cross linked RJ/E ports I would expect that grounding the switch not to make a whole lot of difference.

How did you ground it; off the power supply E or to a ground (planet earth) ?

I have my NetGear GS105v4 grounded & with CAT7 all over the system & their screens cross connected thru the switch,  I did hear a slight difference, but nothing like so pronounced as your description.

Finally,  why not name the switches & cables ??

Hi Mike-B

You asked some good questions! I was posting my experience about interactions in between getting ready for work, and a thought re 'info overload'.

Like you I've trawled most of the streaming and computer audiophile forums over last 6 months trying to absorb best practice, along with general google searches on key topics which led to some more engineering/first principle references. I.e. Linn forum, lejonklou etc. I've found it a bit frustrating as I paid a premium for my US-SSD and set up NAS/network according to most 'best practice' but it wasn't performing as it should.

My CAT5e and CAT6 were both Belkin UTP 'verified' cables.  The earlier Martin Colloms tests recommended UTP, and also noted the effects of switches and linear PSUs.

I started with a metal Netgear GS108v3 close to my NAS on a TP dual 12 PSU, along with my DSL/wifi router, and local CAT6.  Should have been good.

•  The first step was to well-separate the DSL/wifi router which was a big gain but still degrading when connected.  I then put my (now spare) Naim US PSU on this.
• the second was to ground the GS108v3, via alligator clip on chassis to a mains plug on the distribution stack.  At the time I found the ground hierarchy quite sensitive, it needed the best ground available, I've subsequently discovered the NAS has quite a significant earth potential/current (As one might expect a wee computer to have).
• I then bought a Netgear GS605v4 on premise of smaller # ports = less noise(?), more recent revision, and convenience! I would have preferred a GS105 but courier delivery is very inconvenient. This provided the initial 'grey and flat' impression when used as a local switch with Belkin CAT6 on its own TP PSU.
• I then tried a local LaCie USB connected drive on TP PSU to US but this was micro phonic, 'sounded sick'.
• I then retried the local streamer (US) switch with c-stream, this is when I decided c-stream was a positive step in neutrality but the underlying brashness of GS108v3 vs GS605v4 was revealed. Convinced me c-stream was a big step in right direction from CAT6 UTP.
• Just after I put another TP US PSU on my remote Synology DS213j NAS which was a quantum step-jump, along with local c-stream and a second GS605. I' metered the Synology PSU to determine it was grounded so shifted the switch ground back to the Synology (source), the c-stream now tied the switch to NAS ground (on it's own TP PSU).
• final step was grounding the local 'streamer' switch via chassis alligator clip to audio star ground. Something I'd been intending to try for a while but didn't have parts on hand to do.  this fundamentally changed the presentation, brought the full body and emotion and flow, my wife calls it 'rich'.  I have heard a lot of thin/tizzy sounding digital systems.....
• all the ways thru I used USB stick into US as a reference, I had also stripped down my network to 'unplugged and off' and added back successive links and livened devices to map out the successive degradations against chamber music (very revealing)

So I now have two 'subsystems',

1. the NAS+c-stream+switch on TP PSUs, then CAT6 transport to next room. Waiting for long c-stream, will then have to look at the extended grounding chain, may need a short 'breaker'?
2. then the local switch+c-stream+US-SSD+DC1+DAC/XPSDR all star-earthed to same ground plane, digital devices on TP PSUs

Now the 'debugging' phase is over, I'll tidy up with proper earth lugs permanently attached to the key devices in the chain, then 2.5mm conductor to dedicated plug earth. I could have modified the TP PSUs to ground their -ve rail output, but felt grounding directly from the 'signal source' would be better (from turntable days).

Network replay has gone from a very poor second to local USB, to well ahead.  It's finally got all the organic musical feeling I can last remember from my CDS2 (or distant analogue) with the added information, organisation and neutrality virtues of new-generation digital or last generation CDPs.  No obvious digital smear, hifi artifice or grain/edge is now intruding on the music. 

Yes, with CAT6 grounding the switch chassis with made a significant difference even on TP linear PSUs rather than wall-warts.  I attribute this to reducing common- mode noise arising from a floating switch, into the CAT6.

I was really surprised to find 13mV AC (and 3mV DC) between my local floating switch and nearby US. The US has a plastic 1000baseT so remains isolated from the c-stream.

I'm sure there may be uplifts from other products than c-stream, but retailer said people were experiencing good results, and it's a relatively inexpensive baseline.

BTW, mapping the successive degradation from a reference 'local USB' on the US, from no network (best) then adding successive network components, was very informative.  Each additional element added/powered up on the network took its toll even when playing local media.

Thank you 40 below. Wow, that's serious testing you did there.

I agree with the conclusion of your post from yesterday. There are so many variables with streaming that any specific environment is different. Naim did their best to provide a sort of complete solution with their servers and streamers, but there's obviously a limit to the extent of what they can control.

I too am fortunate to have reached a state where network replay is better than the local USB input on my NDS. The aim was to have both equal, so I'm delighted with this result !

I'm still at a loss to understand why my findings with the C-Stream were so different than anyone else's. My dealer was here when with first tried. Listening to a track with not more than a tiny bit of guitar during the intro, I immediately said: "But where's the attack gone ?" He replied: "Indeed, it's softer." I first thought a bit of running in would cure this, but it didn't. I guess, as -- I think -- Frank Abela once wrote: someone's open is someone else's bright.

An area of interest to me is the grounding of the cables. Did you experiment with this ? A friend of mine who works with Cisco products told me my switch would ground the shield of the cables, but the switch manual recommends to ground them, so I don't know what to think. At the moment I don't have any patch panel, where the earthing of the cable normally takes place in professional settings, and I don't know if it would be worth it.

Interesting. In my case, with a streamer, it's different: the USB reference is set with everything else turned off (UPnP, iRadio, you name it) and with the Ethernet cable unplugged. So it should remain stable, mains-borne noise notwithstanding.

Oh, and I don't use the system automation facility. The effect is audible and detrimental as DB found, but I didn't try a shielded cable.

OK, sorry, this is going to put a cat among the pigeons...

I (and others on here) believe that (in absence of network errors) the main influence that networks have on audio systems is through interference, particularly RFI, and we have good engineering reasons for this belief.

Now the effects of RFI are many and varied, it can cause the sound to be dull or artificially bright, it can mask details, it can increase distortions and it can cause intermodulation.  Changing the RFI entering the system (by adding ferrites or changing switches, shielding and grounding etc.) can change the effect of the RFI in many ways. 

This is why different cables and network components have different effects in different environments, and any one person's experience may not be the same as any other particular person's experience of that product.

And here's the controversial bit...

In a system that is being made more bright (aka open) by RFI, altering the RFI can cause that system to sound dull.  Paradoxically, that can then allow subtle nuances to come through that were overwhelmed by the previously over-enhanced transient behaviour.

On the other hand a system that is smoother than the listener likes may benefit from the introduction of a little RFI making it brighter and giving the perception of more detail being heard.

Just as with speakers there's the issue of personal taste.

Someone may like their system to sound brighter than it does, so introduction of the right spectrum of RFI to cause a brightening effect will make it sound better, provided it doesn't also mask too much detail.

And please don't shoot the messenger!

I have just measured exactly the same numbers between my switch to power E. But I noticed if I hold the metal meter probes my body shorts it to practically nothing so I concluded that this was more or less unconnected static. 

My x3 ethernet cables to the switch are Cat7, therefore the shields are all linked thru the switch.

Synology DS214 is a plastic 1000baseT & isolated,  ditto the NDX 100baseT is also isolated, ditto the wireless hub - i.e. the switch has x3 common ground connected arms isolated from everything. 

Which makes me again wonder why you are getting such a change in SQ by grounding the switch who power runs off only L&N & that has unshielded cables.

And finally if you do go all C-Stream (SSTP), does the plastic body GS605 have E linked ports ??

In my case a good switch was certainely not detrimental, but I didn't try "crap-shoot" stuff, so cannot compare.

In fact, it was Simon who pointed to a professional-grade switch (I think my 3560 is a compact form of his 2960). I chose mine over a Netgear switch because I wanted PoE ports and the Cisco wasn't that much more expensive than the more or less equivalent Netgear one. As I bypass the router completely (I use it simply as a bridge), I needed an Wi-Fi access point and wanted to power it through PoE so I could put it where a socket wasn't readily available (and it's one less SMPS). And with this in mind, the combined Cisco solution (switch + AP) was cheaper than the Netgear one. Plus, as I use a small firewall between the router and the switch that came with one more SMPS, I thought I could try using a PoE-splitter to power it instead. Not bothered yet, but I wonder wether the additional load on the switch plus the noise induced by the splitter would be better or worse, sonically speaking, than the SMPS it replaced.

Having a beefy internal PSU that's designed to power 8 PoE-capable ports in addition to 4 additional non-PoE ports, my switch is clearly put to a light service now. But when I read about adding TP linear PSUs to less expensive switches, I'm not certain the decision to shoot for a professional-grade switch in the first place was that ludicrous, even economically speaking. Again, I didn't compare in my own environment. But maybe I'm just trying to make a point.

Hi Huge

Sounds perfectly sensible to me. We'll not shoot you. This time... :-)