Sarum Array Streaming cable

if you find the noise RFI change so sensational - what happens when you disconnect whilst music plays out (as you know the new streamers have large media buffers and only use the ethernet for transfer for a relatively short amount of time) - does it become even more sensational or does the sound take a step backwards?


It’s a bit fiddly connecting and disconnecting around the back of the KDSM, particularly given the substantial connector itself, but I’ll have a go.

Although, once the musical payload has been delivered, I don’t see what that will prove. The buffer will deliver what it has received and then…

1 Like

exactly, it will prove nothing.


Unfortunately the KDSM doesn’t buffer like the new Naim network players but for those that have them, it’s a valid test of how much the connected network is influencing the final SQ. I can see where Simon is coming from here.

It actually would demonstrate a huge amount, and I am sure the Naim designers and engineers would be interested… how does the sound change when you disconnect the cable… does it get better, worse or stay the same… Naim went to a lot of trouble with their new streamers for the Ethernet to have as little affect as possible and focus on media transfer as rapidly as possible as well as use LVDS interfacing … they could hear no sonic shift or at least I was not advised of any in development, and I have no reason to doubt… Trevor appeared pleased along with the Naim development team on how it turned out.
I see you use non Naim streamer, and so that might behave differently and not include the specific optimisations introduced by Naim. This was one of the major engineering innovations with the new streamers…
Also remember Naim are masters in optimum noise shaping to affect sonic performance. They do this with their streamer firmware, by changing the noise profile from the assembly code execution, they can almost create a tone control… this is what was is meant with firmware tuning… although sometimes it doesn’t go quite to plan…
Now this knowledge in part was also used to minimise impact from the new displays and Ethernet connectivity.

but when it has buffered, it was using the ethernet cable. The resulting buffer is due to the streamer and the cable.


Yep - and those bits are safely in the buffer. Network protocols have taken care of that.

Chord Music ethernet

:small_blue_diamond:When a Swedish Hifi-magazine tested ethernet-cables this winter,.so was Chord Music also with in this test.

They agree with what you say @Dave…“and the Music shaping is in a different league”.

And also what @ChrisBell said…

•Whether you unplug the ehternet cable or not,.it has already done its job.
To convey the “signal” as good as possible to this media buffer.

Everything else is just technical reasoning,.that’s how the cable works in practice which is what matters.
You can always discuss if the price-level of £4200:- (in Sweden),.for 1 meter ethernet-cable makes sense.
But that decision is to be taken by yourself,.because at this level,it is the performance you pay for.


1 Like

No you miss the point in my opinion… this is kind of irrelevant, the Ethernet segment and end to end transport flow will reliably convey the data with just about any cable from Cat5 upwards… out of many billions of frames I have had not one error with regular Cat5e.

The effect comes from the serial carrier clock (approx 31MHz for 100 Mbps) travelling up and being received on the cable… for occasional bursts and sustained bursts of frames… this becomes an RF load and can modulate the streamer ground noise floor… each cable will have a slightly different effect.
I have previously posted the TI paper that describes the engineer principles and considerations with this and how to reduce the effect of Ethernet cables modulating the ground plane or coupling into other electronics.

1 Like

Yes, I tried with the ‘it shouldn’t make a difference, it’s data packet processing in the NIC, physical layer, Data layer, OSI stack etc’ but a number of years ago I sat through a Chord demo going from std Cat5/6 cable, through C-Stream to the Tuned one’s on the same equipment with the same material (at the same volume level, nothing else changed) and the difference was real, “but it shouldn’t”.

And now going from Chord Anthem to Chord Sarum Tuned Array on the NDS Ethernet input and reusing the Anthem on the UltraRendu running the SonoreUPnP Bridge, introduces a real difference to the playback stream.
And this isn’t the current Sarum T or even the Sarum Super Array version.
“But it shouldn’t, but it does”



Well refer to the TI paper I have posted many times before, it’s nothing to do with the network data, Ethernet etc, it’s RF loading from the analogue serial clock pulses into a less than perfect transmission line,… and yes it would be surprising if such noise shaping RF loads didn’t modulate the noise floor to some extent… ie it SHOULD have an effect… albeit the effect is very much reduced with the new Naim streamers by very deliberate design.

I think some people get confused because as they hear the effects from the loading of a network lead, the effects are somehow related to the digital media data being transported by the network. But this is almost certainly not the case at all. This is all about good analogue electronics… and is very real, and Naim take advantage of it.

1 Like

I think you are on a hiding to nothing, Simon.

FWIW I sat through a demo of various Chord cables at a HiFi show with a top Linn setup. Various Chord “Ethernet” cables were swapped in and out with an ever increasing cost, my missus then said “what have they changed cos it just sounds the same?” She had no idea that network cables were being changed.
She was right of course it, the system sounded great but the cables made no difference.

1 Like

That was my experience in one of those Chord demos, too. A room full of slightly bemused faces.


Then you’re fortunate, you’ve saved a small fortune.

Pity those poor wretches who do hear a huge difference.

Does it buffer the noise?

Can you do a reliable software update across the network when using Chord cables, I found this failed with AQ, but works all the time with Blue Jeans Cables (BJC)

I really could not hear any difference between Chord Streaming Cables and BJC when I had a listen. Doesn’t mean there isn’t, but simply that BJC works best for me in my mainly Linn system. If I was pushed to chose then the BJC are slighty better.

I attended a demo by Chord Company and was totally bemused. When it came to analogue, I preferred their cheaper cables, which I thought were fine. I’m happier though with Linn cables in my Linn system and Canare with Chord. When I used Naim kit, I liked the cables that came in the box best.

Anyone tried swapping cables & in the same session swapped steamers to compare.
I’m wondering if the change in sound between cables could be different say between Naim & Linn & if so does that not tend to agree the sound shaping theory has merit? .

Mike I would expect the noise shaping will vary to some extent to the platform the cable is connected to… as I suspect a Naim product and Linn product would look quite different from a RF source perspective …
However I am still waiting for someone to say who finds a notable difference with a cable how the sound changes when the cable is disconnected, especially on a new Naim streamer. I find the silence curious.
I can’t hear a difference on my NDX2 myself, but I use an external DAC so possibly not a good reference.


happy those who don’t hear differences or you prefer to hide with their scientific beliefs. They save a lot of money indeed.

I can’t hear a difference on Linn DSM until music stops, of course. This is a BJC Cat 5e between DSM and HP Fast Ethernet switch. FWIW changing the switch made no difference either when I was deciding which to use.

I’m not going to try pulling out the Exakt cable (BJC Cat 5e) between LP12 Urika II and DSM.

Still my ears may be a fault as I have different coloured BJC Cat 5e so I know what’s connecting what. The extraordinary things is I cannot hear any difference between the colours.

I’ve posted several times about the controlled test where listeners could not tell AQ Vodka cables from $2 Ethernet cable that failed a Cat 5e Fluke test.

I’m suspicious of HiFi magazine reviews with no measurements, as they always recommend the expensive cable. If they didn’t then would the manufacturer send further cables to review. I doubt it.

I tested a number of cables in my system, both general Ethernet patch leads and some Audiophile streaming cables: the cable that gave the best result (in my system in my relatively noisy electrical environment) was a perfectly conventional Cat6 UTP Ethernet patch lead. All the STP cables did less well than most of the UTP cables!