SuperUniti review, ethernet cable and shielding

No, he is basically exposing his ignorance, and the willingness of certain members of the audiophile community to believe such things.

Of course the issue for him and others like him, is they have to appear to be ahead of the game, with a better 'higher level' understanding, in order to appear scholerly to us the ignorant.

But alas, computer audio could easily undo that, for instance claiming 1m of ethernet cable will make any difference to the high definition capabilities of the super uniti. Also you cannot 'end to end' between a nas and an su.

Audio guff aside MS has a point, I wouldn't bother with shielded ethernet cable in network audio applications. You're more likely to suffer degradation due to earth currents travelling between the audio device and the network device via the shield. Leave shielded for industrial ethernet applications. Cat5e or 6, not kinked and correctly terminated will do the job nicely.

James

I agree, he is exposing ignorance and regrettably losing credibility.

If there is any susceptibility to RF leakage from the Ethernet cable it's in the link between the switch port and the Naim equipment. The best cable you can use is high quality twisted pair unshielded (UTP) in a regular domestic home network setting. If you have a noisy switch port you can use a RFchoke to impede common mode RF interference if you need to.. The design of the Ethernet cable won't prevent that...  CAT 6a cable is really designed for higher port speed frequencies than our regular consumer LAN equipment supports. It's also more optimised for transit switch racks and conduits where many cables lie in close proximity and so cable cross talk over long length at very high frequencies can be problematic, but that is somewhat removed from our rather  trivial humble home networks... Of course the marketeers like it because they can charge a premium for something that people don't really need or understand.

Be aware, there are many contructs of Cat 6 and Cat 6a cable, and if you are to be critical you need to look at that rather than just a bland a "cat 6" which really can mean many things, for example many Cat 6 cables are not shielded and are the same as Cat 5e cables but with thicker conductors and slightly tighter coupling specification of  the twisted pairs.

I have no opinion on this matter, and no experience experimenting with different Ethernet cabling.  However, one can assume that Malcolm Steward's recommendation is based on listening, not some objective "scientific" criterion.  His website is full of cable/connector tweaks.  So this is just another example of the cable wars.  FWIW, he is not the first to comment on audible differences in Ethernet cabling.

I also comment on lmy listening expierience and ensure things are repeatable and am fortunate being a professional engineer which helps me understand many of the principles rather  than rely  on dubious advice which would probably always leave me chasing my tail. As I said the only part of the equation with Ethernet leads  that can have an audible difference at the physical level is the patch lead and it's length  between the switchport and the network player. This  will be mainly through common mode RFI or leakage from the couplings. The best way to tackle this is a few RFI clamps over the Ethernet patch lead and routing the Ethernet cable away from the mains and interconnect cables.. Now that would have been better advice but only cost a couple of quid at most so probably isn't fashionable. I am sorry any statement of saying use Cat 6 cable to improve sound quality is mis leading and  too vague to be of any use. At best it is masking or modifying an underlying issue or if there is some nascent benefit it it is too vague to be of any practical value.

Having said this, if this sort of RFI is troubling you it is a lot less than caused by a nearby flat screen TV in the room (Plasma or LCD) and light years away from any Ethernet over mains device..

Hi Guys,

Just a little note on the sheilded / unsheilded option. (I'm not going to get into Cat5e v Cat6 v Cat7 debate.)

*IF* your network infrastructure is capable of carrying a shield through from one end of the network link to the other (i.e. if the cabling that you have physically run through your property is shielded *AND* your wall port / patch panel sockets can carry that shield through) then you should use shielded patch leads as interconnects between the wall ports / patch panels and your hardware.

If your network infrastructure is like the rest of the 99.5% of installations and is implemented using unshielded cable (and standard Cat5e wall ports and patch panels) then you should use unshielded patch leads to avoid unterminated shields.

As far as claiming superior audio quality from Cat6 cables - well I haven't read the review in full but the quote attributed to Malcolm Steward made by the original poster is inherently sensible in that you should always use decent quality pre-assembled and pre-tested patch leads. No-one should really need to ever crimp an RJ45 cable!

I have seen some quite awful cables used for networking - for example, telephone cabling being used with RJ45s *GLUED* (yes, really, *GLUED*) onto the ends, cables that have had the RJ45 connectors "crimped" using needle nose pliers and hence have damaged the actual connectors, cables that have been "unwound" for several feet because the user needed a couple of spare wires to connect to their headphone out on their PC or their IR emitter blocks. Most of these cables would hold a data connection (i.e. the network devices at either end would see each other and send and receive data) but in many cases wouldn't actually hold a consistent high-bitrate connection so to that end I would agree with Malcolm ... you've paid for a premium product with what we consider to have a premium level of performance - don't throw that away by using a poor quality interconnect.

I give the same advice when it comes to WiFi and Ethernet over Powerline by the way - avoid both like the plague if you want your kit to provide anything like the performance (and by that I mean stability and consistency) it is capable of - and which you've paid for.

Phil

Hi Phil - if you're using shielded cables within the network, can you terminate the shield at the wall port (if this has an ground connection) and then use a short unshielded patch cable between the port and streamer ? How does the ethernet shield terminate within, say, an NDX if you use fully shielded cable from switch to streamer ?

Just curious as my 5 year old 30m external wall run of Cat 5e is starting to harden up a bit and i'm just considering the options when replacing it. From the advice above it looks like Cat 5e is still the best bet. 

James

Hello Phil………….. Well I must confess that the decision to include a wireless connection on the Naim range has always astonished me somewhat, should a high-end range really have this built-in?

The fact that it exists is bound to lead many to assume that using it is fine and why wouldn’t they think this? When you consider how comparatively cheaply a small bridge can be purchased, probably cheaper than Naim can include this capability, does it really make sense to include this in every streaming product you make?

If we all accept that wired is a superior connecting to wireless, and I can’t imagine many disagreeing, the purchase of a bridge to sit adjacent to the streamer for those who really must use this method would seem eminently sensible/reasonable.

OK, I should perhaps declare a vested interest in that we are unable to use wireless products and with the latest design of boards will be unable to progress beyond our NDX, which is a shame.

When we consider the total cost of the potentially redundant components in a product, power supply, DAC & wireless boards it makes the mind boggle; if only…….

Good job they sound so damned good.

Regards

Peter

Ooops. I made my own network cables by clamping RJ45s myself. I had to do this because I ran the cables in plastic conduits. The molded plugs of patch cables could not be fed through those conduits. I did buy specialist clamping tool to do the job and paid as much care as I could.

Hi Phil, thanks for your words on ethernet over Powerline, I think every device should come with a warning about RFI pollution and that it may cause side effects such as diminish the capability of any nearby high precision audio, radio and video equipment through wide band RF contamination, even for your neighbough!!

I would say however on patch leads as long as the switch port is earthed it is not necessary to use shielded twisted pair STP for the final patch lead. Of course one has to be careful of earth loops. 

Funny however I was with client two weeks ago, and discovered that far more perturbation to connected network equipment is caused by broadcast frame processing rather coupling leakage from the Ethernet patch lead (although in the Naim world this might be mitigated with the NDS). Therefore there might be a case for using a seperate VLAN for audio if your LAN has a lot of plug and play consumer apps on it that quite often heavily use broadcast frames to aid interoperability, discovery  and setup. I will be experimenting with this over the next month.

I also agree for all precision cables unless you know what you are doing it is best to purchase Pre manufactured terminations.

We've all made up our own network cables and it would be naive to say that we never do or never will but if you're running cables through conduit / walls / floors then they really should be terminated (punched down) at wall ports at both ends - then patch leads used to connect the wall ports to the network device.

The "rule of thumb" is that you should never need to crimp on an RJ45 connector.

Phil

As I said above - if the infrastructure is shielded then I'd use shielded patch leads at both ends and carry the shield on through from one end to the other. Shielded wall ports do provide shield continuity.

I'm not using anything "better" than Cat5e myself in my own setup at home...

Phil

If it had simply been my choice then I wouldn't have WiFi in any of our kit - Uniti series or otherwise - the only way to get a truly reliable network anywhere is via wired Ethernet...

Unfortunately the world wants wireless - we don't provide wireless networking on any of our servers and get so many complaints about that but for servers we feel that it is reasonable.

For client devices we include WiFi because customers demand that it is there - we do make no secret that we feel that wired is the better connection to use but at the end of the day the whole "streaming" concept is one which the public are really just getting their head round. Just as there is an education stage between people buying a £150 Micro System from a chainstore and understanding that separates are "better" and that cables make a difference then we're kind of at that stage with streaming / networking - except that now we're trying to educate people that if you are going to have a reliable network then you should do it right. Unfortunately if we excluded WiFi on the kit then we would exclude ourselves from the market and would have no opportunity to even try to educate the HiFi loving public.

I get so many emails saying "It's just not possible to run a wired network in my house" and I tend to look at it that when you make the decision of where to live you then have to accept that that environment then influences what you can and can't do easily. For example I would love to live out in the middle of nowhere however if I lived up the end of a deeply rutted farm track then I wouldn't run my TVR, grumble that I kept bottoming it out in the ruts and insist that it was fitted with mud tyres, high lift suspension, four wheel drive and a low ratio transfer box - I'd have to accept that either I buy a vehicle that's appropriate *OR* I get the ruts filled in. Yes? The "I just can't run a wired network" argument tends more often than not to be more like "Running a wired network would be a load of hassle / expense / decoration and I don't want that grief so I'm prepared to cut corners..."

Of course that presupposes that that user would know enough about networking to know of the existence of wireless bridges and about how to set one up ... whereas for most people it's all still scary enough that they'd simply pass and buy the product that will connect wirelessly.

Yes, I do understand your predicament ...

This will always be the case though - I have a 552 that I have never used the record outputs, the balance control and anything more than the CD and AV inputs on...  

Absolutely... ;D

Phil

Thanks Phil (and Simon) 

James

I’m thinking about getting a streamer and hence how best to provide a wired network in advance of a purchase. My main (only) phone point is in the living room, so that is where my router currently lives, along with the hifi system. I have a pc in the room above that I am thinking of using as a music server in the first instance.

So the extent of the network will be no more than pc-router-streamer. I figured on simply getting a couple of patch leads. A flat CAT5e (if such a thing exists) just draped around the room, hidden behind the furniture but crossing a doorway underneath the carpet. A second standard CAT5e going up, along the corner of the room, through the ceiling and up out into the above room. Anyone see any flaws in the plan?

I read the HiFi+ review and then questioned what cables I should buy, which triggered this thread.

It also seems like a good opportunity to replace my mega-cheap Huawei router, ‘free’ from TalkTalk. Anybody have an opinion on the Netgear N600 ADSL router?

Gav

A few years ago many people involved in this hobby thought all USB cables sound the same.  The reasoning was that USB cables had to adhere to computer industry standards, so how could one cable sound better than another? 

Opinions evolve over time based on knowledge and experience.   My experience and that of many reviewers of audio equipment is that “audiophile” quality USB cables sound better than “standard” USB cable.   Why is Ethernet cable any different is this regard?  Perhaps we’re still in the learning stage and skeptics have yet to see the light or in this case, hear the sound.

Gavin - i'd keep the wired 'audio' network on a separate switch (a little Netgear 5 port job will do nicely) and then have one connection from this to the router for internet access for the server and streamer (for album art look up etc when ripping etc, wireless control points, internet radio). It keeps other network traffic off your audio network and saves you changing your router etc. 

I run my NAS, Mac Mini (UPnP server) and NDX off a single switch with one connection out to the Router. The rest of the family can max out the incoming internet connection and wireless network with no effect on my streaming. 

James

Because you are deluding yourself?

I have never heard a difference with usb cables. The audio sphere is perfect in this regard, its very hard to prove otherwise, so if you want to believe a USB cable makes a difference, and someone is happy to sell you a 500 quid one, its happy days for both parties.

Reviewers cannot be trusted, in any product sphere, they have a vested interest to remaining employed.

Kendrick, I think this has been discussed on this forum before. USB cables are built to a very tolerant standard, so as to keep costs down. By the nature of them they can suffer from reflections and the transformer effect of the cables. Hence USB 'sound' different through the the varying extent they contaminate thier hosts with leakage or resultant RF currents.

Cat 5e is built to a tighter specification than USB. Again RF can be a problem, but can be mitigated by choking or taking care of the common mode interference of the switch port.  Noise in Ethernet cable is very well known about and understood in many more critical applications than our audio systems. The medical industry is a case in point, this is rather old technology now. And of course in the data network industry high speed Ethernet tends to be done by fibre now, not copper (unless very short length, or line speed is 1gbps or less) where RF contamination in the transmission line or send receive cross talk  is not an issue. 

I suspect we will move to fibre eventually in our homes, but I predict we will have  the fun of IPV6 upgrades before then.....

Right now there is more to be potentially gained by looking at the broadcast traffic levels in your audio network, and making sure broadcast traffic is at a minimum. However I know of no easy way for the layman to deal with this, other than have a dedicated audio only subnet off your router (and I mean a proper router),, but even this requires knowledge.. But those with the know how span a port off your LAN switch and look at the amount of broadcast traffic. If high..create a new VLAN with  a stub router and trunk to your switch( or use a L3 switch)  Disable spanning tree, VTP, routing protocols etc on your audio subnet and if your NAS is creating a lot of broadcast traffic, put that in a different subnet as well , (though depending on your file protocol you will need a forwarder for your upnp server). Anyway once you have cleared your audio/upnp only network and taken all that network processing away for your NDX, does it not sound better?

How long before we see audiophile Ethernet switches that minimise RFI impact  - someone is always looking for an opportunity!

Jack they exist... But someone missed a trick, they ate labelled as 'Green' switches, I guess there is more money in the environment than in audio.. But essentially these devices sense the length of the patch lead and reduce the physical layer switching voltage as appropriate, thereby reducing current used to drive the ethernet port and saving money, and for us reducing RFI.. But the marketeers haven't got that angle sorted yet as far as I an aware..

Personally I welcome such products ... as S-I-S has mentioned there are products that do "shape" their response according to loadings and if anyone wishes to market these as "audiophile" or "videophile" products then I wish them the very best...

...however, as with all things, before I spend my own hard-earned on them then I will make sure that they do provide an improvement that is appropriate to their cost within the confines of my own system - as I would hope anyone out there would do.

Phil