What LAN cable are you using in your streaming system?

So long as CAT5 or later, the choice of cable will not make any difference in a streaming set-up.

This is because LAN cables are used for TCP/IP traffic.

The protocols involved are intelligent enought to cope with packet re-sends to ensure that you get entirely consistent data at the other end.  This is different to the fire and forget nature of other digital cables used in audio (SPDIF coax/optical, USB for DACs).

Buy a cheap one in your favourite colour at the right length and never worry about it again.

Cat 5e.

Just to satisfy myself that LAN cables should not make a difference I bought a 10m Cat 5e patch lead and 10m Cat 6 patch lead to try between the switch that sits between my NAS and Mac Mini UPnp Server in the study and my NDX in the lounge, Usually the connection is by 30m or so of Cat5e that runs on the outside wall between the two locations. I couldn't tell any difference between the Cat5 and Cat 6 patch leads but they did sound different to my normal setup. Rather suprised by this I ran this over a few evenings just to make sure the results i was getting were consistent. I then borrowed a cable tester and checked out my own cable and found that when i made it up ii had transposed pins 1 & 2 (pair 3). When i put the cable in a few years ago (for a Squeezebox) i didn't check the connections, as it all worked fine with no drop outs so i'd left it as it was. Remaking the RJ45 connections and rechecking with the tester proved all was 100% and now i can't tell any difference between all 3.

James

My network arrangement is as such:

Modem-->Cat6 Cable-->router-->Cat6 cable-->NAS

router--->wireless bridge-->Cat7 cable-->streamer

Completely no lag even with 24/192 wave files (not even once).

Agreed - the protocol allows an error free link.

But you are introducing a variable though when you connect a network cable to the streamer - you've got an effective antenna, you're conducting noise into the streamer box from other device on the network. Ok the streamer Ethernet port is galvanically isolated by the small NIC transformer but i doubt this is very effective at blocking HF noise. You've then got the network card - various people can hear differences between compressed and uncompressed files on the Naim streamers. Naim put it down to extra noise generated by the processor when doing the decompression raising the noise floor on the power rails. Could the noise signature - both radiated and conducted from the network card vary as its dealing with different network conditions - more collisions, more retries etc (one for Simon here). Of course this is just speculation and would need measurements to back it up but i'd be interested in others (constructive) thoughts.

James

Hi 0rangutan,

I am the network engineer and my mindset also same as you but i tried different brand LAN cable on my streaming system "NDX/282/250/HC/NAPSC" (e.g Supra, Audioquest, AMPConnect & Belden), i can tell you "different cable , different sound".

If you free, go to try different cable but please remember use the same cable as the same time. (NAS to Router & Router to Streamer)

Cheers,

marcobb 

CAT 6a, except for the "backbone" (from one end of the house where the NAS/modem/router are to the other end where the Streamer is) that is CAT 7.

CAT5e everywhere for me...

Phil

cat5/6 the cable spec is fine ,still not convinced that cheap made cables are the same as one that the connection method is better. in that respect you could apply the same logic to audio cables

Acoustic Revive with LAN Isolator

Hi, I am also a network (and media) engineer. The Ethernet cable makes no difference with respect to TCP/IP. ( as long as its a 4 pair).  However the patch lead that connects with your network player has a couple of other considerations with respect to RFI that can couple into delicate analogue or delicate clock circuitry.

A) twisted pair leackage

B) common mode interference.

If a) is prevalent, then different patch leads here will cause varying degrees of RF interference and could 'sound' different.

With b)putting RF chockes around the patch lead near the point of entry into the network player can mitigate this (this what i do on regular Cat 5e ) by increasing the inductance and thereby reactance to high frequency current that is not part of the TCPIP link signals.

Also using a better quality edge switch (connecting to you network player)  or  a switch that is 'green' ie varies the link current to suit patch lead length can help. Remember by default the link current can drive upto 100m and many leads are a lot shorter.

But unless you have a bad RFI (perhaps very noisy mains such as from Poweline Adapters) or a very noisy switch or very poor patch lead) you are in the realms of tweaking.  It is worth routing your Ethernet cables away from interconnect, speaker, or mains leads used for your audio equipment. Using a coil anas oscilloscope you can see the radiation leakage from a network cable.

Also I have found decoupling the DAC from the network player lessens the effect of RFI from Ethernet. Ie NDX->ndac.

There is also a whole other area that can make network players sound different, and that is to do with TCP window sizes and broadcast processing.. But I'll save that for another thread...

Simon

Audioquest Vodka from UnitiServe to switch and from switch to NDX. CAT5e from switch to NAS, which is used only for backup.

Keith

Hi Simon,

To your professional opinion, will something like "Acoustic Revive with LAN Isolator" mentioned by Aleg help?

Interestingly I see that Audioquests webpage for the Vodka Ethernet cables claims...

"For audio applications and protocols, audio over Ethernet offers the virtues of high-speed, low time delay (latency), significant distance capability (1000 feet without an active booster or repeater), and extremely low-jitter, bit-perfect communication."

1,000ft (300m) in a single run of Ethernet cable ... hmmm ... I assume that's just a careless typo and it should be 100m / 300ft.

Phil

audioquest plugs seem  very well terminated   assume cale spec is fine too

Rich,

The plugs look very funky, which would be of little consequence I they didn't sound amazing, but they do. I was happy to believe that network cables and components all sound the same, but the demo at the dealer was astonishing, so it turns out that they don't. The improvement is similar to adding or upgrading a power supply. I realise this is impossible and many people will not need to have heard them to know that this is impossible, but there you go :-/

 Phil,

I ordered the 1,000 foot cable as it seemed good value for money, but in the end I settled for a couple of five metre runs :-)

Keith

My sarcasm detector isn't working very well today - is this a serious post or a wind up?

I made up the bit about ordering the 1,000 foot cable. The rest is true.

Keith

@sckwan ... I have no idea.. Never heard of it before........... 

I checkled at the 1000 foot Ethernet cable posting, be interesting to see thelayer 2 error rate at 100Mbps and 1000Mbps link speed, then we can see about so called bit perfect...... And what on earth has an Ethernet patch lead got to do with audio jitter... (i assume carried by ethernet) I can't believe people write or market this stuff with a straight face.

Simon

The common wisdom is that LAN cables shouldn't sound different but reviews in 6Moons (Acoustic Revive LAN) and HiFi Critic (Malcolm Steward) suggest otherwise.  I have a open mind on the matter, though I was somewhat persuaded by Simon-in-Suffolk's explanation why USB cable can sound different/better in audio applications while ethernet/LAN cable should not.  (Thank you, sir.)

Perhaps Aleg would comment on improvements or changes in his system after using the special Acoustic Revive LAN cable.  

BTW, anyone interested in this topic can watch for an upcoming review of Audioquest Ethernet cables at Audiostream.com.  

Hi, IME Ethernet patch leafs can cause different amounts of EMI and therefore can 'sound' different by thier interaction with delicate digital clocks and analogue circuitry through coupling. However that is quite different from say Ethernet cables them selves are jitter free and audio bit perfect, which IMO is utter poppycock. 

I all honesty I cannot yet comment on changes from the LAN cable by itself, because it is part of a chain of cables (cable length is only 1 mtr and my switch is further away) but I bought it together with the isolator and I am waiting for the moment where I can change the complete setup for connecting my audio PC to my main switch. I will go optical for that and I plan to use the Acoustic Revive LAN-cable for the last section to the audio PC.

Wojciech Pacuła has written his impressions on the use of the Acoutic Revive LAN-cable here: http://www.highfidelity.pl/@main-272&lang=en

Regarding the LAN-isolator however, I can say its effects were immediately noticable. The sound has improved very noticably, much more quiet, easier, better defined.

I had already read good reviews about it on a Japanese forum and it certainly works for me. It is a combination of a transformer isolator and a coil choke for common mode noise. It creates galvanic isolation preventing / lowering electrical noise entering the player over the LAN cable. Upstream switches, routers and NASes are electrically dirty as well and LAN cables can pick up RFI, polluting the attached network player. As I use NAS as my main (i.e. only) storage medium, I thought to give this device a try. And with success I can say.

You can read something about the measurements of the physical effects of the LAN-isolator here: http://www.acoustic-revive.com...io/lan_isolator.html

They are not cheap, but I found a way to buy them directly from Japan, so was able to lower the costs somewhat.

-

aleg

Aleg, thanks for your description of the isolator. Given what you say it does I can clearly see why it would be beneficial to use as it addresses many of the RFI issues i have referred to on this forum, and they show some example of RFI noise in your link. In my view this is a more fruitful area to look at than the cables themselves, where I do believe the snake oil merchants start peddling.

Did you try RFI chokes around the Ethernet cable before you tried the isolator? It has a similar effect on reducing common mode noise.

Simon

Simon

I think I agree with you.

IMHO the positive effects of all these cable variations etc have to do with electrical interferences and not with bit-perfectness (they will all be bitperfect). With regards to jitter I am confused on that, all and everything is nowadays hung-up on their jitter effect and I no longer can get a clear feeling on what is plausible and what isn't.

I tend to read what I believe are credible reviewers and where affordable to me I will give certain things a try and see for myself. It doesn't have to be scientifically explained beforehand though, I notice daily positive effects in audio playback that cannot be explained (yet). And IMHO science doesn't know much about the 'workings of this world' at all, though they believe they do.

Ooooh, did I say that   

Hi Simon

Yes I did try the ordinary RFI chokes, but that didn't have the same effect for me.