HIFICRITIC audio review magazine
HIFICRITIC FORUMS
Current Issue
HIFICRITIC
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

3 Pages123>
Share
Options
Go to last post Go to first unread
Offline Martin Colloms  
#1 Posted : 06 March 2019 13:47:02(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,042

Was thanked: 48 time(s) in 47 post(s)
CAT Network cables: suggestions for a group test on sound quality using the ND555

May I have your suggestions for types, products and and lengths ?

( also network switches )

Martin Colloms
Offline phil page  
#2 Posted : 07 March 2019 02:31:15(UTC)
phil page


Rank: HIFI Guru

Joined: 18/10/2011(UTC)
Posts: 401
United Kingdom
Location: Norfolk

Was thanked: 1 time(s) in 1 post(s)
Well, I’m using Cisco 2960 G, which are all-gigabit switches
Phil
Offline frank23  
#3 Posted : 07 March 2019 21:00:57(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
I think audiophile CAT network cables are snake oil and a total waste of money.

What counts is the power supply of the switch (switch mode or traditional), the grounding arrangements of the device and the cables, on both ends. And even when connected through optical, network equipment elsewhere in the home can have an effect through the power lines.

Maybe if your are lucky a certain CAT cable will lessen the impact of badly designed devices and power supplies, but this "certain CAT cable" may as well be a cheap one as well as an expensive one.

In SPDIF this is different, as there is a timing-master-slave relation between sender and receiver. But in ethernet connected devices, this timing-master-slave relation does not exist. The only differences come from differences in design. And this may be as well a bad sender as well as a bad receiver.

I know, I sound like someone from the beginning of digital, saying that digital was digital, but that was before the impact of jitter was understood. In ethernet networking gazillions of bytes are transported without any fault day in day out across the world. This works perfectly. The only reason this would not work flawlessly in audio is noise and grounding and bad design that makes it through to analogue sections. Not because data does not make it from A to B.

Edited by user 07 March 2019 21:02:01(UTC)  | Reason: Not specified

Offline Pete_w  
#4 Posted : 08 March 2019 09:17:12(UTC)
Pete_w


Rank: HIFI God

Joined: 28/07/2010(UTC)
Posts: 1,037
Location: Cambs, UK

Thanks: 12 times
Was thanked: 11 time(s) in 11 post(s)
Originally Posted by: frank23 Go to Quoted Post
I think audiophile CAT network cables are snake oil and a total waste of money.

...etc...



Yes, 100% agree. That was my point precisely in the Naim '555 thread. If the digits arrive intact and are decodeable, then the digital transmission medium has done its job. If the subsequent analogue circuitry is letting you hear upstream changes in the digital domain, then that's a flaw in the implementation of the digital receiver. It could be letting noise through, or it could be self-generating noise in response to incoming packet timing. Either way, it's absolutely nothing to do with anything upstream from the RJ-45 socket on the back....

That doesn't mean that cables *won't* make a difference in systems which are poorly designed, they may well do some noise shaping, but since "poor design" comes in many flavours, there'll be no consistent results whatsoever.

Offline frank23  
#5 Posted : 08 March 2019 10:44:22(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
Originally Posted by: Pete_w Go to Quoted Post
That doesn't mean that cables *won't* make a difference in systems which are poorly designed, they may well do some noise shaping, but since "poor design" comes in many flavours, there'll be no consistent results whatsoever.


Yes, 100% agree.

BigGrin
Offline en1omb  
#6 Posted : 09 March 2019 11:55:41(UTC)
en1omb


Rank: HIFI Guru

Joined: 27/02/2009(UTC)
Posts: 299
Location: Bristol

Thanks: 6 times
Was thanked: 3 time(s) in 3 post(s)
Chord do a comprehensive range, so it would be good to get a few from them. When I heard a demo comparing them, the top of the range Music one did something quite remarkable. It sounded like a totally different piece of music. The C-Stream is doing good work in my system, albeit with optical isolation in-between Network Bridge and router. AudioQuest do a wide range as well. The Wireworld cable got a positive review on Audiostream and isn't excessively expensive.

SOtM have just come out with a network switch, they also have a network cable.

Edited by user 09 March 2019 11:57:15(UTC)  | Reason: Not specified

Offline Christopher Bell  
#7 Posted : 10 March 2019 04:33:26(UTC)
Christopher Bell


Rank: HIFI Newbie

Joined: 10/11/2018(UTC)
Posts: 6
Location: USA

I use a Chord Music ethernet cable on my ND555 with great results. It leveled the field between CD rips and streaming services like Tidal and Qobuz. It (along with the rest of the Music line) has transformed my system to a level I never thought possible. Difficult to understand how 1m of ethernet cable could make such a dramatic difference... I hope you get a chance to review it.

Edited by user 10 March 2019 04:34:37(UTC)  | Reason: Not specified

Offline en1omb  
#8 Posted : 10 March 2019 15:16:05(UTC)
en1omb


Rank: HIFI Guru

Joined: 27/02/2009(UTC)
Posts: 299
Location: Bristol

Thanks: 6 times
Was thanked: 3 time(s) in 3 post(s)
Clones Audio do a modified Cisco switch. https://www.clonesaudio.com/networking

Seems like an interesting outfit. They do a dc power supply with optional Audio Note capacitors.
Offline frank23  
#9 Posted : 10 March 2019 17:05:03(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
Originally Posted by: en1omb Go to Quoted Post
They do a dc power supply with optional Audio Note capacitors.


This is nice.

I have collected a number of plug-in lineair PSU in the past years. Every time I buy a device now that has a switch mode PSU, I check if I have a lineair one with the same voltage / power rating and replace the plug if necessary. In one occassion, it turned my Hex dac around from "sell" to "hold". But this had nothing to do with network transfer, as my stereo is not hooked up to a network, be it wifi, bluetooth or wired ethernet.
Offline Steveh100  
#10 Posted : 11 March 2019 16:33:18(UTC)
Steveh100


Rank: Administrator

Joined: 18/10/2008(UTC)
Posts: 21
Location: London

Frank and Pete

Interesting povs but somewhat negative when Martin is asking for suggestions for cables to try. You may, of course, be correct but equally, you may be wrong and it may be that the perfect ethernet input does not exist. Without experimentation and some attempt to correlate cables, construction, shielding, grounding rigidity, materials we won't move forward. We can, of course, say billions of bits, or is it bytes, are moved around all days without error as we can say virtually any old cables will work connecting a USB printer but strangely we can hear the differences in USB cables as we can ethernet cables.

It doesn't mean in the perfect system there would be no difference but so far, AFAIK, no brand has launched the perfect system unless you drop to a system whose definition is so low one can't hear differences.
Offline EricE  
#11 Posted : 11 March 2019 19:07:44(UTC)
EricE


Rank: HIFI Novice

Joined: 31/05/2016(UTC)
Posts: 17
Netherlands

Hi Martin,

Consider M-WAY 3DW ReF Lan a Dutch brand with regard to the suggestions for CAT cables.

The information density increases spectacularly. The cables are characterized by an even lower noise floor, tones come to life sooner and they also live longer. Use them before and after the switch.

And make sure that you put the same effort to optimize your Router as you did with your Switch. So linear power supply and absorbing supports. It is even more rewarding! My Magico S5, Naim 500 and Aurender system love it.
Offline frank23  
#12 Posted : 12 March 2019 07:35:30(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
Originally Posted by: Steveh100 Go to Quoted Post
It doesn't mean in the perfect system there would be no difference but so far, AFAIK, no brand has launched the perfect system unless you drop to a system whose definition is so low one can't hear differences.


Our point is this:
Originally Posted by: Pete_w Go to Quoted Post
That doesn't mean that cables *won't* make a difference in systems which are poorly designed, they may well do some noise shaping, but since "poor design" comes in many flavours, there'll be no consistent results whatsoever.


Since "poor design" comes in many flavours, there'll be no consistent results. One could say that this has been the case forever with analog with differing input and output stages. But the difference with analogue is that in this case digital=digital, since it is passive digital, just passing the bits, there is no digital filtering etc done in the switches, just transfer. So any effect that a cable could have is through shaping the analogue noise of upstream PSU's and circuits. The reason this shaping might have any effect is due to the way the input circuit handles this noise.

In an ideal input circuit design and execution, the downstream circuits should not notice anything from any upstream cable. In general one could maybe say that a cable that attenuates noise being generated by the output circuit better, is preferable, since then an imperfect input circuit would not be affected as much.

But a perfectly designed input circuit should work as well with about any cable. Remember, this standard was developed for cables that run through office buildings and can span 100m between sender and receiver and still pump out Gbit data. The standard was designed for cables to be cheap to manufacture. This is no coax SDI, this is just twisted pair, non shielded, that is run through office buildings bundled in dozens for a 100metres.

A harsh environment indeed, and audio transfer is probably the easiest job these cables have ever done. I am interested in sonic differences that Martin encounters, but I would appreciate it if real thought was given to why the assumed sonic differences appear.
Offline Pete_w  
#13 Posted : 12 March 2019 11:53:50(UTC)
Pete_w


Rank: HIFI God

Joined: 28/07/2010(UTC)
Posts: 1,037
Location: Cambs, UK

Thanks: 12 times
Was thanked: 11 time(s) in 11 post(s)
Originally Posted by: Steveh100 Go to Quoted Post
Without experimentation and some attempt to correlate cables, construction, shielding, grounding rigidity, materials we won't move forward. We can, of course, say billions of bits, or is it bytes, are moved around all days without error as we can say virtually any old cables will work connecting a USB printer but strangely we can hear the differences in USB cables as we can ethernet cables.


You may be surprised to learn that I agree, and in fact I'll be as interested as anyone else to see what happens. However, I'm painfully aware that someone will then start attributing a particular "sound" to a cable, and then I (and, I suspect Frank) will scream.

The point that never seems to get made - clearly and unambiguously - in digital cable "tests" is that this is the digital domain. The analogue signal is NOT passing down that cable, the cable does NOT have a sound. From the digital perspective, the cable either carries the bits while generating an acceptably low error rate in the receiver, or it doesn't. As far as carrying the music data goes, that really is the end of the story.

From the system perspective, of course, that's far from the end of the story. Cables distort the signal that they're carrying (that's pretty much guaranteed, it's why you have eye diagrams...) and they may add or remove or pick up or shape noise at all sorts of frequencies. They may even do that in different ways depending on how they're laid out on the floor or tied in bundles. Cables have different materials and architectures, especially when you get to the more wacko end of the audiophile game - and they're all going to behave differently in terms of noise. Let's not consider the question about whether a CAT-6 cable that's designed for heavy noise reduction (naively, let's slap lots of capacitance in) is still a CAT-6 cable because it's possible it won't carry a signal for the distances which are required in the spec... (To pick up your USB point - I have complete galvanic isolation in my USB cable between my streamer and my DAC. Keeping that noise out of my DAC made a huge difference to the sound of my system. But the fact that the b****y thing now won't enumerate and boot up cleanly after a power cut, and needs to have its hand held manually when it starts up, means that actually it isn't a standards-compliant USB cable any more... Cursing )

So the digital signal arrives intact - bits in packets with valid checksums - but with a lot of noise. The receiver can also self-generate a lot of internal noise as it attempts to deal with the bit and packet timing. It's the job of the receiver to sort that out and NOT pass that noise on to the analogue sections within the DAC. The extent to which it succeeds or fails in doing that is, at heart, what this conversation is about.

What Martin & his team, with their analogue ears, will be listening to is the effect on a DAC of an unknown interference source interacting with a cable of unknown behaviour to generate noise in an uncharacterised (by them) receiver which then passes some or all of that noise to the DAC which has unknown noise susceptibility. The fact that they will hear differences is pretty much a given, it seems from the comments in this thread. What will be more interesting is if they use a range of different streamer/DACs and hear *similar* differences. That *might* mean something significant, or it might just mean that all the streamer/DACs are using the same Ethernet receiver chip, which is entirely possible...

Cheers
Pete
Offline Martin Colloms  
#14 Posted : 12 March 2019 11:56:17(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,042

Was thanked: 48 time(s) in 47 post(s)
Hi Frank

Using my ears I find that there is no such thing as perfect in the real world of audio.

The problem lies in the word audio, specifically reproduced sound.

Sound means energy and vibration, and almost nothing audio related on this planet is immune to it.

Our perception can be astonishingly acute in respect of unwanted, vibration induced noise. Many improvements are based on effective noise

suppression, and it was with some surprise that net switches were found to be strongly microphonic expressed in their digital, pure switching

domain.

Noise may be carried down CAT cables to equipment, while the chips, printed circuit boards, and crystal clocks in netswitches are themselves

susceptible.

Yes, input clocks on gear are supposed to be immune to vibration induced jitter, but they are not.

Simple countermeasures may produce surprising results.

The unusual dynamic expression of high end NAIM gear is substantially due to jitter controlling vibration countermeasures.

Martin Colloms
Offline frank23  
#15 Posted : 12 March 2019 19:31:10(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
Originally Posted by: Martin Colloms Go to Quoted Post
Our perception can be astonishingly acute in respect of unwanted, vibration induced noise. Many improvements are based on effective noise

suppression, and it was with some surprise that net switches were found to be strongly microphonic expressed in their digital, pure switching

domain.


Hi Martin, the ear's sensitivity is unbelievable. But I wondered what you meant to say here.

Do you mean that a sound played finds its way back in the digital signal through vibrations? That is, that the digital signal itself when decoded, shows traces of these vibrations?

Or do you mean to say that the noise accompanying the digital signal is related to these vibrations, but the digital information itself remains unchanged?


@Pete_w I fully agree with you. I thought of that myself this afternoon. An analogue cable affects both the signal and the noise it carries. Whereas a digital cable can only affect the noise it carries.
Offline Martin Colloms  
#16 Posted : 13 March 2019 10:27:56(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,042

Was thanked: 48 time(s) in 47 post(s)
Originally Posted by: frank23 Go to Quoted Post
Originally Posted by: Martin Colloms Go to Quoted Post
Our perception can be astonishingly acute in respect of unwanted, vibration induced noise. Many improvements are based on effective noise

suppression, and it was with some surprise that net switches were found to be strongly microphonic expressed in their digital, pure switching

domain.


Hi Martin, the ear's sensitivity is unbelievable. But I wondered what you meant to say here.

Do you mean that a sound played finds its way back in the digital signal through vibrations? That is, that the digital signal itself when decoded, shows traces of these vibrations?

>>>MC: Broadly you could say that is is a form of jitter present when the digital signal is decoded to analogue. The 1,s and zeroes, are unchanged but the micro timing is. Jitter rejection is an imperfect art. <<<<

Or do you mean to say that the noise accompanying the digital signal is related to these vibrations, but the digital information itself remains unchanged?


@Pete_w I fully agree with you. I thought of that myself this afternoon. An analogue cable affects both the signal and the noise it carries. Whereas a digital cable can only affect the noise it carries.


Offline kengale  
#17 Posted : 13 March 2019 13:40:25(UTC)
kengale


Rank: HIFI God

Joined: 25/11/2008(UTC)
Posts: 1,333
Location: UK

Thanks: 1 times
Was thanked: 8 time(s) in 8 post(s)
Originally Posted by: Steveh100 Go to Quoted Post
Frank and Pete

Interesting povs but somewhat negative when Martin is asking for suggestions for cables to try. You may, of course, be correct but equally, you may be wrong and it may be that the perfect ethernet input does not exist. Without experimentation and some attempt to correlate cables, construction, shielding, grounding rigidity, materials we won't move forward. We can, of course, say billions of bits, or is it bytes, are moved around all days without error as we can say virtually any old cables will work connecting a USB printer but strangely we can hear the differences in USB cables as we can ethernet cables.

It doesn't mean in the perfect system there would be no difference but so far, AFAIK, no brand has launched the perfect system unless you drop to a system whose definition is so low one can't hear differences.


The test method will be quite interesting. It will of course need a method of instantaneously changing cables within the system with the listener having no way of knowing at the time which cable they are listening to, and a properly randomised A/B or A/B/X test schedule. It will of course need a third party to be present to carry out these changes, hopefully a non-HiFi-aware person who won't unintentionally transmit to the listener any hint as to which cable is being used or its apparent quality. There will need to be a sufficient number of tests carried out for the results to be statistically significant.

Offline Martin Colloms  
#18 Posted : 13 March 2019 14:17:53(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,042

Was thanked: 48 time(s) in 47 post(s)
sorry but we are not going there again!

Peter Walker, guided by Peter Baxandall attempted to prove my reported subjective differences between 'good' amplifiers were illusory, and failed. They, supported by Tom Ivall editor at Wireless world, campaigned relentlessly to discredit me when subjective sound quality assessment was at last gaining ground.

Over 15 years of HiFiNews reviews did not discuss the sound quality of audio electronics because those that knew better knew that electronics could not have a sound if it measured reasonably well. I made a start on perception in the late 1970s at Hi FI for Pleasure.

We are not measuring the right parameters, even now, but they are difficult to capture.
Good ears make light work of the process.

I can reliably hear the insertion of a top class 0.1ohm power resistor in a loudspeaker connection.
Also different makes and types of distortionless power resistors at particular points in a loudspeaker crossover.

For example, working with a section of programme I know well, if the image focus is sharper, if the depth increases, if the bass lines have clearer pitch and time better, if the vocals are more articulate,then the sound quality is better, for whatever reason.

I cannot prove my perceptions but do my best to reliably report them .

Martin Colloms
Offline frank23  
#19 Posted : 13 March 2019 16:17:00(UTC)
frank23


Rank: HIFI Veteran

Joined: 12/07/2009(UTC)
Posts: 557
Location: the netherlands

Thanks: 2 times
Was thanked: 5 time(s) in 5 post(s)
Originally Posted by: Martin Colloms Go to Quoted Post
sorry but we are not going there again!


ThumpUp
Offline Togil  
#20 Posted : 13 March 2019 19:01:07(UTC)
Togil


Rank: HIFI Veteran

Joined: 04/10/2008(UTC)
Posts: 798
Location: Oxford

Was thanked: 1 time(s) in 1 post(s)

I understand that Peter walker privately accepted that an amplifier could sound better than a standard Quad amplifier but only if it was considerably more expensive ( presumably due to an improved power supply )
Hans
Users browsing this topic
OceanSpiders 2.0
3 Pages123>
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

Follow HIFICRITIC Email HIFICRITIC follow HIFICRITIC on Twitter Follow HIFICRITIC on Facebook