Ethernet Cables in for evaluation

Usually a degree in computer technology includes some courses in analogue system design and noise elimination and I think you are clever enough to have attended one. Somewhere down in the system there are hopefully some analogue devices and these are affected by noise from other parts in the system. If you have a router generating lots of noise there might be a chance it ends up in the wrong place. Now an ethernet connection should be galvanic isolated but there might be better and worse devices in that aspect. On the other hand I would assume there is a bigger chance of having unwanted noise from the ethernet receiver in the streamer than from the router.

 

I think vwestlife is sarcastic and yes, some parts of this thread is really bizarr.

 

I definitely know very little about this topic. It seems all the switch would do is hand off the DAC data packets it receives from the router that are addressed to the DAC. In doing so is it possible the waveform and noise characteristics of the signal received by the DAC may be different than they would be without the switch, since that signal would be generated by the switch rather than by the router? I’m not saying the waveform is necessarily better but if it is somehow different then my ears may be perceiving the change as an “improvement”. If it’s truly impossible that the waveform and noise can’t possibly change across the switch (with its own power supply) as it goes to the DAC then I suppose there is no possible scientific explanation for a sonic difference. I’m just curious what other’s thoughts are. I’m not trying to inject a polarizing viewpoint, just being open minded to all input as I try to learn more.

 

See - the problem is not what you explained is not true - the problem is that given how Ethernet and digital transmission in general actually works, you need to demonstrate that the above causes an audible difference - and while I 100% believe you that you - or anyone else that does - hear a difference, well, that doesn't mean there is one - sample is too small - just one - even if all audiophiles that can afford $10K Vodka ethernet cables got together, the sample would still be too small. Not to mention the actual listening, which is not a controlled experiment, completely susceptible to confirmation bias etc, not to mention the lack of a protocol so on.

But, if anyone can hear a difference and can afford to waste $10K on ethernet audio cables - go for it - I am not fighting that - too much - anymore. Makes you happy that is fine. At least keeps the economy moving somehow. However, I will never stop thinking that in general magical thinking is not great for humanity in general - but well.. only post on this, I promise...

v

 

iSCSI is where it's at.......

 

I am happy you think what I explained is true.

 

Use an old Dlink router. Connect the ground in the router with ground on your analogue outputs. Might produce some unwanted noises.

 

Yes, the twisted-pair wiring used in Ethernet cables is specifically designed to reject noise and interference:

To gain a high level of immunity from interference, twisted-pair circuits use a form of signal transmission known as balanced-pair transmission. Balanced-pair circuits possess the following attributes:

• The impedance, when measured from each leg to ground, is equal. Thus, both conductors are at signal potential (i.e., above ground potential). Compare this with coaxial, where one conductor -- the outer screen -- is grounded. Signals in coax are a form of unbalanced transmission.
• The conductors of the pair are driven in "push-pull" or differential mode. This means that the signal at any point along the circuit, when measured from one leg of the pair to ground, is equal in magnitude and opposite in polarity (or phase) to that measured at the same point from the other leg of the pair to ground.
• Any outside electromagnetic field that couples into the cable causes the same voltage (both magnitude and phase) to be induced onto both legs of each pair in the cable. Hence, this induced interference voltage is in common mode.
• Finally, the terminating equipment at each end of the cable is designed to respond only to differential-mode signals, and is designed specifically to reject common-mode signals. Thus, although the twisted-pair cable picks up unwanted signals through coupling, it does so in a form that can be rejected by the terminating equipment.

 

No, there is no chance

 

You passed the exam :-D

 

I am happy you are happy - or was that sarcastic? YOU ARE SO SMART. ( I really mean it)


v

 

The router might be faulty. Is there still no chance?

 

Yeah I know. I couldn't resist when both sides of this argument are running fast forward without knowing who they are responding to or what their opinions are.

Buying ethernet cables for thousands of dollars doesn't sound sane.

 

I think if you want to believe all this stuff makes a difference that's your right as an audiophile, you just need to completely back away from all the faux-technical stuff and reach the ultimate conclusion all subjectivists do - that science is constrained and simply cannot quantify the huge improvements you definitely hear, so there's no use in mentioning it.

Otherwise, you're going to keep creating concepts like "ethernet noise" and better sounding IP packets in an attempt to validate your beliefs and attract incredulous disbelief.

 

My question really is not one of faux-technical stuff. I'm just curious whether science as we know it today says that it's possible for noise to permeate through an ethernet switch and cable. If the answer is "No, noise cannot permeate switches and cables", then I will accept that and accept as fact that my adding a switch could not have changed anything w/r/t noise. I'm sure someone here is qualified to answer this question. Right???

 

Yes, that's exactly what a switch does.

No, this is emphatically NOT possible.

Think of it this way: Ethernet uses packets of data, millions of them in a continuous stream. Each individual packet of data is analogous to mailing a letter with a single word inside it. It doesn't matter if that letter gets some rain or dirt on it, as long as you can still read the address on the envelope, it shows up in the right order, and the word inside is still legible, then the letter has served its purpose in communicating information. The words are then reassembled into sentences by the person receiving the letters.

With Ethernet, far more than any other form of signal transmission (e.g. SPDIF or analog), the only thing that matters are whether packets are being received successfully. There's error correction built into each packet to tell the receiving device if the packet it received was corrupt or not. It's very much an all or nothing system. Trust me, I've seen Ethernet runs that were too long, or run parallel to power lines, or crimped without the correct twisted pairs, and the result is nothing. No communication. Nada. It's not "hazy" sound, it's NO sound.

Ethernet works perfectly until it doesn't, and that's the end of the story.

 

This is why I quit from bothering with 'audiophile gear' when I saw the hi-fi voodoo BS starting to be applied to common or garden computer parts that were designed by people far beyond the muppets who make over priced hi-fi stuff in sheds all those years. Most of Hi-fi is taking someone's $10 part and slapping a label on it and charging $1000 for it.

"Oh yes Western Digital has deeper bass!" Intel gives much more detail!" "If you place your RAM on a North south axis..."

 

I am insulted and indignant that OP has proposed listening to cables.
A definite scam liberally laced with "snake oil".
Who's he trying to kid? I've got a bridge I wanna sell ya.
Not standing for any more of this ridiculousness.
Hope everyone joins me in boycotting this thread!

 

How does that old joke go? Ah yes:

Hell is where the chefs are British, the mechanics are French, the lovers are Swiss, the police are German, and the network admins are audiophiles.

 

Yes I am I'm long qualified on this, but there is plenty of easily digestible documentation online one can read, no-one needs to explain Ethernet and TCP/IP networks here on an audio forum other than they're nothing to do with the processing of audio.

No

 

I really just wanted a Yes or No answer from one of you digital gurus. I swear I will take your answer as gospel as this isn't my area at all. Just tell me, is it possible for noise to permeate an ethernet switch and cable?

 

Not really true. We have had situations where a switch seemed to be working but every now and then a package was lost for no reason and no retransmission was triggered. It was almost impossible to analyse the trouble situation due to the intermittent behavior.

Do I have to say the switch was faulty.

 

I am curious to hear how it is impossible for noise to travel through faulty ethernet routers and cables.

 

I also had a problem with an optical cable where I was only getting about 40kbps on a 10G link. It was working, but severely degraded. So, not just perfect or not at all. Turns out I was accidentally using a single mode cable instead of a multi mode cable

 

Absolutely, but the problem was with the malfunctioning switch, NOT the cable, which is what this thread is about. I never claimed networking devices are flawless and indestructible, they fail like any other piece of electronics.

It is entirely possible for background noise or hum to travel over an Ethernet cable. However, if that noise is reaching your headphones or speakers then the networked device receiving the data has a severely flawed design, because the Ethernet connection should be completely isolated from the analog stage of the DAC. If it isn't, you've bought a very crappy DAC, and you'd be much further ahead spending your money on a properly designed DAC rather than a better Ethernet cable.

Let me be perfectly clear here, though: the noise I'm talking about does not appear in the data received by the device. The data itself is not being modified and is still being received correctly, there just happens to be an unwanted, stray analog signal being passed through the receiving device and mixed into the final output. This is a flaw of the device, not the cable.

Yes, but the data was still being received, just very slowly, right? It's not like it suddenly started corrupting the data and returning gibberish, you were just having a problem with a ton of lost packets.