'Audiophile' Ethernet Cables.

[4est]

OK, you asked

In data transfer across a network utilising TCP/IP, data is in packets. It either arrives or not. Packet loss can be measured. It either arrives in time to fill the buffer... or not. Timing is not important - it arrives in time, or not.

Equally, *noise* (a word which is bandied around a lot), would have to be severe enough to disrupt packet transmission - in which case the packets don't arrive, the buffer isn't filled and the music stops. This is the beauty (yes, the beauty!) of digital. The "input" bitstream side to the software player is not going to be corrupted unless there is a severe communication disruption. Incidentally, I certainly experienced this with a poorly setup antivirus program that wanted to inspect all incoming NAS data - the music stuttered.

 

On the "other side" of the player, where digital bits of info are fed to a DAC, then of course timing matters. On the analogue side of the DAC, we are then back in the land of traditional audiophiles!

 

So... the ethernet cable either transmits the packets, or not. If the cable is well made, sturdy etc. with good electrically sound connections, then it is as good as any. "Audiophile ethernet cable" - marketing hoopla.

 

Analogue side cable differences, e.g. with different speaker cables - are, on the other hand, definitely plausible and there is good blinded data to show differences exist.

 

Hmm, no I didn't and you didn't tell. At least you didn't tell us anything that any reasonably knowledgeable computer person doesn't already know. I haven't a strong opinion on this, as I have not tested it. You on the other hand appear to be speaking from theory as to why something just cannot be. More times than not, that leads one into a dark lonely place called ignorance IME. I wish you to be right, but as sandyk alluded to, we would see little difference in any digital transmission or process if that were the case. Presently there is more conjecture and argument than facts opposing the large amounts of subjective empirical evidence contrary to your opine.

[Jsmith]

OK, you asked

In data transfer across a network utilising TCP/IP, data is in packets. It either arrives or not. Packet loss can be measured. It either arrives in time to fill the buffer... or not. Timing is not important - it arrives in time, or not.

Equally, *noise* (a word which is bandied around a lot), would have to be severe enough to disrupt packet transmission - in which case the packets don't arrive, the buffer isn't filled and the music stops. This is the beauty (yes, the beauty!) of digital. The "input" bitstream side to the software player is not going to be corrupted unless there is a severe communication disruption. Incidentally, I certainly experienced this with a poorly setup antivirus program that wanted to inspect all incoming NAS data - the music stuttered.

 

On the "other side" of the player, where digital bits of info are fed to a DAC, then of course timing matters. On the analogue side of the DAC, we are then back in the land of traditional audiophiles!

 

So... the ethernet cable either transmits the packets, or not. If the cable is well made, sturdy etc. with good electrically sound connections, then it is as good as any. "Audiophile ethernet cable" - marketing hoopla.

 

Analogue side cable differences, e.g. with different speaker cables - are, on the other hand, definitely plausible and there is good blinded data to show differences exist.

 

As far as I can tell you are only addressing this from one angle, which is the digital data transferred in the cable. What about the ethernet cable being a vehicle for transferring other noise (EMI / RFI) into the audio component that will affect the sound.

 

- You have stated that the analog side of the playback can be impacted by noise.

- Ethernet cables can pick and transmit EMI / RFI noise.

 

So would you agree it is possible that an Ethernet cable into your DAC or Streamer could have an audible difference? I am not talking about a cable to a switch or router, but directly into the audio components.

 

I also agree with others, that you may not need a really expense audiophile grade ethernet cable, but a simple box store cable may not contain enough shielding or be built well enough to avoid the exposure. So you can do so research and find a high quality shielded cable or buy an off the shelf audiophile one.

[Jsmith]

Ok so I wanted to test this for myself. I took my good old BlackBerry, switched off 3G, and dialed my voicemail.

- I placed the BlackBerry next to the Ethernet cable into my streamer and there was no sound.

- I put the BlackBerry on top of the Ethernet cable and I could hear the faintest crackling in the speakers.

- I coiled up the Ethernet cable and with the Blackberry on top the crackling was clearly audible in the speakers.

 

Next I am going to order a 10' patch cord of Belden bonded pair Ethernet and attach a couple of ferrite chokes and test again.

[John R Leddy]

I placed the BlackBerry next to the Ethernet cable into my streamer and there was no sound.

 

This is reassuring. As long as nobody crawls behind my hi-fi shelving unit and lays their BlackBerry on my cabling I should be OK for sound quality. No need to worry about purchasing different cables; what I have will do just nicely. Good enough for a mixing console, so good enough for me. Thanks for reporting on your test. It's what I already thought, but it's good to have confirmation. Don't know how the cable doubters will respond though.

[eternaloptimist]

Hmm, no I didn't and you didn't tell. At least you didn't tell us anything that any reasonably knowledgeable computer person doesn't already know. I haven't a strong opinion on this, as I have not tested it. You on the other hand appear to be speaking from theory as to why something just cannot be. More times than not, that leads one into a dark lonely place called ignorance IME. I wish you to be right, but as sandyk alluded to, we would see little difference in any digital transmission or process if that were the case. Presently there is more conjecture and argument than facts opposing the large amounts of subjective empirical evidence contrary to your opine.

 

I replied to the wrong quote (intended for JSmith). Feel free to ignore.

As you say, any reasonably knowledgable computer person should know this - so shouldn't have to spell it out.

Yes, I am relating theory - theory as as explanation of a phenomenon backed by evidence.

Far from being led to a dark and lonely place, I am trying to avoid the large stadium called "gullible".

I too have an open mind - but not so open that my brains fall out!

Big claims are being made regarding the audibility of difference in digital network cables. Plenty of conjecture and hypothesis but crap evidence - anecdotes are not quality evidence. Fun for a hobby... but not evidence.

 

Doesn't it strike you as fascinating that the cable manufacturers will produce beautiful graphs of alleged frequency responses etc. in the realms of where only bats or microwave detectors can hear (ok, some hyperbole here...) but no reports of independent blinded testing? The onus is on the manufacturers to produce something that resembles evidence, not on me to disprove it....

 

I think I will bow out now as this is getting silly. If you want to fork out $$$ for this, it's your money, enjoy!!

 

@Jud: Albury is in Eastern Australia on the border between Victoria and New South Wales. Come on down!!

[4est]

Doesn't it strike you as fascinating that the cable manufacturers will produce beautiful graphs of alleged frequency responses etc. in the realms of where only bats or microwave detectors can hear (ok, some hyperbole here...) but no reports of independent blinded testing? The onus is on the manufacturers to produce something that resembles evidence, not on me to disprove it....

 

Yes, and on many levels I agree. I would completely if it were that I have not always seen concrete evidence supporting what I've experienced.

[sandyk]

I will run naked down the Main Street of Albury if in an ABX test, a fancy Ethernet cable (or USB cable for that matter) can be reliably picked from a solidly built generic cable built to standard specification! By the way, this is not a sight anyone should have to witness, so I hope I am not wrong!

 

Come on you guys who have posted strong preferences for a particular brand of USB cable. Take the test, then head down to Abury in Au. , get the cameras rolling and post it on You Tube !!!

[eternaloptimist]

Come on you guys who have posted strong preferences for a particular brand of USB cable. Take the test, then head down to Abury in Au. , get the cameras rolling and post it on You Tube !!!

 

Fantastic!!

Incidentally, I would love to host CA visitors! Love a good debate!

[junker]

Data over ethernet is packet based. If there is ever a problem from say a solar flare or SKYNET dropping an EMP in your vicinity then the dropped packet gets resent. It's part of the protocol. If is was streaming that would be another story, but image if little bits and bytes were getting corrupted all the time over ethernet then pretty much our entire internet infrastructure would fail.

 

I say if your Cat 5e/6 cables are working properly then spend some money on cables that would actually make a difference.

[sandyk]

If is was streaming that would be another story, but image if little bits and bytes were getting corrupted all the time over ethernet then pretty much our entire internet infrastructure would fail.

Unless I am mistaken, nobody has reported ANY data loss or corruption, despite hearing differences between different Cat. cables.This is more of the doomsday garbage type of scenario that seems to get trotted out whenever people report hearing differences that are presently unexplainable. Planes would fall out of the sky, automobiles with ECUs would no longer start, the older cars that didn't have ECUs would get stuck at failed traffic lights, mobile phones would no longer work (what a blessing for some !) etc. etc.

[souptin]

...doomsday garbage type of scenario... Planes would fall out of the sky, automobiles... would get stuck at failed traffic lights, mobile phones would no longer work... etc. etc.

...And many years later the starved remnants of humanity watched in amazement as somebody calling himself "sandyk" emerged from his concrete, lead and ers paper shielded bunker and said "Oops, sorry guys. Maybe I took things a bit too far with that last ripping session"

[Jud]

Data over ethernet is packet based. If there is ever a problem from say a solar flare or SKYNET dropping an EMP in your vicinity then the dropped packet gets resent. It's part of the protocol. If is was streaming that would be another story, but image if little bits and bytes were getting corrupted all the time over ethernet then pretty much our entire internet infrastructure would fail.

 

I say if your Cat 5e/6 cables are working properly then spend some money on cables that would actually make a difference.

 

Hi, junker. The problem with thinking this way is you get yourself in the "digital" box and forget these are electrical wires. So they can carry noise into the analog part of your audio system just like your analog cables. The electrical noise traveling down the wire into your system has no way of "knowing" the wire is also carrying analog signals representing packets of digital information, so it just runs right along that wire as it would any other. (Yes, manufacturers take steps to prevent this noise from getting into analog circuitry, but these steps are not perfectly effective.) Optical cable, now you've got a different story (though optical transmission is subject to its own problems).

[Jud]

...And many years later the starved remnants of humanity watched in amazement as somebody calling himself "sandyk" emerged from his concrete, lead and ers paper shielded bunker and said "Oops, sorry guys. Maybe I took things a bit too far with that last ripping session"

 

Now that is funny [hat tip].

[Jsmith]

Data over ethernet is packet based. If there is ever a problem from say a solar flare or SKYNET dropping an EMP in your vicinity then the dropped packet gets resent. It's part of the protocol. If is was streaming that would be another story, but image if little bits and bytes were getting corrupted all the time over ethernet then pretty much our entire internet infrastructure would fail.

 

I say if your Cat 5e/6 cables are working properly then spend some money on cables that would actually make a difference.

 

Just like the other posters you are addressing only one possible issue, which is packet loss. You have stated nothing about the other potential issues such as EMI / RFI noise being picked up and transmitted through the Ethernet into the audio circuitry.

 

Now I am sounding like a broken record, but will everyone please look at things holistically. It does no good to keep repeating the same one point over and over again as the end solution when it addresses only part of the problem.

[Paul R]

Ooooo... Get your running shoes on, at least where it concerns USB cables. Might be safe for a while though, plane fare down there is a bit excessive.

 

As for Ethernet cables, I doubt anyone challenges that the digital data gets there unscathed. Like others, I think unintentional electrical noise can hitch a ride on the cable though. Just set a bluetooth keyboard down over an ethernet cable and at least here, you can sometimes play tunes with the noise. Depends upon the cable and the system it is going into. Can do almost the same thing with a power cable too. Speaker cable seems resistant to bluetooth for some reason.

 

Paul

 

 

 

Again, this is digital side... The likelihood of there being a genuine difference rather than a psychoacoustic cause for the described difference is much, much lower given known physics. Hence the burden of proof is much higher. Where is the blinded data? Case reports and anecdotes are fun and I enjoy making them myself. I do not, however, claim any of my hobby-based musings / observations as evidence. I am very open to the possibility of psycho-acoustic effects.

 

Incidentally, am willing to believe there are purple aliens on the dark side of the moon... but it is so unlikely that I would want to see spectacular proof! Speaker cable differences - likely to be small but plausible based on physics. Ethernet cable differences, eg from NAS to PC making a difference. Highly implausible, would need rigorously demonstrated evidence - much more than forum anecdotes. I will run naked down the Main Street of Albury if in an ABX test, a fancy Ethernet cable (or USB cable for that matter) can be reliably picked from a solidly built generic cable built to standard specification! By the way, this is not a sight anyone should have to witness, so I hope I am not wrong!

[macbrush]

I absolutely agree with you on this one. Personally, I am not sure about ethernet cable, in fact, I just randomly checked among those 100 switches that I managed, and couldn't find a single transmission error, some of them have been transmitting dozens of terabytes of data since last reset. But I do know that cables pick up noise, and sometimes carry them down the chain to the analogue part of the system, sometimes they're audible and easy to fix, on the same token I don't see why they won't affect SQ very subtly.

 

Now I want to talk about USB cables. I am not a "golden ear" or anything, not by a long shot, in fact, my left ear was slightly damaged in rave parties back in the old days (I doubt my right ear is in any good condition either), so I doubt I can ever hear music as most of you do. So I want to talk about technical side of the stuff. Until recently, I thought cables are cables, they're all the same, and one day, I discovered that cables really aren't made equal, especially USB cables. I am a Mac user, and one problem plagued me for absolutely years, USB drives eject themselves randomly, not because of power issue I assure you. I tried absolutely everything, and couldn't figure out what was wrong. Then about a month ago, I bought a LaCie USB 3.0 drive, it comes with a good quality 1m USB cable, but the drive kept ejecting itself in my own computer, I couldn't figured out why since everything look okay, so I tested the cable with another drive and it worked perfectly, put the cable back to the original drive, and the problem came back. One thing suddenly hit me, cables in the old SCSI era, often you have to mix and match the cables, swap things around the chain to make things work. So I brought the drive and cable back to office and started experiment on mixing cables and external drives. I found that combination of cable, drive, USB host adaptor and OS have an impact on reliability. Windows is much more resilience than OS X in terms of USB bus error, but still noticeable, it won't eject the drive most of the time, but you still can see errors in logs. Anyway, out of dozens of USB cables, all good quality cables, at least from the look of them anyway, and all passed cable tester with flying colour (I am not that sloopy), then at the end I found a couple of them works perfectly in all combination, most of them short, about 1 ft, but one of them was a 6 ft long cable and the other was a 3 ft cable, so length probably matter, but definitely not main deciding factor. Needlessly to say, all those "good" cables are now at my home. :-P

 

Personally, I am confident to say that cables weren't made equal, at least for USB cables and probably Firewire too since I had some similar issues with firewire too, admittedly to much lesser degree. I probably won't buy any USB cables with a crazy price tag, most likely I won't hear any difference anyway. But from now on, I will get good quality non-audiophile cables and test them thoroughly before letting them into my system. BTW, good looking cables come from reputable manufacturers do not necessary translation into good cables. Professional heavy duty cables are a good starting point though, IMHO, if you're on a budget.

 

Cheers

Kenneth

 

Ooh, Albury where? :-D

 

'Cause I don't know about Ethernet cables, but I definitely believe I hear differences in USB cables, even feeding the async USB input in my DAC.

 

I also have some thoughts about possible ways to improve the discriminatory value of A/B (or A/B/X) testing that I keep threatening to put down in a thread I started. Time for me to buckle down and do that in the next 2-3 days.

[mwheelerk]

Now I want to talk about USB cables. I am not a "golden ear" or anything, not by a long shot, in fact, my left ear was slightly damaged in rave parties back in the old days (I doubt my right ear is in any good condition either), so I doubt I can ever hear music as most of you do. So I want to talk about technical side of the stuff. Until recently, I thought cables are cables, they're all the same, and one day, I discovered that cables really aren't made equal, especially USB cables. I am a Mac user, and one problem plagued me for absolutely years, USB drives eject themselves randomly, not because of power issue I assure you. I tried absolutely everything, and couldn't figure out what was wrong. Then about a month ago, I bought a LaCie USB 3.0 drive, it comes with a good quality 1m USB cable, but the drive kept ejecting itself in my own computer, I couldn't figured out why since everything look okay, so I tested the cable with another drive and it worked perfectly, put the cable back to the original drive, and the problem came back. One thing suddenly hit me, cables in the old SCSI era, often you have to mix and match the cables, swap things around the chain to make things work. So I brought the drive and cable back to office and started experiment on mixing cables and external drives. I found that combination of cable, drive, USB host adaptor and OS have an impact on reliability. Windows is much more resilience than OS X in terms of USB bus error, but still noticeable, it won't eject the drive most of the time, but you still can see errors in logs. Anyway, out of dozens of USB cables, all good quality cables, at least from the look of them anyway, and all passed cable tester with flying colour (I am not that sloopy), then at the end I found a couple of them works perfectly in all combination, most of them short, about 1 ft, but one of them was a 6 ft long cable and the other was a 3 ft cable, so length probably matter, but definitely not main deciding factor. Needlessly to say, all those "good" cables are now at my home. :-P

 

Cheers

Kenneth

 

Interesting. I have always been a proponent of good quality cables of all types. I say interesting because I had an issue several months back (Feb 2012)when I replaced my 1 TB Western Digital Drives with 2 TB Lacie drives. I experienced the same self ejection issue. One of the drives continually and the other less frequently but no less bothersome. I was online with Lacie support constantly over a four month period. I tried a couple of different stock USB cables and finally bought a different better quality cable. Problem hasn't been seen since.

[wgb113]

These are good enough for me:

 

http://www.belden.com/pdfs/Prodbull/CB006.pdf

 

It's performance over generic category cables has been both demonstrated and measured. Do I claim to hear a difference? Nope. I just don't worry that I'm missing anything from a speed/performance standpoint.

 

Bill

[PewterTA]

Galvanic isolation sure gets rid of the emi/rfi issue..... if your computer is decent and you are careful with keeping interference to a minimum... Ethernet cables make no difference...

[Jsmith]

These are good enough for me:

 

http://www.belden.com/pdfs/Prodbull/CB006.pdf

 

It's performance over generic category cables has been both demonstrated and measured. Do I claim to hear a difference? Nope. I just don't worry that I'm missing anything from a speed/performance standpoint.

 

Bill

 

Right now I am using some Belden Bonded patch cables as well. I smacked some ferrite cored on the end just for good measure.

[dnoyeB]

I'd like to add something to this discussion.

 

Transmission over Ethernet is packet based. On multiple levels. The Ethernet layer has its packets. That layer has its own error detection. The error detection is good. On top of the Ethernet you will have an IP layer packet. This also has error detection. On top of the IP layer is either TCP or UDP data. These both have error detection as well. TCP will flag errors and retransmit the bad data. UDP will just flag the error and leave it up to the application to resend or not.

 

Audio requires pretty good packet stream. There is often no time to retransmit. This is why you use a buffer. If a bad packet comes, you keep playing while you wait on a retransmission of the data. During this time you will be eating away at the buffer. The packet comes, and fill your buffer again. So you should be transmitting data over the network faster than it is being used in order to maintain this buffer. The only way to drop quality is to drop enough packets to empty the buffer. Your audio device is fully aware of this packet loss and can easily tell you about it if they so chose.

 

Ethernet cable uses twisted pairs of wire. This is a poor mans way of doing coaxial. This provides a level of immunity against noise. Additionally, you can get 'shielded' Ethernet cabling. For more protection.

 

Incidentally, there is one real advantage of fiber over Ethernet. No conducted or radiated noise is picked up with the laser. I think most of you probably have experienced this during the TOSLINK vs. RCA digital data days...

 

 

Audiophile network cable is snake oil. There is no entity in the computer world that can or will accept bad data. That is why network cables are made to a standard.

[realhifi]

Ethernet cable uses twisted pairs of wire. This is a poor mans way of doing coaxial. This provides a level of immunity against noise. Additionally, you can get 'shielded' Ethernet cabling. For more protection.

 

Sounds to me like noise COULD affect a lesser made cable? Or possibly a better connector could help in noise rejection?

[Jud]

 

Incidentally, there is one real advantage of fiber over Ethernet. No conducted or radiated noise is picked up with the laser. I think most of you probably have experienced this during the TOSLINK vs. RCA digital data days...

 

 

Audiophile network cable is snake oil. There is no entity in the computer world that can or will accept bad data. That is why network cables are made to a standard.

 

You've somewhat contradicted yourself here. If we only have to worry about data and not about the fact that there are electrical connections between electrical components that can transmit varieties of electrical noise, then what would be the "advantage" of fiber? (Incidentally, fiber has its own disadvantages, e.g., bandwidth.) Simplification for the sake of modeling is necessary, but let's not oversimplify to the point where we ignore something that may be important, i.e., let's not concentrate on the data connection to the point we forget plain old electrical noise.

 

Is there a lot of noise? Unless you're in fairly unusual circumstances, no. But when electrical noise in the system can, for example, affect the clocking in the DAC, and at digital stream resolutions where jitter numbers in the picoseconds may make a difference, then the potential for even low amounts of electrical noise or other electrical effects to impact sound quality is something at least worth thinking about and discussing, rather than dismissing out of hand as "snake oil."

[junker]

Boulderdash. Poppycock. It's this kind of voodoo - and $120k speakers - that does audiophiles, and the industry, such a disservice.

 

There are numerous cables and interconnects that will make a difference. This is certainly not one of them. You guys are trying way too hard.

 

If you are concerned with noise run a proper length of quality spec Cat 6 cable, call it a day, and buy speaker wires, interconnects, USB cables, and even AC cables.

 

Run, run away from "audiophile" ethernet cables, or any companies that promote their products are such. These types of charlatans are dead to me.

[junker]

Snake oil. Advantages are less resent packets on long runs, or in extreme RF environments (i.e. server farms, etc).

 

Anyhing work it's salt will re-clock the digital signal that is pulled from a buffer with jitter dependent of the re-clocking device (i.e. ESS Sabre, etc.)