$340 Audiophile Ethernet Cable Tested

An anonymous reader writes: Ars Technica has posted a series of articles attempting to verify whether there's any difference between a $340 "audiophile" Ethernet cable and a $2.50 generic one. In addition to doing a quick teardown, they took the cables to Las Vegas and asked a bunch of test subjects to evaluate the cables in a blind test. Surprise, surprise: the expensive cables weren't any better. The subjects weren't even asked to say which one was better, just whether they could tell there was a difference. But for the sake of completeness, Ars also passed the cables through a battery of electrical tests. The expensive cable met specs — barely, in some cases — while the cheap one didn't. The cheap one passed data, but with a ton of noise. "And listeners still failed to hear any difference."

Comments

Passed data with a ton of noise?

[Mr D from 63]

What is this data passing noise?

Re:Passed data with a ton of noise?

[Mr D from 63]

The fact that they even wasted time thinking about doing a listening test is enough data I need to know they don't know WTF they are doing.

Re:Passed data with a ton of noise?

[chipschap]

"The cheap one passed data, but with a ton of noise. "And listeners still failed to hear any difference."

Well, duh. It's digital, not analog. As long as bits are not irrecoverably lost, how is it going to sound any different?

Audiophile ethernet cable .... there's one born every minute.

Re:Passed data with a ton of noise?

[Firethorn]

Okay, digital data is supposed to be easy 1 and 0 communication. But when you get down to the physical media, said binary digits are represented by physical phenomenon. So +3.3V = 1, 0V = 0 type stuff.

Voltage, resistance, EM waves, magnetics, etc... You're actually back in the world of Analogue, and here you have to worry about noise.

When you're moving data as fast as you can, or storing it as densely as you can, interference becomes more likely. For example, you'd think that +3V =1 and 0V = 0 would be easy, but when you're flipping the signal as fast as you can, you end up with the cable possibly acting like a transformer or capacitor. So the voltage might run a bit higher, a bit lower, a bit faster, a bit slower, etc...

Radio transmissions, Solar noise, close by electrical cables, other data cables with parallel runs, etc... The world is 'noisy' even if you're using wires.

That's why you have error correction in digital communications. So the 'occasional' bit can become flipped and the system transparently recovers it, and you get your transmitted data, identical from the other side.

Re:Passed data with a ton of noise?

[bobbied]

THIS! But testing Ethernet cables passing TCP is pointless anyway... Where is the Bit Error Rate specs here? Digital data is either correct at the bit level, or it's not...

Re:Passed data with a ton of noise?

[erice]

What is this data passing noise?

What they should have meant: the raw bit error rate was high and there were many retransmits.
What they probably meant:: The square waves weren't very square.

Re:Passed data with a ton of noise?

[Mr D from 63]

Hey, lets sell some audiophile routers!

Re:Passed data with a ton of noise?

[Anonymous Coward]

They did the listening test first, before the cable quality check. And yeah, they were really, really skeptical of the cable, but they'd had enough of the 'if you've never listened to it, how can you know?' kook crowd - they wanted a to do a real test just because people don't test these types of claims. They either write them off as ridiculous (as they nearly always are) or they buy into them completely.

So here: Objective, blind test. No difference as far as anyone they tested could reliably tell. (They did have one person in the test correctly guess which cable was which. Out of two who tried - the rest didn't even try.)

Re:Passed data with a ton of noise?

[Mr D from 63]

Sure, noise is the right term, when used properly. To imply that the noise is in the data, as they do, but 'you can't hear it', is quite ridiculous.

Re:Passed data with a ton of noise?

[trabby]

But man! The lack of oxygen in the copper allows you to hear unicorns farting rainbows 50ft away from the mic in the recording!

Re:Passed data with a ton of noise?

[goarilla]

Indeed this is stupid, even Ethernet frames use CRC codes.

Re:Passed data with a ton of noise?

[jandrese]

They did it because that's how the company that makes the cables describes their test environment. They claim "clear unmistakeable" improvements in the audio quality in the setup Ars used, but only if you plug the cable in in the correct direction (denoted by an arrow on the connectors).

In fact the company claims that they determine which way to face the arrow by plugging the cable in and listening in both directions and choosing the best.

Re:Passed data with a ton of noise?

[Mr D from 63]

They did it because that's how the company that makes the cables describes their test environment. They claim "clear unmistakeable" improvements in the audio quality in the setup Ars used, but only if you plug the cable in in the correct direction (denoted by an arrow on the connectors). In fact the company claims that they determine which way to face the arrow by plugging the cable in and listening in both directions and choosing the best.

The direction of the cable? Are you shitting me?

Re:Passed data with a ton of noise?

[ArcadeMan]

And audiophile microSD cards!

Re:Passed data with a ton of noise?

[bluefoxlucid]

That needs to go away. We need an Ethernet protocol extension with BCH or Hamming code support.

Re: Passed data with a ton of noise?

[Anonymous Coward]

Nope. Common in audio to have directional cables. Guess no one ever told the golden ears that all audio systems use alternating currents.

Re:Passed data with a ton of noise?

[slashmydots]

I was thinking the same thing. Invalid data is not accepted at the recipient with any protocol that uses Ethernet that I've ever heard of.

Re:Passed data with a ton of noise?

[Firethorn]

Yeah, if you're in an exceptionally noisy environment with a long cable run, it might make sense to buy a more expensive cable that features things like additional shielding, thicker gauge wires, etc...(or just go fiber).

Otherwise the ECC does it's job and there's no practical difference between the cheapest wire possible and the most expensive.

Hell, audiophiles can't tell when the testers are using coat hangers as speaker wires, and that's an actual analogue signal!

But the cases where a decently constructed commodity cable won't do but the 'premium' will are generally limited.

In many cases with 'premium' cables, you end up with cable and connectors so heavy that that causes the very problems than the cable was supposed to solve!

I'll admit, I've spent double the money on 'better than the cheapest I can get' cables on occasion, but that's usually because I had broken the cable through use/abuse that was there originally, so getting a 'tougher' cable made sense.

Re:Passed data with a ton of noise?

[lsllll]

Sometimes I wonder if anyone actually RTFA. From TFA:

Even the most rabid speaker cable true-believer audiophiles will admit that digital is digital—at this point, almost everyone has accepted that the bits will arrive, or they won’t. However, the audiophile contention is that some amount of electromagnetic interference or noise is transmitted up unshielded Ethernet cables, through the Ethernet port, and into the computer’s DAC (the digital-to-analog converter), which then makes itself apparent to the listener by coloring the sound in some way.

So, the contention is not that these cables will differ in a "DATA" setting, but that the cheaper cable may introduce unwanted noise into the circuit after the DAC. Now we can argue whether that does or does not happen (I believe it to be highly unlikely), but the argument is not about digital noise.

Re:Passed data with a ton of noise?

[goarilla]

Yes error correction instead of only error detection would be nice.

Re: Passed data with a ton of noise?

[Paco103]

It's common to have polarity, but not directionality. It's always been common to have them denoted in some way (usually a white stripe on the jacket, or gold/silver wire, etc). This denotes your "common" and your "signal" (alternating). Polarity is different than directionality. The arrow would indicate that a signal travels through better one way than another, which would mean your cable is acting as a diode. If that doesn't fail spec I don't know what would.

Re:Passed data with a ton of noise?

[Joce640k]

THIS! But testing Ethernet cables passing TCP is pointless anyway... Where is the Bit Error Rate specs here? Digital data is either correct at the bit level, or it's not...

It's almost as if you didn't read the article before posting that.

Hint: The problem perceived by audiophiles isn't caused by data errors.

Re:Passed data with a ton of noise?

[Dutch Gun]

There's nothing wrong with spending a bit more money to get a quality cable, so long as what you're looking for is better physical quality that meets the expected spec, and not "magical fairy dust" quality.

Re: Passed data with a ton of noise?

[Joce640k]

And don't forget that cables need to be burnt in for a month or so before they achieve full sound.

(Moving them or bending them in any way de-aligns the atoms and means another burn-in period...)

Re:Passed data with a ton of noise?

[viperidaenz]

don't remind me

Re:Passed data with a ton of noise?

[sconeu]

CompUSA used to sell "Music CD-R"s. I think the only difference was that the MAFIAA got a cut of the sales.

Re: Passed data with a ton of noise?

[omnichad]

I think the directionality is based on extra shielding that is grounded only on the far end so extra electronic noise isn't brought into the DAC. Doesn't mean it's not completely bogus, but they have great reasoning behind it.

Re:Passed data with a ton of noise?

[Mike Van Pelt]

I got into this with an audiophule type a few years ago. He, with a completely straight face, asserted that double-blind testing was an inherently flawed methodology for evaluating the objective marvelosity of some silly audiophule crap he was touting. (This obviously being some entirely new definition of the word "objective" that I was previously unacquainted with.) In that case it was 12-gauge solid copper speaker cables at $$$$$/foot. I said "And that is different from $0.12/foot Romex... how, exactly?" He started going on about how these things were oxygen-free rectangular cross section, hand-forged by the Kebler Elves with tiny silver hammers... and then summarily dismissed double-blind testing when I suggested it.

Re:Passed data with a ton of noise?

[mwvdlee]

The makers of the cable didn't know WTF they were doing; they claimed it was better for music.
The testers merely were merely thorough by explicitely testing the claims, asinine as they might be.
By doing so, they are taking away any dumbshit rebuttal the makers might have had otherwise.

Re: Passed data with a ton of noise?

[DrLang21]

(or just go fiber)

Using fiber is a difficult expense for true audiophiles. The fibers typically used can contain a lot of impurities which distorts the color of the light signal, introducing noise into your audio. This is why when using fiber for audio, true audiophiles only use diamond fibers extruded through Emerald dies enchanted by a wizard after he puts on his robe and hat.

Re:Passed data with a ton of noise?

[Anonymous Coward]

Audio CD-Rs and "data" CD-Rs differ by a flag in the metadata that is already on the disc as you buy it. There were standalone audio CD recorders which only recorded to audio CD-Rs. No other devices give a shit. And yes, that flag meant that the MAFIAA had been paid off.

Re:Passed data with a ton of noise?

[Coren22]

You can actually hear the whore snorting coke in the recording studio bathroom, it is insane!

Re:Passed data with a ton of noise?

[EvilSS]

Sure, noise is the right term, when used properly. To imply that the noise is in the data, as they do, but 'you can't hear it', is quite ridiculous.

They being who ever wrote the summary because it is not implied in the articles (yea, I know, who the fuck RTFAs).

Re:Passed data with a ton of noise?

[publiclurker]

One of my cars only plays back music CD-R's. Any other media refuses to play even though it works fine on all other devices. I don't know if that is due to dumb luck or something on the media, however.

Re:Passed data with a ton of noise?

[ArmoredDragon]

The reason why it's not there is because ethernet is supposed to be just a dead simple layer 2 multi-access data protocol. Adding that kind of shit just contributes to latency (any kind of error correction involves additional parity bits and more processing.) And yes, I'll grant you that the added data and processing for error correction is tiny, but multiply that by a billion in large scale networks and you can see where there's a problem

If you need error correction, use UDP and handle it at the application layer, that way you aren't negatively impacting every other application that doesn't need error correction.

Re:Passed data with a ton of noise?

[LinuxIsGarbage]

While computer CD-R drives can write audio tracks to standard CD-R blanks, some set-top CD Recorders require "CD-R Audio" blanks. CD-R Audio blanks were more expensive so RIAA could get their cut, but otherwise used identical technology.

Re:Passed data with a ton of noise?

[Darinbob]

Ethernet at the MAC layer has digital data but the wires of course are all analog. If it wasn't analog there wouldn't be a bit error rate. The point was that even with a lot of noise the digital signal was still extracted correctly. The noise on the cheap cable was very bad, falling well outside the specs, and yet it still transmitted the data. The "noise" either causes the packet to be dropped or the packet is successful. The specification is about the limit of what is allowable so that it works even in a worst case scenario for what it is designed for, and they weren't doing anything close to having difficult operating environment for ethernet.

Re:Passed data with a ton of noise?

[Darinbob]

The argument is about people wanting to spend extra money without feeling bad about doing so.

Re:Passed data with a ton of noise?

[Baloroth]

One of the almost-plausible arguments is that a poor quality Ethernet cable can pick up analog noise which can leak into the actually-analog speaker output, since the Ethernet cable and the speaker cable go into/out of the same system. Like I say, it's almost plausible: noise from within the computer really can leak into the speaker output, and RF signals really can propagate through the Ethernet cable into the computer. The question is: does that actually make a difference, especially one that is in any way perceivable? The answer, as TFA found, is a decided "no".

Re:Passed data with a ton of noise?

[Aighearach]

Actually, in pro audio ethernet is used with proprietary protocols, handled by black box ASIC chips with special switches. I deal with this crap in the studio. Where I am they use it mostly for the personal mixers providing monitor outputs, but some places use it for inputs too.

Also, warning for those about to embarrass themselves making fun of "directional" cables, that means it is grounded at one end, and you put all the grounded ends in to the same device to avoid ground loops. If you don't know what it is, it must be brain-numbingly stupid... right? Ignorance is bliss.

We use normal cables, sure. But we do buy expensive ones with nice plug shielding, because musicians may or may not even be sober at work. Expensive cables isn't just for fancy looks, that build quality can make a real difference. Plus, it might be a lot easier to get the bean counters to agree to buy premium cables, than to get them to agree to replace equipment. They might tell you, "use the spares until they fail" in which case you'll regret not having spent their money on the "over"-priced ones.

Same with instrument cables. No, a brand new expensive guitar cable does not sound better than a cheap one. But after 300 shows, the cheap one craps out during a show or session, and the expensive "hifi" one didn't because it has premium long life rubber and better plug strength. So it does actually sound better once you factor in the way it sounds when equipment fails and you can't even hear the instrument. Most of that benefit is in the middle price range, of course. A $1000 guitar cord probably has metal mesh "shielding" that substantially increases cable strength, but the $300 one already has plastic mesh that will provide more than enough abrasion protection.

Also... some commenters don't know this, apparently, but a "placebo effect" is a real effect. It doesn't mean it is a scam, it means the people were successfully tricked into getting healthier faster, or in this case, to have more fun. If you "trick" them into thinking the artist is more artsy, they might enjoy it more too. Pretty snooty to claim they're not really enjoying the subjective aspect of their choices as much as they claim to... especially if you're also claiming that due to the placebo effect they really are enjoying it more!

Re: Passed data with a ton of noise?

[Aighearach]

If you wonder if ground loops might be a bogus concern, you've never worked in pro audio. ;)

Ask anybody working on the stage before or after a concert. They can set you straight on this one.

Re:Passed data with a ton of noise?

[Anonymous Coward]

I guess you just need to sell it without the fraud then. "Placebo Effect: $350" but I didn't see that on the ingredients or packaging.

Re:Passed data with a ton of noise?

[Aighearach]

I'm not sure about other systems, but most modern linux burners can set the flag. You can also flag them as audio masters, which is really useful for when you want to stick them into pro audio equipment.

IME if you set the flags right you can even get them to play in drives that attempt not to play any kind of CD-R. :)

Re:Passed data with a ton of noise?

[Ian A. Shill]

This. A thousand times this.

i have mod points, but i haven't been able to mod for weeks, but +1 insightful if i could now.

Re:Passed data with a ton of noise?

[ArmoredDragon]

I could see why they might do that, which is the same reason why jumbo frames only begin to be practical on gigabit networks. (There's just too much latency on anything slower than gig.)

Then again, 40gbit is typically optical, so there's practically no noise to begin with.

Re:Passed data with a ton of noise?

[Anonymous Coward]

You are stupid - nobody was talking one way or another about the durability of the cables. And yes, the placebo effect in the hifi market is a scam because it is not about helping people have more fun, it is about bilking them out of money. Learn to think you dunce.

Re:Passed data with a ton of noise?

[adolf]

You're full of shit.

CD-Rs are not bin-sorted in this way. The data that differentiates a "music" CD-R from any other CD-R is mechanically molded into the disc itself at the same that the disc is manufactured.

And so, by the time the molding is done, BEFORE the dye is applied, the sputtering of the reflective coating is complete, and the protective lacquer is applied, BEFORE a batch is ready to test before (possible) batch bin-sorting and silkscreening: The music CD-Rs are already irrevocably music CD-Rs.

The designation is part of the ATIP data. It is as set in stone as any non-recordable, molded/stamped CD ever was -- despite the rest of the contents being recordable.

Re:Passed data with a ton of noise?

[thegarbz]

Actually they knew exactly what they were doing. The Audiofool crowd love talking about how things sound despite things like specs, measurements, the laws or physics or anything else being against them. They started off by doing a blind trial. Right there and then it should have been over. The tests I think were done more for shits-and-giggles.

Re: Passed data with a ton of noise?

[thegarbz]

Almost but the idea is not to terminate it at the "far" end but rather the "near" end. The noise should be brought back to the low-impedance source. It's completely bogus in any residential or living room but not completely unrealistic when talking about long cable runs.

Re:Passed data with a ton of noise?

[lsllll]

How so? I do not believe audiophiles who spend extra money on an ethernet cable necessarily feel bad about doing so. I would actually suspect that they feel good about it.

In addition, my response was to the argument about data transmission correcting for error due to a bad cable, and that the argument still stands regarding possible noise inserted after the DAC. Like I said before, I believe that's highly unlikely, since noise after the DAC means that there's noise on the MB and there are plenty of more sensitive circuits on the MB susceptible to noise, but to summarily dismiss that theory (or via test results from 20 subjects who were primed anyways) is just wrong.

Re:Passed data with a ton of noise?

[Sardaukar86]

If you "trick" them into thinking the artist is more artsy, they might enjoy it more too.

Auto-tune proves your point. Justin Bieber, Kanye West, Brittany Spears.. the list of auto-tune 'artists' grows in number and nausea every year.

Re:Passed data with a ton of noise?

[ihtoit]

is there a "+1 Irony" mod? Because the Arctic Monkeys sound like shit anyway.

Re:Passed data with a ton of noise?

[ihtoit]

the differences between music and data CDRs:

1. the MAFIAA get the $2 per unit price difference
2. there is subchannel data on audio CDRs that if missing, will not permit standalone audio CD recorders to burn to them. For PCs this makes absolutely no difference whatsoever, since PCs simply ignore the subchannel data.
3. dye technology on audio CDRs limit the write speed to (last time I checked which was admittedly half a dozen years ago) 8x as opposed to data CDRs capable of 56x write speed. This is applicable to standalone as well as PC burners. It's the disc. Speaking of which, is Ritek purple still the go-to dye for DVDR?
4. Some stack players won't read data CDRs. If they don't find that ACDR subchannel data they won't read the rest of the disc.

Re:Passed data with a ton of noise?

[Anonymous Coward]

Gigabit (only BaseT, not fiber) ethernet has error correction inherent in the trellis coding used at the phy layer.

10GBaseT Ethernet (802.3an) uses a [2048,1732]2 LDPC forward error correction scheme. Fiber base GigE typically does not, except alternate encapsulated carriers like SDH VCAT or OTN (G.709).

G.975 for 10GBaseT over submarine cables uses a Reed-Solomon code with a 7% overhead. 10G WAN Phy has no FEC, 10G on G.709 OTN has FEC inherited from G.709.

In general 40G and 100G systems have Reed Solomon error correction on 25GBaud carriers, and not on 10GBaud carriers. Of course on G.709 there is FEC in the transport.

So really it's a mixed bag, but to say ethernet has no error correction is a false-hood, especially if you are talking about the most common layman GigE. It's likely in future that all higher baudrate fundamental carriers will have RS-FEC builtin, as they are running into limitations of the physical optical channel that can be mitigated with FEC.

Re: Passed data with a ton of noise?

[omnichad]

Yeah - that's why I said there's great reasoning. But it doesn't justify the cost - anyone can make a cable following that principle for a few dollars.

Re: Passed data with a ton of noise?

[omnichad]

The noise should be sent to the switch where no audio is playing. That's what I'm considering the far end. The noise won't do anything there.

Re:Passed data with a ton of noise?

[Anonymous Coward]

I'm going to go off on a hunch and say you have no effing idea what you're talking about. Let me explain.

"Actually, in pro audio ethernet is used with proprietary protocols"

Is it a digital protocol? If so, then who cares? It's either a 1 or a 0. As long as you can reliably tell which it is, the quality is unimportant. If they're sending analog down it, then they're idiots because Ethernet wires by necessity of fitting into the RJ-45 connector standard does not make a good medium for transmission of analog data.

"that means it is grounded at one end, and you put all the grounded ends in to the same device to avoid ground loops"

Ethernet by definition isn't a grounded protocol. if you attempt to ground one end, you're going to fundamentally screw up 0 detection on the other. Both ends need to know where 0 is, and this is achieved by using a floating ground. Granted, since we're talking Ethernet anyway, any ground would be an earth ground and relatively close to each other. That statement is literally non-sense. You'd never use an earth ground on one side and a floating ground on the other. You never ground on one side and not the other. Plus, a "ground loop"? What the hell are you talking about? Ground is where 0 voltage is defined. To imply a loop, that would mean 0 on one side was connected to something other than 0 on the other, which would lead to a feed back loop which would cause all electronics connected to it to let the magic smoke out. You'd burn everything out if you tried to do such a thing. If you're concerned about grounding non-connected items, you use an earth ground, as it's a constant and not relative. As a power engineer I literally have no idea what you're talking about. You want all grounds in the system to be tied to a constant voltage. It doesn't matter what that voltage is, just as long as it's constant across all connected nodes. if they're not all connected and you want different ground voltages, then you have to use electric isolation, usually in the form of either magnetic or optical isolation. But Ethernet doesn't support either of those.

The only valid point you make is better made cables. Everything else you said is what smug "in the know" fools say. Either it's digital or it's analog. Ethernet is a digital standard so as long as you can tell if it's a 1 or a 0, that's the end of the discussion. If you're talking analog, why the hell are you using a digital standard cable for it since they're terrible for analog signals?

Re:Passed data with a ton of noise?

[Daniel_Staal]

Except that the $300 cable isn't grounded on either end, and shows a high level of crosstalk. So the arrows on this cable are just to make people think it's worth the $300.

This cable is better than a $2 cable: It's well built, and meets the specs - barely. But you can get $10-$20 cables that are as well built, and meet the specs with less margin for error (these literally tested as 'within the specs' by less than the margin of error on the testing device) easily.

Re:Passed data with a ton of noise?

[whoever57]

Actually, in pro audio ethernet is used with proprietary protocols, handled by black box ASIC chips with special switches. I deal with this crap in the studio. Where I am they use it mostly for the personal mixers providing monitor outputs, but some places use it for inputs too.

I seriously doubt that pro audio provides a large enough market for ASIC chips, unless the chips were designed years ago. These days FPGAs have taken all the low-end market for ASIC and high-end ASIC have vast NRE costs.

As for proprietary protocols, why? The manufacturers might claim this, but how do you know? Using proprietary protocols when there are well established industry standard prototcols only increases price and very rarely provides any performance improvement. Again, with the size of the pro audio market, does this make sense?

Re:Passed data with a ton of noise?

[ChrisMaple]

Changing magnetic fields induce voltage along a wire. Two "ground" wires, traversing two different paths through magnetic fields between two pieces of equipment, can result in a current flowing around the loop formed by those two wires. Add more pieces of equipment, and each "ground" can be effectively at a different AC potential. This causes noise in analog signals if the signals are referenced to "ground".

... you use an earth ground, as it's a constant and not relative.

That is a fantasy. There is no one universal earth ground.

Re:Passed data with a ton of noise?

[TechyImmigrant]

That needs to go away. We need an Ethernet protocol extension with BCH or Hamming code support.

The 1970s called and wants its error correction scheme back.
Try LDPC or turbo codes or or Reed Solomon if you like it old school, but not too old school maybe mixed with soft decision decoding, or fancy DDFSE schemes.
But never meddle in the ways of MIMO, for therein madness lays.

Re:Passed data with a ton of noise?

[Sun]

The direction of the cable? Are you shitting me?

There's shitting you here, alright, but it's not jandrese that's doing it.

Shachar

Re: Passed data with a ton of noise?

[TechyImmigrant]

Such as the A, A not, B, B not. It is impossible to induce both a positive and a negative noise onto a wire at the same point in time, so as long as both signals are true it's a one, if not it's a zero. More wires, but zero errors.

Because it's massively inefficient. 4 wires to get data through which could each carry 80% of the goodput with 20% error redundancy. So that's 3.2 times better than your 4 wire scheme.

Re:Passed data with a ton of noise?

[gweihir]

This is _analog_ noise. Unless it exceeds a certain high threshold, it has no effect on data packets. It does allow one to gauge cable quality though. For example, if you have high, but still irrelevant, noise ion a 10m cable, you may be able to predict that you wil get actual digital errors on a, say, 100m cable.

Re:Passed data with a ton of noise?

[gweihir]

That is complete BS. The Ethernet data cables are differential and isolated via transformers. They are also twisted-pair. Nothing is going to leak from them. More likely a person touching the keyboard will already cause more EMI.

Re:Passed data with a ton of noise?

[gweihir]

Sure, if noise gets high enough. But at that point the cable will be close to complete failure. Below that level, you get a perfect digital transmission, and that is that.

Re: Passed data with a ton of noise?

[whoever57]

If you wonder if ground loops might be a bogus concern, you've never worked in pro audio. ;)

Ground loops are a real concern in many audio hookups, but not for these cables, which have no way for the shielding to connect. Only the twisted pairs can connect and none of these carry ground.

Re: Passed data with a ton of noise?

[TechyImmigrant]

Consult the relevant pages from the best book on electronic noise and how to get rid of it.
Noise Reduction Techniques in Electronic Systems by Ott.

Re:Passed data with a ton of noise?

[serviscope_minor]

That needs to go away. We need an Ethernet protocol extension with BCH or Hamming code support.

Ooh good call. But I suggest you go with the much more modern and effective LDPC codes. You can then call your suggestion 802.11n or 10GBase-T. Maybe you could go for simpler weaker codes like trellis coding. I think 1000Base-T might be a good name for your system.

Re:Passed data with a ton of noise?

[serviscope_minor]

The reason why it's not there

There is literally no reasy why it's not there by reason of it being, in fact, there. 1000 Base-T uses 5-4 trellis coding, already and the 10G over copper specs use LDPC codes.

Re:Passed data with a ton of noise?

[serviscope_minor]

As with many things there's a grain of truth, which is enough to get someone hooked. Oxygen free copper is a real thing and you can buy it in bulk. The main property is that it has far fewer small inclusions of copper oxide (hence oxygen free). While this does lower the elctrical resistance very marginally, that's not what it's for.

The problem with oxyide inclusions comes when working with compressed hydrogen. The hydrogen diffuses through, and slowly strips the oxygen from the copper creating tiny pockets of very high pressure steam. This causes serious embrittlement problems.

So unless you're operating your hifi deep on jupiter, in the native atmosphere, you probably don't need oxygen-free hifi cables.

Re:Passed data with a ton of noise?

[Alioth]

The question is: is the signal-to-noise ratio good enough? If so a cheap cable that passes the data is every bit as good as an expensive one, so long as the packets arrive intact at the other end.

Ethernet already does a lot to counter noise. The signals are differential pairs (so instead of having ground and signal, you have signal+ and signal-). The wire pairs are twisted, which keeps them in close proximity. Interference tends to be common mode noise (so for two wires close together it will affect the signal in each wire almost the same), and differential amplifiers are designed to only amplify the difference between the two wires and will therefore reject common mode noise. Each end also has an isolating transformer, and each end has proper termination (to avoid things like reflections which can bugger up signal integrity). It takes a significantly terrible out-of-spec twisted pair cable to make ethernet stop working.

Incidentally, the signalling for 100baseTX ethernet only has a fundamental frequency of 31.25MHz (naively people would expect 1MHz per 1Mbps but this is not so). 100baseTX uses a 3 level (in other words +1, 0, -1) non return to zero signalling (in other words, a 1 will cause the signal to change level and a 0 will cause the signal to remain at the current level - or it might be the other way around, it's a long time since I did this stuff). Each 4 bits is encoded into a 5 bit symbol designed to prevent long runs of 0s (which would cause the signal level to remain constant for too long). Lots of people call an ethernet connection a "broadband" connection, but it's not, it's baseband (hence the "base" in 100baseTX).

Re:Passed data with a ton of noise?

[Mr_Silver]

The fact that they even wasted time thinking about doing a listening test is enough data I need to know they don't know WTF they are doing.

That's absurd. If something claims to be awesome at doing X then the best way to disprove it is to test it doing X. You then publish the tests and the results.

Saying "well that's clearly stupid so I'm not going to test it" doesn't prove anything because you haven't actually debunked the claim - however right you might be.

Re:Passed data with a ton of noise?

[Mr D from 63]

The method of debunking the claim was quite stupid. In fact they were not, in the end, able to completely refute it because they too the wrong approach and played the game of the vendor.

Re: Passed data with a ton of noise?

[omnichad]

Why? Can't you just tell me what you want to say? I already know what I'm talking about.

Re:Passed data with a ton of noise?

[Opportunist]

Because then you know it doesn't work. Would you put some faith in the healing effect of something labeled "sugar pills"?

Re:Passed data with a ton of noise?

[iluvcapra]

I got into this with an audiophule type a few years ago. He, with a completely straight face, asserted that double-blind testing was an inherently flawed methodology for evaluating the objective marvelosity of some silly audiophule crap he was touting.

This has been a consistent argument from audiophiles for several years:

In short, for the important stuff, like "Do amplifiers or cables or differing storage media sound different", "blind testing" of any kind, single or double isn't likely to work because there are simply too many characteristics present and changing, and (if only because of the way human perception works) it's virtually impossible to isolate them and make sure that all of the testees are hearing the same test of the same thing in exactly the same way.

In short, "we believe in high fidelity but only in a purely non-falsifiable experiential sense." You can talk all you want about your error rates and THD+Ns but all they want to hear about is the "clarity," the "smoothness of the tone," and the "space around the instruments."

And these aren't crazy things to talk about, but insisting that a physical thing, a $300 ethernet cable, can actually create these things in a way that a cheaper one cannot is a kind of fetishism.

Re: Passed data with a ton of noise?

[TechyImmigrant]

>Why? Can't you just tell me what you want to say?
Alright. It's not just ground loops. Everyone who attribute electrical noise only to ground loops doesn't understand noise.

Ground loops matter, but the way the current and signal flow in the wire matters too. Ott's book pages 58 and 59 show 10 configurations, 5 ground loop configurations and five non-ground loop configurations, all with different noise suppression characteristics.

Cutting a ground loop will cut ground loop currents and the associated noise, but there's another 25db of noise suppression between the worst performing and best performing non-ground loop configuration.

The Wiki page is about ground loops, so talks about ground loops. That doesn't mean ground loops are the only consideration or even the most important consideration. If your problem is about noise in wires, then there's more to it than ground loops.

Re:Passed data with a ton of noise?

[Lord Flipper]

Same with instrument cables. No, a brand new expensive guitar cable does not sound better than a cheap one.

Although I generally agree, the fact is there are degrees of what constitutes "expensive" and of course, the nature of the guitar pickups, local environmental factors, and, last but not least, little things like anti-RFI on the guitars own guts, and the integrity of the amplification unit... all bear a huge responsibility as far as the presence, or lack of, "interference."

Having played electrics, more on than off, since about 1965, I've seen my fair share of musical instrument hype. And at times I've tried all sorts of devices, wiring schemes, ground lifters, and yup, expensive cables. And it usually boiled down to: physically moving around to find a sweet spot.

But, I am currently doing my limited "plugged-in" practice in a room with funky old electrical wiring, playing on a number of single-coil Fenders, through an old Ampeg that dates back to somewhere between '68 and '72... and the situation, powered up, whether the guitars volume pots are full open, or closed, is... whisper quiet. No 60-cycle hum, no RFI. Nada.

After all these years, frankly, it's a bit startling. The only actual circuit change? Klotz cables (their higher-end model). Silly pricey. No phony "gold" connectors.No ultra-thick wrapped copper (as in bogus "Monster" stuff). As a matter of fact, they don't even tangle like normal cables. And I'm using one to run into a small (5 or 6 unit) pedalboard, that has nothing but 9-volt batteries, and no ground lifting whatsoever, and another Klotz running out of that to the front-end of the VT-22, with precious little attenuation, and no "noise suppression" or "gates," at all.

Different strokes...

Re:Passed data with a ton of noise?

[Agripa]

Ethernet by definition isn't a grounded protocol. if you attempt to ground one end, you're going to fundamentally screw up 0 detection on the other. Both ends need to know where 0 is, and this is achieved by using a floating ground. Granted, since we're talking Ethernet anyway, any ground would be an earth ground and relatively close to each other. That statement is literally non-sense. You'd never use an earth ground on one side and a floating ground on the other. You never ground on one side and not the other. Plus, a "ground loop"? What the hell are you talking about? Ground is where 0 voltage is defined. To imply a loop, that would mean 0 on one side was connected to something other than 0 on the other, which would lead to a feed back loop which would cause all electronics connected to it to let the magic smoke out. You'd burn everything out if you tried to do such a thing. If you're concerned about grounding non-connected items, you use an earth ground, as it's a constant and not relative. As a power engineer I literally have no idea what you're talking about. You want all grounds in the system to be tied to a constant voltage. It doesn't matter what that voltage is, just as long as it's constant across all connected nodes. if they're not all connected and you want different ground voltages, then you have to use electric isolation, usually in the form of either magnetic or optical isolation. But Ethernet doesn't support either of those.

Ethernet is transformer isolated on both ends but if a shielded cable is used, then the shield connects to chassis ground on both ends. The problem occurs when chassis ground at each end differs which is not uncommon when dealing with the distances that long lengths of cable allow.

If a current flows through the shield then it couples to the ethernet signal lines but this would be unlikely to affect ethernet. When dealing with audio gear or sensitive instrumentation however, the ground current can cause problems whether it affects the ethernet interface or not and the common solution is to use an unshielded cable or ground the cable at only one end.

Professional audio gear uses balanced instead of single ended connections for audio to prevent ground loops; this is just an extension of that. It is also why USB and Firewire and not suitable in many cases unless a galvanically isolated interface is used.

Re:Passed data with a ton of noise?

[Agripa]

Idiot. A ground loop is *not* going to affect the quality of digtial audio delivered via an ethernet cable

This will almost always be the case however the problem is not with the ethernet signals being affected; it is with sensitive equipment like audio equipment at either end being affected. Balanced audio signals are commonly used where a single point ground cannot be guarantied and grounding a shielded cable at both ends would defeat this.

Re:Passed data with a ton of noise?

[Agripa]

The article covered this. There was no continuity between the connectors through the shield but they did not test to see if the shield was connected at only one end:

It is also possible—I have no handy way to test—that they've tied the shield to one end only, though this would be highly nonstandard for network cabling."

Re:Passed data with a ton of noise?

[ihtoit]

it's relevant for those who still use DVD for home theatre.

Thank you, come again.

Re: Passed data with a ton of noise?

[AgNO3]

They did this to disprove the company claims. Can't very well follow scientific method to disprove a claim unless you do the test that is directly related to the claim.

Re: Passed data with a ton of noise?

[Aighearach]

The thing is, audio isn't a digital circuit. It is a mixed circuit. The digital noise absolutely can and does leak into the analog signals.

Re:Passed data with a ton of noise?

[Aighearach]

Just walk into the studio and pop the fucking box open and look at it if you want to see if it is an ASIC.

http://yamahacommercialaudiosy...

Not a very good angle. I don't have any at home to take a pic.

You can forget about "these days" they've been making these for years. And yes, it is a huge market. BTW, FPGA is great for short runs, but when you're a major audio company selling large numbers of expensive units then ASIC is way cheaper. It isn't close.

And there are multiple companies with proprietary protocols in this space. They all use ASICs because microcontrollers are too slow, general purpose computing isn't responsive enough.

As to why, because incompatibility is brand lock. If it just pretends to be different but actually isn't, then plugging other machines in will work. Well, that is my cynical assumption. Also, in the case of Aviom (the one that is in almost every pro audio studio in the country) they use the ethernet physical protocol, and a proprietary daisy-chained network layer with redundancy. This allows for fairly robust behavior without any network administration required. You just plug the stuff in the same way you would with any other digital audio connection. And it shows up on the mixer in the intuitive place. Also, it stuffs a bunch of flags and crap into the packets. You can buy various controllers to remote manage stuff without learning networking, just how to stab buttons and spin rotary encoders. And there is a special switch that is also like a receiver and provides n digital audio outputs that can be selected from the 64 ethernet channels. So a standard studio setup you have maybe 8 individual monitor devices with ethernet, instrument in, headphones out with EQ, then a 16 port switch, then a 16 output receiver, which plugs into the mixing board. So you've actually got a dozen or so of those ASICs in each setup. And if you want to understand the network topology, just look at the cables. That is the topology.

In the end, why? Because you have one wire from each instrument position to the switch, and then only one wire from the switch to the sound shack. Then it goes back to one wire per track. That is actually a huge gain. If it required a network admin, or some kind of "computer guy," it would not be a gain.

http://www.prosoundweb.com/ima...
These things are ubiquitous. Sad but true. I preach standards at these guys all the time, and they laugh at me, "I already spent $10k on the Avioms, too late for that" ("Plus, everybody knows how to use them already")

Re:Passed data with a ton of noise?

[Aighearach]

I can't even get people to stop using those Morley Volume Wah pedals from the 70s. They needed two internal voltages (well, 3) so they used dual half bridge rectifiers with different outputs, and so the different parts of the load each only pull their power off one direction of the AC. Creates huge amounts of hum. People just wave their hands, "oh you can filter that out." Yeah, in theory, but then, why do I spend so much time listening to it coming out of the speaker until somebody bitches?

It isn't usually the old wiring, it is the old crap that people plug in. Most old equipment was well designed and doesn't introduce much hum. Some does.

Yeah, in a private vanity audio room you can just move stuff around. But in a pro audio situation, that is lost money, and the equipment changes all the time as you bring in different musicians. Having directional cables and planning where the noise goes prevents problems when somebody plugs in their precious POS pedal. It is almost always analog effects or old cheapo digital effects and not the preamps or instrument. Modern cheapo digital is no problem, because these days you just use an off-the-shelf switch-mode power supply. And even if they do build in their own power supply, 2 more rectifier diodes is cheaper than doubling the capacitance. Once upon a time though, diodes were expensive and power supplies were noisy. Yeah, everybody tried to blame the house wiring. But no.

On the road, ground loops are serious business because stage wiring is hell.

I know a guy with over $20k in audio cables, most of them are Klotz. He says the same thing, "they don't tangle."

Re:Passed data with a ton of noise?

[Aighearach]

Because then you know it doesn't work. Would you put some faith in the healing effect of something labeled "sugar pills"?

If the label says "sugar pills" then the placebo effect will likely not manifest except for illiterates.

Re:Passed data with a ton of noise?

[lucien86]

Then I suggest you do a course on network infrastructure. Ethernet is a low level basic transfer protocol, the error correction is all done in the layers above.

Re:Passed data with a ton of noise?

[bobbied]

I would say your "specs" are a bit different than the home audiophile's.

I fully get durability may be YOUR primary concern, in which case you want a cable that will work after being used and abused. For you, the connector quality, how fast the wire work hardens and breaks, and strain relief build are very important. But for most audiophile's, the issue is "how good does it sound". They are all about "my system sounds better because I used better components" and because they don't muss around with stuff all that much, durability is not that important.

Vendors prey on this "if it costs more it's better" mentality and sell essentially the same cable, off the same assembly line that meets the same specs in different packaging. That "monster cable" 10Ga stranded speaker cable really isn't twice as good as what you can buy from a home improvement store or even cheaper from an online retailer that sells 10Ga stranded wire. You can pay $100 or more for that name brand 10Ga stranded, or do the same thing for $10 (plus shipping). Audiophiles are routinely "sold" stuff (cables, power conditioners, components) with brand names that sound better, but where the specifications of the cheaper option are the same or better. This is most true with cabling..

Re:Passed data with a ton of noise?

[bobbied]

OR people who don't take the time to read the package....

Re: Passed data with a ton of noise?

[bobbied]

The thing is, audio isn't a digital circuit. It is a mixed circuit. The digital noise absolutely can and does leak into the analog signals.

then the equipment is badly designed...

Re:Passed data with a ton of noise?

[bobbied]

Oh, and one more thing, before I go..

Ethernet cables are transformer isolated at EACH end of the cable. Unless you are using CAT-6 shielded wire, the ground loop thing is generally rubbish as CAT-5 carries no ground or shield.

Now if you ARE using CAT-6, then you may need to worry about breaking the ground loop by disconnecting (isolating) grounds between pieces of equipment, I suggest you just break all CAT-6 grounds at the remote equipment and let the local safety ground in the wall plug serve for that. Now if that's a problem, you have a larger grounding system issue and will need to create a "signal ground" using some really low impedance cabling, start putting in isolated ground receptacles and paying very close attention to both the electrical code (for safety), isolating equipment from the racks they are in and establishing a set of "wiring rules" that avoid ALL Ground loops.

I've only seen ONE installation where the grounding of equipment was over complex, and it had multiple studios (8 total), where we had the top FM station for the market and produced and fed state and national news casts for smaller stations in the network. This was all on the 2nd floor of a building and we had a "signal ground" system of a dozen ground rods, wired together using multiple runs of 00Ga stranded welding cable. EVERY power outlet was isolated, every piece of equipment had a single low impedance connection to the signal ground and how all this happened involved a pretty complex set of rules. It was actually MY job to install all that stuff, so grounding is a subject I'm fluent in, or at least I was..

Re:Passed data with a ton of noise?

[ArmoredDragon]

There is literally no reasy why it's not there by reason of it being, in fact, there

No, it does not exist at layer 2.

1000 Base-T uses 5-4 trellis coding, already and the 10G over copper specs use LDPC codes.

...Which resides in Layer 1 of the OSI model.

Re:Passed data with a ton of noise?

[FreedomFirstThenPeac]

But the fact that they did not know reminds me that we are not them.

Re:Passed data with a ton of noise?

[Some_Llama]

i've been listening to this ethernet cable for about an hour now and i don't hear any "noise". Should i be holding the cable directly against my ear or some distance away to allow the "noise" to come to full effect?

so far all i've gotten is a tired arm and some sort of rash on my ear.

How much did they charge to test it?

[gurps_npc]

I am thinking $340 thousand to test it seems reasonable. After all, when I test a $3.40 cable, they easily spend 34 thousand to test it.

Mission accomplished

[Anonymous Coward]

They sold at least 2 of them now. I'm betting someone just won a bet.

Re:Mission accomplished

[Mike Van Pelt]

Argh, I tried to mod the above as "funny" and the mouse slipped just as I clicked. Somebody mod it as funny, please.

Don't buy the cheapest cable

[lgw]

This comes up whenever audiophile cables are discussed, but it's worth repeating: don't buy the cheapest cable.

There may be no useful difference between a $10 cable and a $1000 cable, but very often there's a real difference between a $10 cable and a $1 cable. Even for digital data, really cheap cables often don't meet spec, and can cause frustrating intermittent problems. You don't need anything exotic to avoid that, just avoid the bottom tier.

An example from my living room: I use a 45 foot HDMI cable to plug my TV directly into my HTPC (for reasons of convenience that aren't that interesting). The spec calls for thicker-gauge wiring for HDMI cables over 30 feet (IIRC), and you'll quickly see the price jump between cables that meet that spec and cables that don't. Don't buy the cheapest junk possible, that's all it takes.

It used to be that Dayton Audio was the only "solidly built, not too expensive" brand I knew about for cables, but Amazon changed that - now there are a bunch of options, including some sort of Amazon store brand that seems to be fine.

It's worth paying a bit more for solidly-built cables that meet spec (and especially for Ethernet cables, for some guard on the cable that keeps the clip from snagging or breaking off it you need to pull it through a tangle). Anything beyond that is a bit silly.

Re:Don't buy the cheapest cable

[Blaskowicz]

For mini-jack to dual RCA cable (line level audio) you can absolutely go for the cheapest cable. Sometimes a more expensive cable will have a complicated RCA connector that breaks down.

Re:Don't buy the cheapest cable

[Tailhook]

This goes for life generally; people that sort everything by price and buy the cheapest one are a liability to the rest of us.

Re:Don't buy the cheapest cable

[bobbied]

Oh for Pete's sake.... You Looney audiophiles need you heads examined..

Buy the cheapest digital cable that works reliably and don't spend a dime more. If offered a $45 HDMI cable over a $2 one, save your money and go cheap, heck by 3 of the cheap ones incase it breaks while installing it, you will be money ahead and you won't hear the difference EVER.

Now, on ANALOG cables, they used to have a point (albeit a barely measureable one). Case in point is speaker cables. For audiophile like performance, speaker cables ARE something that you can demonstrate matter. However, MOST people don't have the ears or the listening environment necessary to justify what some vendors confidently say you need. Most people are fine with 18Ga Stranded, especially for short runs and middle grade equipment, and I guarantee that 12Ga stranded will not be noticeably different for ANY home installation I can imagine, at least not different enough you can hear it.

When will this hold over from the "monster cable" audiophile fad be over? There was a time when that gold plate *might* have mattered to your analog cable so paying 20X the price was justifiable from the technical perspective, but now with digital cables all you are getting is fleeced...That and the justified scorn of those who really know what's going on. Well, that and bragging rights about how much you spent, if that matters to you..

Re:Don't buy the cheapest cable

[Anonymous Coward]

So the cheap cable didn't pass the spec test. What did they test for? The spec that was advertised. What was that? Cat6. What do you really need for GigE? Cat5e. Don't buy the cheapest cable? Bullshit. Buy the cheapest cable. It's GOOD ENOUGH.

We make fun of audiophiles, but the superstitions are strong in the IT field too. Gigabit Ethernet over Cat6 cable for audio streaming, and "don't buy the cheapest cable" is "insightful".

The one thing that does matter in Ethernet cables is shielding: Don't use shielded cables for connecting devices which depend on analog signal quality. You'll get ground problems, and shielding isn't necessary for short runs and outside of environments with heavy electromagnetic interference. No, your speaker cables do not produce heavy electromagnetic interference.

Re:Don't buy the cheapest cable

[Legionary13]

Agreed, and I have some data to support this. It’s bogus, but you can’t have everything. After noticing that my Ethernet switch and its wall wart were getting hot I ordered a replacement plus a few cables. They were brand name cables bought from Amazon. Not only is the switch faster but the cable to my main computer is red - bound to make a difference. On Wednesday, the first day the new equipment was in use, a cricket match started. For those who have not come across it, this is both a sport and a rain dance. Although I was working I kept a tab open with a cricket feed - and everything was happening far faster than anybody could believe. The match ended in less than three days - many fail to finish in the five days allowed. The red cable is probably the key component.

Re:Don't buy the cheapest cable

[lgw]

I've had cheap longer mini-jack cables fail - just break inside the insulation. I've had cheap RCA cables break, short, and most annoyingly have the center-pins break off and get stuck in my equipment.

Yeah, avoid the $40 job with the weird connectors, but a $4 patch cable can save a lot of headache over a $1 cable.

Re:Don't buy the cheapest cable

[Mr D from 63]

It is stupid tests like this by people who don't really understand the fundamentals of digital data transmission and processing that gives those audiophiles that don't know better excuses to keep the faith.

Re:Don't buy the cheapest cable

[lgw]

If offered a $45 HDMI cable over a $2 one, save your money and go cheap, heck by 3 of the cheap ones incase it breaks while installing it, you will be money ahead and you won't hear the difference EVER.

I hope you don't work with technology in any way. Sure, buy the cheapest cable that meets spec, but remember the first rule of engineering: the vendor is a lying bastard. There's a reason the cheapest cable is the cheapest cable. Paying $45 for a 6-foot HDMI cable is silly. Paying $45 for a 50-foot HDMI cable isn't.

Also, for HDMI specifically, the different numbered specs matter depending on use case. If your doing "4K" video, you'll want the HDMI 1.4 (or above) cable. If you want high color depth for a specific application, you'll want at least 1.3.

Sure, cheap is good, but as always in life, avoid the cheapest crap in the store.

Re:Don't buy the cheapest cable

[Anonymous Coward]

No, we're the people who are responsible for the availability of the $5 cable that you want to buy. Do you understand what would happen if nobody bought the cheapest item? The seller raises the price until the item is no longer the cheapest. But that makes somebody else's product the cheapest, so now that needs a new price tag. And so on.

Striving not to pay more than you have to is a necessity in a market economy. People who knowingly pay more cause misallocations of resources.

Re:Don't buy the cheapest cable

[Fjandr]

The cheapest cable is frequently not even close to good enough. I've seen numerous $0.99 Chinese cables that actually have a loop of the conductor that obviously was laying flat against the side of the mold when the injection process happened. As a result, it sticks through one of the molded connector walls. Just a bit of rubbing and you've got bare conductor exposed.

Re:Don't buy the cheapest cable

[avandesande]

Too bad none of the big box stores have the 10$ cable, usually it's 30$ gold plated thing that is a rip-off. I've had good luck get cables for 1$ @ thrift stores.

Re:Don't buy the cheapest cable

[occasional_dabbler]

Sadly I have no mod points tonight, but I have noted your ID and you will receive all of my next batch, even if you post 'moo cows' BTW you owe me a keyboard.

Re:Don't buy the cheapest cable

[Minwee]

This comes up whenever audiophile cables are discussed, but it's worth repeating: don't buy the cheapest cable.

I'm not sure if you read the follow-up article, but this bears repeating.

The cable that was used for comparison was the cheapest cable. In fact, it didn't even pass the Cat-6 certification tests done by Blue Jeans Cable after the even had finished.

But even with that nobody could tell the difference in the final sound quality.

Re:Don't buy the cheapest cable

[Legionary13]

Another spinoff: get a red cable, get funny/

Re:Don't buy the cheapest cable

[Slashdot Parent]

Even for digital data, really cheap cables often don't meet spec, and can cause frustrating intermittent problems.

I have also experienced this with uber-cheap HDMI cables. Tried to connect components with the cheapest HDMI cables possible (they came for free with some of the components that I bought) and there were problems. I forget what it was, but I think the TV wouldn't talk to to the Blu-Ray or the Tivo or something. I don't remember, but anyway, shit wouldn't talk to each other until I replaced the crap HDMI cables with monoprice cables. Still very cheap, but at least they are tested and meet the relevant specs.

Re:Don't buy the cheapest cable

[Slashdot Parent]

Buy the cheapest digital cable that works reliably and don't spend a dime more.

You and grandparent are saying the exact same thing. The point that GP was making is that some of the cheapest of the cheap cables don't even meet the relevant specifications and that that can cause problems.

You don't need cables made out of gold, but you often do need them to conform to the specs. I've had this problem with cheap as shit HDMI cables where my components wouldn't recognize each other until I replaced the cables with monoprice cables. So it's not like I had to spend a ton, but I did have to get actual certified cables.

Re:Don't buy the cheapest cable

[omnichad]

No - because they pick up so much noise. And the wire can literally break inside from being too cheap. I know, because I threw out my cheap one.

Re:Don't buy the cheapest cable

[rthille]

Pretty sure the HDMI specs don't say anything about the physical characteristics of the cable, only the electrical ones.

Re:Don't buy the cheapest cable

[Darinbob]

Even if you don't meet spec you will likely never see any problems with the cheap $1 cables unless you're in an extreme environment that's going to exacerbate the errors. The spec is like the "best by" date on food; it doesn't go from fresh and good to putrid poison on the date listed.

Increase the amount of electromagnetic interference and the cheap cable is going to notice it more, then wrap the cable up into a tight coil, stick it in a high temperature attic, save money by getting programmers to crimp on the connectors after 5 minutes of training, and use a hair dryer during operation, then you're going to see a more hostile environment and you may have to resend some packets. But use that cheap-ass cable that's way out of spec between two audio components in a specialized listening room then you're likely to never see a problem. Chances are any problems come from the cheap cable actually breaking rather than dropping enough packets to make a difference.

Re:Don't buy the cheapest cable

[Darinbob]

But you do pay more when it's a necessity. Which is why there are different categories of ethernet cables. In an environment and usage where Cat 5 cable is good enough, then it's wasted money to use Cat 6. But if you're doing gigabit of 10gigabit traffic, then you don't want Cat 5 or even Cat 5e.. There are other factors too, you may need a cable that doesn't degrade and become brittle because you're in a high temperature environment, you may need waster resistance, you may have fire safety issues, etc. Getting the most expensive cable there may be a good idea, but you're not going to buy those at the local consumer computer store either.

Re:Don't buy the cheapest cable

[adolf]

You don't need cables made out of gold, but you often do need them to conform to the specs. I've had this problem with cheap as shit HDMI cables where my components wouldn't recognize each other until I replaced the cables with monoprice cables. So it's not like I had to spend a ton, but I did have to get actual certified cables.

In days past, I've had issues with HDMI cables. And I don't mean some hodge-podge "the colors were much richer with the expensive wire, and the blacks were more black" I-just-spent-$900-on-a-bottle-of-snake-oil type of issues.

I mean real, tangible brokenness: Green screens, sparklies, failure to sync culminating in 15-minute-long cycles of hard-cycling the cable box, subsequent angry wife -- that sort of thing.

And I'm not even talking about long runs. 6 feet, tops.

So I went to Monoprice, as one does, and ordered a handful of differently-colored (because wiring is easier with colors) heavier-gauge ("premium") cables. And lo, the cables showed up and they were colorful and the wiring began.

Much to my surprise, the new Monoprice cables worked even worse than the old and crappy cables that barely worked previously.

My solution (being of an engineer's mindset) was to look at the new ones vs the old ones. The Monoprice cables were apparently well-constructed with heavy wire, and had ferrite beads. The old ones also appeared to be well-constructed, but with much smaller wire and no ferrite beads.

I decided to minimize these differences. And so, using a sharp knife and a hammer, I removed the molded-on ferrite beads from the Monoprice cables. And the thus-modified the Monoprice cables have worked marvelously and flawlessly to this day.

So is it a matter of meeting spec, or is it a matter of a product that actually works? Was the gear non-compliant, or the cable, or both? (Further: Which end, source or destination? Both ends? Which cable? Are both cables non-compliant?)

Footnote: Consequent to this, for my own purposes, I've taken to buying cheap used HDMI cables through Ebay or Amazon from folks who appear to be hard up for cash and are able to take accurate-looking photos. Haven't been let down yet, with my own pile of gear. For my day job, I order from Monoprice as a rule, and haven't had any issues with their "premium" cheap cabling outside of my own system...even for lengthy 45' runs with wire the size of a thumb. It just works, just not for me with my stuff. Deductively, my gear is the problem...but my gear is the expensive part and I'm not looking forward to replacing it any time soon.

tl;dr, If the cabling doesn't fucking work in an application, it doesn't matter how much it cost. And if it does work, it *also* doesn't matter how much it cost.

Re:Don't buy the cheapest cable

[lgw]

The actual spec is behind a paywall, as with most tech specs, but Wikipedia says.

cable of about 5 meters (16 feet) can be manufactured to Category 1 specifications easily and inexpensively by using 28 AWG (0.081 mmÂ) conductors.[107] With better quality construction and materials, including 24 AWG (0.205 mmÂ) conductors, an HDMI cable can reach lengths of up to 15 meters (49 feet).[107]

You may be right, and this is just the physical consequence of the spec, but 28 AWG is quite thin wire. (One poster said his long cable has a booster, so maybe that's another way, but that's not "cheapest" either).

Re:Don't buy the cheapest cable

[adolf]

For mini-jack to dual RCA cable (line level audio) you can absolutely go for the cheapest cable. Sometimes a more expensive cable will have a complicated RCA connector that breaks down.

Some cables are worth more, because of reliability.

My preference is the same as the broadcast standard (where failure isn't an option): Canare. Made in Japan, generally sold in bulk lengths or reels. They've got stuff for 75-Ohm video, speaker, a couple of variations on balanced (for microphones or other), and a cable made for electric guitars which I find particularly reliable.

It terminates like a breeze, isn't alarmingly expensive (though it's certainly not cheap), and interconnects made with it (even plain-old RCAs) tend to last for decades. It survives trucks being driven over it and touring duty of being used and abused on a daily basis, and lays flat by default so that it only becomes a trip hazard if you really try to make it one.

In contrast, I recently sold for scrap a pound or two of cheap/free-in-the-box dual-RCA cables. I was doing some work on my little home theater, and needed 24 sets (12 pairs) of RCA leads to integrate a new project I'd built.

Of course, I had a box full cheap/freebie 3-foot RCA interconnects, all used but neatly-sorted and wrapped, and most of unknown heritage. I dug into that and started hooking up my new kit.

Most of them had only one channel that worked. Some didn't work at all. And most frustratingly, I found that of the many dozens (hundreds?) of feet of cheap RCA cables in front of me, I only had exactly enough that worked reliably to get things connected: Some seemed to work but only did so part of the time, which made me question my own soldering on the project and made trouble-shooting a multi-hour process instead of a hook-it-up-and-go sort of thing.

A wire that doesn't work (or worse, one that only works sometimes) is worth less than no wire at all.

And wire that tends to work tends to cost also more than wire that tends to fail, because (if for no other reason) copper is expensive.

Does that mean that one should spend $340 on an Ethernet cable? No. Does that mean that one should spend $60 or $600 on a short RCA patch cable? No. Does that mean that $10 or $20, well-placed, might garner a much more reliable solution than a freebie RCA? A thousand times, yes.

Re:Don't buy the cheapest cable

[pz]

Definitely true that it's worth buying cables that you trust for reliability. I have worked in research labs all my adult life. We use gobs and gobs of BNC cables. I've watched countless researchers who don't know any better waste hours and hours of their time chasing down cable / connector problems. I use only ITT / Pomona BNC cables and have never, ever had a failure. Naturally, more of my budget goes to cables than others, but time is the precious resource.

Re:Don't buy the cheapest cable

[PraiseBob]

FYI - the gold plating doesn't improve the sound any, it's what the gold doesn't do that is important. Gold doesn't corrode or rust. It makes zero difference for the first few years, but after 10 years in a slightly humid environment, some cheaper plating materials will oxidize, while gold keeps a consistent connection.

Re:Don't buy the cheapest cable

[ihtoit]

yep, the one weakness on ethernet: those fucking tags.

The number of times I've sent cables back because they didn't have tagboots on...

Re:Don't buy the cheapest cable

[ihtoit]

I made my own. Solid-core RG58 to BNC or RCA for the win. 600-foot maximum run length per segment.

Re:Don't buy the cheapest cable

[drinkypoo]

It's awesome how you really are the stereotypical dead broke neckbeard on slashdot.

Getting just five used cables this way will save enough to pay for breakfast for two in a nice restaurant in Mendocino. You don't get to do fun stuff by throwing away money. I've saved hundreds of dollars just by hauling crates of cables around with me through my various moves, many of which came from thrift stores, yard sale, and flea markets.

If I need a cable right away, I don't look to see which flea market occurs next. I just buy it. But mostly I don't have to, because crates. Sorry you haven't got a place to store crates yet.

Re:Don't buy the cheapest cable

[Trax3001BBS]

It's worth paying a bit more for solidly-built cables that meet spec (and especially for Ethernet cables, for some guard on the cable that keeps the clip from snagging or breaking off it you need to pull it through a tangle).

You mentioned everything needed but one, the wire needs to be Copper and not Aluminum

 

Anything beyond that is a bit silly.

http://www.costco.com/WireLogi... see the arrows? Signal is best if the arrow points to the input :)

Re:Don't buy the cheapest cable

[gweihir]

Actually, on analog, gold plating _decreases_ the sound quality, as any crossing from one metal to another one does induce unavoidable noise. Audiophiles generally do not know that little physical fact, because they do not understand audio technology.

Re:Don't buy the cheapest cable

[Vlad_the_Inhaler]

Wait until a laptop power cable fails, at 240v. A smell of burning then the insulation burns through.
I was really happy to be right there at that moment, it took a couple of seconds to turn the thing off at the wall.

The Ethernet cable I had simply fail during normal operations was harmless in comparison.

Re:Don't buy the cheapest cable

[Vlad_the_Inhaler]

There are chemicals you can apply to plastic to make it less brittle, chemicals which are banned in most of the developed world because of their carcenogenic side-effects.
The computer magazine I read conducted a test of various components at the start of the year and had a very big surprise. I believe product lines were dropped.

Re:Don't buy the cheapest cable

[mark_reh]

And when you're out to eat at a restaurant, never buy the cheapest wine- go for the second cheapest! That way the girl you are trying to sleep with will think "he knows about wine so he must be cultured and that makes him a good choice for mating", or she will think "while he is conscientious about finances, he doesn't pinch pennies and that makes him a good choice for mating". It's a win-win!

Re:Don't buy the cheapest cable

[bobbied]

I meant "gold plated" as in "over designed" and not necessarily actual gold plating.

Re:Don't buy the cheapest cable

[bobbied]

I don't agree... You use the cheapest cable that operates as you want it to...There is absolutely no value in buying a "better" cable over the cheap HDMI cables in the bargain bin if both work.

If you don't need HDMII 1.4, and 1.1 suits your purpose, by the cheap 1.1 cable. If there are two 1.4 2 Meter cables and one is $40 and the other is $1.99 buy the cheap one. If the vendor is selling crap, take the cable that doesn't work back for a refund.

If you are worried about future applications of the cable, DON'T bother unless the "future" is pretty close. Electronics and advanced cables *always* get cheaper over time and it's rarely cost effective to buy more than you need now, over replacing it later.

And yes, one of my primary jobs is system integration. I design and build complex racked systems in the telco space which are high availability (5 nines) call processing platforms. So I professionally make cost verses durability decisions. In a home AV system, for the digital cables, cheap is just fine. What suffers is durability, but most home AV systems are not high vibration, and usually get upgraded often anyway.

Re:Don't buy the cheapest cable

[bobbied]

Dude, I'm saying you buy the CHEAPEST that WORKS and forget all the hype, especially when working with digital cables. So if you are looking at two HDMI 1.1 cables and one is $1.99 and the other is $40, you buy the $1.99 one.

Re:Don't buy the cheapest cable

[mattventura]

(Note: I'm in no way trying to defend such audiophoolery as buying expensive ethernet cables)

The reason why stuff can matter even in digital is because it does have to get converted to analog at some point, and if things aren't completely isolated, then interference in a digital portion of the system can leak over to an analog part. Probably imperceptibly, but still there.

Which makes me think: Why isn't anyone swindling the audiophools with some overpriced fiber ethernet gear?

Re:Don't buy the cheapest cable

[bobbied]

Why isn't anyone swindling the audiophools with some overpriced fiber ethernet gear?

Who says they are not? (grin) Actually I'll bet they already do, then charge them huge prices for "monster" fiber optic cable with FC ends on them... If not, want to start a business?

Re:Don't buy the cheapest cable

[lgw]

Oh, sure, if you don't value your time at all, that's the best strategy. Just keep trying cheap cables until one seems to work, and return the rest. Personally (and professionally), I have better things to do with my time than screw about with sketchy cables, but to each his own.

Re:Don't buy the cheapest cable

[bobbied]

If you don't know enough about what you are buying to KNOW if the cable's specs are good enough to work, then you deserve to be fleeced. If you are not willing to take back something because it didn't' work, you are asking to get fleeced... If you don't have time to fuss around, then just pay or installation and let the installer choose what you get (and will pay though the nose for).

I'm cheap. Personally, I use the "junk" HDMI cables at home and I get them for about $3 each when the local electronics retailer puts them on sale. Now I don't have ANY equipment that needs anything better than HDMI 1.1, but we've already established that for home use, I'm decidedly cheap. I work hard for my money and I'm not going to throw it away on something I can do cheaper.

Oh, and I've actually NEVER had a cable issue with any of the cheap HDMI cables I own. Like it or not, most of these things come off of the same few assembly lines in China now so for the same spec, one is as good as the next for the most part.

Re:Don't buy the cheapest cable

[lgw]

Sounds like you are, in fact, buying mid-tier cables on sale, not buying junk cables.

Re:Don't buy the cheapest cable

[rthille]

From HDMI.org (not certain that's the official standards site...)
http://www.hdmi.org/installers...

  Like all HDMI-enabled components, cables must be tested to meet the Compliance Test Standards set by the HDMI Licensing, LLC. Cables must successfully pass a signal of a certain strength (Standard cable must deliver a signal of 17Mhz; High Speed must deliver a signal of 340Mhz) to pass compliance.

The HDMI specification does not dictate cable length requirements.

Re:Don't buy the cheapest cable

[lgw]

But the laws of physics do. To maintain signal strength over a long run, you either need thicker wire or a booster (as well as wire pairs properly matched for impedance for the signal, as with any transmission line longer than 1 bit). Reliable, long (over 30 feet) HDMI cables do one or the other, as you can use very thin wires for a short run, so everyone does. Sure, it's not like ethernet, where collisions need to be detected within a certain latency, placing a hard limit on cable length, but you're still moving electrons in a wire.

Let's be specific here

[jones_supa]

The reason that the listeners could not hear a difference is not because the cables did not differ in quality, but because Ethernet is digital and has the capabilities of error correction and retransmit. The chip might have extra signal processing as well, to do noise reduction for example. In the test, these kind of characteristics were enough to fully compensate for the flaws of the crusty cable.

There's still many scenarios in which you can benefit from better EMI shielding and conductivity, even when talking about a digital application.

Re:Let's be specific here

[gweihir]

Bullshit. Unless you get dropped packages, it will make zero difference. If you get dropped packages, the cable is very near to complete failure.

Re:Let's be specific here

[gweihir]

None with regard to Ethernet used in a "normal" environment. Seriously, what you claim is religion, not technology.

Re:Let's be specific here

[jones_supa]

There's buffering which avoids that problem.

wait, what?

[Chris Johnson]

ETHERNET cable?

Oh come on.

The sad thing is, people will assume by this that people can't hear a difference between, say, speakers or playback devices. Way to just idiocrafy the world a tiny bit more, guys :P

Re:wait, what?

[Khyber]

So, what, you never heard of the Denon Link cable?

Re:wait, what?

[Minwee]

The idiots in question had this to say about the quality of their digital data:

"Extremely high-purity Perfect-Surface Silver minimizes distortion caused by the grain boundaries which exist within any metal conductor, nearly eliminating harshness and greatly increasing clarity"

"Sound appears from a surprisingly black background with unexpected detail and dynamic contrast."

"All audio cables are directional. The correct direction is determined by listening to every batch of metal conductors used in every AudioQuest audio cable."

If that's your starting point for idiocy, the only next step is to start painting equipment with magic symbols to repel gremlins and evil noise fairies.

Re:wait, what?

[Anonymous Freak]

They tested the $340 one because they weren't willing to pay for the $1000 "Ethernet audio" cable...

Audiophools are dumb.

Re:wait, what?

[gweihir]

That is an impressive collection of utter fail. Audiophiles seem to have an even larger percentage of complete idiots than the general population.

Re:wait, what?

[iluvcapra]

They tested the $340 one because they weren't willing to pay for the $1000 "Ethernet audio" cable.

Under normal circumstances a manufacturer would provide a sample for a media outlet. Audiophile gear manufacturers don't do this, for some reason -- reviews in audiophile mags usually seem to come from enthusiasts who've already bought-in, literally.

Right on time

[notdsk]

Slashdot.. the only place to read news... the next day

Why does ./ link to reviews from tech troglodytes?

[Anonymous Coward]

I mean it's a digital signal! If it fails the crc on receiving the protocol asks to re-transmit. The transfer might be faster, but definitely not more lossy.
There is no ethernet cable in the world which is sufficiently bad, that there are enough retransmits for mere audio to stutter or stall.
The whole review is asinine.

#thepowerofmarketing

[linkchaos]

In other news...... Buy our overpriced cables so you can listen to MP3s and pretend your really an audiophile. #thepowerofmarketing

Re:#thepowerofmarketing

[linear a]

You can upgrade the encoding. I'm re-ripping all my stuff at 112 now to replace those crapply old 96K MP3s.

Re:#thepowerofmarketing

[JaneTheIgnorantSlut]

The more expensive cables allow me to hear all of the rich colors of my music.

Whenever something is being sold with a made-up ..

[bobjr94]

...category from the seller (Cat 7 or Gaming Class RAM, JDM Type 2 Racing Intake), just walk away. Any tech person will know this cable is just a stupid waste of money. It's likely about a $8 to $10 cable, they don't need to sell very many to make a profit.

sometimes it seems to me

[The-Ixian]

that audiophile either means "wealthy" or "sucker" or maybe both...

I get that you want the best possible sound... and in some cases the placebo effect may actually help you enjoy your music more... but are there really enough of these people to base a business on?

I suppose, if you don't have to do anything except throw some gold plating on a connector and you are already in the cable business.. why wouldn't you?

This reminds me of another product, I think I saw on /, a while back... it was just a little plastic riser that kept your cables elevated off the floor and separated from each other... the cost was something like $100 per "device"...

Well... whatever floats your boat...

Re:sometimes it seems to me

[AthanasiusKircher]

I get that you want the best possible sound... and in some cases the placebo effect may actually help you enjoy your music more... but are there really enough of these people to base a business on?

If you're effectively making a cable that costs maybe $10 to manufacture, but selling it for $340, you don't need many "audiophiles" to make a significant profit. If you have a few hundred of them, you're already making 6-figure profits. (Obviously some cables may cost a little more to manufacture, but certainly not anywhere near as much as they are charging.)

It's kinda like wine. There have been studies that show that if you serve cheap wine in expensive bottles, people like it better. There have been studies that show that many wine prizes are awarded so haphazardly that you might as well choose them at random. There have been studies that show that actual wine judges at a major competition could barely rate the same wine with consistency above random chance on consecutive days.

And yet, people still will pay hundreds of dollars for some bottles. Recent studies have even shown that people literally get a better "pleasure" response in their brain when they are told a wine is expensive, compared to when it is supposedly cheap. It's more than a casual "placebo effect" -- it's something that people will pay hundreds of dollars to experience, even if most of that effect comes from the act of paying the hundreds of dollars rather than the product itself.**

I'm sure most audiophiles have a similar experience -- they literally receive more pleasure when they listen through an expensive cable. They want to pay more for that experience. So why not let them, I suppose? It's not like faith healers or psychics, who might do real damage with their charlatanism... the only damage these cable dealers could do, I suppose, would be with some obsessed audiophile who goes and throws his money away on expensive cables while his family starves. Maybe there's a couple people like that in the world, but it's certainly not a common problem.

And these sorts of "tests" won't convince anyone. I'm not sure what the point is anymore. It's like James Randi going after Uri Gellar -- true "believers" don't give a crap what the tests of "skeptics" say... they'll just keep believing. Let 'em enjoy their magic cables.

[**NOTE: To be clear, I am NOT saying all wine is the same. There are a lot of different varieties and flavors. But I do believe you should just buy what you like. There are $5 wines that have easily beat out $100 wines at blind tastings. So, if you like a wine and discover it's only $5, keep buying and enjoying it. If you like the $100 wine, and you like the taste enough to pay $100, fine.]

Re:sometimes it seems to me

[russbutton]

that audiophile either means "wealthy" or "sucker" or maybe both...

It means neither. There are plenty of wealthy suckers out there who have to have "the best" because it proves their wealth and how they're just better than everyone else. They're the same people who drive luxury automobiles, wear a $3500 Apple watch or show off their trophy wife. That doesn't make them audiophiles.

A real audiophile is someone who loves hearing music as real as they can make it and also values what a dollar can bring. It's the DIY guys who are your "real audiophiles". None of the ones I know use anything more than 14 gauge zip cord for speaker wire, but they build some really awesome loudspeakers for a fraction of retail. They also never set foot in Best Buy. Pay a visit to the Madisound web site, the Linkwitz Labs website, or Parts Express.

If you think good sound can be had from ANY set of headphones, then that's just ignorance.

Re:sometimes it seems to me

[olterman]

Audiophile basically means freedom of choice: you can pick whatever you want. Other than that, now it also seems to be about "warm sound" or just being different. For example: you get bored of CD/SACD/etc. sound and you want something different. People buy C cassette decks from the 70s and claim they are "a lot better than CD" just because they sound different. Even boomboxes give that retro feeling or bring up memories. Rarely is this about "having the correct sound": placing the same speakers in different position will create different sound. It's not about science either. It's a spiritual experience.

Personally, it would be nice to see graphs of different products where X axis is the price and Y axis is the "measured quality". Somehow I think the graph will look inverse U: costing "too less" won't have the essential parts in them and "costing too much" include the actual scam products.

Re:sometimes it seems to me

[serviscope_minor]

[**NOTE: To be clear, I am NOT saying all wine is the same. There are a lot of different varieties and flavors. But I do believe you should just buy what you like. There are $5 wines that have easily beat out $100 wines at blind tastings. So, if you like a wine and discover it's only $5, keep buying and enjoying it. If you like the $100 wine, and you like the taste enough to pay $100, fine.]

I keep hereing this, and where can I find these cheap wines? Or is it that plenty of $100 wines come with a lingering afterburn too?

Mod parent up.

[khasim]

I've had similar arguments with telco people. If the DIGITAL part of the system is not dropping (or delaying) packets then there is no problem with the DIGITAL part of the system.

Swapping cables that are not causing dropped/delayed packets for other cables that won't drop/delay packets is useless.

And testing the digital portion is very easy.

If you think you hear a difference, it is probably your imagination or the analog portion on either end.

Re:Mod parent up.

[Mr D from 63]

And testing the digital portion is very easy.

Yep, you just need a superwhamadyne gozinta gozouta comparator.

Re:Mod parent up.

[viperidaenz]

Unless the cable you used in the digital part of the system spews out so much EMI it could have a effect.

Re:Mod parent up.

[Aighearach]

Keep in mind that the "directional" cables are grounded at only one end, and you can't guarantee that digital and analog will have separate ground paths. They won't be separate, actually. So the noise from the digital system really does leak into the analog side. Most of that can be filtered out, but it isn't always easy to filter it just enough but not too much, in varying conditions that are only partially under control of the sound team.

For home use, perfect filtering should be easy, and problems are limited to start with. But in pro audio this shit matters. It is not a pure digital system.

Re:Mod parent up.

[CoderJoe]

From the cable tear-down, it does not appear that either end had any of the shielding connected to the plug shield.

With XLR cable connections, you have your shield, which IS connected at both ends, and your signal ground, which is isolated from the shield.

Re:Mod parent up.

[Sardaukar86]

And testing the digital portion is very easy.

Yep, you just need a superwhamadyne gozinta gozouta comparator.

Or, you know, Wireshark. But whatever spins your wheels.

Re: Mod parent up.

[driblio]

30ms of latency doesn't matter of you have a 3 second buffer

Re:Mod parent up.

[iluvcapra]

Keep in mind that the "directional" cables are grounded at only one end, and you can't guarantee that digital and analog will have separate ground paths.

The ones the audiophiles sell don't generally lift the ground on one end; also this is an ethernet cable which means it's electrically isolated, it usually doesn't have a shield and "signal" doesn't flow unidirectionally down it.

It makes sense to lift the ground on an XLR cable because in that case the cable shield is connecting either the audio or chassis grounds on two pieces of gear, but again we don't automatically lift the ground on the sending or receiving side, because it usually depends on wether or not both sides are audio ground, wether the ground is lifted in the box on one end or the other, and where this cable connects relative to where the ground stake is. Ground lifting is something you do once you've built the room, you don't just let the manufacturer do it.

Re:Mod parent up.

[Agripa]

From the cable tear-down, it does not appear that either end had any of the shielding connected to the plug shield.

I suspect the reviewer was not looking for this during the teardown and missed it if it was present.

With XLR cable connections, you have your shield, which IS connected at both ends, and your signal ground, which is isolated from the shield.

And equipment with XLR connections often has a "ground lift" switch to safely disconnect the ground and XLR ground isolators are available for equipment that lacks it.

On Stage With The Amazing Randi

[westlake]

You begin with a lecture to your Vegas audience of confirmed skeptics about the pseudoscience of high end digital audio cables ---- and afterwards claim with a straight face that confirmation bias didn't taint your so-called experiment.

The entire affair was inexcusable pop-science crap and wholly unworthy of Ars.

Re:On Stage With The Amazing Randi

[jandrese]

Found the audiophile.

Re:On Stage With The Amazing Randi

[JBMcB]

The inexcusable part is - as lousy as this test was - it was a lot more than any cable company has done as far as proving their cables make an audible difference.

You'd think a company hawking a $6,000 RCA analog interconnect would be eager to show how much better it is than even their own lower end cable.

Forgot the important part

[linear a]

Ya gotta paint the use magic marker to color the plugs green (not the contacts) to get the effect! Fools.

Noise also wasn't heard. Checksum still worked.

[medv4380]

The cheap cable may have had signalling noise on the line in some way. Improper shielding right next to an electric cable will do that. However, until the noise reaches a level that the error handling can't compensate you shouldn't notice it at all. The odds of you consistently passing Ethernet frames that pass the checksum so you'll hear the noise is unimaginable unless deliberately setup to do so. It's a simple checksum fail - resend frame. It'll cut down on your bandwidth with re-transmissions, but not on the quality. The fact that people don't hear the quality difference on WiFi is all the proof you need. The day my Ethernet line has more errors than my WiFi signal is the day I replace the cable.

Re:Noise also wasn't heard. Checksum still worked.

[bbn]

There is only a very small window where a bad cable gets you bit errors, but not so many that the networking becomes completely unusable. If just 1% of the packets are dropped, your effective TCP transmission rate will drop do 1 Mbit/s or lower.

If you are doing a file download from your NAS you will be doing a lot of 1,500 bytes ethernet frames. That is 15,000 bits per frame. If just one if those bits fail, the packet is dropped. Less than 1% of the packets can be dropped before you will notice that something is seriously wrong with your network, that means less than 1 error bit in 1,500,000 can be accepted. That is so close to zero, that in most cases you will go from perfect network directly to "this does not work" as soon as the signal degrades below the point where there are no bit errors.

Problem with blocking

[PerlPunk]

It looks like there is some unaccounted for variance in their design: "The listeners would be asked which audio sample (electronica, male vocal, female vocal, or instrumental) they wanted to audition. The requested sample would then be played through one cable, then we'd swap and repeat per the test protocol."

They should have either made people listen to the same audio sample or made everyone listen to all the samples.

Audiophoolery

[tlhIngan]

I've always wondered about people who buy these kind of cables. I mean, they're expensive cables, but what do they plug them into? Do they spend $340/$4000/$10000+ on a cable only to plug them into a cheap $15 D-Link switch?

I mean, what are the "audiophile" switches out there? Do they buy those $10,000 Cisco Catalyst switches? Or do they prefer HP ProCurve? Or do they just plug them in any old switch or whatever came with their $20 router?

It's just like power cables. You're telling me that the power, which came from a power station hundreds or thousands of miles away, travelling through copper wires, then coming into your house wired with regular Romex style house wiring, that some special cable used in the last 6 feet really matter? Or do they rewire their house with special audio quality wire? Do they buy special electrons from their power company? Or paid to have their house wired using the special cable? Are you telling me that after hundreds/thousands of miles, the last 6 feet really matter?

Re:Audiophoolery

[Fire_Wraith]

I use only locally generated solar power from unidirectional panels with audio-rated Monster cables to preserve smooth integrity of the power, connected to special reserve batteries with a non-volatile charging mechanism to prevent the introduction of harmful battery fluctuations that might induce noise artifacts into the components. I then had my entire house lined with a special AudioQuest aluminium sheeting to block any gamma rays or neutrinos that might strike a cable or component and cause stutter in the electron flow through the cables and cords.

Re:Audiophoolery

[Fjandr]

You need to upgrade your house shielding to prevent muons from wreaking havoc on the up/down quark balance in the helium-injected conductors. You do use helium-injected wiring, don't you?

Re:Audiophoolery

[Princeofcups]

If you have that much money, you are not doing your own wiring. There are audiophile "contractors" who will come in and install the most expensive setup possible. Since most expensive is better, right? These are the things install to justify the cost.

Re:Audiophoolery

[houghi]

I mean, what are the "audiophile" switches out there?

It depends on the salesperson and the budget.

At one place I worked some people wanted to have 4 PC's connected to 4 networked hardware machines. No Network, no Internet. Just the 4 machines on one side of the room and 4 machines on the other side.

Due to safety issues, running a cable from oe side to the other was not possible. So they asked their IT who asked their Network specialist from Cisco. I never heard exactlu what the solution was. When I came in all I heard they would not do it because of the price. I believe it was 3 or 4.000EUR.

So what I di was buty two switches for 100EUR together. Connected 4 machines to it. Connected 4 machines to it. The connection inbetween was a simple connection to the wallport that was already available on each side and then have IT patch them from one connection to the other.

I am sure other companies would have payed the premium price of several thousands of EUROs that were solved by not even 100EUR and I was not even looking for the cheapes ones. Just ones from the store that was open on my way home.

Re:Audiophoolery

[argStyopa]

"Or paid to have their house wired using the special cable?"

I really, really, really would like to find out how one starts such a contracting business.
I would be *delighted* to rewire your house with "cost is no object" cables. My service invoice will be comparable.

Re:Audiophoolery

[Knuckles]

RTFA really, you are talking about a totally different thing than this cable is concerned with - rightly or wrongly, most likely wrongly. It's not about the digital data. Such a cable would be used to, connect, e.g, the digital out of a CD player to an external DAC. What these cable guys are worried about - beside their wallet - is the analog connection. I'd be surprised if it made a difference, but you are barking up the wrong tree. As for your question about the price of the equipment typically connected by such cable: Yeah, you can easily get a CD player and a DAC for several thousand dollars each. And yeah, those things in an adequate environment do make a difference in sound, and if you never heard a really good hifi set you don't know what you are talking about if you think otherwise. So if you think about it that way, if you spent, say, USD 50,000 on a very good hifi set that does make a difference, are you going to worry about a cable for 340, how stupid it may be?

Re:Audiophoolery

[keko]

I've always wondered about people who buy these kind of cables.

It's from the analog days. When audio was transmitted the analog way, it made sense to have good sets of cables though the entire circuit, from source to speakers, and whatever you had in the middle (e.g. if you were into precise equalization). Main reason was cable attenuation due to intrinsic conductor resistance, that could *really* affect analog signals even in short distances, and of course signal/noise ratio.

So, the notion of "premium audio needs premium cables" is now somehow written into collective unconsciousness.

With that in mind, there's several reasons to buy a >$300 premium ethernet audio cable in 2015:

#1 you're ignorant
#2 you're ignorant and you've got money to burn
#3 you're ignorant and you want to show your ignorance off
#4 you're out of your mind. also you're ignorant.

See? There's plenty of people like that in the world. Best businesses margins are based in extreme stupidity.

Re:Audiophoolery

[gweihir]

And this is not the only long-term scam that rakes in a lot of money. There are a lot of basically stupid people that still manage to get a decent or even good salary. Society is broken that way.

Re:Audiophoolery

[olterman]

I don't understand many things. One is why do people buy expensive cars when they have will never get the real benefit. They have (or should) obey the speed limits. In some cases the cars might be faster but less reliable. Owning an expensive car gives the feel of power and muscle. Owning expensive audio rig gives the feel of golden ears. Just owning "personal" gear gives the warm and fuzzy feeling of being different in the era of perfect sound in the pocket. "Audiophoolery" is popular because of the many variables and personal tastes. It will never go away.

Re:Audiophoolery

[Fire_Wraith]

Certain expensive cars do provide benefit. It's not just top speed, it's acceleration. Having that ability to go from slow to fast, very quickly, provides an advantage in many situations, that if you've never driven one regularly, you probably wouldn't even think about. Also, you can take that car to a racetrack and let loose in a controlled environment, so at least you do have an option for it. (And that's aside from the fact that some of those cars are also rather visually impressive).

Moreover, some of those cars really aren't -that- much more expensive. You don't have to spend hundreds of thousands of dollars to get a car with some serious power and muscle.

The problem with audiophile gear is that it's largely not scientific - so far as anyone can tell, at least with these cables, it's a bunch of snake oil. A proper car analogy would be something like suggesting that you should only use CarQuest Tires for your car (which cost $10,000), and Monster Sparkplugs and cables because pseudo-scientific mumbo jumbo that's not really going to have any discernable effect other than in someone's mind. Now, replacing a 300hp V6 with a 500hp V8? Yeah, you might see a bit of a difference there.

Jitter

[tepples]

As long as bits are not irrecoverably lost, how is it going to sound any different?

Jitter, the relative timing of when bits arrive. Audiophile tests led to its having been identified as a measurable problem in S/PDIF installations, which led to use of a circular buffer to mitigate jitter. True, S/PDIF is connection-oriented, and networks using Ethernet are packet-switched with such a buffer being implemented in the network stack and in the application. TCP in particular retransmits packets corrupted by noise. But if the retransmitted packet doesn't arrive in time, emptying of the buffer causes an audible interruption. And if the buffer is enlarged to prevent this, this can affect perceived responsiveness in interactive applciations such as music production and video games.

Re:Jitter

[bbn]

You can not have jitter at the bit level and still receive the signal error free. It is ethernet and you will get your jitter simply because there is other traffic on the cable, so there is no guarantee that the line is free for immediate transmission of your packet. Even with QoS you will have to wait until the current transmitting packet is done.

But jitter is a non issue because all applications that receive audio via ethernet will have buffers to deal with it.

And last, because ethernet is a packet network, you will need to buffer your audio so you can collect enough bytes to send a packet. If you had no buffer, you would be sending packets with 1 sample (4 bytes assuming 16 bit stereo). The overhead of that would be astronomical plus you would indeed die the jitter death.
 

Re:Jitter

[tepples]

And last, because ethernet is a packet network, you will need to buffer your audio so you can collect enough bytes to send a packet.

True, but you usually don't want a 48 ksample buffer in a LAN environment because that'd introduce a second of latency.

Re:Jitter

[Anonymous Coward]

Most LAN environments are, at best, gigabit ethernet with jumbo frame support. Your 48k packets are too jumbo for that. (In DBZ terms, they're "over 9000!") A packet carrying a frame is not going to carry the entire buffer's contents all at once. That would make each packet way too valuable and a retransmission would "break the bank" so to speak. It would incur a huge overhead and introduce way more latency than a bunch of smaller failure-friendly packets would.

Also, did you know that WAV (and AIFF, with some minor differences) puts samples into "chunks" of 64k each? The playback buffer on a WAV file is 2/3 of a second for 2x16x48kHz audio.(AIFF's chunks are split differently, and it uses 4/3-second playback buffers, one per channel.) But you never noticed that, because a properly written playback system doesn't require filling the buffer before starting to play it, and latency on even crusty old PATA is fast enough to make this process nearly instantaneous. Likewise, a network-fed buffer won't have any immediate noticeable issue as long as the packets aren't delayed appreciably. Latency matters, while buffer size is secondary and likely doesn't matter to the big picture.

Re:Jitter

[leenks]

Only if the data was sent at 48ksamples per second. I would hope that on a gigabit network the data was sent to saturate the buffers as fast as possible and playback can start as soon as the buffer is full.

Re:Jitter

[tepples]

If you're listening to music, and it's got a 1 second buffer, you'll never know.

Unless the pause and rewind buttons start to get laggy.

this can affect perceived responsiveness in interactive applications such as music production and video games.

none of this is important when listening to music or watching a movie. And at 48k sample rate, you aren't talking Telco anyway.

True, music production and video games aren't telco, but they need low latency for the same reason as telco.

Re:Jitter

[gweihir]

Nonsense. TCP/IP always gives you a lot of jitter. The target application compensates for that. Audio is so slow in comparison to the data-cables and buffering is always used, that unless the cable is very near complete failure, you are not going to hear anything.

Re: Jitter

[iluvcapra]

"Also, did you know that WAV (and AIFF, with some minor differences) puts samples into "chunks" of 64k each?"

Um, no. The data chunk of WAV and AIFF files may be page-aligned, it may start on a 4k offset, but the chunk itself is contiguous.

Forward error correction

[tepples]

Wouldn't a layer of forward error correction using a BCH code, as used in Compact Disc Digital Audio and QR code, help with real-time UDP streaming?

Re:Forward error correction

[tepples]

ECC inside individual packets? Pointless. Every network standard worth its salt has at least a CRC32 covering the entire frame.

CRC32 is just error detection. It doesn't actually correct errors. Or are you claiming the packets would just get dropped along the way by intermediate routers that check the CRC?

Re:Forward error correction

[serviscope_minor]

1000 Base T uses trellis codes.

I'm not sure how QR codes would help however...

Re:Forward error correction

[tepples]

I mentioned QR code as a well-known application of BCH (aka Reed-Solomon) code.

Re:Forward error correction

[serviscope_minor]

Oh I see. I parsed the and "and" as listing multiple different FECs, not as listing different applications to BCH codes.

MP3, Vorbis/AAC, Opus

[tepples]

An upgrade from 96K MP3 to 112K Opus would probably provide substantial fidelity improvement to your CELTic music, as it's 2 codec generations later.

Latency

[tepples]

Fewer retransmits lets you push the buffer length lower for less audio latency. This can become important for real-time applications such as gaming and music production.

Re:Latency

[gweihir]

If you get retransmits because of the cable, then your cable is close to complete failure. Before that you, cables are not a significant source of retransmits. At all.

Where is the direction arrow?

[rlh100]

These can not be very good cables because they lack the direction arrow that the Belden audiophile Ethernet cables have (had?). This was so you would know which way to plug them in. Packets flow from hub/switch to the device.

And if you believe this, I have a bridge to sell you.
It is orange and you will make your money back in picture postcard royalties.

Networks stack Cables

[clockley(571021718)]

The speed of TCP error correction on the laptop and router and laptop have a greater effect on the audio quality than the cables. Why? Less likelihood of drooped frames by the software when errors do occur. (Even the premium cable was out of spec.)

Re:Networks stack Cables

[clockley(571021718)]

affect

Directional arrows aren't as silly as you'd think

[Theaetetus]

These can not be very good cables because they lack the direction arrow that the Belden audiophile Ethernet cables have (had?). This was so you would know which way to plug them in. Packets flow from hub/switch to the device.

And if you believe this, I have a bridge to sell you. It is orange and you will make your money back in picture postcard royalties.

It's in the caption of the very first picture:

Audiophile-grade "Vodka" Ethernet cables, from AudioQuest. They even have directional indicators!

But, surprisingly for Ars, they missed the point of those directional indicators. The article on electrical testing hints at it:

Finally, the braided shield inside the cable drew some comments. "There is no continuity from the body of the one connector to the body of the other, indicating that the shield has not been terminated to one or both of the connector," noted Denke. "Our 6A uses an absorptive shield—that is, the cable is shielded but the shield is not terminated at either end. Alien crosstalk is the crosstalk which occurs between cables, as opposed to the internal crosstalk which occurs between the pairs in a cable. This may also be why there are unterminated shields on the Audioquest cable—I’m not really sure what the reason is there, though I had thought that the shields on Cat 7 were required to be tied to ground. It is also possible—I have no handy way to test—that they've tied the shield to one end only, though this would be highly nonstandard for network cabling." (emphasis added)

It's highly nonstandard for network cabling, but highly standard for audio cabling - it's called a telescoping shield and is used to prevent ground loops and audible (60 Hz) hum. Typically, you leave the shield connected at the low-impedance source, and disconnect it at the high-impedance load... as a result, the cable actually does have a directionality, but on the shield, rather than the signal lines. I can guarantee that's the intent with these cables and why they're marked with directional arrows, and it's pretty surprising that Ars and Denke missed it. Maybe they were stuck thinking "network" cable rather than "audio" cable.

That said, because these are network cables, that telescoping shield is irrelevant. You're not going to get ground hum into your amplifier from your network card, the way you would with a shield on an analog audio cable. They're simply not connected, and if they were, you'd have much bigger issues - like that hum causing all sorts of problems on your PCI bus. This is why network cable shields are typically connected at both ends: ground loops are irrelevant.

Blindfold test

[PopeRatzo]

I'm starting to wonder if all the loud music when I was younger damaged my ears. Every time I turn on the radio, everything sounds like shit.

Re:Blindfold test

[JustAnotherOldGuy]

Same here. I find that if I turn the radio all the way down to the point where I can no longer hear it, that's when it sounds the best.

Obvious fail is obvious...

[Joce640k]

I think I see the problem: They didn't allow any burn-in time for the cables before doing the listening test - they just plugged them in and started listening as if that was going to sound correct.

A real listening test needs at least of couple of weeks for the atoms in the cable to settle down after moving/bending it in any way.

Re:Obvious fail is obvious...

[mwvdlee]

Poe's law applies here.
I honestly don't know if you're being sarcastic or audiophile.

Re:Obvious fail is obvious...

[I4ko]

I hope you are sarcastic, and all that read you, understand that you are. Even a 1Gbps Ethernet is running at mere 125Mhz. And for 8 years working in a telecom only once I had an Ethernet cable fail due to what I can partially atribune to wavelength fatigues, instead of the much more likely contact oxidation. This is a 2 ton glass CRT that needs to sit for 2 months after being installed to level out temperature stresses.

Re:Obvious fail is obvious...

[idontgno]

Or maybe the Futurama meme version.

Re:Obvious fail is obvious...

[thegarbz]

You've never talked to an audiophile before have you.

Re: Obvious fail is obvious...

[asliarun]

That is funny and actually true for many audiophiles. The burn in myth.

But i find it interesting that no one called out ars on their shoddy experiment. If you are going to bother going through a scientific ish experiment, at least do it in a better manner.

Someone who buys a "high performance part" would be doing it on a system with other components that are well built. DACs and preamps and power amps and sources with well built power supplies, components with matching impedances, high quality speakers or headphones that offer neutral and accurate audio reproduction. Such as Sennheiser 800s or even studio monitors.

And you use well mastered and well recorded audio that has enough instruments, enough detail, enough dynamic range, enough variety - that you are testing accurately and comprehensively.

If you are going to test a high end car part for example, no matter how hokum it sounds, you will still test it on a high performance car, on a track, and driven by pro or amateur racers. In other words, enthusiasts.

You will not likely put the replacement part in a corolla and ask someone to drive it on neighborhood lanes.

Apologies for the car analogy. But i find it disingenuous that no one has sarcasm and derision when people spend stupid money on cars, parts, components, etc. And there are ricers and there are serious performance enthusiasts, and there are people who will pay a million for a vintage.

But there is a special kind of sneering that happens only with high end audio equipment. I submit that there is a lot of snake oil, as it is in many other pursuits and hobbies as well. But getting accurate audio reproduction is extrely difficult and fiddly. The component setup is extremely fragile in terms of how small inocuous component changes do make audible differences. Good or bad. And that is what some take advantage of.

You can choose where you want to draw the line.. i.e. how much of an enthusiast you want to be. But as in other pursuits, for the true enthusiast, there is often no point of diminishing return. There is only the pursuit.

So be gentle, please!

Re: Obvious fail is obvious...

[tehcyder]

. But i find it disingenuous that no one has sarcasm and derision when people spend stupid money on cars, parts, components, etc.

Speak for yourself.

Re:Why does ./ link to reviews from tech troglodyt

[Minwee]

There is no ethernet cable in the world which is sufficiently bad, that there are enough retransmits for mere audio to stutter or stall.

I think you'll find that this very expensive cable has sufficiently poor quality that it will impact the reception of data.

Oblg. old. Audio Joke:

[UnknownSoldier]

There is an old joke in the audio industry.

* If you want to make a million, spend a million.

Works for crap such as Monster Cable, Bose, Beats, etc.

--
Married Audiophile Joke:
"When I die I hope my wife sells my speakers for what they're worth rather than what I told her I paid for them." :-)

Re:yup, that is exactly how to do it

[Coren22]

Personally, I would have run Gigabit transfers over the two cables and compared.

Re:Directional arrows aren't as silly as you'd thi

[omnichad]

I'm going to both agree and disagree with you. Yes, everyone missed that this introduced noise making it to the analog side is the reason for the shielding they use. It's not about missing bits that will be retransmitted anyway.

But on a cheap DAC, I get all sorts of PCI bus noise going through my speakers. Have you never owned a cheap computer where moving a wired mouse produces noise through the speakers? Or gotten coil whine from a GPU to introduce noise to the speakers? Obviously an audiophile will own better equipment, but there's theoretically a chance of interference.

Of course the solution for that is just just use optical cable to your speakers to prevent any possible analog noise.

Re:Directional arrows aren't as silly as you'd thi

[Anonymous Coward]

Have you ever looked at a typical RJ45? It is a plastic plug mating with a plastic socket.

There is no standardized way for a plug to pick up ground from the socket. There is no ground pin in the Ethernet RJ45 pinout. Neither side of the cable can possibly have a grounded shield.

Re:Whenever something is being sold with a made-up

[Anonymous Coward]

Category 7 and 7A are not recognized by TIA/EIA and are not real categories and are not used by any networking standards. TIA/EIA is currently defining a Cat 8 cable specification for use by 40Gbps being defined in 802.3bq

Of course there's hardly any difference.

[JustNiz]

They used the cheap $340 ethernet cable. They should have used the $10k one. Literally anyone can hear the difference with that one. i think its because it has electrolytes.

http://hothardware.com/news/10...

OK, let me settle this once and for all...

[Anonymous Coward]

I was an audiophile, worrying about the last 0.00001% THD and IM and noise and everything else, until a transducer Engineer (read: a professional who designs microphones and speakers for a living) casually mentioned that the very best, cost-is-no-object, hand-assembled microphones and speakers distort at least several percent.

Each.

Which means that no matter how great the ELECTRONIC portion of your signal path, the beginning and end have distortion measured in numbers with nonzero digits to the left of the decimal point. Doesn't matter how clean your electronics or special magic cables or whatever are, they can't remove the distortion that's already there in the original signal (unless your source is purely electronic) nor prevent your speakers/headphones from adding even more right before you listen.

After that, I paid attention mostly to noise floors and ignored everything else. I keep a lot more of my money and I have a lot more self-respect because I know I'm not wasting said money imagining that I can hear something that I cannot.

It's your money, spend it how you like.

Re:OK, let me settle this once and for all...

[gweihir]

Incidentally, you could have learned the same by having a look at a datasheet of a high-quality speaker or mic. But I guess audiophiles do not look at actual facts, as that could shatter their fantasy-world.

slashdot cheap shots

[russbutton]

Every time I see an article on slashdot about audiophiles, it's ALWAYS about stupidly overpriced cables. I think they're as silly as anyone here does, but dontcha think that this whipping boy is getting a little old? It's kind of like pointing fingers and laughing at Kim Kardashian who is famous for nothing more than being famous, or Donald Trump's hair.

W. C. Fields was right when he said there's a sucker born every minute and some of them throw their money away on fancy cable. So what?

I'll bet that most everyone here who dumps all over the fancy cables also has a crap audio system of their own, or thinks that they know what good sound is because they bounced for expensive headphones. Sorry guys, but listening to iTunes on headphones doesn't qualify you to declare your superiority about anything audiophile. A real audiophile isn't someone who lusts after $50,000 loudspeakers or buys the stupid cables we've been talking about.

A real audiophile is someone who builds his own power amps or loudspeakers. A real audiophile gets some microphones and goes out recording music at live events, brings it home and gets a faithful recreation on the home system. A real audiophile is someone who combines a love for music and a love for getting their hand "dirty" with the engineering of sound reproduction. That has little to do with stupid cables or excessively expensive gear of any kind.

If you want to experience startlingly realistic audio reproduction and not have to take out a 2nd mortgage on your home, check out the DIY designs at Linkwitz Labs.

Re:slashdot cheap shots

[ihtoit]

I'm not an audiophile (hate labels) but my idea of recording live is a stereo microphone on a boom made from a ten foot fishing whip, fed into the line input of a minidisc recorder. For the win. :D

Re:slashdot cheap shots

[russbutton]

Dude... A minidisc recorder? You can still get media for them? They haven't been current in 20 years. That's like running your web server on a Sun Sparc 2. Do you still run the Netscape web browser?

I don't know about the one-point stereo microphone though. Since you're obviously a headphones guy, have you ever looked into doing binaural recordings? Get yourself one of those foam heads and a couple of mike elements. Now hang THAT from your fishing rod and you'll be cool.

For the win...

Re:slashdot cheap shots

[ihtoit]

I have a dozen minidisc recorders and a case of blanks, so I'm good until they stop making AA batteries. You can't get media anymore, they stopped making blanks about two and a half years ago.

Are vinyl records still "hip"? Still got loads of those and spare cartridges for my player. As far as I know you can still get vinyl.

BTW, I still run my home automatics (lights and socket timers) on an Epson HX-40, which also happens to have been my first Z80-based computer (yep, got that one even before my first ZX81). My home webserver runs on a Dell Latitude CP, which is a Pentium II with a battery that lasts for frickin' ever. What more could you possibly need for a media portal?

Re:slashdot cheap shots

[russbutton]

I stopped recording on my Revox A77 open reel back in the 90s. I got some excellent recordings on it, but it had limited headroom and was a bit noisy. I now record onto a 6 year old laptop running Windows 7 and a USB audio interface at 96khz/24 bit resolution. I use inexpensive Behringer cardiod condensor mikes and then do a little EQ in post editing and get some very good results. Most of what I record is my wife's string quartet and sometimes my brass band, Horns a Plenty.

www.hornsaplenty.org

I still play vinyl as I've been into audio since Nixon was president. I started playing big band trumpet back then as well.

Being an audiophile means you have a deep emotional connection with the music and you want to recreate it as clearly and failthfully as can at home. It's not about having the most expensive gear. It's about playing back music in such as way as to recapture the experience and emotion from the original performance.

I think if I were to purchase most everything I have today, it would run me about $8k in total. No excessive cables. I run 14 gauge zip cord, but even then I run about about 200 feet of cable, so even with Parts Express 14 gauge zip cord, that's still a chunk of change. I run four 25 foot long cable runs on each side as I use 4 channels of amplification on each side as per the spec for the Linkwitz Orion loudspeaker design. The Linkwitz designs give you sound reproduction that is truly unsurpassed by anything anyone makes, at any price. I like the big Avalon and MBL loudspeakers, but they cost anywhere from $20k to $50k, not to mention the way expensive amps needed to drive 'em. The Linkwitz designs are things that us ordinary human beings can afford without having to sell your house. Do a Google search on Linkwitz LX521. You can't get better sound and this is something anyone who has a full-time day-job and a real love for music can afford.

Re:slashdot cheap shots

[ihtoit]

I had an A77! It was a Mk.III given to me by my old French teacher, stood me in good stead for six years until I lost it in a move. They're apparently making a bit of a comeback as retro audio gear goes, I wouldn't go out of my way to find one tho... I like my minidisc too much..

MOAR BUFFERS isn't always the solution

[tepples]

So I shoot someone in a video game, and I hear it one second later. Or the video and audio in a movie are out of sync by a second. Is either acceptable, and if not, what's the workaround?

Re:MOAR BUFFERS isn't always the solution

[leenks]

No, neither is acceptable in those cases, but the topic is audio playback in an audiophile environment, so data on a gigabyte network doesn't have to travel at 48kbit/sec - it can fully saturate the buffer almost instantly if necessary before playback starts as the whole listening experience was probably started only after several hours of procrastination and cleaning of equipment took place. The system would have enough time to transfer the whole FLAC 192kbs album before playback needed to start! For a video game and many applications, sure, the latency is usually acceptable (I'm a semi-pro musician and make use of soft synths, so I'm fully aware of the effects that even 30ms of latency vs 10ms can have, let alone the 500ms or so that pipe organs can sometimes exhibit).

Hard hitting stuff

[nysus]

What's next? Earth isn't flat?

You don't understand "ground loop".

[jeffb (2.718)]

You should consider the possibility that electrical engineers who regularly use the term "ground loop" just might understand the concept better than you do.

Run two identical wires down to your "green wire buried in the dirt".

Run some current along one of the wires.

Observe the voltages at the near end of each line. You'll find that the line carrying current is no longer at "zero volts" relative to the line not carrying current. That's because those ground lines are not perfect conductors.

if the electric noise is detectable

[superwiz]

that's enough. All it takes now is a few secret deals with high-fidelity audio manufacturers to intentionally degrade performance of their equipment when electric noise is detectable (even if it can be compensated for). The argument will then automatically become that you can't hear the difference because you are not using top-quality equipment in the 1st place. And then the manufacturers of the cables will be able to peddle it to everyone buying top-quality audio devices just because the devices will seem to need them. I am actually curious (no, I have not read the article) whether the cheap cables can still sustain the required rates. In other words, can they still sustain 1Gbps transmission between 1Gbps eth cards? Cat5e definitely CANNOT. It will top out at around 350-400Mbps. If these cheap cat6 cables have too much noise they can't guarantee 1Gbps. It may not matter to those using them with household devices, but it definitely matters to people have quality of service contracts which require them to pay when they can't supply a promised level of performance.

Re:if the electric noise is detectable

[Alioth]

It's entirely irrelevant when we're talking about 44.1kHz audio, which is what we are. The cable would have to be ghastly to the point of basically not working at all to bugger up 44.1KHz audio, even uncompressed (which is only 1.5Mbit/sec)

Re:if the electric noise is detectable

[superwiz]

You are not getting it. It's not about bandwidth. It's about latency. Correctly lost information requires round trip. If the drop rate is so high that it requires a few round trips, it can cause some of the info to arrive too late (even though a lot of the info would be arriving on average). Never underestimate the bandwidth of a truckload of backup tapes moving at 60mph on a highway, just don't forget it's latency.

Re:Why does ./ link to reviews from tech troglodyt

[superwiz]

There is no ethernet cable in the world which is sufficiently bad, that there are enough retransmits for mere audio to stutter or stall.

Oh? Why not? Most people will be able to hear 1/50th of a millisecond of missing data. Regardless of bandwidth, if your transmission drop requires a round-trip re-request of data, the latency of the connection can delay the arrival of the data past the point where it is needed to be played.

Wong place to spend money

[Air-conditioned cowh]

Digital audio has no "sound". It is where it becomes analogue and what happens after that that matters. Assuming no data errors, the only way anything in the digital chain affects the sound quality is by the interference and jitter it induces on the DAC and analogue components. This can be quite noticeable (and measurable), if you've ever heard the background mush from a cheap MP3 player or some mother board sound outputs through headphones, but a decent DAC would be well isolated and the clock would be at the DAC, and preferably be the audio master clock and not just phase-locked to something.

Re:Wong place to spend money

[gweihir]

That is utter BS. A dedicated crystal produces no better or worse clock than a PLL clock derived from a crystal input. You are calling all digital engineers clueless hacks here, when that really applies to you.

Re:Wong place to spend money

[JustAnotherOldGuy]

^^^ THIS

It's amazing how little people know about stuff like this, and yet how certain they are that they do.

You're correct- a dedicated crystal produces no better or worse clock than a PLL clock derived from a crystal input. If they thought about it for a moment they'd realize this, but that would require thinking about it for a moment.

Re:Wong place to spend money

[gweihir]

Indeed. And it would require some actual understanding of how things work.

Re:Wong place to spend money

[iamacat]

If there is any packet loss, retransmissions are needed. No problem for a single, audio-only link where the receiver can buffer data. If you also have video or multiple audio links, simply dropping the packet is probably a better option that breaking video/audio synchronization. For non-interactive content one could design a protocol with very precise clock synchronization, where each device will play a given frame of buffered data simultaneously. Have never heard of anything like this implemented in consumer electronics though.

Re:yup, that is exactly how to do it

[gweihir]

And that has what relevance for audio? Right, none at all.

Re:Networks stack Cables

[gweihir]

Seriously, you have no clue how TCP/IP works. Case in point: TCP has absolutely no error correction. At all. Your statement is complete and utter BS.

TCP has retransmission, but for that to be needed over a single cable hop, the cable needs to be close to complete failure. It basically does not happen. Basically all bit-errors are introduces by broken Ethernet cards and switches. A lot of the drops are caused by overloads and are intentional drops under software/firmware control.

Re:Wtf?

[gweihir]

It is a system interface test. Quite standard. Requires some actual engineering knowledge though to understand why these are sensible.

Re:This test was flawed

[serviscope_minor]

Using error correction to achieve 0% loss over cheap ethernet cables is cheating.

Cheating how? This is a test of network cables. If the cable gives you 0% errors by any usable meant then it works perfectly. If I want my 60 kbit MP3s transferred losslessly over the network, I really don't give a crap if it's using trellis codes or retransmits.

Not this shit again...

[JustAnotherOldGuy]

If there's a more gullible group of people on this planet than audiophiles, I've never met them. I suspect it's the same on other planets as well.

I mean, who else would pay $400 for a wooden knob that supposedly makes the sound better? Or colored pebbles in a jar that (again, supposedly) "purify the tonal balance" of the room the music is played in?

Anybody really surprised?

[DiEx-15]

I'm not. Selling an expensive POS for something that is at it's very nature subjective only proves it's just for bragging rights.

I'm sure that there are a few Audiophiles that will screech that X makes Y sound better, or X is better for data transfer. However, no amount of empirical data can change the immutable fact that "sound quality" is a subjective observation. What one person thinks is "sounding great" might not for another.

Sure, there is some general agreement that an audio track that sounds clear is way different than one that sounds like a jack hammer got at it and tossed into a blender. Again though, there is going to never be a consensus with people that hear the same audio file but the only difference is something extremely trivial (and ridiculously expensive).

Re:This test was flawed

[TsuruchiBrian]

I was hoping the facetiousness of my post would be obvious.

Re:Directional arrows aren't as silly as you'd thi

[Agripa]

Unlike Unshielded Twist Pair (UTP) which uses an RJ45 style connector with a plastic body, shielded twisted pair (STP) uses an RJ45 style connector with a metal body that connects to the cable shield.

Re:Directional arrows aren't as silly as you'd thi

[Agripa]

.That said, because these are network cables, that telescoping shield is irrelevant. You're not going to get ground hum into your amplifier from your network card, the way you would with a shield on an analog audio cable. They're simply not connected, and if they were, you'd have much bigger issues - like that hum causing all sorts of problems on your PCI bus. This is why network cable shields are typically connected at both ends: ground loops are irrelevant.

The ground loop formed by using ethernet over shielded twisted pair is not irrelevant if it corrupts the chassis ground which is also the shielded side of any singled ended signals like RCA audio. It would be very unlikely to affect the ethernet but it could sure screw up sensitive single ended signals.

Re:This test was flawed

[TsuruchiBrian]

woosh

Re:This test was flawed

[TsuruchiBrian]

Yeah, and it's so amazingly ridiculous that it must be completely serious. Jesus...

Re:Error correction?

[Some_Llama]

I have had the same rare chance of failure with Ethernet cables but weirdly enough i just happened to encounter such a failure last week, the cable looked fine, and it was transmitting but i was getting 40-50% packetloss, when i swapped out the cable the problem went away.. we had just gone through a cubicle shuffle and i believe the cable was handled roughly in the move and hurt the wires in some way.

First time i have encountered an Ethernet cable "fail" without an obvious show of defect (broken plastic covering/wires pulled free of their crimping, etc).

Re:Whenever something is being sold with a made-up

[Some_Llama]

RAM has latencies that make overclocking possible by upping the speed of the ram past it's normal rating to match the OC of the cpu, so yah, gaming class RAM is legitimate if it has lower CAS latencies.

Re:Directional arrows aren't as silly as you'd thi

[Some_Llama]

the company that sells these say that the arrows are determined by testing the cables to determine which end to put the arrow on from the basis of best audio quality, this would denote that it is not from some special shielding or construction but pure bullshit.

Re:This test was flawed

[Some_Llama]

nope, your sarcasm has evolved to a near perfect form. I frequently have the same problem :(

smiley faces help but defeat the purpose :)