Audiophile Network Cables..... REALLY...REALLY!

[blacsno]

Originally Posted by cleeds

Antonmb, I agree completely. The "test" also had has numerous other shortcomings. For example, it says "interested potential test subjects screened for suitability basically making sure the person ... didn’t appear to have an agenda." Those could be some of the most valuable subjects to test - experienced, well-trained listeners - and any bias they might have would be negated by a properly designed and conducted blind test. That is one of the very raisons d'etre for blind testing in the first place - separating bias from result! Oops!

Instead the "test" organizers promoted their own bias. Prior to being selected, subjects were instructed that audiophile network cables were "magical" and pseudo-scientific. Oops!

The organizers had planned to test 20 subjects. But, after testing only seven, "According to the test criteria, this many misses counted as a meaningful result, and the test was terminated." But the same statistician also concluded - after the test was complete - that gee, once he really really thought about it, the test's equating an inability to detect a difference with a "fail" introduced another variance into the test. Ooops!

I could go on, but I won't. The test proves nothing. Blacsno says the very notion of audiophile network cable fails his "giggle" test. That's ok; some people think South Park is funny.

I don't know if audiophile network cable has value or not. But neither giggles nor pseudoscience will resolve that for me.

At least your suspicious of the 1/2 off "Jones town" kool aid on the clearance rack.

Ken

[blacsno]

Originally Posted by BlueFox

I think one issue is some people confuse a correct CRC with musical accuracy. Data with induced jitter will still give a correct CRC, but not sound right. Sort of like a turntable that has a very slight speed variation.

YOU'RE absolutely RIGHT! The thought of a bling bling cable will fix this is nonsense! If the turntable "CPU(s)/Router" is off speed "CPU(s)/Router processing errors no matter how miniscule" the music is incorrect end of story.

Ken

[blacsno]

Originally Posted by chessman

Ken, I am ok with confirmation bias as long as my ears consistently fool me.

You do, however, make several spot on points. I will have to be satisfied with being confused.

I'm just trying help as a member of this fine site. There are many aspect about this hobby I just don't understand. Time here on Audio Aficionado has greatly help.

Ken

[Orb]

I think a good real world example is when Hi-fi News reviewed both the Melco N1Z and N1A.
When used with its network setup the measurements showed minimal improvements, in fact the only improvement was over jitter that was still very low.
The review done with the N1A went even further and to try and create a situation where the network functionality showed a slight improvement he had to change his network infrastructure to a more noisy/basic setup (using the router hub rather than a quality switch).
This suggests an Ethernet cable is not really the challenge here, although it does make sense buying a well built-shielded one but this does not need to be incredibly expensive.

Just to add the Melco's do show a big improvement with its USB setup when compared to a PC/MacBook (using USB).
Unfortunately neither of these reviews with their measurements are online (did a quick search).
Cheers
Orb

[Antonmb]

Originally Posted by AudioNut

People who don't believe that there are differences in wires probably don't believe there are differences in ropes either . . . There really aren't any differences are there Tony?

Thanks for pricking my bubble Julian. I do believe there are differences in most cables (and ropes), although I'm not entirely sure about ethernet. What I also believe is that, if you're going to claim scientific validity for test results, then the test should be rigorous and scientific and not just staged, biased theatrics. I also know that you can't test for everything, even with rope. You can test all you want, but often the only way to know for sure how a product will perform, is to put it in the hands of users and let them work with it over an extended period of time in their system - and then trust the judgment of experienced, trained users.

[BlueFox]

Originally Posted by blacsno

YOU'RE absolutely RIGHT! The thought of a bling bling cable will fix this is nonsense! If the turntable "CPU(s)/Router" is off speed "CPU(s)/Router processing errors no matter how miniscule" the music is incorrect end of story.

Ken

I think you misunderstood what I was saying. The price of a cable that will not introduce jitter is another matter. I agree it probably won't require three, or more, figures.

[Glisse]

I would not rely on the Ars Technica site, or its personnel, to dismantle my power amplifier and tell me whether it represented good value or not. This is a click/bait exercise for them, which seems to be successful.

No offence to the OP, but the people who seem most vocal on this issue (the value or not of cables) tend to have seldom done any comparisons between cables. And by comparison, I mean listening, not cutting them apart.

There is no accreditation system for ethernet protocol standards, just because a cable has 5e/6/6a/7 printed on it doesn't mean it will meet those standards.

The role of shielding, specifically the differences in requirements between network appliances and audio systems is significant. Forget about jitter and transmission errors for a moment. Cables such as Cat 6A with shields connected to ground at both ends may dump noise into your system: it was interesting to see that AQ had floated the shield ground at one end of the Vodka cable. How many people here are using network appliances, such as switches, routers, NAS etc with grounded connections? I suspect very few to none. Where does the shield noise then go?

As an aside, get rid of the cheap SMPS used by switches, NAS, routers etc, that are on the same circuits as an audio system. This is the primary source of much of the electrical noise piggybacking on ethernet cables - RFI/EMI being the rest.

Ars did not include an AQ Pearl cable as well. The Vodka is AQ's second from the top ethernet cable - very much overbuilt with serious connectors, thick silver coating over premium copper wire, etc, etc. Pearl is not much more than Blue Jeans Cable (who just terminate Belden or Canare raw cable, they are not cable designers). Yet Pearl is guaranteed like all AQ cables. Guess this would not be so "newsworthy"?

On my own anecdotal experiences, and I now use ethernet connections in all of my audio systems, about 25% of the time I can hear a difference between ethernet cables, the rest none. There is no "rule", or definitive yes/no scenario here. I don't have any expectation bias one way or another. But I very much doubt that if I listened via headphones connected to a Dell laptop that I would hear any difference between anything at all. That the test itself was rigged was hardly a surprise, given those involved and the agenda.

Like anything in audio, to see what works, and what doesn't, JUST LISTEN

[aqman]

It would seem any test of an Ethernet cable should include a measure of number of dropped packets or number of packets that had errors. Since even dropped packets are re-sent this should not cause any data errors this would not be a problem unless cable was really screwed up. Therefore if it sounds different it would likely be noise coupling into the system somehow. Then it seems you are also testing the Streamer or whatever's ability to not pass along or be affected by this noise.

In any case, Shielded twisted pair cable is available from mono price for a few dollars. So what is Audioquest or whoever really selling you? Do they even know? Or care?

I remain skeptical, but admit I haven't compared Ethernet cables in an audio system.

[Masterlu]

I know they use their own DBS Dielectric Bias System using 72v



http://www.positive-feedback.com/Iss...st_diamond.htm

[Orb]

Originally Posted by aqman

It would seem any test of an Ethernet cable should include a measure of number of dropped packets or number of packets that had errors. Since even dropped packets are re-sent this should not cause any data errors this would not be a problem unless cable was really screwed up. Therefore if it sounds different it would likely be noise coupling into the system somehow. Then it seems you are also testing the Streamer or whatever's ability to not pass along or be affected by this noise.

In any case, Shielded twisted pair cable is available from mono price for a few dollars. So what is Audioquest or whoever really selling you? Do they even know? Or care?

I remain skeptical, but admit I haven't compared Ethernet cables in an audio system.

Just to clarify, packets are only retransmitted based on the higher protocol and not Ethernet, this means it would rely upon TCP, but unfortunately the streaming itself uses UDP.
So application and session control would use TCP, but the actual audio stream would be UDP and without same correction integrity.
TBH this should not be too much of an issue for a wired connection (maybe some influence if using a cheap and basic hub router with everything going through that *shrug*), but can have adverse effect when using wireless (for many reasons including shared contention and signal quality-strength).

Cheers
Orb