Archimago's musings on ethernet cables and audio

I’ve seen so much stuff spouted on here about ethernet and how music is different from data, how optical fibre is “audibly” better than copper, yada, yada, yada.

Found this old (but very relevant) article comparing ethernet cables and their effect (or lack of it) on audio and felt compelled to share it:

11 Likes

Could you summarize his conclusions?

Different ethernet cables make no difference to the sound

16 Likes

Ill grab my popcorn whilst waiting for the Placebo Club for Advanced Auditory Capabilites to arrive.

22 Likes

:joy: :rofl:

2 Likes

:rofl:

/10char

I’m watching the worm can slowly emptying.

1 Like

This statement is not correct.

Unfortunately most audiophile ethernet cables are shielded with metal connector housings.

This can break the ground isolation that unshielded UTP ethernet provides and can provide ground connections between the networked devices, something generally not wanted for home audio use…

That’s a nice potential pathway for leakage / ground / RF currents … a potential pathway for things to sound “different”… not in the technically good way either…

Archi’s measurements may not have revealed this issue but ground loop issues are always system/setup dependant.

Best to play it safe with UTP cable. Something like Belden’s 10GX Cat6A cable is a bit special in that it is shielded but it’s shield is ‘floating’.

1 Like

Another objectivists vs subjectivists thread. This gets no one anywhere…

1 Like

As a response to @ricgf request for a summary of the article and the cables tested therein, it is correct.

However, I agree with your comments, and this is a subject I have raised several times before.

Many audiophiles don’t understand the different ethernet cable categories or their construction (UTP, S/UTP, FTP, STP, SFTP) and why it’s important that the right cable is used.

I wholeheartedly agree that UTP is the best choice and the solution of using fibre optic to eliminate “noise” carried by ethernet wouldn’t be needed if the right cables had been used in the first place.

Cat7 and Cat8 have no place in audio, but it’s the usual snake oil marketing. If 6 is good, then surely 7 and 8 must be better?

2 Likes

Some people would do well to understand that transformers and chokes don’t block all noise. Some people need to understand that the quality of the switch changes how much noise may or may not be present on the differential lines.

There are reasons to use CAT5E or higher quality cable.

https://www.ti.com/lit/an/snla107a/snla107a.pdf?ts=1615717330282&ref_url=https%3A%2F%2Fwww.google.com%2F

Some people do understand that but still have doubts that this has an audible effect on sound quality. If it doesn’t affect the analog output from the DAC, the amount of noise on the Ethernet cable or between the PHY and the MAC is a diversion (as far as sound quality is concerned).

1 Like

Ethernet also works in differential pairs to further reduce noise, it doesn’t just rely on transformers and chokes.

Numerous tests have shown that ethernet noise does not manifest itself in the analogue output of the DAC, so chasing it down to the nth degree of elimination is a waste of effort.

Cat5e (100MHz) is good enough for most applications and is recommended for up to 90m backbone and 10m patch leads at 1Gb. Some Cat5e is tested to 350MHz and can support 2.5GBASE-T and 5GBASE-T.

Cat6 is rated for 10Gb over 55m - this is what I used all over my house and for the runs out to my office.

There’s almost no difference in cost compared to Cat 5e and it allows a degree of future-proofing.

If you really want total future-proofing, Cat6A gives you up to 100m for 10Gb at a modest premium over Cat6.

3 Likes

All you have to do is sit down and listen.

1 Like

it is much better opening a propiatory thread instead of trashing product support threads!

1 Like

I love these threads and all the talk of “noise” but alas I am a human and therefore have to make do with human ears and hearing abilities. Most, if not all of the “noise” that is mentioned in these threads is easily measured by various pieces of testing equipment but very little of it falls within the range of normal human hearing abilities, e.g. jitter in the pico-second range. Yes I know that it’s a noble goal to try and remove as much of this “noise” as possible but inaudible “noise” is just not something to lose any sleep over.

7 Likes

Hooked a live one here !! :slight_smile:

1 Like

There are two types of people:
The ones who believe there is no difference in cables and therefore searching for technical facts to proof their believe.
And the others who are explorers and researchers with a curious mind. This people will go to a hifi-dealer, borrow a cable, and listen at home with their own ears.

1 Like

Hi Jürg, that’s a very black and white view and as a scientist I find it a little disparaging - an exploratory and curious mind goes hand in hand with being a scientist in my view. I accept that there are areas where cables can make very minor differences (and those differences can be easily explained by science); however, ethernet (provided the correct type of cable is used) is not one of them. Numerous studies have proved this.

There are lots of reasons why people might subjectively believe they can “hear” differences - the role of science is to evaluate whether those differences are in fact, real or imagined. The only way someone can prove they are real is by reliably and repeatably detecting those differences in controlled tests.

The proclamation “I’ve heard it, so it’s real” doesn’t cut any ice with me…

As an aside, here in the UK, we have an advertising regulator, the Advertising Standards Authority (ASA) which has in the past taken Hi-Fi product manufacturers to task for making claims about cables etc. which they could not prove were true.

17 Likes

Ethernet is terrible for audio. Everyone should be using token-ring at 16Mbps. :stuck_out_tongue_winking_eye:

7 Likes