Ethernet Cables

06-18-14: Kijanki
New computer with faster CPU and memory can do more harm than good with Ethernet connection, often producing more electrical noise. Speed doesn't make any difference as long as it is adequate (pretty much any computer) since timing is irrelevant with Ethernet.

The key word is, CAN. All I can say is new computer sounds MUCH MUCH better. Most obvious improvement is sound more continuous ... appears old computer doesn't have enough cpu to keep up.

BTW, I also hard wired my computer for audio to router.

I agree with Kijanki's statement, and I also agree with Joe's emphasis on the word "can." There are countless unknown, unpredictable, and complex design-dependent effects going on inside of a computer which can affect the amount of noise coupled onto its ethernet interface. And in addition to "how much" noise is present, the susceptibility of the audio system to its effects will be dependent, in unpredictable ways, on its frequency spectrum.

One of those many factors, btw, is whether the computer or the router or other device it is connected to is old enough to be using 100 megabits per second ethernet, as opposed to gigabit ethernet. Although I would not want to speculate as to which ethernet speed is likely to be more problematical for an audio system, if in fact there is any difference.

So I don't think any generalizations can or should be drawn as to whether an older or newer computer is likely to be more suitable for use as an audio source. And if a difference is perceived when comparing results using an older and a newer computer, I don't think there is likely to be a good basis upon which to attribute that difference to a particular aspect of their designs, such as CPU speed.

Regards,
-- Al

Joecasey, you're right - one computer CAN sound better than the other but it is not related to speed (or amount of RAM) but to amount of electrical noise produced. Whether computer can or cannot keep up does not change the sound because timing is not involved. If it is too slow (not likely) you will hear dropouts rather than change in sound quality.

Your upgrade to AQ cable worked because you replaced unshielded cable with shielded one. Chances are that you would get exactly the same improvement from generic inexpensive shielded cable.

I truly have a hard time believing that these "special" ethernet cables make a difference. I've NEVER seen a single packet error (out of countless millions sent -- just type 'ifconfig' on a Unix-based system and it will show you the number of packets transmitted/received successfully, and the number with errors) on my Cat 6-based home network. Ethernet is designed as an interference-rejecting connection, and higher-level network protocols (TCP) have error detecting capabilities built-in.

Regular ethernet cables are acceptable for large data centers that need flawless, rapid communications. Why do we expect/think that adding some additional fancy shielding makes a significant sonic difference for our audio applications? Even gigabit ethernet (well within the capabilities of Cat 6 cable) is overkill for CD-type audio streaming. More fancy ethernet cables, such as Cat 6A or 7 allow for higher frequencies and can do 10Gbps+, that's true.

But to believe that an expensive, custom ethernet cable makes a difference for a short, packet-based digital transmission system over a short distance is quite a stretch.

OK, time to raise shields for the onslaught of the people who will criticize me for not being a "true believer".

Michael

Kijanki,
Seriously, I would agree with you if, say my server application requires 4GB of RAM, old computer has 8GB and new has 16GB. Also old computer has enough CPU to stream data faster than DAC requires. But this was not my case.

As far as cables, I don't worry WHY any longer but just use sounds best.

I do believe in noise. Try some RCA and XLR caps for all your unused digital and analog outputs and inputs. Probably the best ROI in all my years in this hobby.