Relationship between Ethernet Switch and SQ

Point/Counterpoint on these forums has become so tiresome.

So many of these threads turn into pissing matches when two sides each take absolute but opposite positions.  Many times, when I look into the stuff myself, the reality seems to fall somewhere in-between whether the issue being argued is "scientifically possible" and whether it is "likely to be audible in a hi-fi system." 

A by-product of these arguments occurs when manufacturers prey on these debates and develop (expensive) "solutions" that are marketed as miracle cures for whatever the alleged issue is that is somehow limiting a listener's satisfaction with their hi-fi system.  How many "audiophiles" have a box of no longer used "miracle cures" on their shelf?  Just go back and read posts from 5, 10, or 15 years ago about audiophile doo-dads that were considered "almost mandatory" yet are either no longer being used or have been replaced by different (and usually much more expensive) doo-dads.  Examples might include certain fuses, DBS wire biasing, Shakti this or that, crystals, certain USB and Ethernet filters, etc., etc.

This Ethernet cable issue seems to have some basis in reality in that yes, running an Ethernet cable next to certain types of power cables/lines can cause EMI and, yes in certain cases, this can cause minor to major degradation of a digital signal.  However, the specific situation causing this issue is mostly unlikely and can wholly be avoided by separating the Ethernet cable by even a small distance.  In addition, such minor data loss is mostly corrected through Ethernet internal error correction and retry protocols.  As well, it seems the Ethernet spec (IEEE802.3) has specific requirements for isolation and resistance to current and voltage spikes to maintain performance in variable EMI and RFI scenarios as well as for electrical safety.  As a result, it seems the risk of typical Ethernet cables properly run in a residential environment affecting an audio signal (while not impossible) is probably somewhere between minimal and non-existent.   Here is an interesting discussion of the issue, and here is some related stuff on the Roon Labs Community forum.

"But I hear it in my system!"  This starts round two of these debates pitting auditory abilities and system capabilities on one side, against possible psychological factors affecting what people believe they hear on the other.  There is no winning these arguments since the two sides cannot agree on suitable measurable metrics, or even agreeable protocols for conducting listening tests.  On one side, "If you cannot accurately select the cable/tweak/etc. statistically better than 7 out of 10 times, then how can it be making any difference?"  The other side says, "DBT doesn't work, and only through long-term listening in your own system can you accurately hear what something is doing."  The real problem is not about measurements and protocols but rather that everybody wants to be "right."

The only solution seems to be agreeing on no solution.  IOW, share observations and experiences, and let go of the need to be right.  If somebody wants to spend an exorbitant amount of money on a rock that is sold to improve SQ when placed in the same room as an audio system, then simply say, "cool, enjoy your rock."

 Why would we be concerned about noise along an ethernet cable.  Noise doesn't come into play until the signal's past the DAC?

To quote Hans Beekhuyzen: "But it does. It clearly does". 

Long before we could explain why the sky appears blue, we all observed it with our senses and agreed that it was blue, but we didn't know why. The funny part is usually when humans ask they are too young to understand the answer, and by the time many are old enough to understand the answer they don't care anymore. You know, that innate curiousity doesn't exist anymore. 

This is where we are at- many of us are agreeing that ethernet filters do make our kits sound better, therefore we conclude they must be absorbing and filtering out electrical noise. At some point we will be able to prove it and explain it to those who question it before observing the outcome using our senses. But for many we don't care about the reason, we just know that it does. And many probably wouldn't understand the answer anyway. 

What am I missing? 

A better soundstage, and a smoother, less digital, and more enjoyable sound. Yes, I am a recent believer. 

Digital transmission (external to a chip) is typically an analog waveform which “simulates” the ones and zeros over a cable. Most believe it is either a one or a zero but the fact is the analog waveform is a voltage approximation (the better the approximation, the better the signal) can be affected by noise, cable length and the veracity of the signal transmission. Its critical to do your own research because there are parts of the chain where the signal tranmission is fragile and susceptible to external interference and the physics of analog signal transmission absolutely apply.

There are other parts of the chain where there is little to no effect. You need to understand which is which because so many argue one side can/can’t be affected/influenced by applying their knowledge and understanding of one side of the equation to the other side where they believe their logic should apply equally. It doesnt work that way. I’m not smart enough to explain, let alone understand, the intricacies but I am smart enough to understand when I clearly hear an improvement to discount the opinion of those who swear there can’t be.

Trust the science and trust your ears but most importantly, we must have a credible constant to compare against….and our memory and how we think something sounded is fallable in seconds.

So if I hear a tree fall while walking in the woods, should I assume that a tree really did not fall because 1) I did not see the tree falling and 2) I expected to hear a tree fall while in the woods and so therefore I heard one fall.  There is a third option:  There are no trees in the woods…. (There is no spoon.). Oooooh.

@tonywinga 

To further refine: if two tress fall, one after the other, and with the first one you had an ethernet filter in your pocket and the second one you did not and yet, they soundeded identical….well, then the filter makes no discernable difference.