That’s all dealt with in the DAC surely. That’s it’s job, A Digital to Analog converter.
I know with a Meridian DSP they keep the signal in the digital domain until the last moment so they can properly finesse the signal via dithering, Apodising and upsampling crossovers etc
There is much more to worry about with a lossy analog audio chain and maybe the noise is introduced at this stage.
1 Like
There are reasons which I explained earlier why more than the “bits” matter for ethernet, and lets not forget that different network tweaks sound different. If you want to pretend something isn’t true despite both theoretical and empirical indications to the opposite, there is not much I can do about it.
But its neither hard nor expensive to experience this yourself. Spend around 75$ on a couple of TP-Link fiber media converters and a short fiber cable, and test it. I am not saying you will think it sounds better, but if you have a reasonable transparent system you will hear a difference (in my system they gave added clarity and details but also a thinner and more sterile sound).
Btw, I just bought a Pink Faun LAN Isolator, which is 4 isolation transformers, one for each signal in a TP ethernet cable (in other words, a well implemented galvanic isolation). The isolator certainly don’t change the bits (if it did I would return it). But its not hard to hear the difference it made. If you live in Stockholm, Sweden, you are welcome to come visit and I will demo it.
You can listen to what Hans has to say about it here: https://www.youtube.com/watch?v=tPWS3Lk6y1g (but I guess he also imagines things, like all we who hear things you can’t explain).
Are we talking about Ethernet, or not? I need an answer, so we are on the same page.
Both the theory and empiry of Ethernet is well known for over 40 years: It just works. I’m not pretending anything. Those who are pretending are those who believe listening are the same as empiry, even if they actually hear a difference - it’s just anecdotal observations. Me buying equipment and listening for myself will not change that fact. An if I would hear a difference, I would like to know WHY. And no one has explained to my why there can be different SQ over Ethernet. Not in this thread.
Which theoretical and empirical indications says the opposite? Your theory about noise? Please point me to it, because I want to read and understand.
The only empiry I have seen regarding Ethernet and SQ shows no difference.
Of course we are talking about ethernet, I have said that time and again. If measurements don’t show difference that is audible it only tells me that those measurements aren’t accurate enough (which is old news, measurements can at best give an indication of sound quality or transparency).
But here is the thing: I often experiment and listen to audible changes, and if I hear changes and they stay there over several days, and I can confirm it by switching back, I tend to believe my ears over measurements or what others say. Even if I can’t explain it in theory (in the case of ethernet its easy to explain though).
I know, listening to your HiFi to hear how it sounds is a very strange way indeed according to some, but you should try it. 
But since you have all answers, can you tell me why a LAN isolator gives very audible changes to the sound? Or are we all affected by some mass-hysteria and imagining everything you can’t explain?
Definitely. And then to post opinion positive or negative no matter as long as it is honest.
In my set up and listening experience keeping network cables of same Maker before/after the Switch gains additional improvement of SQ.
2 Likes
The system is global. You have Ethernet frames but not only … the components which search for frames will find frames. If you integrate POE, you will find electric power. And if you don’t have power, you don’t necessarily need all of these pairs.
Cables carry everything they can carry and even, we set priorities (QOS).
The same blog (archimango) but not the same words:
The system is global. You have Ethernet frames but not only.
I also have Med 1005 filters (galvanic) on the cables coming out of my nuc …
and 2 Meraki in cascade (a mechanical diffusion filtering system: a single switch for the final connection to my USBridge).
It’s just … better and cheaper than a Melco.
It’s like bananas …
if you buy bananas. They were transported by boat … but the boat transported many things.
Sometimes even spiders in bananas. 
1 Like
So you say, the human ear is far better in detecting differences than the best measuring equipment 
A good way to cure audiophoolie is education.
Why is the wrong question. It can’t. But your brain can.
You become defensive again.
I’m not sure about that either, but Magnus could be right, of course. The problem is that you’d need to do a double blind test in order to prove all this. And that objectivity is what’s missing in most parts of these kind of discussions.
1 Like
One thing often missing in most audio forum double blind testing discussions is the importance of details of the testing procedure itself…
There’s lots of self proclaimed experts but from what they write, it looks like they are oblivious to the several complexities that the brain potentially adds to these tests… these people seemingly only look at these things only from the side of expectation bias (obviously valid) but there are potentially other brain complexities involved… confounding variables.
Throwing stats out there about testing results without specific details on the testing procedure… just like everyone learnt in high school science experiments, garbage in = garbage out.
1 Like
Hi Bill,
My EE 8Switch comes with (apparently?) a ‘medical grade’ switching PSU, so I’m going to stick with for the time being.
Glad that PSU has worked out for you 
1 Like
Medical grade eh! Well who am I to argue 
1 Like
1 Like
The brain ALWAYS “adds complexities” to listening tests. Not only potentially. Every test method has (its own specific) disadvantages. You won’t find a single peer reviewed publication (on this or any other subject) in which the author doesn’t discuss the pros and cons of his own methodology. Nor will you find a single researcher who doesn’t use and evaluate more than one test method.
Oh, this is very interesting.
Also interesting… these authors who also get involved on forums should talk more about the cons in their posts… But instead we only get things like the below with no details of the procedure behind the results, because it apparently requires 1100 pages to summarise the procedure behind these results:
Hi all
I am reading this topic since the beginning, but I wonder and I am sincere:
For me, cables, from RCA to speakers are like 5% of SQ in a set up
In % how much do you think Ethernet cable is responsible in this chain?
Because to me it is like looking for the graal as difference between good mp3 and cd’s are barely noticeable for most of people, so I wonder how even the slightest improvement could be heard.
I’m not sarcastic, just want to get what is the point
I think You’ll be the one to give an answer…for Yourself.
(basis on calculated 5%)