Not to pile on more, but I just read this review, and it’s conclusion;
…it became clear that the S100 offerednoticeable improvements in detail, transparencyand overall ease of listening when compared to the Netgear and a couple of lengths of Ethernet cable. There was tighter bass and just a sense of more space, air and focus in soundstage pictures compared to that basic set-up…
Audible differences of the nature described above (and some of the posts here) requires bitstreams to be altered. (As I’ve said many times before) it’s trivially easy to objectively prove that Ethernet switches do not alter bitstreams. And if one thinks noise is the culprit, theres no way for noise to travel down fibre, problem solved.
(I’m sure I’m not going to change the minds of those who imagine they can hear a difference,) but trying to ascribe analog-like qualities to the effects of an Ethernet Switch is beyond absurd; modern day “Flat Earthers” who utterly refuse to believe all scientific evidence to the contrary.
Looks to me like its not just confirmation & expectation bias, clearly theres false consensus effect at work here. Andrew Everard just lost all credibility now that we know for a fact that his hearing can’t be trusted, pity.
If bitstreams cannot be altered by switches then a change in sound must be the result of either this statement being incorrect or it is due to some other factor.
You do suggest one mechanism ( could there be others that you haven’t considered?), noise and you reasonably aver that noise cannot travel via a fibre interface. However it is only fair to Andrew Everard to point out that in normal circumstances he does use a fibre based network and that he therefore says in this review
" I’ll stick to my belief that one can obtain similar gains…with some DIY network fettling". In that context he concludes that the S100 is for those that want “a simple, easy to install upgrade”. Presumably for the majority of us without a home fibre network. I do not see the two as necessarily being in contradiction.
It is, of course, unfortunate that neither in this review nor in any other that I have read nor in the manufacturer’s literature is it revealed exactly what Melco are claiming that their switch actually does . Only then can it be verified or dismissed. All we have currently are reports of the subjective outcome ( and some measurements which may or may not be relevant) . If we both agree that there is nothing that we can think of that it can be doing that does not mean it is doing nothing.
In the meantime I shall carry on enjoying what I cannot be hearing .
Pete, I think you hit two extremely important points, and then the most important conclusion of all regarding enjoyment: (1) the S100 might well be achieving what one would/could do with a variety of individual network improvements; and (2) the lack of engineering explanation for what the box-makers purport to be doing hampers any reasonably objective or credible discussion and assessment. We are all left with these subjective assessments. I for one thoroughly believe you and Ian both when you say you hear these differences. Believing you (or Andrew Everard) and asking why are not mutually exclusive. Wanting to share an understanding of why helps to improve our understanding of how things work and how to improve them.
Like you and Ian, I can enjoy my system without knowing every “why.” But when it comes time to spend money on a change, then the explanation of what the change is and how it contributes to improved fidelity matters a great deal to me. I agree with Ian about trusting one’s own ears; they’re all we’ve got, and they are unique to us. But I also believe in another old saying: “trust, but verify.”
Greg, I am pretty much in agreement with you about verification which I believe is a prudent position. However I also need to take a pragmatic view on whether or not I buy something.
I am in my 70s with a few health issues. I therefore regard myself as now having a finite window left to enjoy things. Do I wait until Melco reveals its engineering explanation followed by the inevitable debate which, knowing this hobby, will take years if ever to resolve or do I take the position that though I haven’t any idea why it does, or even should, work it nevertheless adds significantly to my enjoyment of music?
I took the latter decision. For all I know it may prove to be a bogus innovation in the end but, frankly, I may not even be around to learn that. Meantime I will have had a lot of fun.
Pete, I totally get that. I just turned 65 this year, and while the “window” still seems to expansive, it is decidedly more finite than it once was. And so I see things similarly. If something makes sense at some level—engineering, convenience, conceptually, aesthetically, etc.—and it’s in the budget, I might spring for it on the theory that I only have so much time, and I might as well enjoy it. When I make those decisions, a company’s reputation plays into the calculus. Melco has a good one, and makes good stuff. What you say makes sense to me.
It’s similar, but not identical, to the question: “How many more times do I want to swap 45kg power conditioners–which, by definition, allocates time away from relaxed enjoyment—before I decide I’ve heard (and lifted) enough?” The answer to that one is now zero.
Pete, noise that could potentially creep into a system over an Ethernet cable without affecting the bitstream cannot possibly have the sort of effects claimed by Andrew because of the layers of isolation/abstraction between the incoming Ethernet bitstream and what the DAC eventually receives for conversion into analog.
As for other factors that I may not have considered, (as I mentioned before), if Ethernet Switches/Cables had the sort of effects that some Audiophiles claim, then the whole world as we know it today wouldn’t work properly; Ethernet based systems carry far more sensitive and critical data flows across millions of “non-audiophile” systems everyday without problems. Even the most expensive Ethernet Switches used in these millions of networks are based around the same handful of merchant Ethernet chipset suppliers as used in consumer Ethernet switches (like the Broadcom chip used in the Melco S100 for example).
And I’m not just speaking theoretically by the way, I’ve actually made the effort to objectively validate in my system that changing the Ethernet Switch (or Ethernet Cable for that matter), does not make any difference. This is why you will not see any “Audiophile Ethernet Switch” manufacturer providing proof of efficacy, because they can’t.
Ironically, the disaggregated and configurable nature of the dCS Vivaldi system makes testing these sorts of things incredibly easy. For example, streaming a track off your UPnP server, and with your Vivaldi Upsampler in Clone mode, its easy to show that changing your Ethernet Switch or Ethernet Cable makes ZERO difference to the AES/EBU bitstream received by your Vivaldi DAC. Not only that, the bitstream is identical to the originating .wav file stored on your Server.
So, is the difference you’re hearing actually a result of the Ethernet Switch/Cable change, or is it in fact from biases affecting perception? Of course some may not care either way (seems you’ve got a few decades over me of not caring ).
I’m trying not to waste much time on these debates, but this part is troublesome. I see no reason to view the functioning of printers, web browsers, or data storage having a meaningful analogue with dynamic audio playback equipment. Noise anywhere in the wired signal chain, from a USB port, ethernet, ethernet renderer, mechanical vibration, ground loops, mains, etc makes its way all the way through the analog components, amplifiers and speakers shaping the listening experience in real time. I think at this point all of this should be well established, except some still believe that network audio is the exception. For those that bother to test in earnest, this belief is easily put to bed.
I was not speaking specially about ethernet, and I won’t play the “appeal to a scientific paper” game. The principle is fundamentally the same for all situations where a wire connects a digital device with a DAC. As was the case with all other wired data transmission schemes, the designers presume that what works for a printer is just as good for audio reproduction, and haven’t yet figured out that their design could have some flaws that are apparent in an audio environment.
When the bar for noise isolation is raised to the level necessary for high resolution audio playback, unoptimized ethernet is audibly imperfect to a degree that is easy to detect if you try.
Well, I am specifically talking about Ethernet as you can plainly see from my posts.
If it’s so well established, and “easy to detect” as you claim, then there should be plenty of evidence to back up your claim, why won’t you provide any pointers to them at all?
I come from the audio engineering world where I’ve encountered this format of argument copy and paste dozens of times about literally anything in the audiophile world. Would it surprise you to know that there is no established scientific evidence that power cables impact audio quality? Indeed, there is no established evidence that digital cabling of any kind, digital sources, analog interconnects, speaker cables, vibration, have any impact on audio quality whatsoever. In fact many dispute that there should be any audible difference at all between 16 and 24 bit, any sample rates, or PCM vs DSD, or that there is any difference between DACs with greater than 20 bit performance.
And yet, by simple empirical testing it is obvious that all of the above have notable and worthwhile sonic impacts. Slowly a consensus forms because enough end users have done their own listening, and that is how the market moves. If you have spent more than $2000 on DAC and $50 on cabling, you probably own gear that someone else insists is unscientific snake oil, so if you want to play the evidence game you have to start by throwing out most of hifi. Otherwise you have to judge with your ears.
You’re mistaken. There are very clear measurable parameters that differ between various power cables, with measurable differences in the consequential analog output. Whether those differences are actually audible is a separate (subjective) matter of ongoing debate. Likewise for the rest of your arguments.
But back to Ethernet; I take it then that you’re unable to provide any tangible evidence whatsoever of noise carried over Ethernet cables affecting just the analog output, and not the digital bitstreams.
I thought a lot before further replying. It is easy for this discussion to devolve into little more than a mutual tar baby, each side increasingly attached to their chosen ground with no room for common understanding. I like this forum too much to want to see any of that here. I understand Ian’s point of view, and I once held to it. But I have since moved more in the direction of: "I can believe you are hearing it, but help me understand why.” In that spirit, I am replying.
Anup is correct. There is actually ample evidence of measurable performance characteristics of all sorts of audio cables and components. And there are more than a few engineering explanations of why one cable might sound different from another. See some very extensive discussions here. Maybe everyone doesn’t agree with Galen, but he has his reasons for his cable designs, and he publishes them. And he’s just one. [There is also a fair bit of evidence of why the internals of some components are such a mess that the best cables in the field can’t make them any better. ]
Ian, what do you mean by “simple empirical testing”? I assume you are aware of the many A/B/X tests that showed many audiophiles could not hear the differences between many things they thought they could. And perhaps you are aware of Schiit’s testing that suggests the same? Now, the methodology of some of those tests can of course be criticized. That’s the nature of tests and science. But to me, it casts doubt on the notion of “simple empirical testing” for which there is anything like a scientific consensus.
I’m more than willing to entertain a hypothesis that states: “the conductivity differences between silver and copper make one of them a preferable conductor for 110 Ohm digital cabling.” Once we agreed on cable geometry, we could then test whether the square waves transmitted and received by a certain type of digital transceiver over a certain distance differed as between the two. They might. It’s unlikely they would differ enough to prevent bit-perfect playback, which is actually good for our experiment. Then we could set about to listening. Given the limitations of cable swapping, and the known tricks our aural memory can play, it might take a bit of finagling to agree on a test protocol, but I bet we could. I expect some people would hear no difference, while others would, and of the latter, some might be able to repeatedly, and reliably identify the differences. I would view that as a Eureka! moment because it gives us something new to ask and understand about how to audibly improve our cabling. If people are hearing audible differences between digital cables like Ethernet cables, someone should be asking why (even if we can’t yet measure it), and want to track it down. Such an experiment could do so. The people selling such cables should, it seems to me, embrace such an experiment.
And what does “how the market moves” mean? What you are describing is that a herd can be led by influencer opinions. I’m not saying that all such things are like the emperor’s new clothes, but some of them might be.
It’s interesting that your argument is that, in the absence of any scientific evidence, we should pay attention to or care about none. Anup is not “playing a scientific paper game.” He’s asking: “where’s the evidence that something is happening here?” What is it about your Shunyata Ethernet cable that could contribute to an analog outcome? Is it shielded? In your system, that might be a good thing, but in mine, a problem. [BTW, I tested that difference in my system. I tried Supra shielded cables. They were a bad thing. I reverted to unshielded, compliant, Cat6 as recommended elsewhere here. That was a good thing. I tried a couple of super-expensive versions of that same spec. They made no difference. Maybe you or someone else would hear a difference with better ears. Don’t know. But if you did, I would still want to know the cause, to the point of one of your earlier posts, because cumulative similar improvements might become audible to me or everyone.]
This takes me back to the point I made earlier: you may not care about why one cable sounds different from another, but at least two other people do or should: the maker, and anyone else who might be interested in knowing whether it will sound different in their system (someone like me, or any other colleague, who is intrigued by your suggestion that a conductor choice in a power supply cable to an Ethernet switch should alter the analog sound of your system, and wants to understand why before forking out the dollars and/or rewiring their system)
It’s not denial to say “I’d like to understand why something is happening.” That’s not denial; it is the best form of skepticism. I’ve never said “you’re not hearing it.” I am saying, “let’s explore what could be causing that.” My dealer wants me to demo $2k Shunyata clock cables. Boo-yah, I say! But then, I start to think: that’s five clock cables, $10k all in, pulling out the Vivaldi stack, and the Cybershaft, etc. Nobody is holding a gun to my head, but I might still wonder: is it worth the time and/or money? Money, if you have it, is easy. Time and labor, less so. And using them up detracts from the musical enjoyment Pete mentioned. So, I would ask: what can I expect the Shunyata cables to do that my BJC or Geistnote or Van Damme can’t? My system is fairly resolving; what can I expect to hear and why? This is not denial; it is wanting to learn.
Anup noted:
Fiber won’t fix everything, but one of the simpler and cheaper ways to test/eliminate your noise is to get fiber between your DAC and everything upstream. You clearly already know that. But it, or something similar, is an important thing to do in order two help eliminate that noise we know can matter.
P.S. I haven’t copy-pasted anything here other than quoting you and Anup. Ironically, I’ve read just as many (copy-pasted?) denials by people in the camp who assert that “science has to catch up to listening.” That is probably true, but it doesn’t mean that the science is irrelevant. And it doesn’t mean that a bunch of mutually reinforced subjective impressions overcomes the science.
I think part of the problem is that most Audiophiles don’t have a good grasp of how Ethernet (and TCP/IP) really work. Perhaps the difficulty stems from not understanding packet-based transport built around layered abstraction since most Audiophile have only been exposed to physical layer digital interfaces like S/PDIF.
And then along comes USB which has genuine issues around electrical noise, the lack of galvanic isolation, bad driver circuits etc., and people just conflate the two and assume Ethernet is the same, reinforcing this misguided view across scores of audiophile forum echo chambers.
Further “legitimising” the view are the unscrupulous “Audiophile Ethernet” manufacturers profiteering from products with no evidence of efficacy whatsoever (despite some having access to appropriate tools!). Over time the misguided view has become “well established”, “easy to detect” etc. etc., but with literally zero evidence from anyone anywhere to back-up such claims except “trust my ears”.
Sadly, even seasoned Audiophile Journalist are not immune.
A well respected UK Audiophile Magazine (subscription based in print only) recently published an incredulous purely subjective piece on Audiophile Ethernet Switches/Cables having sound quality impact.
I wrote to the Editor, including illustrations for their technically competent writers to verify for themselves how it doesn’t actually cause a difference, and that they should clarify to their readers wth a follow-up.
Disappointingly, his response was to decline the suggestion, and then to propose that the sonic impact was due to “noise and vibration”, with zero evidence other than “we heard a difference”
On hindsight, I should have expected that, after all, we’re now living in a Post-truth world where basic scientific rigour seems unnecessary and alternative facts prevail
Hi all, thanks all…totally confused audiophile who does not understand lots of the network specs.
Simple question;
I have a Rossini Apex Player.
I bought the Innuos ZENITH MK3
I have a 2 gigabit cable provider set up.
How would you set this up? I am thinking ethernet cable from the direct wall outlet which goes direct to the router/modem to the Innuos lan plug. The my Silver Dragon moon Audo USB from the Innuos out to the Rossini? Run sense off my MacMini?
Thoughts and much appreciation to anyone who cares to pity and share a lost expensive music lover…
quote=“ARDDCS, post:75, topic:1589”]
I am thinking ethernet cable from the direct wall outlet which goes direct to the router/modem to the Innuos lan plug
[/quote]
Yes.
This will work well but is not the best sounding IMO. Rossini network input is best and this is connected to your Innuos by the network ( Ethernet) Innuos port labelled “Streamer”. You can then use dCS Mosaic rather then Sense. The Ethernet cable from Innuos to Rossini should NOT be shielded and simple CAT5E unshielded will be fine.
Mosaic or Sense are control units and are normally used via your phone or tablet. Mosaic is available in iOS ( Apple) or Android versions. AFAIK Sense is only available in iOS but I am happy to be proved wrong. Your phone/tablet (wifi) and Rossini should be connected to the same LAN group.
I am guessing that you mean an ISP ( Internet Service Provider). You need this if you want to use streaming services. You do not need a gigabit line for audio streaming.though. Good if you have family members watching different 4K TV channels at the same time though.You will need to decide what service you want to use. Hi-resolution files are streamed by Tidal and Qobuz. Once you subscribe to one ( or both) you sign in via the app you chose. Other streaming services are available and Mosaic provides for Deezer and Spotify subscriptions too. I expect Sense offers something similar.However Sotify does not offer lossless streaming and neither have hi-res.
If you want to have your own stored music library ( UpnP), that is another issue and I won;t go into it here unless you ask.
Thanks for taking the time. I use Tidal and Qobuz. Will want to use Sense when the unit comes, tomorrow and hook to my Rossini is UPnP. I have over 1000 rips cd files I will move to the MK3 and rip thbalance of my collection. So I will have local files, Tidal and Qobuz high res and DSD files . Just wan to make sure the connections I mentioned will be the best for sound
No and this consideration is not relevant for USB. Unshielded cable only concerns network connection for technical reasons in this particular context and use of shielded network cable from server to Bartok has the potential of adding noise rather than the opposite.
This means that the correct type of network cable between these components can be both cheap and effective . So cheap that it should be easy for you to buy a length to compare USB and network connection .
).
]
