I have seen a few threads on network switches, but not much on here.
So anyone using them?, are they even needed with DCS?, as some manufacturers like naim, for instance, are getting good results from them.
Now it could be that they need something like this, as the design is, lets say lacking and so benefits from one, now i use a Cisco switch, mainly as its very cheap, plus i need an extra port, as the router was full, now adding this, to my ears did nothing for the sound quality, but maybe a better switch like the melco S100, ER, or EE, ect might make a difference?, so come on guys, what are your views on it all.
Cheers
Duncan, I’m not trying to be a horse’s hind end, honest, but a search for network switch yields over 50 results including this one of course. Maybe start there and then ask narrower questions?
Yes Greg has a point. It has been well covered previously.
In terms of data transfer audiophile network switches ( I think it is this we are talking about not just network switches in general) can do nothing better than other switches or routers ( which are part switches) can do.
Unfortunately, however, although I cannot understand why audiophile switches do make a difference and as the user of a Melco S100 a significant one ( read improvement for difference IMO).
So, I conclude that it is not possible for such switches to make any difference. But they do.
Yes i did a search and found a few, but they didn’t really answer my question, as most went off in all sorts off different directions.plus mainly the same few guys.
Just wanted to get an idea on how many off us use a separate switch like the melco S100, etc. As like I said I use a cheap cisco one, but didn’t hear any difference
I’ve used and still use some very good Cisco switches in my network. For my audio systems, I isolate with a combination of fiber, GigaFoil 4, and EtherRegen. I view these as noise reduction/problem prevention devices, not devices that improve any other aspect of audio reproduction. Switches themselves have no apparent impact. Switches simply cannot change the sound of the bits passing though them. If they do, they’re doing something wrong or something else before them was. I’m not of the “1s and 0s” school. I know that switches can inject noise into a system, and I suppose a bad enough switch could contribute to packet loss that might cause a problem. I know that “1s and 0s” are voltage swing interpretations of digital information. This is a process that can be done badly, but if your DAC is receiving a bit perfect file, and its Ethernet port is properly constructed and isolated, it’s inexplicable from an engineering perspective what sound difference a network switch can cause. I’ve been an ahdiophool long enough to know that sound differences have often been heard before they’ve been understood, measured, and quantified. I am open to the possibility that something is being heard that we don’t yet know how to measure. But even testable hypotheses are lacking.
When one looks at what’s inside of so-called audiophile switches, one sees zero explanation for alleged SQ improvement. They are generally not as well made as a Cisco Catalyst. That’s not to say that Pete and others don’t hear what they hear, but my hypothesis is that what is happening is that the audiophile switch is fixing a previously undiagnosed problem with the switch and/or cable it replaced. Not that there is anything magical about an audiophile network switch.
Whenever I hear an SQ difference, or hear one reported by another audiophile, I always ask: “why should this be?”. There must be a reason. If the reason is anything other than expectation bias, there has to be an engineering explanation. I am amazed how many times people tell me they “don’t care what the reason is; it just sounds better.” The reasons that SOtM and Melco give for the “sound” of their switches are not engineering reasons. Even a lot of the stuff proferred by Uptone Audio on behalf of ErherRegen is more hypothesis than proven engineering (a solution in search of a problem as someone put it), but at least they offer some design detail that makes sense, e.g., fiber isolation (as opposed to most of the others).
I believe in using quality switches and employing fiber isolation. The latter is more luxury than may be necessary. Properly certified and unshielded Cat6 cable seems all that dCS equipment requires. A GigaFoil 4 (hasn’t worked for everyone here in the Community, but I have two working flawlessly) or EtherRegen can supply the isolation one needs to protect the DAC from upstream noise, if any. Other than that, I’d be looking at my power, power cables, audio cables (as opposed to digital), and the most my budge or patience can tolerate for the best equipment I can sensibly afford. I’d always rather pour as much into the speakers and audio electronics as possible.
I use a SOTM SNH-10G with external Sotm power fooding and with some options they propose.
Cable are Audioquest Vodka.
It works great, for me this means reliable, no down time and the sound of my system pleases me… that’s all I can say, but I didn’t spend time in comparisons of network gear.
Well been demoing a melco S100 the last week, and can certainly say it adds more space and clarity to the streamer side off my system.
Its not massively different, but it is different and a better different, its more noticeable on certain tracks, one album that really stands out, is nils lofgren, aqustic live, more open sounding, more air around the instruments and his voice is cleaner, just makes listening to all thats going on a bit easier, if that makes sense.
This is going through a melco N10, the rossini connected to the player port on the N10, using AQ diamond for the N10 to rossini and catsnake 6 for the rest.
Try it ( with and without) on some large scale orchestral music that has wide dynamics and lots of tonal colour. I hope that you will find the difference more marked.
Where I first most noticed its impact was on a recent recording of Ravel’s La Valse ( Javier Perianes, L’orchestra de Paris , Qobuz 24/96). A few bars in there is a crescendo. Without the S100 it was clouded and rather limp. You could hear the orchestra becoming louder through tonal changes but the volume itself didn’t increase much. Put in the S100 and it was now all there. Something that occurred sonically where its existence can be verified by looking at the score . Then there was the added openness, timbral accuracy, deepening of soundstage etc.
In short, the more the recording has to offer the more you will notice - as you did with the audiophile favourite live album by Nils Lofgren.
Oh, somethtng that isn’t mentioned in the instructions. The router and Roon ( if you run it) go into port group 5-18. The N10 internet connection goes to a port in group1-4.
Well further testing off the S100, yes it does add more space around instruments, but it also takes away dynamics, to the point off i have started to hate what it does, quite a turn around really, as to start off it got me, as i could clearly hear more easily things, but as said the more i listened too it and swapped back on tracks to just the N10, it started to come apparent that it just killed it, rather than improved it, some tracks its much more apparent on, been listening to norah Jones new album and that just sounds wrong in comparison, the piano sounds very different, hear voice, is nice and clear, but again sounds wrong, plus as i said the dynamics on the complete song, just takes more off a step or two back.
Its actually turned my rossini into something its not, and for that the S100 is out, and the Cisco is going back in.
I plan on auditioning a Telgartner M12 Switch Gold in the near future. I will insert it after my opticalmodule and before my Upsampler. I want to keep the long length of fiber optic in the loop.
Well i thought i will try once more tonight, still crap.
Looks like others have also found it the same as me, some saying it needs a much better power supply, as the standard one is very noisy and just crap, well for me going down this rabbit hole even more, with a different power to try and fix the problem, is a no go from the start.
It should be good from the start, and not need a ÂŁ500 or even ÂŁ3000 power supply to start to do what it should do out the box.
So back to the dealer it goes, even he said most return it, as they dont hear the difference a ÂŁ2000 bit off kit should do, says it all really.
Oh, there is much fun here, as well. It’s just much more respectful.
.