I’d guess a lot of DCS DAC owners have added third-party network servers to their setups. I’ve tried four recently, and every one of them brought noticeable sonic improvements. After adding the Rossini Clock, it got me thinking—DCS should really consider getting back into the network server space. I’d definitely pay a premium for a DCS-designed unit, especially if it included a proprietary connection to further reduce noise. Just a thought!
(This is likely not going to be a popular post among subjectivists 
)
IMHO, dCS should absolutely NOT get involved in the Server business.
If a Server is exclusively serving music files and not doing anything else, it really won’t make any sonic difference whether you’re using a $2,000 Server or a $20,000 Server when streaming via Ethernet.
At the Ethernet interface board on dCS platforms, there’s absolutely no difference in the IP Packets flowing in between different servers; because all IP packets are error-corrected, buffered, and re-assembled into exact replicas of what’s on the file on the Server storage, and then separately synchronously Clocked out for digital-to-analog conversion.
No doubt many audiophiles and vendors pushing expensive Servers don’t like this view, but it’s an objective fact.
It’s the same reason why a Qobuz Server sitting hundreds, if not thousands of miles away is able to stream music directly to your home without you hearing glitches. It’s the underlying basis for most of the world’s modern economy.
12 Likes
Fighting talk on a HiFi Forum!
I think it is fact - but there is nuance. If we accept that sound can be degraded by poor choices of equipment (eg cables, switches, hard disc drives) then conversely we must accept that having these technologies optimised to the Nth degree will be beneficial. So my point would be that its not about these jazzy servers adding anything, its about them compromising the data less.
I’m not a big believer one way or the other by the way, just applying logic to the “data is data” argument.
I also think perhaps the point is wider than a server - i think @JohnDcsMe may be referring to a Streamer (like the likes of Aurender and Lumin make) where there is storage on board and optimised streaming technology. This also introduces more opportunity for less compromise - whilst (yes) likely introducing some of their own.
Just follow this logic: To dismiss the idea means you ignore the fact that a lot of DCS DAC owners will buy network servers (PhoenixNet, Silent Angel, etc.). They heard something that had them write a check. I did. Moreover, it’s just good business. DCS is leaving money on the table.
Sorry Matt, when it comes to IP Packet based streaming, data is data.
Network based Server streaming, whether local to your home or otherwise, is based on completely standards-based Internet protocols that operate exactly as intended. There’s no “nuance” to how packets are received and re-assembled into data thats identical to their source files.
If that wasn’t the case, the Internet as we know it could not function.
Those are not Network Servers. PhoenixNet and Silent Angel are Ethernet Switches.
By Server I mean a compute platform with a CPU and Storage that serves files.
5 Likes
I think this could go down a well trodden rabbit hole, but while I agree of course about the nature of TCP/IP and its inherent qualities, I think there is a little more going on in a streamer (and a DAC for that matter) than just IP networking. If there isn’t, we are all wasting our time!
1 Like
Don’t conflate the Digital-to-Analog conversion stage with the Data networking stage within the dCS platforms, which is the point of this discussion. They’re distinct.
As far as the incoming network data stream is concerned, like I said, a dedicated $2,000 Server will operate just as well as a $20,000 Server. Absolutely no different from the perspective of the received IP stream.
Yes, you’d absolutely be wasting your time (and money) if you dropped serious money on a Server.
5 Likes
…something that does really good metadata lookup and has good metadata editing would be handy though!
/me ducksandrunsforcover hugging my copy of MP3Tag…
P
4 Likes
I think you should ask your dealer to line up that demo - take along your (quite rightly) sceptical ears and see if you still hold the same opinion afterwards. I did, and I don’t, and I’m definitely not in the more money than sense camp.
And as I’ve said - if you accept that a poorly designed solution can make a negative difference, then it’s kind of impossible not to accept that a well designed solution can make a positive difference - ergo good quality server makes a big positive difference.
We can perhaps agree to disagree. But thousands have heard the big difference a good quality music focussed server can make - and I suspect that includes dCS themselves.
This said - as per the OP - do I think dCS should enter the server market. No I don’t. It’s very niche - high cost, and best left to those affiliated to the manufacture of the components that matter (eg Melco)
This thread has largely veered off course and has turned to a discussion of whether or not network servers make audible differences.
The subject is whether or not dCS should make one. This is a commercial decision. Is there a market gap to be serviced and at a price that is attractive to buyers?
Perhaps making costly network servers may have been more viable in the recent past, however many are moving to using streaming services only with no demand for storage or UPnP facilities.
In any case, I am sure that the smart people at dCS may have already considered such a move.
1 Like
Generally true but not if the product is a true, standards-based commodity, e.g. bits.
To introduce some levity into the discussion, and in a friendly attempt to articulate what I believe is Anup’s position, I quote from Borat’s singing of the American national anthem, with the words of the Kazakhstan national anthem:
“… Kazakhstan is the number one
exporter of potassium.
All other central Asian countries
have in-fer-ior po-tas-sium…”
: )
2 Likes
Besides the technical viewpoint which I agree with Anupc, I would not be interested in a separate server. I want a one box solution, like the Bartok.
2 Likes
You’re probably unaware that I actually have a Melco N1ZS that I use with my Vivaldi stack, both connected via a main Ethernet Switch (all fiber ports). I also have a Synology NAS, and a separate Roon Server (M4 Mac) both located in my storeroom elsewhere with their own dedicated Ethernet Switch, but connected over 10GbE fiber to the main Ethernet Switch where the Vivaldi stack and Melco are.
Let me assure you, to the Vivaldi Upsampler’s streaming board, other than the different IP headers between the Servers, there’s absolutely no difference to the IP Packets coming in from any of the 3 platforms when Serving the same file. Zero difference.
If you just think about it for a minute, you’ll realise how silly it is to suggest that the Server makes a difference;
I’m sure you agree, in part thanks to TCP/IP, even a cheap PC/Mac can download files off the Internet and present it completely error-free to an attached screen. No errors, no glitches. no changed data etc. Likewise you’re probably doing your Banking transactions on your Mobile phone, happy in the knowledge that you’re not going to encounter any data errors or changed balances etc.
Have you ever once cared whether the Server at the other end was special??
The ONLY way two different Servers can sound different when serving the exact same file is if one of them in some way alters the bit stream (wrapped in packet). Otherwise, both Servers will deliver the file error-free to the dCS, and they will sound identical.
7 Likes
I find this more than a little condescending. Having a different opinion and a different experience isn’t silly.
I left it saying I think we can agree to disagree - I’m still there - shame you’re not. Ah opinions on the internet eh..:
1 Like
It wasn’t meant to be Condescending at all.
Also, it wasn’t an “opinion”, it is science fact. 
4 Likes
Lost in translation perhaps, but condescending it was and it remans. Sorry, but the ever so superior air and the little flexes don’t add to the debate, they just come across as dismissive of other viewpoints and experiences.
Its a scientific fact (actually an engineering fact) that TCP/IP works in the way it does, everything else around how data is read from disk and how much noise there is in an environment…well…there are variables that the server/streamer market are there to serve, and justify investment in my case. And interestingly in yours as well - feels like from what you’re saying you’ve wasted your money on Melco, and should just have raided eBay for a 2004 IBM Thinkpad, a USB SATA disk and a couple of D-Link Hubs.
I absolutely value your opinion on lots of things on this forum - but we aren’t going to agree on this - the discussion on this matter is nowhere near the binary position you’re presenting to be, but you’re feeling somewhat entrenched in your views on this matter, not remotely open to other opinions or experiences, and life is too short to argue with people on the internet - so lets leave it here 
I started this chat not to make an argument that a network switch like PhoenixNet or Silent Angel or any of the third party devices are good or bad. I experimented with four of them over a three week period and heard a much quieter background, so I spent $4700 on some gear. Whether you like this type of ancillary devices or not, you have to acknowledge that a lot of money is being spent by the DCS Community on this type of gear. So IMHO, from a pure business perspective, DCS should offer a device that competes. It’s money left on the table. That’s all.
1 Like
I respectfully disagree with this position.
Whether or not there is money to be made is just one small factor in entering a new market. If entering said market decreases the market value or position of a brand then it is a bad decision. On Wall Street, there is a Lot of money to be made in the SPAC market, but this product (mostly) defrauds unsophisticated retail investors, and so many firms choose not to enter it.
Similarly, if dCS enters a market where there is no or questionable scientific basis, e.g. audiophile fuses, it would be forced to make questionable marketing statements about said product/market, which could decrease its credibility in its core market, the DAC market, and hurt the company immensely.
In contrast, dCS, IMHO, has wisely invested in added features to its DACs, especially its preamp capabilities, that enable it to make solid claims in this space, and thereby convince some consumers (including many forum users like me) to forgo a preamp entirely and increase spending on the DAC. To me, this approach makes a lot of sense.
10 Likes
No personally, I do not think dCS should get into the server market.
I do love this forum, but I find these types of circular conversations very frustrating. There is no room for subjectivity with properly implemented TCP/IP system. Since dCS has properly implemented ethernet magnetics (as long as the last leg is unshielded cat 5e or 6 and under 100m), the data stream arrives error free. Nothing before the chain can make it sound any better when using ethernet.
dCS has said as much many, many times in this forum.
7 Likes
Not at all. At the time it was introduced, it was the only NAS with passive cooling/without a fan, and so sits quite comfortably next to my dCS stack. My Synology NAS serves a different primary function, as does my Roon Server.
Wasting my money would be if I upgraded the Melco to the N1 or N5. They’re absolutely unnecessary.
On this point of Servers not having a sonic impact, it absolutely is quite black and white.
The irony is that even if you ignore the science/engineering behind it, your own daily lived experience tells you so, but you choose to ignore it
3 Likes