PS - I guess the thing with chasing the idea of eliminating noise is that everyones situation will differ in terms of how much noise is really present.
I live in an apartment in central London so in theory that makes me a prime suspect for loads of noise. But in all honesty I added the Baaske to try and eliminate or safeguard against something I wasn’t even sure was an issue for me!
Also the noise will probably vary throughout the day depending on many other factors…
Somewhat related I guess is how you segment your network. Again - this is an idea I picked up from Sooloos users.
If possible, you should put your network player and end-point on a dedicated switch. This will limit network packets to those related to just those two devices. So the network adapter in your Bridge is not working filtering packets to/from your printer, pc, fridge etc…
Like this:
Wireless Router - Switch - Roon Server
- GISO Isolater - Bridge II / DS DAC
Hope that makes sense 
Mark
markus46 saidIf possible, you should put your network player and end-point on a dedicated switch. This will limit network packets to those related to just those two devices. So the network adapter in your Bridge is not working filtering packets to/from your printer, pc, fridge etc…
Like this:
Wireless Router - Switch - Roon Server
- GISO Isolater - Bridge II / DS DAC
100Base-T Ethernet is plenty more than enough to handle the data packets of audio files as well as those of other kinds of network traffic. Therefore I don’t think it’s necessary to have a network path dedicated to the route between your music server and music renderer (in Markus46’s case his Roon server and his Bridge II/DS DAC). Given the ubiquity of WiFi the chances are that most network traffic in your home isn’t over wires.
On the other hand I’m an audiophile too so I understand the quest for ‘perfection’. I just wouldn’t want other readers to be spending time and money on dedicated Ethernet cables unless they think there’s a problem to be solved. The question to ask yourself is what else is on the end of that network cable running to your renderer.
Peter
Sorry Peter but I’m with Mark on this one.
My rack contains the AV gear including TV as well as the Hifi gear.
Cat 7 ethernet to dedicated switch for NAS and Renderer only (on iFI Ipower).
Cat 6 ethernet to separate switch (on stock psu) for my Oppo blu ray, AVR, PS4, Sky Q and TV.
Don’t have a clue whether it makes an iota of difference but for an extra 20 quid it seemed a reasonable idea to get another switch.
Cheers,
Alan
Aha! I’ve done similar things myself, such as buying (for the vicinity of my hifi and AV gear) only a switch with a metal case (to reduce RFI output) and replacing its switched mode power supply with a linear one.
So I’m being cheeky when I ask, “Did it [the use of a dedicated switch] sound better?”.
To be honest I’m as much an equipmentophile as an audiophile. The audiophiles are spending their free time listening to music!
Peter
To share my experience:
connecting my Bridge2 via TP-Link MC110CS optical transceivers made very nice improvement in SQ. However I found out that when P10 (powering all audio gear) was connected to network switch (the same one as bridge) the SQ improvement was not so noticeable (if ever) - thus most of the time my P10 is now unplugged from ethernet as I don’t want to buy another optical transceivers (powered by additional two PSUs…) just for the P10 to have network.
In next step I’ve replaced standard switched PSU bricks to my NW switch and also to optical transceivers with linear PSUs - I was hoping this would help to get P10 connected back to network without worsening the SQ - nope. Linear PSUs made either no, or only very minimal change to SQ (probably due to the P10 is anyways filtering the noise the switched PSUs has been putting back to power line).
Now I have connected Bridge2 via the MC110CS to dedicated network card in RoonServer - i think this helped SQ a bit in comparison when B2 was connected to home LAN (about 20 devices connected) via the same MC110CS - but not in such big extent as the first having B2 connected via MC110CS.
My most important lesson I learnt here is - do not galvanically connect anything more than essentially required to audio chain (including network, not just audio).
Peter saidPeter that is ABSOLUTELY correct. Granted it is possible to saturate a 100 Mb network, but it's hard. A good rule of thumb is that networks start to get saturated at around 40% capacity. That's the point where retransmissions start to add to the traffic, which is a problem that compounds itself. There are a lot of variables that can change that. But it's a good rule of thumb.100Base-T Ethernet is plenty more than enough to handle the data packets of audio files as well as those of other kinds of network traffic…
Peter
So when HD movies clock in at around 5Mbs… if you are going to be watching 8 HD videos simultaneously through that router, then sure, get a dedicated switch for your endpoint. But short of that, there is zero need for it.
Wow! I just realized what I just posted was incorrect. It was correct if you were using hubs for you network. But nobody uses hubs anymore. Duh! I’m getting old…
Everyone uses switches now (vs. hubs), and collisions are nearly impossible on them. Subsequently they can operated pretty much at 100% of their rates capacity. So 8 HD movies was wrong. Instead - with a 100 Mb switch, It’ll start to matter around 19 or 20 simultaneous movies.
And in all fairness to the line of thinking, a separate network for your endpoint would have had merit back when our LANs used hubs and were 10Mb. But those days are long behind us.
Great thread ! I had bought TP link optical Ethernet devices over 6 months ago but they were T1000 and didn’t work.
I have been meaning to get the correct 10/100 units and optical cable to isolate the BII /DS from my Mac and router.
I read about the better power supplies but now After reading here about the battery option I will order that as well.
Can anyone confirm I want the dual or single optical cable?
I have a long toslink cable from my Mac pro 2013 I sometime use for listening to simulcast dead shows, and decided to try it with Jriver and unplug the Ethernet.
While it did seem smooth (since it was isolated from Mac), it wasn’t as resolved as from the Bridge II
Incidentally I did try listening to LPs with the DS powered down and Ethernet unplugged and thought it may be a small amount better. The real test would be to have a Friend unplug the digital gear while I’m at my listening seat to hear the benefits, if any.
I truly believe the isolation Will be benefitial in removing the last traces of digital hardness.
Peter saidIn a spirit of exploration I’m going to challenge that(!)
100Base-T Ethernet is plenty more than enough to handle the data packets of audio files as well as those of other kinds of network traffic. Therefore I don’t think it’s necessary to have a network path dedicated to the route between your music server and music renderer (in Markus46’s case his Roon server and his Bridge II/DS DAC). Given the ubiquity of WiFi the chances are that most network traffic in your home isn’t over wires.
On the other hand I’m an audiophile too so I understand the quest for ‘perfection’. I just wouldn’t want other readers to be spending time and money on dedicated Ethernet cables unless they think there’s a problem to be solved. The question to ask yourself is what else is on the end of that network cable running to your renderer.
Peter
As you and scolley have noted, there’s plenty of bandwidth in a typical home network to handle orders of magnitude more traffic than that of an audio stream. So if there is an improvement to be had by isolating network traffic between server and end-point in this way, the reason for the improvement must be due to something other than approaching the limits of available bandwidth?
I can’t say there is an improvement because I’ve never set up my network in any other way. However here is a review of an audio server (MD600) and end-point (MS600) that touches on the audible changes that occur with network complexity. http://www.hifi-advice.com/blog/review/digital-reviews/network-player-reviews/meridian-sooloos-md600-and-ms600/
I’m not sure if your reference to dedicated ethernet cables is a typo, but to be clear - I’m discussing the isolation of network traffic via a dedicated switch. And speaking of the expense, a 4-port switch is a mere drop in the ocean compared to cost of almost any other component in our stereo systems. But I do appreciate your desire that other readers not waste their hard-earned on devices and techniques that can’t possibly work:)
If you search the forums at meridianunplugged.com you will find reports from people that have tried this. Some report an uptick in SQ. Some report no appreciable difference and others (of course) poo-poo the idea without having tried it. I imagine that there might be some discussion of that at computeraudiophile.com as well but I’ve not searched.
What else is on the end of that network cable running to my renderer? I’m probably going to fall into a carefully planned logic trap here, but the answer is a GSIO LAN Isolator. And on the end of that device is a 4-port switch.
Cheers
Mark
markus46 said… if there is an improvement to be had by isolating network traffic between server and end-point in this way, the reason for the improvement must be due to something other than approaching the limits of available bandwidth?
I can’t say there is an improvement because I’ve never set up my network in any other way. However here is a review of an audio server (MD600) and end-point (MS600) that touches on the audible changes that occur with network complexity. http://www.hifi-advice.com/blog/review/digital-reviews/network-player-reviews/meridian-sooloos-md600-and-ms600/
I’m not sure if your reference to dedicated ethernet cables is a typo, but to be clear - I’m discussing the isolation of network traffic via a dedicated switch.
…What else is on the end of that network cable running to my renderer? I’m probably going to fall into a carefully planned logic trap here, but the answer is a GSIO LAN Isolator. And on the end of that device is a 4-port switch.
In my reference to a dedicated Ethernet cable I meant an Ethernet cable providing the path between one’s source of music files (music serving computer, NAS, broadband router [for streaming]) and one’s music renderer, but which had no other data traffic on it. For example my setup looks like this…
Home ‘office’: broadband router / WAP / music server / switch (for extra ports – I’ve a lot of kit in that room) > long, single network cable (through walls, etc.) > Lounge: switch / DSjnr / Sonos Connect / smart TV / Blu-Ray player / Apple TV box / WAP
(WAP is WiFi access point. I have one in the home office and one in the lounge for improved coverage.)
That long, single network cable provides all data for all of the devices in the lounge. But as I’m not worried about network bandwidth I’m content with my arrangement. In the context of my setup a dedicated Ethernet cable would mean an extra one between the switch to which my music server’s connected and my DSjnr alone, but my assertion is that I don’t need one and that others need not spend their money on similar.
This is only peripheral to galvanic isolation, though. Whether one were to use a dedicated (as defined above) network cable or a cable for all purposes (my arrangement) there will still be the potential for injection of electrical noise into a DS/jnr through its network cable. And so we’re back to optical isolation or a transformer isolator for that last Ethernet leg.
Peter
Hi emailists,
You need the duplex cable. I use the following one which was cheap and worked immediately:
https://www.amazon.co.uk/dp/B00J0T6D08/ref=pe_385721_37986871_TE_item
C2G 1m SC-SC 10Gb 50/125 OM3 Duplex Multimode PVC Fibre Optic Cable (LSZH) - Aqua
Just check the connectors match those on your FMCs as they come in different sizes (SC, LC etc).
Also FWIW I use a cheap 6 quid digital plug timer to charge the battery power supplies. Set to charge between midnight and 7am when I’m not listening to music. Just saves messing around switching the power on/off for listening.
Cheers,
Alan
@Peter - You are indeed correct. My concern is that people not waste their hard earned money. Yet you are also correct about my objections being based on assumptions of inadequate bandwidth. So I must say that I was not referencing the galvanic isolation that a dedicated switch might impart, and the sonic benefits that might impart. My understanding of networks relates to the movement of data. I have little/no understanding of network devices as electrical devices. So I must say, if it has that benefit, I would be the last to know. However, I will pass along this little tid-bit…
I know a guy (on another forum) that owns a high-end audio concern. When I mentioned to him that I was trying out the TP-Link media adapters, he wanted to know what power supplies I was using, and followed that with the following comment…
“My experience with the optical isolators has been neutral at best. In many cases they actually made it sound worse. While I like the idea of optical isolation the reality of it is that you’re at the mercy of the hardware in that little box and its associated power supply. I’m finding that those things can inject more noise than a quality switch.”
I mention it only because his concluding thought might indicate some value in a quality switch. I note that he did not say dedicated. Hmm… maybe that’s implied?
Hi @scolley
Interesting. And only goes to show what a minefield this all is. I put my NAS and Rendered on a dedicated switch so that traffic from other items didn’t impact them negatively. But taking your bandwidth comments on board this may or may not be a factor. Whereas not directly attaching 3 or 4 other devices to the switch might actually be an accidental advantage from the galvanic isolation perspective! It’s no wonder we joke about ‘nervosa’ as trying to think thru all those factors could drive you round the bend…
I’ve considered the Audiophile switches - Paul Pang and co. But concluded (rightly or wrongly) that better power supplies were the main benefit they brought to the table. I therefore ended up going a normal switch and the iFi Ipower as a more reasonable (70 bucks rather than 200+ bucks) compromise.
I’m not aware of any ‘audiophile’ FMCs yet but put them on Stereophile, Audiostream etc with a complimentary review and I’d probably be tempted to part with that hard earned cash. A fool and his money indeed 
Cheers,
Alan
Parhaps PS audio’s next product should be an audiophile optical Ethernet interface designed by TED. Or perhaps another method of isolation?
ACTUALLY - the next bridge with MQA support should have isolation built in.
That would be the best solution - even if a dongle had to plug in the back if there’s not enough room on the board .
I can relay my experiences with galvanic isolation since I’ve been down that rabbit hole. I definitely found benefit in using optical isolation between my generic switch and my Mac Mini with FMCs powered by UpTone LPS-1 with splitter feeding both modules. I didn’t, however, find any noticeable benefit in optically isolating the direct “bridged” connection from Mac to the DS Sr so I removed it from the system. My guess is that it already been isolated from the switch and the last run of optical fiber was redundant. And, yes, the LPS-1 feeding the modules had a slight benefit in SQ compared to standard wall wart. I also use Cat7 ethernet cable throughout, which I also think slightly helps SQ. All of these little things add up to a pretty significant improvement in SQ of Roon with Tidal to Bridge II compared to no optical isolation and no direct bridged isolation and with Cat5.
AMG - are you saying the Ethernet goes direct from the Mac to the bridge ? My Mac pro has two ethernet ports but I connect the Mac to a router with wifi built in and that handles the modem and connection to the BII. I would think the Mac is just as noisey as the router and should be isolated as well.
scolley saidI'm sorry for any confusion. My suggestion re dedicated switch for server and end-point devices was not to do with galvanic isolation. It was to do with network traffic isolation which while not exactly on-topic was (I thought) somewhat related and potentially of interest.@Peter - You are indeed correct. My concern is that people not waste their hard earned money. Yet you are also correct about my objections being based on assumptions of inadequate bandwidth. So I must say that I was not referencing the galvanic isolation that a dedicated switch might impart, and the sonic benefits that might impart. My understanding of networks relates to the movement of data. I have little/no understanding of network devices as electrical devices. So I must say, if it has that benefit, I would be the last to know. However, I will pass along this little tid-bit…
I know a guy (on another forum) that owns a high-end audio concern. When I mentioned to him that I was trying out the TP-Link media adapters, he wanted to know what power supplies I was using, and followed that with the following comment…
“My experience with the optical isolators has been neutral at best. In many cases they actually made it sound worse. While I like the idea of optical isolation the reality of it is that you’re at the mercy of the hardware in that little box and its associated power supply. I’m finding that those things can inject more noise than a quality switch.”
I mention it only because his concluding thought might indicate some value in a quality switch. I note that he did not say dedicated. Hmm… maybe that’s implied?
Agree 100% with the guy you quoted re a quality switch.
Here is a link to the exact same conversation we are having re galvanic isolation from another forum:
LAN isolator anyone? - The Hitchhikers Guide To Meridian - Powered by MrTechGuy.co.uk
It’s runs to 28 pages so we have a bit of catching up to do
If you read this thread, you will come across quite a few posts regarding the audibility of changes when a LAN isolator is introduced at the end-point. From memory, I think the discussion is around passive devices like the GSIO, not active devices.
When my local dealer was still using a Mac Mini [he is now using and selling Aurenders], he recommended using a Netgear Wi-fi extender. From there he would plug in to the DAC.
I never heard a comparison, but I do trust him. I know that even with the Aurender, he recommends using a USB to coax converter. He told my buddy who bought an Aurender that he didn’t like their USB interface. In the store he either uses an Audio Research Dac 9 or one of the Moon DACs.
While I am quite fond of the Dac 9, I don’t like the Moon DAC I did hear. I was familiar with the whole system, and I was hearing a glare in the midrange. My only thought after listening to that setup was I would of had listener fatigue in a very short time. My system I can, and do listen for 6 hours or more, at a time. My butt gets tired, but my ears are fine.
Has anyone tried a Netgear extender, and then running the Ethernet cable to the Bridge? The only downside I can think of is that you have to plug the extender in. You would also need decent Wi-fi. We just went from Time Warner to Spectrum, and I get 60Mbps wired and close to 40Mbps over my Wi-fi.
Once I get my Raspberry going, I am going to try an extender, then Ethernet to the Raspberry and USB to my Benchmark in the bedroom system. I’m hoping I can get it to stream Tidal.
I'm convinced that changes in sound quality when network connections are changed (or when servers are changed) are a consequence of changes in electrical interference injected into audio equipment through network connections. The sources of interference are many (aerial effects of data cables and PSU cables to network components, induction from other cables and equipment, switched mode power supplies and probably more). So descriptions of changes in sound quality when network topology is changed need not be doubted but equally may be no guide to the effects of network changes in your own setup.In my reference to a dedicated Ethernet cable I meant an Ethernet cable providing the path between one’s source of music files (music serving computer, NAS, broadband router [for streaming]) and one’s music renderer, but which had no other data traffic on it. For example my setup looks like this…
Home ‘office’: broadband router / WAP / music server / switch (for extra ports – I’ve a lot of kit in that room) > long, single network cable (through walls, etc.) > Lounge: switch / DSjnr / Sonos Connect / smart TV / Blu-Ray player / Apple TV box / WAP
(WAP is WiFi access point. I have one in the home office and one in the lounge for improved coverage.)
That long, single network cable provides all data for all of the devices in the lounge. But as I’m not worried about network bandwidth I’m content with my arrangement. In the context of my setup a dedicated Ethernet cable would mean an extra one between the switch to which my music server’s connected and my DSjnr alone, but my assertion is that I don’t need one and that others need not spend their money on similar.
This is only peripheral to galvanic isolation, though. Whether one were to use a dedicated (as defined above) network cable or a cable for all purposes (my arrangement) there will still be the potential for injection of electrical noise into a DS/jnr through its network cable. And so we’re back to optical isolation or a transformer isolator for that last Ethernet leg.
Peter
So descriptions of changes in sound quality when network topology is changed need not be doubted but equally may be no guide to the effects of network changes in your own setup.
Of course!
This is only peripheral to galvanic isolation, though. Whether one were to use a dedicated (as defined above) network cable or a cable for all purposes (my arrangement) there will still be the potential for injection of electrical noise into a DS/jnr through its network cable. And so we’re back to optical isolation or a transformer isolator for that last Ethernet leg.
Agreed. I hope I didn’t come across as advocating a dedicated switch instead of galvanic isolation.