Thank You again for Your detailed answer. You are destroying my certainties: I2S is inferior to USB, PS audio DAC is low quality, Tidal is worse than Qobuz. I dare to think that Your criticism is excluding any Rose responsibility so I will try to use an MSB platinum V dac (I hope You consider this a good machine) from my brother with different hi res material and with I2S and USB connection and see if something changes. Thanks for Your time! Francesco
Well, I am not arguing whether Tidal is better or worse than Qobuz (although they do pay artists much less), just that I personally would not give them money 
PS Audio has great marketing, in person of Uncle Paul but it is mostly based on telling people interesting fairy tales
But in any case, this is just basic troubleshooting. Since whatever it is that affects drop-outs is dependent on upsampling settings, not where you get the files or how, the problem is most likely to be in the Rose, connection, or DAC (or some handshaking between them). Easiest way to find it isd to change one thing at a time. Start with the connection. If that does not change anything try a different DAC. Could be MSB, could be the cheapest $50 Topping, as long as it is known to work. If stutter goes away, youâve found the problem.
As for I2S, the real I2S is a perfectly fine technology. The problem is that the only thing it had ever been designed for is moving data inside a device, over carefully (one hopes) designed traces on a circuit board. Between chips on that board that run off of the same system clock. There are no provisions in I2S for any kind of external connectivity. Thatâs why you get different manufacturers using even different connectors for it (whatever they could design in the cheapest) and even if they do use the same HDMI connector, there is no standard for pin assignment or signaling, so you end up having to fiddle with configurations, custom cables, etc. etc. to even get it to work. And once you do, you are pushing some signal that has never been designed for that, over a cable that has never been designed for that either. It might even work, but always suboptimally. And the other problem with it is that itâs synchronous â timing is in the signal, and over a length of a cable (again, that has never even been designed for that) you get some extra jitter. Generally not enough to have any adverse effect, but still. Modern DACs will usually try to reclock it properly, but sometimes it does not work.Itâs basically the same as S/pDIF, just capable of doing higher resolution signal. With USB data is sent asynchronously, buffered by the DAC, and fed in based on DACs clock. Pretty much any DAC does have a clock that is good enough, and because it sits right next to it, inside the device, effective accuracy is better than from a fanciest audiophile master clock sitting somewhere and sending signal over a cable (where the clock signal also gets more jitter and stuff).
If you look at any pro equipment, or even something really high-end, you do not find I2S connectors there (with possibly an exception of some proprietary SACD player interface, and that is due to Sonyâs licensing requirements; it used to be the only way to get a high-resolution signal, or DSD, without using playerâs built-in DAC, was to use some hacly I2S-like connection because otherwise Sonyâs lawyers would be all over you. With downloads itâs a moot point). For that matter you wonât find SPF cages there eitherâŠ
Iâd say you should try it both with a different connection and with a different DAC if that doesnât help. Changing the connection is probably easier than the DAC, only one cable, so Iâd start with that.
1 Like
The fiber optic cable is the cheapest component! If you read a little the discussions here and on other specialized forums you would find that the single mode connection is the most recommended. In fact, the connection kit sold by Hifi Rose for connecting the RS130 to the RD160 is single mode, as is the RS130 to RSA720 connection kit.
Try to document yourself a little at least here on the forum and you would learn from the experience of other users.
Otherwise you are on the wrong path!
Thank You Boris for Your crystal clear lesson about USB and I2S! I will follow Your advice and try to make one change at a time, so to hopefully identify the problem. I am sorry but also grateful for the amount of time You have dedicated to my issue (and to quench my ignorance). Francesco
Thank You Paul for the extensive bibliography You provided to amend my ignorance. I hope to be able to get into it and understand everything since I am far than an expert. Still I do not think it is possible for me to switch from multi to single mode fiber cable. I have just moved into a new restructured home and my multimode cable is passing through walls for a path of tenth of meters and simply it is impossible to substitute it or pass a single mode fiber cable unless to risk a divorce. So I have to deal with it and I hope it is still possible to have a top audio and video performance. Francesco
Youâre welcome! Hope you will be able to isolate the problem.
As far as fiber goes, you will find here lots of people absolutely convinced that there is some audible difference between brands of fiber, let alone types. These are the same people who are swapping fuses in their gear, and back in the days used to put green marker on their CDs (after âdemagnetizingâ them of course!).
Single-mode is âbetterâ in a sense that if you need to connect to a device 100 miles away thatâs what you need to use; multi-mode canât go that far. Assuming normal domestic distances of maybe possibly tens of meters at most, the only thing single-mode gives you is bragging right of having spent far more than necessary to achieve exactly the same result.
If it could have any effect, somebody would be getting a Nobel prize right now for discovering heretofore unknown laws of nature, not peddling âmusicalâ cables. And every manufacturer of those would have been taken over by various spook agencies for use in spy satellites and sigint, not allowed to sell the on the open market.
If you posted a topology of your network it would be easier to understand your situation. But even so, have you thought about using a switch between the âdomesticâ network and the âaudioâ network? From the switch you could start with single mode fiber in RS130.
As for the âbibliographyâ it is a small part that was handy to me, I think you should at least read the posts here on the forum regarding fiber optics.
Regarding the supremacy of one or the other (single mode vs multi mode) I didnât establish this, I was interested in what happens in pro audio, where almost all recording studios use single mode fiber and in high-end audio all high-end manufacturers recommend it and use it on their top products, from MSB Technology, PlayBack Design, Soulnote, Innuous NazarĂ©, Linn Klimax, Matrix Audio, Lumin, Aurender, Esoteric to Teac.
Who am I to argue with them? You know how it is, if 10 people tell you that youâre drunk, you have to go to sleep!
P.S.
ADOT, a company owned by Melco (now DELA), was one of the first to bring fiber optic kits to market for high-end audio. First with multimode fiber, then they moved to single mode.
Why? Read here: ADOT - MC01 Audiophile Fibre Network Connectivity
[Edited for clarity] Why add a switch to âisolateâ your audio network?
In a switch, your port is isolated. Its not a hub.
So you gain nothing.
If youâre trying to isolate traffic, you use a router and create a subnet. For a home network, really more work than its worth.
What makes you think your home network is at a saturation point?
Note this is a different issue than using fiber. Long runs, plus future proof, fiber can make sense.
None of those are pro grade. Most arenât even âhigh-endâ but just like Rose, mid-market brands trying to appeal to audiophiles by offering quite pointless but shiny âupgradesâ. Or outright snake oil sellers.
Merging (is Merging pro enough?) or Burmester, or dCS (are those high-end enough)? Look around, not a single SFP cage in sight.
That something owned by Melco would sell âaudiophile gradeâ (mind you, not pro-grade or anything, and when you see âaudiophile gradeâ you know itâs a scam) is not surprising. Too bad that their write-up is very much like Direttaâs marketing â a whole lot of meaningless words strung together with absolutely no meaning.
If it makes one happy to throw money at nonsensical doo-dads, instead of buying some good recordings, thatâs their choice. But it is really quite laughable when people not too aware of physics start reposting marketing materials (and not very good at that) as if they were reality. Believing that an âaudiophile fibre networkâ would have any effect on the sound is like believing that buying a pair of Air Jordans will make you a great basketball players.
1 Like
Boris is right. The difference isnât in the fiber. Its in the transceiver.
And that is where you may have issues. Less so if you focus on 1Gb/s speeds. The faster the network the more heat will be generated. Too much heat, you shorten the lifespan of the component.
1 Like
That, too. Although copper transceivers tend to get even hotter than optical ones.
But in general case even 1Gb is wasted on a music streamer, thereâs nothing you could stream to it (until someone persuades audiophiles that DSD4096 is the beesâ knees), and while modular SFP construction makes lots of sense in switches and other network infrastructure equipment, itâs a waste of money in a streamer (or even a pro-grade DAW or interface, which is why you do not find them in real equipment where people pay top get the job done right, not for audiophile bragging rights).
What actual benefits does fiber provide for moving maybe a 100Mbps (with video) stream to a playback device is a mystery that will never be solved!
Boris, Iâll just add my two cents here.
Since Iâm always tempted by T+A products, I know perfectly well that SFP connections arenât necessary for audio. I once had a phone call with them in Herfurt. They said that audio components and networking are, in most cases, closely related, and that an SFP connection isnât required.
As a company, T+A elektroakustik doesnât use SFP (Small Form-factor Pluggable) connections as standard for network connectivity or data transfer in its high-end audio components (such as the SDV 3100 HV, SD 3100 HV, or the R-Series).
Network connectivity:
T+A uses classic RJ45 Ethernet ports for LAN connections or WLAN modules in its network streaming clients, receivers, and DACs (e.g., MP 3100 HV, MP 2000 R).
Audio Connections:
Components within an HV or R series are connected via proprietary HV-Link/R-Link cables or classic analog XLR/RCA connections.
Focus:
T+A focuses on galvanic isolation and interference-free audio transmission through high-quality copper cabling, rather than fiber optic SFP modules, which are primarily used in IT network environments for long distances.
In summary:
SFP ports are not currently standard in T+A audio devices.
And thatâs saying something, considering even T+A doesnât use SFP connections. But as you wrote, even Dieter Burmester from Berlin knows that it offers no added value for his high-end devices.
Thank You Boris, as usual, very clear also about fibers. I will proceed as You suggested by attempts. Francesco
Sorry Paul I am new here so I am just learning. Thank You, I will read the posts as You suggested. Maybe to use single mode cable from the multiswitch could be the easiest way (besides the compatibility): I will try. Francesco
Yes, T+A are correct in that you donât need fiber.
Thereâs a lot of unnecessary hype around the network and re-clocking.
If you have commercial grade networking kit⊠youâre ok.
The only thing about fiber is that its now to a point of where homes are now being fitted out w both CAT 8 and Fiber runs to each room. Or just fiber. (Note: Long CAT 8 runs will degrade the bandwidth capabilitiesâŠso you could see CAT 6e wiring but because the price difference is negligible people just go CAT 8. )
So if youâre going all fiber then having an SFP or SFP+ port can make sense.
Its more of a future proof thing. I wouldnât be surprised to see more of it as the cost of the switches go down. The only advantage of the âaudio gradeâ switches is that they run quieter. The commercial switches are more concerned w cooling and their fans are louder. They are meant to be in machine rooms.
1 Like
I use Burmester as an example because while something like T+A (or Lyngdorf, or quite a few others) are great, properly designed devices, they arenât quite at the level of outright luxury goods like Burmester. A typical Rose customer can at least consider getting a T+A with some effort. Burmester is quite a different level, price-wise.
They still do know how to design things properly, rather than just throwing useless bling. SO, no SFPs there.
Youâre welcome good luck. Let us know what you find.
As @Smegel mentioned, it would be good to have some information about your network setup â how large the distances are, how many devices are on it, etc. before making and solid suggestions.
Some things to remember though. Ethernet is galvanically isolated by definition (well, unless you are using Cat 8 cables, which can give you ground loops) so unless you are concerned about a lightning hit traveling a through your network no fancy switch or fiber will give you any better âisolation.â
Due to the way Ethernet (and UP networking in general) works, âaudiophile switches,â especially ones with master clock inputs, cannot do anything, and do not do anything useful. They are literally solving a problem that does not exist.
A lot of marketing material you read is just that â pure marketing, aimed at the target audience of people who are not necessarily technical but can afford whatever is being peddled. For example, that ADOT thing @SPaul provided a link for. It had (probably) been run past their legal department and is not âactionableâ in a sense of being a total lie. Mostly because it does not really mean anything, but also because they are using a lot of smart-sounding fancy science-y words like âjitterâ that are technically correct in some very narrow sense of existing. They just neglect to mention that even if it is present (and if you are using e.g. multi-mode at lengths it has been designed for, it is not there to any meaningful degree), it quite literally does not affect anything coming out of the other end of the device. It is, once again, a solution in desperate search of a problem.
Or even some decent home or SOHO kit. For most people professional network equipment is not only overkill but will likely create more problems.
Cat 8 (if it is real Cat 8 of course) not only costs more but is more expensive to run. Unless we are talking some Beverly Hills McMansion, pretty much all houses, even Texas-sized ones are perfectly within range of Cat6a for anything a regular consumer will need in the foreseeable future.
With the vast majority of home devices using good old built-in RJ-45 connectors, fiber really makes sense only for the backbone between switches.
To put it cynically, if you donât know what SFP (vs. SFP+ vs. QSFP) is, you donât need it.
Most reasonably sized SOHO switches are passively cooled, too (well, 90% or so of âaudiophileâ switches are just some TP-Link, at best, usually itâs the cheapest one they could find, repackaged in a shiny box with gold-colored plating on the connectors to look expensive) so even that isnât much of an advantage. If your house is large enough that you need to install aggregation switches with high-speed backbone links, you have a closet to put them in, tooâŠ
1 Like
You can get commercial grade kit that isnât overkill for home.
You just canât get it from your box store. Usually a smaller shop or if you want to spend too much call CDW.
CAT 8 doesnât cost that much more. Go price out a 1000â roll. Then compare to what youâre spending on your home computer and AV equipment. Fiber is actually going to be cheaper, but your runs would have to be purchased individually unless youâve got the gear to cut and repair fiber. Not something that should be left to armatures.
Want expensive⊠do a 500â run from a condo unit to the telco closet where the builder required that it be put into its own conduit and run separately from their telco runs. That was CAT 5 done in 1998. That was expensive and lucky for me, I was able to snag some of the existing copper wire where as AT&T was putting in optical switches. This was so I could actually bring in a higher speed internet connection for a SOHO.
And if you thought about it⊠youâre running fiber because youâre future proofing your home.
So when you come out of the wall, you go into a switch where you have some RJ45 and one or two uplinks that are fiber. Now you have fiber to the room and then RJ45 for your other gear.
Again the costs are relative. And if that bothers you⊠remember you also pulled some copper wire so you leave the fiber unused . Pulling copper and fiber at the same time is going to be roughly the same cost. The cost of the fiber is relatively cheap when you consider the cost of labor, patching and painting.
Knowing you⊠youâll run w wi-fi 7 and call it a day. However⊠if youâre doing any SOHO or work at home and actually work in R&D in IT⊠(cough, cough) having this as an option can pay off if youâre doing anything that requires speed.