The theory is common mode noise. See above.
Just curious: have you yourself tried different Ethernet cables? You might be surprised that the audible differences are very comparable to those between different analogue cables. At least in my system they certainly are.
3 Likes
I have been the happy owner of a Rossini Player & clock for a month now, and this is also my first experience with streaming.
I listen to CDās, Qobuz and a few DSD tracks. I have a very good and stable Internet connection: 2.5 Gb/s on Optical Fiber directly into my router (the one the Internet company provided), which is also my only switch and wireless access point. DSD tracks are on a new MacbookPro M1 Pro with minimserver and are wirelessly streamed to the router. A measly Amazon 15 euros 5 meters Ethernet cable goes from the router to the Rossini. It sounds great. My other cables are Hijiri (a Japanese brand) and are not cheap. I can hear the difference in speaker, interconnects, and power cables.
What keeps me from buying expensive Ethernet stuff is quite simple, the kind of consideration typical of the renown Maximegalon Institute of Slowly and Painfully Working Out the Surprisingly Obvious. It goes like this:
The streaming guy at DCS says that we need unshielded, certified cat6 cables, where the keywords are unshielded and certified. Now, if Ethernet stuff actually made a difference, why should he say that it doesnāt? The company would lose sales, because potential customers would not be experiencing the full potential of their 30k or even 100k boxes! And those around him at DCS should be totally crazy in letting him say so! They are not saying that power cables, usb cables, and interconnects do not make a difference. Just the Ethernet cable. I think they know Ethernet streaming and their gear better than I do. I am sure they have the budget, contacts, logistics, and time to test cables that I will never hear of. I also build my own computers and know how Ethernet works, so I have some problems accepting that the Ethernet chain can influence SQ. Finally, as a professional mathematician with good engineering and Physics background I have no problem understanding the maths and some technicalities behind digital transmissions: it all confirms what DCS say.
Nevertheless, someday I am going to test some mid class Ethernet cable. I suspect it wonāt make any difference whatsoever. But if it does, it has to be HUGE. Not subtle. HUGE. Otherwise, I know itās not real. I know the feeling of hearing something āthat only the chosen ones can hearā, and then say that the difference is obvious, while it isnāt obvious at all, just because it makes you feel such a hi-end guy ā¦ Been there, done that.
Nope, this time it has to be SO HUGE that I donāt have to search for it: it has to come out and look for me and call for my attention.
I will let you know.
A.
6 Likes
Sure, always try before you buy. My advice for your listening test: try a VERY high-end Ethernet cable too. I recently bought a CrystalConnect Monet and itās crazy good. If itās really true that āit all makes no difference at allā, this cable will sound exactly the same as a 5 dollar Belden, correct? You will be laughing your pants off. Really.
1 Like
While searching for an ethernet cable, I came across the Polish manufacturer GFmod Audio Research
Audio. They offer several types of ethernet cables. Initially, I decided to use the cheaper GFmod Advanced LAN MKII cable. After plugging it between the switch and BartĆ³k, the sound of the system improved in every aspect. Naturalness and analogy, tangibility are the most noticeable aspects of improving the sound of my system. Encouraged by such good results, I borrowed a more expensive cable: GFmod Advanced Pro LAN Ethernet. To my surprise and joy, it turned out that the sound of the system improved even more. More information about the recording, the space in which it was realized, the music scene expanded deeper, the colors of the instruments shone. I was very skeptical about the influence of ethernet cables on sound quality but this cable is worth buying - I donāt know how it is possible to rent a cable outside Poland, but I think it is possible. I encourage you to test - and you will not have to send it back. The improved wording will justify this expense. I hope that I am not breaking the forum rules by praising the product - but the effect I have obtained makes it impossible to stop myself. I read this thread with interest and thatās why I decided to do this experiment with a cable replacement.
2 Likes
Thank you for your answer. Unfortunately I am not going to invest 6900$ in an Ethernet cable (I need 5 meters), and my ethics prevent me from testing things I am not going to buy anyway. What I am trying to say is that if I hear a difference between a regular UTP cat6 cable and this one, and this difference is so significant to justify that expense, which is close to the amount I paid for the Rossini Clock, then there must be something totally wrong in the Ethernet card of my Rossini. And my question stands: if this cable transforms the Rossini (or Vivaldi, or Bartok) into something so good, why donāt the guys at DCS say so? They would be shooting themselves in the foot!
But sooner or later I will test a middle tier cable. If it makes a difference, Iāll be there to take note, and Iāll tell you.
A.
2 Likes
Thank you, unfortunately I was unable to find an English translation, and I donāt trust Google with technical thingsā¦ Any hint?
A.
The Ethernet card in all dCSāes is pretty standard fair. Itās fine, but nothing special. Costs are around EUR 120,-. I myself am not using the streaming functionality from the dCS, but an Aurender W20 which sounds way better - as it ought to considering the price. My point is: one streamer is not the same as the other, and the same goes for Ethernet cables. Whether one is willing to spend a lot of money on either is a different thing of course. But I can assure you that thereās a lot of SQ to be gained with an even better streamer than the built-in one of dCS. The same goes for better Ethernet cabling, switches and their power supplies etc. They all have a very audible impact.
Iām not sure I agree with the āas it ought to considering the priceā.
As it has been said many times already, there are so many levels of abstraction in the streaming protocol, that the data (data, not music) either arrive correctly to the DAC or not. I can accept that a good ethernet cable design can limit noise, but itās not noise that can modify in any way the data reaching the DAC. Maybe itās noise that can somehow make it to the output and subtract to the overall quality of the music. But I digress.
If the Aurender is used as a streamer, with no signal processing, there is no single reason why it may sound different than the internal streamer of a Vivaldi. Otoh, you may like the processing that is done in the Aurender more than what you can have in the Vivaldi, and thatās totally subjective*. Otherwise, the next step could be to ask for a dedicated Internet connection just for music. And I suspect that someone might say that the difference is totally audible, and would be glad to pay tenfold for the magic hi-end connection. No. Itās data. Itās what I have been working with all my life, that and mathematical analysis. This is not simply a digital signal. Itās data, once it makes it safely to the final buffer, itās all the same. Jitter doesnāt matter. You can even ask a resend and put things in order afterwards. It will sound exactly the same. You can save it on your hard drive, and from there master a CD. If the mastering succeeds, it will be the same music. Identical. As someone else already said, I can pay many thousands euros for a speaker cable, because I can afford it, because it sounds better than anything else I have tested, and because there are plausible reasons why it makes a difference.
But I donāt buy the ājust listen with your own earsā mantra. Nothing personal, and I thank you again for your time and your answers, I just donāt want to end up with tuning pebbles, magic clocks, resonant bells and whistles and all the junk that the ājust use your earsā party usually collect in their listening rooms.
Now I have to order the 1.5mt Geistnote clock cables and convince my wife that I need them to replace the generic 0.5 mt blue ones I have. She hates long cables, and she is going to hate the color of the new ones.
Letās hear if the clock cables make a difference.
A.
*Also, if you let the Aurender build the digital signal and then go into the DAC via a digital cable, there may be a difference, but I assume from what you write about ethernet cables that you are using an ethernet cable, so you are using the Aurender as a pure streamer connected to the DAC via ethernet, and not as a ātransportā
2 Likes
The English version of the site is indeed missing, which of course is a problem. I am convinced that writing a e-mail in English will not be a problem for the company. However, it seems to me that there is not much technical information on the page describing the cable. There you can find information about what materials were used for the construction of GFmod Advanced PRO LAN and the philosophy behind this idea.
1 Like
I appreciate your answer and scepsis. Everyone is entitled to their opinion. A few notes though:
-
Aurender and dCS have closely cooperated in designing the W20 - especially for their Vivaldi class (which I own, but also when I had the Rossini the upgrade was striking). dCS Even created a special Aurender solution to couple the clocks of the Aurender and the DAC in case of an isosynchronous connection. Itās also in their manual. Now, if both parties would think thereās nothing to be gained with an external streamer, why bother at all?
-
And yes, jitter is one of the very audible influences in digital audio. How do I know? Because I have the Vivaldi clock, which in itself is a big improvement on the Rossini clock - also when connected to the Rossini Player. I tried it myself at home. Question: if the additional Rossini clock didnāt matter to you, why did you buy it? Surely not just to please the neighboursā¦?
Now comes the funny part: I have added a second clock to my system, which is a 10MHz reference clock. Itās connected to the Vivaldi clock. Now, despite your technical reasoning really everyone can easily hear the improvement here when I switch it on or off. No audiophile ears are required whatsoever.
As a note though I can imagine you are referring to the relevance of jitter in Ethernet-based data traffic only. At least in theory you are absolutely right that jitter should not be relevant there. Although my experience with high-end switches (with specialized FiFo buffers) seems to indicate that might not be irrelevant either - to the ears at least (Iām sorry to say that to you). -
You are right; data is data. Iām working in IT for decades myself. However in audio there are phenomena that are outside of the data that clearly influence SQ. Besides jitter (indeed totally irrelevant in generic data traffic), but also common mode noise in the analogue domain. The latter is very easy to hear. Want to know how? While playing a track through your buffer, just unplug the Ethernet cable. You will very likely hear a clear difference/improvement. If you donāt, donāt spend any money on a better switch or Ethernet cable. You will NOT get a better sounding Ethernet connection than none at all. That is the benchmark. No Ethernet connection means no common mode noise entering your system. Just try it and let us know what you found.
-
You really should try much better clock cables than the standard ones provided by dCS! Just try them and be amazedā¦
By the way: the Aurender is used as a transport. Itās connected to the DAC through a very high-end USB cable. Yes, that quality compared to a standard ā¬ 2,- USB cable is very audible to anyone that hears it. I have tried it with the most cynical listeners that I know and they couldnāt believe their ears! 
Cheers!
V.
3 Likes
Hi Vincent, thank you for your answer and clarifications. I think we agree on most points.
As I wrote, I was under the impression that you were using the Aurender as a pure streamer, connected via RJ45, not as a transport connected via USB. This changes things a bit.
- dCS and Aurender worked together and they probably built the best solution for an isosynchronous USB connection. It reduces jitter which is indeed (2) a very audible influence in digital audio (but not in digital data transfer).
About the Rossini clock: it matters a lot to me, especially after a fortuitous incident a few days ago (more in that in another post). So yes, we agree on the importance of a jitter free (achievable only in dreams) audio transmission.
You got it right about my skepticism on jitter and etherrnet. I think they do not belong in the same sentence, or even on the same page.
Then you say āNo Ethernet connection means no common mode noise entering your systemā. Totally agree. And I will do the test you suggested: I find it intriguing and clever.
I never installed the stock cables. I bought some better ones, but only 0.5 mt long. Now I am going to try the 1.5 mt Geistnote and see (hear) what happens. Yes I believe clock cables can make a difference, even though they do not carry any music.
Finally, it may well be the case that a well designed and built USB cable minimizes both jitter and noise, so I have no problem believing that it can make a difference.
A person I know, who was once an engineer at JVC and now has a business of his own, refuses to add a USB port to his hi-end DAC because he says that the USB port is the best way known to man to inject noise into a sound system.
So turns out we agree more or less on everything!
Thank you again for a very nice exchange of opinions.
A.
2 Likes
Thanks for your response. Indeed we do agree on most points. Good to hear that you already started upgrading the clock cables from the beginning. The standard ones do work, but thatās about it.
As a sidenote: the dCS USB connection is asynchronous (as opposed to isosynchronous), which means the DAC is dictating the timing of the data. Thatās a good thing, especially with the two high-grade clocks connected to the DAC. The special connection that I described above is meant for isosynchronous connections, i.e. SPDIF, AES/EBU and the like.
Do let us know your findings when disconnecting the Ethernet cabling. Iām really curious!
Cheers,
Vincent
2 Likes
Hi all, I am going to place an order for the Blue Jeans Ethernet Cable from their website. Is the one labelled
Cat 6 Patch CordsāBJC C6P
Bonded pairs, Cat 6, 250 MHz; Rated CMR for installation
Every cable individually tested, with report
the right one? Is it actually UTP?
Thank you
A.
1 Like
Certified Cat 6 is available in both shielded or unshielded varieties. I have looked at the Blue Jeans website and am not able to see exactly what they are offering. Note that the cable is custom made for Blue Jeans by Belden so there isnāt a Belden catalogue item to assist in identifying if UPT or not. I would email them before ordering to check or buy Cat 5e which is definitely unshielded and is perfectly up to the task you are putting it to.
2 Likes
Thank you thatās what I will do.
Edit: I did, and here is the (pretty fast) answer about the cat6 cable:
That is the correct cable- we donāt use any shielding.
ā Jeff
Blue Jeans Cable
A.
1 Like
Ok itās done. I ordered and received the cable yesterday.
So now I know that noise can pollute my system via an Ethernet shielded cable.
I did not anticipate a difference of this proportion, which means that I probably live in a very noisy place. On the other hand, I think since cellphones and wireless networks have been around, a non polluted place is hard to find.
Itās audible, itās pretty much obvious. A great investment.
Thank you to the Forum members for their great advice.
A.
2 Likes
Iām in the same boat. I have a new Bartok on order and Iād like to lay the foundation before it arrives. Iāll be going from ISP modem into a Google wifi router ā ethernet switch. In the switch Iāll plug in my mac mini and cat 6 hardwire into the Bartok as a Roon endpoint.
Sanity check - does the following meet the recommended guidelines for a switch and ethernet cable? Amazon cable matters CAT6 and TP-Link TL-SG105 5 Port Ethernet switch.
Iām new to the home networking bit so any guidance will help!
The Cat6 cable on Amazon may be OK for connecting the router to switch but may not be for connecting the switch to Bartok. CAT 6 is available in both shielded( F/UTP) and unshielded(U/UTP) types. For the connection to Bartok the cable should be unshielded. Unfortunately the description on Amazon omits which type it is. NB: CAT 5e is unshielded and is adequately specified for audio data distribution should you prefer to buy a cable of that type.
I will not comment on the switch as there is better expertise available on such things from other members here.
The Amazon Cat6 cable pictured is an unshielded one. Perfect for a dCS DAC.
2 Likes
@PAR thanks for your reply - I just realized that CAT6 is overkill for the Bartok which only has a gigabit port. Iāll look for CAT5e per your suggestion.