Here is my set up: PC running JRiver-ethernet cable 1-verizon fios router-ethernet cable 2-BuffaloTech NAS, and verizon fios router-ethernet cable 3-wall (in my study)-ethernet cable 4 (in the wall to the basement)-ethernet switch-ethernet cable 5 (in the wall to my music room)-wall (in the music room)-ethernet cable 6-Directstream DAC Junior.
Therefore, total 6 ethernet cables, 4 and 5 are long runs in the wall, not upgradable at the moment, assume that the builder used generic Cat5 cables, is this case, is it worth to try better ethernet cables for 1, 2, 3, and 6, since they add all together only accounting for a small portion of the ethernet signal path?
Another question for my education, what is the exact signal path in my setup? I use MConnect on my iPhone to control JRiver. After I press the play button on my iphone, what happens next in order for the 1/0s stored on my NAS to get to my from DSJ? maybe 1/0s on the NAS to the verizon fios router, to my pc, back to the router, and to the DSJ? One reason I asked this is, if not to replace all ethernet cables 1, 2, 3, and 6, if only replace two of them, where to put them in the chain potentially will make a difference?
I wouldn’t think so. Ethernet cables can make a difference. I have a cable coming out of the wall into my music room, feeding a network switch. My NAS and DS DAC are both connected to this switch. Upgrading from generic cat5e to Audioquest Vodka for the two cables (NAS–>switch and switch–>DS) made a difference; not gigantic, but clearly audible and (to me) worthwhile in the context of a revealing system. But you have so many runs, including the two in the wall, that I am doubtful. But I have been wrong before; the only way to know for certain is to try.
Seems that some feel that the last cable in line connecting the DSJ (6) is the most important. I agree with magister that the amount of change is small re a single cable swap and your system is a little more complex than mine.
Thanks Wglenn and Magister. Could you also help answer my second posted question regarding the signal path? Currently, my PC/Roon and NAS are directly connected to the router located in my office, will it improve sound if I move the NAS to my music room, having NAS and DAC directly connected to a switcher as in Magister’s case?
Honestly, the best answer that I can give you is to give it a try to see if there is a difference. There are so many variables here that I do not understand, including the signal path, that it bears just experimenting to find out. We are used to thinking of signals in audio progressing in a linear fashion but that is not the case with digital transmission which is bidirectional, packeted, monitored and corrected along the way. Way above my pay grade but there may be others here who could provide more intelligent insight.
I also use audio quest vodka cable (cable connected to DS DAC) with good results.
To get all out from good quality Ethernet cables It is worth of trying isolation of network connection from main switch to DAC e.g. With Optical media converters (removing EMI and RFI generated by other devices in network). Also I have noticed if cable runs are long then by adding a switch just before dac makes better results. Good LPSs for media converter and switch also makes a difference IMO.
In my case NAS and Roon server are connected to main switch before optical isolation, so after optical isolation before DAC, there is only one switch and galvanic isolator.
I think you guys need to do some research on how Ethernet actually works and how it is implemented.
IMHO, there is no chance in hell that the wire you use affects the audio quality of bits that are packetized and sent out through a collision-based protocol into receivers that completely buffer and isolate the packets.
EldRick said
I think you guys need to do some research on how Ethernet actually works and how it is implemented.
IMHO, there is no chance in hell that the wire you use affects the audio quality of bits that are packetized and sent out through a collision-based protocol into receivers that completely buffer and isolate the packets.
Capacitive noise coupling? Differences in RFI shielding (e.g. to control emissions of noise that may affect the rest of the system) Ground loops? (Via the power supplies if not necessarily the signal lines?) Does changing the number of collisions affect the noise that the Ethernet injects into the rest of the system? I can hear them on an AM radio... I'm not claiming that there's an audible effect or that there isn't, just pointing out that your statement is a little condescending and overly simplistic in that you aren't systematically ruling out some other fairly obvious and arguably plausible mechanisms.
EldRick said
IMHO, there is no chance in hell that the wire you use affects the audio quality of bits that are packetized and sent out through a collision-based protocol into receivers that completely buffer and isolate the packets.
I stand by what I heard when I upgraded my ethernet cables. I cannot explain it, but a) I trust my ears and b) I have been at this hobby long enough not to have pre-formed opinions based on price or brand. I've heard expensive stuff sound bad cheap stuff sound good. If the cables had not been an audible improvement, I would have been delighted to send them back and save the $$$.
Like cabin, I found that adding a good LPS for the network switch and the NAS helped.
I use a Fiber FMC connected to a fiber switch and then a $11.75 6ft Blue Jeans ethernet cable connected to the RJ45 source. The best and effective way to isolate the source from any noise in the network domain - like NAS, PC, etc and lot cheaper than the audiophile cables (though I fell into the trap and bought one earlier).
Here’s the thing. The wire quality cannot help make the bits better - and many people find it hard to grasp how changing it can affect sound quality. I am among them. We know digital cables can make a difference, and that often has to do with timing, but in the packetized world there are no timing issues.
In the case where people hear a difference with Ethernet cables, and there are plenty that do, all I can guess is it must have something to do with shielding or noise. Other than that, it would be one of those mysteries—things we hear but cannot explain.
If you think Ethernet cables or wifi can affect the data, I’d stop using internet banking NOW. I do my banking on my phone and I’m still waiting for a rogue “0” to be added to my account … and my money is equally precious as my music, as without money I could not afford to go and listen to music.
Some manufacturers suggest wifi in preference to cables or vice versa. That just confuses the issue, which I do not see as an issue at all.
Over the last couple of years I have eliminated all expensive cables and several expensive components without any detriment. Most of my cabling is balanced and comes from the pro store, as do my speaker cables, as pro cables tend to be well insulated given how they are used. I wanted a CD digital source. I asked the largest Japanese manufacturer of transports if there was any performance difference between their $40 transport and their audiophile $3,000 transport, to which the answer was basically “no”.
The only thing left that could be considered an expensive accessory is a PS Audio P3 power regenerator, which I acquired some years ago and did seem to give a cleaner sound, and has remained in place ever since.
This all somehow reminds me of the Buffalo Springfield lyric: except in reverse
“Nobody’s right if everybody’s wrong”
I am in something similar to Paul’s position: I know enough people hear something here to suggest that something is happening, and I have come to the conclusion that it probably relates to the sort of factors Ted describes. Much like improving the power supply on my Cisco switches; those had a much bigger impact than expensive Ethernet cables [that I also tried and returned; I suspect my old ears aren’t what they used to be]. But I do expect to hear a difference when I implement my incoming Mivera Audio Superstream with its dedicated switch and fibre connections. I anticipate that decoupling the Ethernet electrical path will make an audible difference in front of my MSB Select DAC [which remains Roon-unready] as compared to the LANRover.
I am a network guy, and when it comes to data over IP, there is zero reason why the cable should affect the data. But just like a lousy video cable to your monitor can add noise to the picture you see [without affecting the quality of the underlying data displayed], a good cable can improve it. I think there is probably an audio analogue to that.
BTW, anyone know off hand if the inputs on the DS, including there Ethernet input for the Bridge, are galvanically isolated? Hope so.
And therein lies the probable root of this whole discussion.
Note: differences are not the same as improvements.
I too have gone the simplification route with my system over the years: it’s plain to me that no additional devices sound better than No Additional Devices.
And therein lies the probable root of this whole discussion.
Note: differences are not the same as improvements.
I too have gone the simplification route with my system over the years: it’s plain to me that no additional devices sound better than No Additional Devices.
Agreed. “Difference” ≠“better” [and that is why in a separate thread in a separate forum, I have asked Paul why he thinks the DMP/DS combo sounds better than the Bridge/DS combo on identical file material.
And I also agree that simpler, i.e., fewer devices and connections, almost always equals better, though some of Paul’s observations about the insertion of a preamp in certain systems caused me to rethink a few things. Right now, I have a Tortuga LDR passive pre in my speaker system. It accommodates the BDP-105 as an SACD player and the NADAC as DAC for now, so it serves a valuable switching function that I will no longer need when the DMP arrives, but whether it makes anything sound better is up for grabs. But other than that, if I can make it simpler, I sure try to. And in my case, network audio is simplest of all.
Having putzed with a few servers and noting differences among them I can see where the DMP may outshine a computer server. In particular, the MAC based servers do not sound as good to my ears than PC based ones. I attribute this to the ability to control the software environment, to some extent, offered by Windows and the BIOS/UEFI environment. Having a superior, purpose built device like the DMP to extract and deliver the goods without the extra “noise” of unrelated software processes ('housekeeping" if you will) could give you a better result. I found that there was a wasteland between a $1K PC server and the $10K+ range for commercial servers where the perceived differences were minimal or mostly absent despite expensive hardware differences. I have not had the opportunity to try the Aurender W20, though, but at it’s price… My experience is neither extensive nor authoritative and YMMV.
Back to the OP’s first post, before you spend the kind of money that an Audioquest Vodka Ethernet cable can cost you, I would recommend that you look at whether fiber might be an alternative or supplemental approach. If you do, then a fiber/Ethernet link or switch is another or additional way to address it.
Bob said
It's possible to implement fiber for about $100, though I expect one would want to have a very good power supply for the transceiver nearest the DAC. There is also this for $126: http://www.emosystems.com/product/en-70e-ultra-compact-network-isolator/
I haven’t personally tried either, but am thinking about one or the other
Interesting suggestion. Thanks. I see this companion product, somewhat more expensive, the data sheet for which specifically identifies “Audio” as one of its intended applications. May have to try one just to hear what it’s like.
My experience is neither extensive nor authoritative and YMMV.
