@zoom25’s point was that once the Roon core is isolated via ethernet and separate power circuits, there should be absolutely no difference in SQ as long as the server has adequate horsepower.
2 Likes
well that depends…
Please do go on…
I proposed above how you can troubleshoot airborne interference by moving computer and other non-audio devices far away. Only the endpoint stays in the audio room which is connected via ethernet (RAAT).
As for ethernet noise leakage, I also proposed the use of FMC above and how you can apply it to determine if the difference in SQ between computers or Roon vs. Roon Server is due to electrical noise or networking issue (digital).
As the Lumin gentleman who mentioned power line contamination, let’s put the computer into another circuit. Furthermore, run the audio gear through a UPS or isolation transformer to further separate power interference on the mains.
This way we can address airborne interference, ethernet isolation, and power isolation. If there remains a difference, it must be digital in nature.
What else is there to consider???
For our internal testing we only run Roon Server on Linux, simply because we do not have another spare Windows PC for this task.
By default Roon sends bit perfect audio, so the digital bits are the same. (There are also products that claim to improve Ethernet packet jitter, but I believe 99% people here would dismiss those as snake oil.)
Yeah, what is jitter for an asynchronous, out-of-order protocol?
I order 24 CDs per day from Amazon, that is enough for continuous play, but they arrive in batches by Fedex, so that is 24 hours jitter, or about 100,000 trillion picoseconds of jitter. And double that on the weekends.
4 Likes
And how about you?
A somewhat relevant discussion by Charles here, one of the most qualified and respected DAC designers on the planet:
“That has been nearly a decade now, and it has become more and more clear that the audible differences are caused by various factors that are not well understood, yet seem to be dominated by EMI and RFI noise coming from various sources.”
1 Like
Do you use Volume Leveling in Zone Settings? In my setup Volume Leveling makes the sound quality worse. Decreased dynamics and decreased focus.
Setup: Roon Rock NUC -> Chord 2Qute -> Leema Quasar -> ATC SCM19
I spoke with Caelin of Shunyata last week in London and essentially he said much the same thing. To paraphrase what he said:-
The age-old debate about how “power conditioning cannot fix the 1000 miles of cables before it gets to your house so it’s all BS” is not really the issue. We should be way more concerned about the EMI and RFI that is coming from inside our house and the its of computer and hifi equipment.
This resonates with what Hansen is saying.
1 Like
Agreed. The part about about EM/RFI has been echoed to me by a couple other well respected USB DAC designers before also.
But the “caused by various factors that are not well understood” is quite refreshing for me to read. It shows that the best guys at this stuff are still trying to work out why people hear the differences that they do… not everything is fully understood yet.
That shouldn’t stop us being able to share observations of course, even if explaining the causes/reasons is a difficult task.
1 Like
I try my best to stay neutral in audio debates, while keeping an eye for all kinds of potential factors that may influence sound quality. For this particular type of product, we never tested it, but I know a few customers who have positive experiences with it. I always respect my customers without exception, and I never dismiss their reports as expectation bias. However, I’ve seen far more people sell those to second hand market soon after they acquired it. Even for something like FMC, there are many people who dismiss it because they assume that Ethernet isolation in all devices must be 100% perfect.
1 Like
The biggest reason I didn’t go for Roon earlier was that it really sounds different compared to JRMC I used earlier. In short, it’s like described earlier. Roon has slightly thinner sound with more high end while JRMC has fuller sound with beefier low end. The difference is surprisingly clear. Which one is better is fully dependant on rest of the equipment chain. Currently Roon works well for me but in the end it was the UI which hooked me. There’s simply no returning to the old way of handling the library.
I use single computer setup and my PC is connected through Chord SilverPlus USB cable to TeddyPardo U2S USB bridge. I also have Intona USB Isolator and Audioquest JitterBug doing their magic between the PC and rest of the setup.
Disclosure: We share the same US distributor as Sbooster.
Having spent the evening comparing my PC based Ethernet Roon system to my MacBook Pro with USB connection, using Audirvana, I have to agree with the OP.
Dac is Holo Spring L3, with Bottlehead Crack and HD650 cans.
Jeff
1 Like
During the install was not asked anything. I am able to use Roon Server and Roon co-existing on the same machine. Roon which installed on the same machine as Roon Server does not act as remote only as you implied and also stated in the last two paragraphs of https://kb.roonlabs.com/Sound_Quality_in_One_Computer
Hi guys. Very interesting thread. Especially becasue I’ve been recently wondering whether migration from Roon to RoonServer has any impact to the SQ or not. Well, I’ve just done it tonight. And I’m a little bit surprised. SQ is better (wider scene, more air, better instruments separations) and I don’t think it is only placebo, however I didn’t make a blind test. I play from the single PC (Win 8) dedicated for audio purpose only, totally passive with a linear power supply, via DIY USB cable without power line to the Gryphon Kalliope DAC (it has a very good USB input with the big battery of capacitors). And my audio setup has two active power filters to minimaize any noice from the power line and whole circuit. Now I use only my Ipad pro for remote control and I’m very happy!
1 Like
When I compared JRMC and Roon, I also tested RoonServer installation against the regular installation. I wasn’t able to hear any difference between the regular installation vs RoonServer. They both sounded identically different compared to JRMC. For the sake of convenience, I’ve since had the regular installation on my PC. My PC is not completely dedicated but I do use Fidelizer and I’ve done all the basic optimizations recommended.
On my 2015 i7 Macbook, A+ CPU usage is <2%, after the first ~10 seconds or so it takes for buffering.
RoonServer alone CPU usage is 5-10%, even after 30 seconds.
Roon Remote app seems to add another 5-10% CPU usage on top of that, when Remote app is open.
Ignoring the Remote app (app is closed) so GPU usage not a factor (I think?) and just looking at RoonServer running alone, I wonder why even with DSP OFF and Audio Analysis OFF, RoonServer’s CPU usage is 5-10% compared with A+'s <2%
A+ is probably coded in a native Mac OS / OS X environment where as roon is in effect ported from a development other than native on the Mac
That could be the difference…but I’m not a software developer…so I could be off the mark totally, but I’m sure one of the roon boys could elaborate.
For the Roon Remote app yes. But I wouldn’t think so for the RoonServer app? We could both be wrong. Let’s ask the Guru @brian