And the server isn’t on that box, just the player right?
And using the Squeeze server in Roon?
I could easily believe that a roon endpoint would do that, my best guess is in order to deal with the mS timing they guarantee across multiple end points with RAAT, that the endpoint busy waits rather than just releasing the thread, where it could easily take 10mS to be woken up.
That would result in it using 100% of a CPU, which appears to be what your seeing.
Can’t you impact how this timing is handled by adjusting clock priority for the endpoint?
I have two different boxes. My CX is appointed as a dedicated server and my EX is appointed as a dedicated player. I was looking at the system health of my server only. All of the local files are stored on my server via SSD.
OK that makes sense then, Roon server is not exactly light weight, but CPU of the server shouldn’t impact the playback quality of the streamer, unless they are sitting on top of each other or connected together by something more than a network cable.
Edit: Though I just checked my server and it’s bouncing around with Roon Appliance using <1% CPU, so YMMV.
And just playing something locally it seems well behaved (no busy waiting I can see, though roon bridge might behave differently I guess.
Short version, no.
Dropping it’s priority would just make other “Ready” processes more likely to run instead of whatever is running.
So I redid the test, but looked at my player’s CPU/temp at the same time. The player didn’t even fluctuate from 0%, whether it was on Roon or Squeeze. I suppose that’s why Antipodes opted for the two-box solution – to take the load off of the player.
Maybe the conclusion here is that there needs to be minimum CPU load on the server side too, so both sides are important.
???
That’s just not how network connections work.
All the server does is pump bytes into a blackhole, there is no timing involved, when the buffer is “full” it just waits for the streamer to consume some of the data it’s already sent. If it couldn’t do that fast enough you’d get obvious dropouts, not a slight degradation in quality.
It is certainly possible for the Server to pollute the Streamer, but to do that there needs to be a way for it to transmit it’s noise. It could be any physical connection, even the the network cable (though my experience with testing that was that it wasn’t a large source of noise at least in my setup).
Or possibly if the streamer is in close physical proximity to the server through RF interference.
My streamers are all at least 12 ft from my server (some considerably more than that), and the only one with a physical connection is also optically isolated.
At least you know you’re not crazy! Thanks for doing this and honestly it’s why I like being part of this community. This is a topic that anywhere else would have led into a obstinate refusal of any issue, quickly followed up by blame and ending up in an argument.
When you trust your ears and you can repeat and test in a methodical meaningful way then you can generally identify the source of a given issue.
I know that this is where I’ll get optimum network performance but this shouldn’t affect actual processing but I do have to my my Server to the same side of the mesh endpoint that the bridge sits on. I was going to do this over the holidays and I got scared after noticing random reboots from my PC that I may have “too much” draw going through one AC conditioner. I moved the laser printer to another area of the house inside the service closet of the apartment and plugged it into a mesh endpoint and the random power outage went aw
Likewise, which is all best practice and all meaningful but then again, I didn’t do it because I was addressing any issue, it was more preventative and theoretical.
Yes, wasn’t really a part of any plan on my part, server happens to be in one corner of my office and my primary setup is on the opposite wall, the optical isolation was a cheap check initially, and later I got an optical Rendu at a good price in an auction.
squeeze material UX… 
you talking about this material UX?
Yep… trying it out on the antipodes instead of roon. I like the fact it combines Spotify and Qobuz together. Can’t say anything about the sound quality yet.
The UX is terrible though.
It def takes some time to get used to and certainly not as pretty as roon. That’s why i spent months going back and forth to compare and hoping that the delta was small enough to settle with roon. Once, i heard the differences and knew what to look for, i really couldn’t go back. It really would have been settling for me
I have no doubt that if the sound quality difference is noticable to my ears and prefers squeeze I’ll move to it.
I’m still getting used to the Antipodes sound vs my Auralic and figuring out how to best use it…
Ill be keeping an eye on your progress with LMS and Material. After trying to find a good streamer with its own good app and UI it seems that i either stick with the lumin app as others seem very similar, or look at LMS at it is the basis for others (innous) and has the sound quality.
My experience with Roon is that it depends on the point in time you compared and the gear you used. They are working on SQ so it varies between releases. Initially when 2.0 was released it went down but somewhere recently it went up again I have the impression. I also used to have loss of detail and clarity when I used DSP but I couldn’t tell much difference besides the bass shelf I added a few days ago. Either my hearing is going or something changed. I also had an update on my Sotm SMS-200 Ultra Neo that mentioned SQ so it also depends on how well Roon was implemented by a manufacturer I guess and maybe their commitment to get the best out of Roon as well.
Anyways, when reading I’m tempted to go LMS again but there are just to many drawbacks for my ecosystem in general.
well Antipodes say this on their server dashboard page:
“Squeezebox Server is license-free and Antipodes is committed to its on-going development as the best sounding playback solution available on Antipodes music servers.”
if you’re curious, this is what they say about roon
“Roon Server is fully integrated with Antipodes music servers. Users need to obtain a license from Roon to operate this application.”
I’m kind of wondering whether or not Antipodes doesn’t like Roon. It’s not unheard of as many vendors have given up support because they couldn’t get Roon to work to their satisfaction. The best way to reduce noise is to turn down the CPU to just what’s needed but if you do that and a protocol that required more CPU comes along. Just shooting shit at the wall here.
I trust you guys hear what you hear but I couldn’t and I don’t know if I’m alone in the comparisons I’ve done. But Innuous pretty much states the same thing when using Roon as a core and an endpoint in their MK3.
My understanding is either they use the roon bridge, which is closed source and they have no control over, or they can use the SDK, which I’m going to guess is just source code to compile something equivalent, but likely supports more platforms that just linux.
So I’m not sure there is much the manufacturer can do.
I do know at one point one manufacturer had accidentally bypassed some of their internal processing when playing via Roon, resulting in a lower quality output than with their native app.
EDIT - It was Auralic
Part of the issue is there is just no way top know if a streamer manufacturer is playing with the bit’s it’s sent before it sends it to the DAC.
There is no way to know what RAAT does that the other options don’t, pretty much all the other protocols just download a compressed or uncompressed image and play them, so it’s not like they should vary, reading a stream over HTTP.
So all we can do is deal in anecdotes and experiences.
I’ve tried the NS1 over both UPNP/DLNA (using the native app) and Roon with no obvious difference to me, and it could well vary by streamer.
I don’t know if you caught my commentary but I did the same with the dCS Bridge. I agree it’s just unknown without understanding what both the vendors and Roon does. I just think it’s no coincidence so many vendors have dropped out of the Roon Ready program. I mean they even wound up changing it to simply Roon Tested which speaks volumes of manufacturer PR face saving if you ask me. Both on Roon and the Manufacturers side. I couldn’t blame someone not wanting to participate in a Roon Ready program if they were then going to sound inferior and potentially take a black eye in the media for it. Conversely, Roon just wants to say it works with as many vendor equipment as possible.