For those who say Nucleus does not change the sound quality is not having the same experience I have had. I have a highly resolving system and looking to optimize the sound quality from my stored music. My current end point being a Bryston BDP2 network player. This weekend I took home a similarly priced end point $3200+/- and the Nucleus + for demo. I first listened to the new ( but broken in demo) endpoint ( after 24 hours of warm up ) and in about 3 seconds knew it was less than half of the Bryston experience. Disappointed I almost did not take the Nucleus + out of the box as I figured I already had a dedicated ROON server to ROON spec
(OptiPlex 9020M
Intel Core i5-4590T CPU @ 2.00Ghz
8GB Ram
64-Bit OS, Windows 10
100GB Solid State Drive)
so what is the Nucleolus + going to do?
Fortunately I did put it on the network and holy crap what an improvement! Not a maybe or could be but a holy crap it is. My dealer wants me to also demo the new Antipodes CX+EX $10,000 combo for which the price does not so much scare me but the fear of doing anything that changes what I am hearing now. My wife ( who has incredible hearing ) comes into the room and says I do not know what you did or how much it cost but do not change it.
If youâre connecting these devices (Roon core and endpoint) to each other by network, then the upstream devices shouldnât be impacting the endpointâs SQ. You should be able to use any networked roon core and have it sound the same feeding your endpoint provided it isnât under specced.
Are you putting these Roon core devices near your audio rig and on the same circuit? Are you using shielded network cables by any chance?
zoom25 , that is exactly what I expected but I promise you the results were substantial . The Nucleus is in the basement. Picture attached shows the original Dell computer 
Shocking discovery! I wonder if anyone experienced the same jump in SQ when changing the Roon Core / Server, when it is NOT directly feeding the DAC with streamer duties.
This is why I said changing Roon Core devices or switches/cables (at least UTP) âshouldnâtâ result in different SQ. Although, it is possible in some setups and environments for these to make a difference by direct or indirect mechanisms. I went through that for many months and then made a few changes and since have had things sound identical regardless of the Roon Core device used or cables or switches/routers.
Iâm not sure which new network player @Billt1 got, but he should try that with the BDP-2 and see if it makes any difference.
I think the end goal with these things should be to have the setup be completely immune to outside changes. With regards to power, I use a Torus isolation transformer for my audio gear. I think money can be better spent here than on another Roon Core device if itâs not really required.
@Billt1 thereâs a simple test you can try yourself or even blind with the help of someone else. Although, give it a shot by yourself. Iâm not sure how long the buffer lasts with Roon on other devices, but itâs a few seconds here on the BDP-1. Get accustomed to a track and its flow and then pull out the cable, do you notice a change in SQ when the network cable is out? For that duration, the network player should be playing by whatâs in the buffer.
If you leave the cable out permanently, the playback will stop completely. However, if you do it for a brief duration and plug the cable back in quickly, the playback shouldnât stop. Do you notice a difference in SQ during the plugging-unplugging-plugging transition?
Iâve used this test for testing out different cables and network switches/routers and their power supplies (linear vs. stock). I can quickly swap out between all these things and even not having a single thing plugged in for a brief duration. The sound never changes between devices or when there is nothing plugged in. IMO this should be the end goal.
Try it and report back 
zoom25 I did not name the Demo endpoint or will I as I did not speak well of it. My experience was with the Bryston in place and only swapping the Nucleus in for the Dell . The house is brand new and I did make allowance to have a Torus wall mount installed but have not done it yet . I will continue the research and report but believe me I am as surprised as anyone the Nucleus made a sonic difference.
Oh, okay. So it was with the Bryston BDP-2 that you tried the Nucleus and not the other endpoint.
Regarding the Torus, Iâve been told that they always prefer to use a Torus box in the audio room near the audio rig for SQ purposes over a Torus wall mount panel far away. A Bryston BIT20 or Torus RM20 should be perfect.
If youâre using a BDP-2, then Iâd advise you trying out the ethernet pull-in pull-out method. See if you notice a change in that. If you do notice a difference when the ethernet cable is plugged out with the music playing, then you definitely should try to figure out the culprit. The idea is when itâs playing from the buffer without the cable, the BDP doesnât see or get impacted by any external Core device.
If your rig is too far away to make the change, get someone to help you out for a few minutes.
Torus, off topic but. When I built my house I had a sub panel installed within 5 feet of outlets for both my upstairs and downstairs systems with much slack on all wires so the WM AVR 60 could easily be installed. Frankly the line noise measures so low (20mv) I havenât felt the need to spend $10,000. New house in a new neighborhood with a transformer at the foot of the driveway
Nucleus + does make a difference when I compared against a fully tricked out MacMini, I said very much the same thing, although attributed it to a zero noise floor from the server, I caveated the difference as subjective ( after all ultimately what we all hear is subjective) but the difference is bloody huge! (Subjectively) 
Question: Are you guys connecting these Cores to a router/switch or running a bridged connection directly to your endpoint?
router/switch
Core to router/switch to endpoint
Yeah, thatâs what I figured. I donât get it. Itâs the router that regenerates the data and then pushes it to the endpoint. The only thing I can think of is that Nucleus w/ Rock has a better protocol for sending data across the network? Although, I highly doubt that and havenât ever seen anything pointing towards that. All these Roon Core devices should be pushing the same data to the endpoint.
I could understand if the Nucleus was swapped out for another Roon Core device that was on the same circuit or near the audio system? Although, that doesnât look to be the case with your setups.
Iâd still welcome either of you to try the test I mentioned above with any of the Core devices (Nucleus or NUC or something else). Does anything change for the duration when the ethernet cable is pulled out? Iâd love to hear your responses or from anyone else. Is there anyone elseâs endpoint that changes in SQ when the ethernet cable is pulled out?
There may be a little more to it than this (maybe not though). Quite a few people noticed SQ improvements in some past Roon updates that optimized CPU usage of Roon Core (for all platforms).
With ROCK, the Roon Team have the opportunity to optimize things much more because ROCK works on recommended hardware - and with Nucleus possibly a little more still, since theyâve selecting ALL the hardware involved.
It doesnât mean ROCK/Nucleus will always sound better than other Roon Cores but it may have the best opportunity to sound best, since OS processes can be much better controlled by Roon OS running on ROCK or Nucleus.
Maybe?
I can see the OS changes and updates as a potential improvement. Itâs possible that newer updates make the workload on the endpoint easier than before. Thatâs a valid point. However, if all these Roon Core devices are up to date along with the OS of the endpoints, Iâm having a trouble seeing why the performance might differ.
Either way, the simple test of plugging and unplugging the ethernet cable with various Roon Core should be a useful one.
Computer A sounds identical when playing off the buffer alone vs. with the cable plugged in. Similarly, Computer B sounds identical when playing off the buffer alone vs. with the cable plugged in. Is it still possible that Computer A and B can still sound different (or rather make the endpoint sound different)?
This is an important question to think about.
If the answer to the above question is that YES it can sound different, then the only mechanism I can think of is that a connection to a particular Core and endpoint causes an immediate change to the endpoint that retains itâs properties with the cable plugged in vs. out. However, the properties ONLY changes when another Core is plugged into it and the performance of the endpoint changes.
I hope that makes sense. 
Noted, but as I explained just above, the Roon Team have more control over ALL the OS processes with ROCK (Roon OS), than say the OPâs âOptiPlex 9020M Intel Core i5-4590T CPU @ 2.00Ghzâ Roon Server running on Widnows 10âŚ
Reducing CPU loading, reducing energy consumption of processes, turning off processes that arenât required all seems to be beneficial to SQ - as I also explained just above, quite a few people have noticed past Roon updates that optimize these areas in the past have resulted in better SQâŚ
Maybe?
Since weâre talking about subjective SQ, it can be very system dependant. Some of these differences are impossible for me to notice on a speaker system but more obvious to me with closed back headphones (where the all important room is a non factor).
I get the software updates making could have made the workload easier on the ENDPOINT. However, I still donât understand how it should matter how hard the OS and computer is running on the Core side? As long as all these Cores running same Roon version can send the same PCM data in a timely manner, it should not matter how hard the Core is running. The noise or workload of the Core should be isolated.
Even as a theory, Iâm simply not getting it on this one.
I only bring these past software updates up to say that maybe (or not?) reducing CPU loading, reducing energy consumption of processes, turning off processes results in better performance and improved perceived SQ?
So putting aside my updates example, if the Roon Team can control all processes and CPU loading with Roon OS running on ROCK or Nucleus, then maybe this can perform better than a Windows 10 machine running Roon Server?
This is beyond my expertise though. I am careful not to make any claims but to ask âmaybe?â.
None of this is life and death stuff of course. But I do (sadly) find it interesting to read and learn about (from experts that do know this stuff).
With the ethernet cable plugged out, my Klimax DS plays itâs buffer until empty around 15 to 20 seconds. The sounds remains the same as with the cable plugged in regardless of which source is being used.
The only conclusion I can draw, is the source in use is ultimately affecting/contributing to the sound quality (lack of emf, noise, processes etcâŚ) prior to the renderer in use in the chain - everything else being equal location, switches etcâŚ
I know it was suggested above, but pulling out cables and listening as buffers are drained seems like an odd way to judge cable effects.