This made me smile that Dolores and her Limerick accent has traversed the globe, then a bit sad that she is no longer with us.
.sjb
Audiophiles spent mega bucks on music servers like the Wadax Atlantis Reference ($59K), Taiko Audio Extreme (âŹ24K), Antipodes Oladra ($25K), etc. All those audiophiles have problems in their heads or ears? Bits are Bits right? They shouldnât sound different right?
An expensive watch does not make a day have more hours.
An expensive car does not make earth rounder.
And expensive computer does not change the sound produced by a well build Dac.
Expensive streamers have many advantages, but none of them concerns sound quality.
Itâs science. Enjoy music 
Expensive streamers have many advantages, but none of them concerns sound quality? Why do you call them advantages if they donât concern the sound quantity? They should have no advantages, except cosmetically. The advantages are they look better, are heavier, quieter, and with more bragging points?
An expensive watch does looks better; an expensive car does rides better; an expensive computer has a larger CPU that can compute faster⌠⌠it is all quantum physicsâŚ
Exactly âŚ
Exactly, and fasting computing does not produce a better sound quality. You donât need a fast CPU to achieve the perfect sound quality, if we understand as perfect the exact signal it was recorded. Itâs digital. A faster streamer might find your tracks faster, might handle a trilliion track library effortsly⌠you get the point.
If you alter the original recording/signal by using filters, then a faster and more powerful computer might be needed, but itâs more a I can / canât use it with my computer than a sound quality problem.
âexpensive car does rides betterâ. Expensive streamers might make your music listening experience better. Better software/interfaces, more searching speed, more capacity, but do an expensive car reduce the distance between start and finish? Same with digital audio. Same packets. Same distance.
And the forum such users populate is quite frankly one of the worst I have read posts on. Whatâs best forum is full of FUD and more naked emperors than anywhere else. One user claimed that removing the filters from Roons DSP menu even when DSP was not active changed the SQ!! The herd followed claiming he was right.
Wow, Iâm dumb but I still cannot swallow that
These do not always hold true, some of the most expensive cars have the worst ride, comfort wise, they may handle a curve, but canât handle rough surface, are loud, and are not even comfortable to sit n, IMO. Same goes for the computers, Some have a name that you are paying for but are way under powered and use sub-optimal parts. As far as the watches go, that is a subjective opinion that they âlookâ better, some of those âExpensiveâ watches are truly hidious⌠again IMO, YMMV
As has been stated elsewhere, it boils down to what you perceive in your set-up to sound good to you. The caveat is that you wont know for sure unless you try it yourself.
There might be an explanation for this (one that is almost always left out of discussions): Roon and Roon DSP are software, so they can have bugs.
This means that it is not excluded that the user disables the DSP, the user interface shows that the DSP is disabled, but in reality the DSP is not disabled in the code (or not fully disabled).
The situation can be the same for discussions related to bit-perfect: the fact that a software / protocol / DAC input etc. itâs bit-perfect in theory, it doesnât mean itâs in practice. It is also in practice only if there are measurements that demonstrate this (updated after each release).
If you âthink it sounds betterâ that is nothing more than an indication that you are allready in the extreme nitpicking area and most possibly even in the imaginary area of audio evaluation.
The link you posted is more than 5 years old and I donât think itâs up to date anymore. I mean (at least) this paragraph:
âAll of that switch flipping happens before playback startsâonce itâs running, the signal path is doing its thing, and not interacting with the switches in the user interface at all anymore.â
The application does not behave like that. The switches take effect after playback starts.
If the paragraph above is in accordance with the current design of the application, then we are exactly in the situation mentioned by me: the application has a bug in the DSP settings. 
But it doesnât and it didnât back when they posted it on the forum. I tested it at the time itâs bitperfect and had no affect at all. How did I know it was bit perfect well I pass MQA to my DAC without Roons decoder and this still worked which it would not if DSP was active in any way as it would break MQA signalling . I also ested it on my RME DAC with its bit perfect test signal it passed with them present and inactive and with them removed.
The post was a direct response to people imagining a difference by flipping the switch at the time it was written.
I had converted my 10th gen I7 Nuc to a ROCK, but I use Roon at work too, so the fact that I couldnât install ZeroTier made me convert my Nuc to a Linux box and install Roon Server. Then I could setup ZeroTier and even Resilio Sync to sync my music folder (syncs to NAS, USB attached to NUC and even my work computer for off site back up).
Sound quality? I couldnât tell the different TBH.
If you are using USB directly from your NUC there may be any number of reasons it may sound different. But, if you are using RAAT connected endpoint there is no way the platform for Roon core, Nucleus or whatever can change the sound.
Almost all of modern DACs with usb input are asynchronous with buffer. That means being roon player bit perfect, neither the computer core usb output or the usb cable used (not being a faulty one) matters.
So no, even using the usb output doesât change the output of your modern DAC for almost all users.
All those concepts of noisy usb outputs, jitter, etc, are either things from the past or another audiophool myths.
I donât know why but this reply just really has me laughing. âIâm dumb but even I canât swallow thatâ LOL 
Thank you
If you are indeed hearing a difference, then it will most likely be due to a difference in the USB configuration. I suggest that your Ubuntu was configured for Synchronous USB which is subject to jitter while ROCK optimises the USB configuration to Asynchronous. This means your DAC will not be dependant on your NUC hardware for its clock when converting back to analogue. Note that ROON do not recommend connecting a DAC directly via USB but rather it should be connected via a ROON ready Streamer (via ethernet) if you are trying to obtain optimal audio quality.