DEVIALET now Roon Ready.

[David A]

Is there a difference between say an NUC with i7 chip, a PC with same i7 chip running Linux, a PC with same i7 chip running Windows to run as Roon Rock (sorry, got my terminology wrong!   , newbie here)?   Assume they have same ram, disk storage and doing the same thing with a mid sized library, no DSP, no upsampling.   Is the sound quality affected by using a slightly difference platform?  Will adding a linear power supply to the hardware improve sound quality?

OK. I think the answer to your question is yes, there can be a difference. One of the reasons is simply that the PC is often going to be doing a lot more stuff than the NUC. For a start the PC is going to be processing video output whereas a lot of the NUCs or similar servers run headless,  no video card, and you use an app and/or a browser based interface on a tablet or another computer for control. The PC will also often be doing other tasks such as regular email checks, etc, while also serving up music. I think it's safe to say that the general feeling of those of us running dedicated servers of some kind or other rather than running Roon on our regular computers is that there are benefits to running a dedicated server because it's usually doing one task only, serving music, and that reduces problems arising from other tasks running in the background, or even the foreground at times while you're relegating music playback to a background role. With a mid sized library, no DSP and no upsampling you could probably get by  comfortably with an i3 which is what Roon's own basic Nucleus uses.

But then, if you're running Roon on your normal PC and you choose to use an iOS or Android tablet to run the Roon app to control things, you can turn your monitor off, shut down all of the other apps you're running on your PC when you listen to music, and you're probably not going to be much worse off in theory than you would be running Roon on a NUC dedicated to just running Roon.

Adding a linear power supply may improve sound quality, but only if it's a better quality power supply than the one it's replacing. Not all linear power supplies are equally good, not all switched mode power supplies are equally bad. Just because something has a theoretical advantage doesn't mean that advantage will be realised in practice. How well something is implemented can often be more important than whether it's theoretically the best way to go.

How big a difference you will notice can depend on a range of factors including others that you didn't mention.  Then there's the issue of whether the increase in sound quality you get by going down the path of using a NUC or a dedicated server and perhaps replacing the power supply with a better power supply and perhaps doing various other things is worth the cost of doing those things. Only you can answer the value for money questions. I can say that I think you can get improvements by doing some or all of those things you've mentioned but if you make the wrong choices in what you buy you may get little or no improvement. I can't tell you how big an improvement you will get or whether you're going to think you've got value for money out of it. There are some things in audio you can only find out the hard way, by trying them yourself.