Things other than SQ are probably better outside of Roon, but the user experience is just so much better with Roon. I'm hoping my K50, where it separates out the player and the server on 2 different cpus helps though. I tried Squeeze, but that sh!t is just not user friendly at all. I have like 14tb of local music, and I need a good UI. I'll sacrifice some sound quality for ease of use.
I couldn't agree more. Best UI and features by FAR. This is the real trade-off ditching Roon.
I'm interested in the Taiko Extreme and Pink Faun, but I'm kinda skeptical that they will be that big of an improvement over my K50, but I don't know. Yes, auditioning them would be best, but that's not possible where I live unfortunately.
Indeed. "Better" doesn't always necessarily mean "best."
I can likely tell you whatever you'd like to know about both. But the fact remains they are bespoke solutions (i.e. built with commercial off-the-shelf components like MSI/Asus motherboards, Samsung NVME/SSDs, etc.) from very small companies based in the Netherlands. Given a long enough timeline, you will likely be on the hook for supporting it yourself. All the SQ in the world isn't worth that frustration and wasted money if you aren't a very technically capable person.
Also, either server will end up costing around 25k. Yeah...
I'm guessing 60-70% of most, but not all, bespoke server cost is R&D (and because they probably think they can get away with it). After performing all the no-brainer stuff like disabling unused ports, processes, protocols, circuits, hardware, etc., they basically take "educated guesses" at what might improve SQ and test theories over and over again. It's an absolutely painful and painstaking process testing and gathering all of these little SQ nuggets, let alone developing custom solutions to leverage them. In other words, all of these "SQ improvements and lessons learned" from from trial and error equate to a bespoke server builder's "secret sauce."
For reference, I've attempted to build five of my own audio streamers, in various sizes with various components, and I can't even touch the SQ from my Pareto server. This tells me three things for certain:
1. I absolutely suck at building high performance audio streamers (I even tried cannibalizing most of the components from the Pareto)
2. I lack the time, patience and inclination to chase down all the little tweaks that could result in better SQ.
3. After basic common sense measures are implemented, I firmly believe the lions share of SQ improvements come from software, firmware, driver, and electrical optimization. BIOS tweaks in particular appear to be the most important and seem to yield the greatest bump in SQ. But they are simply beyond my skill level to decipher and would require too much time to learn.