bits is bits

@curiousjim I’m just guessing here, but maybe the difference is in the equipment Qobuz uses before the tunes hit the interwebs. ..,

I did not post the results, but yesterday I did a deep search multi-layered query about the differences between Qobuz vs. Tidal vs. Amazon vs. Apple vs Spotify and the "differences in sound quality" and "why we can hear differences". 

What came back was more different about each than what I would have expected, starting with different mastering versions of music, and then differences between FLAC and other non-FLAC versions that get transmitted for starters, and what actually occurs between each of these service providers before packets get transmitted across an IP network to your home and streamers/dacs, next.  

Each of you can do your own queries and debate from there with yourselves one way or the other, but I walked away interpreting that all bits and are not the same, and each of these service providers are not the same in what you receive on the receiving end. Some of them are messing with it on their end, before transmission and will leave it at that.  Qobuz did get praised for being the must untouched FLAC files and untouched service though, and most here seem to already know this. It might be worth retesting 3-4 of them and validate what differences you can hear first hand. I can hear differences on my system, for the five I’ve tried. So, "Bits /= Bits" in this case on what was originated, sent, received for all of the service providers mentioned so far. Check again, listen again, and compare again fwiw.   

I don't get why someone would have multiple streaming services and use each one independently. Based on reports that each streaming service has unique sound qualities, and the idea your streaming setup should be optimized for each player, this doesn't make sense. I have my entire music library on one server, this include Tidal and Qobuz, Roon provides the music player so there is no unique or differentiated sound from my Qobuz or Tidal music files. I've also optimized my streaming setup for Roon and Roon only. 

I'm not going to debate whether these streaming services offer unique sound qualities since I don't use their music player software. I'd only say a music file is bits is bits, the differentiated sound quality would be due to the music player software these services offer and how your streaming equipment interacts or works with them.

“Most audiophiles have come around by now to the realization that TCP is indeed bit-perfect - period.”
@devinplombier 

This part of the equation was never in doubt ~ the bits get there just fine. It’s what happens after that where things truly get interesting.  

As you pointed out, the issue revolves around parasitic noise propagating through conductive cabling, potentially coupling into the analog stages of components with insufficient isolation or grounding design, resulting in audible artifacts.  

The key is finding an isolation device that intelligently removes parasitic noise before those bits ever reach your streamer or server. Do that, and you’re golden.

@lalitk 

Doesn't a run of SFP fiber between listening room and utility room serve that purpose?

For the sake of completeness, the SFP converter on the system side should probably be equipped with a low-noise power supply, since it is located on the "sensitive" (read: non-isolated) side of the fiber run.

“Doesn’t a run of SFP fiber between listening room and utility room serve that purpose?”
@devinplombier 

Yes, it would. I personally didn’t care for using SFP fiber and converters in my system. For context, my streamer doesn’t have a native SFP port, so I ran a fiber optic link with converters on both ends between the router and the streamer. The result, though technically clean, made the sound feel sterile and dry. 

I settled a while ago on Telegartner Opt Bridge. It converts incoming copper (Ethernet) to an internal optical domain, then back to copper, thereby breaking ground-loops and removing parasitic noise. I am also using a LHY LPS to power the Opt Bridge.