I wonder if I have a weak link in my setup that is causing the streaming sound to not be as good? I can only imagine it’s my router. I have a Amplifi mesh router that has some units that plug in a couple places through the house and boost the signal and it works really good. So I have the Nucleus + plugged via ethernet (audioquest diamond ethernet cable) then another audioquest diamond ethernet goes to the DCS Rossini DAC/Clock. I had some extra CDs laying around (my whole CD collection burned up in a wildfire so had been streaming only. I bought an Audiolab 6000 cd transport to play a few CDs I had that could not be streamed from anywhere and have been totally shocked how much better it sounds than streaming. It’s very good as is the vinyl much better than streaming anything. I don’t know that much about how much difference the router connection makes. I just did what DCS suggested I do to hook this DAC up. I also feel like the native files sound better. I had bought a few things on band camp and to me they sound better than streaming the same thing from Tidal or Qobuz. That goes through the router so I wonder why the native files would sound better. I wonder if any of you experts out there have any thoughts on why the CDs sound so much better. I’m getting to the point of only wanting to play physical media it sounds so much better. If I find something I like my thought now is to try and get the item in hand so I can hear it for real. Hopefully I’m not starting any wars here I am just curious if maybe my setup can be improved for streaming or if anyone out there hears this the same way as I am.
You are on a very good path. That’s the way! (no joke here).
Any kind of anyone’s setup can be improved no matter how good it is at the moment (audiophilism 101 
). There are a lot of influence factors which you can work with to improve (or not), especially in the streaming, since getting the music from someone’s servers to your ears it’s a more complicated chain than simply playing your discs.
I’m in the same boat, tough it’s my own fault entirely, my disc players are deliberately a few classes above my streamer of choice.
Some time ago I made essentially the same point. I felt that streamed files from Qobuz didn’t sound as good as downloaded files (from Qobuz) stored in AIFF format on my NAS.
Most people who responded said that they couldn’t detect a difference, and some said that there couldn’t be a difference.
Since that first experience, I’ve changed the network configuration, replacing my Cisco managed switch with a Cisco gigabit unmanaged switch. Also, for my primary listening area, I’ve changed from streaming via a dCS bridge (to my Nagra DAC) to a direct usb connection between the Nucleus and the DAC. Now I think that streamed Qobuz sounds fine.
Of course, the whole thing could be in my mind, and I have no interest in arguing otherwise. Nevertheless, I think it may be worth trying some changes in configuration before writing off streaming.
As a user of a similar DAC i was really surprised by the audible differences between my previous Auralic Aries G1 and my current Aries G2.1. If ever there was a glass ceiling…
The psychological effect? I don’t mean that facetiously. Our minds play a huge part in how we hear things, if our mind is aware we’re listening to a format we assume to be better and especially if we’ve taken that extra effort to take a silver disc out of it’s jewel case and place it into a CD tray to play it, is it not entirely unsurprising that we may hear/experience it differently?
If the same PCM data is put though the same DAC and played at the same level it should sound identical, regardless of the path that data has taken to arrive at your DAC, after all it’s the same data arriving at the DAC in both cases. The most likely explanation, if it’s not a psychological effect, is that the version hosted the streaming service is i) a different master or ii) has been volume levelled in some way.
Different master is indeed a prime suspect, but there’s also a possibility that the streaming solution is dropping packets somewhere. If sound samples are being lost, you’d expect the streaming software to say so, but maybe it doesn’t. Maybe it interpolates between the samples to fill in gaps. I can see how that might not work well.
The stream will be buffered at the TCP/IP level and any lost packets resent. If that wasn’t the case the internet would fail to work reliably, and banks would be going back to paper 
Legality aside, you can easily test this out by downloading a stream from Tidal/Qobuz to your desktop a couple of times and then comparing the resulting FLAC files, short of you cutting the internet off on one of your attempts the files will be identical. Streaming a FLAC file is effectually just regular downloading with some buffering.
Roon also implements it’s own buffers in RAAT before any audio is sent to your DAC. Again you can easily test this by pulling the ethernet cable out of your endpoint and music will still play for 20 or so seconds.
So, while there could (in theory if using a one-way, synchronous digital connection) be unrecoverable packet loss between your transport and your DAC – which would most likely to show itself as a pop/click. That won’t be the case for the TCP/IP stream from Tidal/Qobuz.
I’m going with different masters or volume levelling/boosting at Tidal/Qobuz. But if we really are comparing the same PCM data then it’s psychological effect.
I personally cannot detect differences although my (audiophile) friends claim they do (physical media sounding better…).
As for potential explanations the only one I have read and don’t find altogether unlikely is that processing the digital stream requires some computing power that could inject noise. That is (so the theory goes) why WAV might sound better then FLAC (more effort to process as you need to uncompress). A CD transport on the other hand could induce noise due to the physical rotation. Go figure…
There is a point where common sense ends and marketing ploy starts. I think a good DAC can handle noise and jitter well enough to make these variables irrelevant.
Burn a cd from a FLAC download, and see if it sounds better than streaming it. You don’t really list your setup, so l´m not really sure what’s being compared here.
Any time the signal path changes it will sound different.
You may simply prefer the path from your CD transport. It just means it is time to start swapping gear on your streaming transport side. It took a long while to get my streaming path as good, now better, than my CD path.
Too late for that! 
The whole thread is click-bate - how can a $500 CD player sound better than a $28k DAC? Of course, it’s the network switch/cable - either that or buyers remorse !! Never seen this discussion before
I think this discussion is a variation on the “bits are bits” theme, which has been discussed a few times before.
Yeah maybe I’m just crazy in that case why am I even striving to hear better sound. I should just tell myself this sounds as good as it gets and stop worrying about it.
It’s not clickbait and to bad some of you guys are such sourpussies for a little discussion. It’s not a 500 CD player, it’s a 600 CD transport into a DCS DAC.
Just listen to what you prefer
… and/or enjoy the difference
It’s interesting, and part of the fun
You are not mistaken. Recently I came to the same conclusion. PCM stored on CD is optimal format for physical media. Streaming is a different beast. The differences are noticeable, yet no immediately obvious. Streamed file like FLAC contains all the data required, but the other side of equation is that sound producing software has to calculate frequency response and it can be different than originally intended. Simple example multiply or divide two numbers (the same numbers) in Excel and Access. The results will not be identical. There is always small difference. This is a nature of computing. The same song in one format will sound different playing through various software applications. Out of curiosity, I loaded FLAC, AIFF and WAV files albums on my iPhone (Plexamp). I was stunned how spacious and detailed AIFF and WAV files sounded in comparison to FLAC. The issue here is not format itself but how that file is processed by an application.
Unless you use some kind of DSP, there is no such calculation involved during playback.
I think you are referring fractional numbers here. FLAC only supports integer formats, and integer arithmetic is always exact. As a matter of fact, decoding FLAC performs only integer arithmetic by design.
Any application supporting FLAC will decode same file in the same way and will produce the same result. If you download a zip file from the web, you can unzip it using any app you want, you don’t need to know which app was used to zip it.