I’m confused of what you mean by this and a “flac master file”. In case with “flac master file” you mean when a digital recorder first puts out a flac file, then I’d see no compromising in converting it to wav or aif, only advantages (if any difference).
But the flac discussion wasn’t my main point, I mainly wanted to say that when comparing disc with streaming, they should both use PCM decoded format, not one of them flac.
I think finally a disc vs. streaming comparison needs as similar environments as possible…in terms of file format, but mainly in terms of HW. The more or less same (lens) technology prior to a DAC, used with a disc vs. a network streaming interface. This same environment (say a PSA transport, which in my understanding is a streamer streaming from a disc…and the Airlens streaming from network) should isolate its stream as good as possible from the drive or network noise.
Besides PSA, a good comparison could be Wadax, offering a CD player and a Server at highest possible effort. MBL seems to have no network streaming experience /product yet, so to compare an MBL CD player to another manufacturers streamer is no good idea for a meaningful comparison imo.
So if in a meaningful comparison CD resolution from a CD drive still sounds better not only than streamed CD format but also than streamed hires of the same album, then network streaming still has serious issues (I don’t mean online streaming, which again can have its own special issues, among others depending on the used provider).
Hyperion, my main source for downloads, describe their files thus:
Most of our newer albums offer Studio Master FLAC & ALAC files. These are 24-bit (or 20-bit in the case of a few older recordings) and at whatever sample rate (44.1 kHz, 88.2 kHz, 96 kHz, etc) was used by the sound engineer at the time of recording and mastering the album (hence ‘Studio Master’).
Besides being so much more convenient, I can’t see how compressing to 16/44 can be of any benefit, if not harmful. As so many labels provide their studio masters, it makes ripping CDs a bit pointless.
The Grim MU1 and the MBL Transport are connected to the MBL Dac with Identical Stealth AES/EBU cabling. (The Dac has two identical AES/EBU inputs). I get why people have a hard time believing crappy low res CDs sound better than DSD…but with a great transport (retail 26k) the shiny CDs do sound better or at worst the same as DSD. @paul172 has also experienced this phenomenon with a different CD player in his system.
I also have an MBL N31 CD/Dac with a built in galvanically isolated Roon endpoint/streamer built by MBLs Jurgen Reiss. The CDs sounded as good or better than streaming with the same device. So I have also experienced what you speak off with the same companies galvanically isolated streamer built in the same chassis…
MBL actually does offer Roon Endpoints that are galvanically isolated with their own power supplies into both their Noble N31 CD/Dac and MBL 1611 F Dac.
I believe your premise that one would need a streamer from the same manufacturer without basis. You might as well say to be absolutely certain the listening chair should be from the same manufacturer. That is just as silly.
But thanks for playing!
Here is also another possibility with my comparisons. Most of my comparisons have been with music recorded prior to 2000. I like older jazz from the 60s and female vocalists. I am not sure when recording equipment started using 24 bit versus 16 bit to capture the original in digital. I would be interested to see what @Paul would think would be better… A recent recorded Octave artist using DSD through an Airlens feeding the PS Audio DAC MK 2 or the PS Audio Transport spinning a SACD or a CD…or those pesky vinyl pressings at 45 RPM My guess newer recorder materials might sound better in a streaming format…but most older recorded stuff doesn’t.
I have! 
Now, I am confused. No recording system outputs compressed FLAC or ALAC files.
What they more correctly should write is that their Studio Master FLAC & ALAC files are compressed versions of the 24-bit (or 20-bit in the case of a few older recordings) and at whatever sample rate (44.1 kHz, 88.2 kHz, 96 kHz, etc) was used by the sound engineer at the time of recording and mastering the album (hence ‘Studio Master’).
That would make sense.
Yes certainly, CD resolution recordings can’t get better by streaming than with a buffering CD drive (assumed the majority of the surrounding is equal).
I wasn’t aware of this. It would also be a great comparison. My only doubt in optimal comparison conditions then would be, that Roon still seems to be one of the worse options sound wise, so the choice of another SW would make it more fair.
No problem with our different opinion about the relevance of as much as possible equal preconditions for comparisons and about the applicability of the electronic HW-listening chair analogy.
Just to explain my logic behind the drive/streaming comparisons…not sure if we all agree about the following (my understanding):
Non buffering disc drives are not competitive anymore, buffering is state of the art to get rid of unwanted disc drive influences.
a buffered disc drive is more or less the same as a streamer, just with a drive unit and without the noise issues of a network connection.
this means, if a disc drive has the same surrounding technology inside the unit as the streamer, it only depends on the quality of the disc drives‘ drive isolation and the streamers‘ network isolation, which of them sounds better.
as it seems more difficult to isolate a streamer from network/hard disk noise than to isolate a disc drive from drive noise, a streamer using the same resolution source files, can only be as good as a buffering drive using the same rest of technology inside, but not better.
if or by how much hires files on a streamer sound better than CD resolution files on a disc drive at equal preconditions, depends on the ability of the streamer to isolate everything from network noise at least as much as the CD drive can isolate from drive noise.
It is a great question/observation and one I am anxious to answer.
If you are ready to do this, please provide accompanying relevant preconditions used 
(file format, streaming SW, separate power cirquit for network infrastructure than for audio, same resolution files). I assume you then use same power cords/signal cabling anyway.
I can understand how this could be the case, so long as you are generally comparing CDs to rips of CDs that are converted to DSD for storage and/or to CD transport digital outputs that get converted to DSD on the fly prior to or by the DAC.
If that makes sense; FWIW.
I don’t think the HD FLAC files are compressed.
Linn say much the same, save their studio masters are now 24/192. FLAC or ALAC. I have no idea what ALAC is.
https://www.linnrecords.com/downloads
The goods news is that the Mk2 DAC has two I2S inputs so a direct comparison can be made, and the AirLens can be fed from the Innuos Zenith, which is very low noise b3cause it buffers local streams.