One quick question:
It is often said that mastering is the most important thing, and that comparing, say Spotify against hi-res FLAC is meaningless unless the mastering is the same for each. This may be true, but in reality, how often is the low-res Spotify file a differently mastered version to the CD file, or even the Hi-res FLAC.
It seems to me, given the acknowledged cynicism of the music industry, that most of the time, the same source file will be supplied to whichever customer is asking for it. It seems unlikely that a label would remaster a file to work better via a compressed streaming service, when it could just as easily hand over the 16/44 CD file and let Spotify compress it how it chooses.
So, would it not be reasonable to assume that the 320Kbps Spotify file is the same mastering as the FLAC (particularly if the FLAC is the red book CD file), unless it can be shown to be different?
Many popular titles have been remastered multiple times and often sound radically different. Take Kind Of Blue as one example, there are at least four CD masterings: 1984 Japanese CBS (non-pitch-corrected, rare, collectable and apparently very good indeed - I'd like to find this one), 1986 Columbia Masterworks (non-corrected, rather rolled-off in the treble), 1997 (speed-corrected, rather bright), 2009 50th anniversary/current (speed-corrected, warmer than 1997). These differences in mastering are more significant than any difference in bit-rate (though obviously one would choose lossless every time if possible). As such it is essential to know one is comparing like with like, and fairly complex if not as in this example none of the examples are bad as such.
A similar story with say Pink Floyd, depending on the title there are three, four or more different masterings, some are superb, none are actually bad. Some artists are less lucky, e.g. Genesis, the first issue Virgin/Charisma CDs with the covers with a grey border are pretty respectable, everything later is just horrible. Similar story with Blue Note, the original 1987 issues are lovely, as is the Connoisseur Edition, whereas the RVG Edition varies from a bit odd to absolutely hideous. It's a minefield, and mastering is where many of us here start our system from, I'm certainly OCD in the extreme about it and often spend an evening comparing several different issues deciding which to keep. It's fascinating just how much difference this aspect makes, and it surprises me just how little thought many who spend huge amounts on turntables, DACs etc give the subject. It is *way* more important IMO. It's also astonishing just how many masters exist for many titles, some have been done ten or fifteen times, and it is most unwise to assume newer is better!
PS Spotify is interesting as there are often multiple issues of some titles, i.e. it often has multiple masterings available, though identifying them is often very hard unless any bonus tracks are a clue. It's confused further by holding many pirate issues of many jazz titles etc which could be taken from anything and screwed about by anyone. I'm no longer a subscriber, but this was certainly the case a couple of years back.