[Another one from an email. I have edited my reply just a little for clarity]
“Do you do Mid/Side type processing when you master? Like, compress it separately maybe?”
….. it depends. It all always depends. MS is The Big Thing atm, like sidechaining has been, MB compression before that… it will pass. Just another tool. But yes, sometimes I do if need be.
FWIW IMO one of the most important things to remember in audio is that the best way to achieve a coherent, powerful, tight sound is probably not to split it up into different bits, do different stuff to them, and then put it all back together. Probably in 80%+ of cases I prefer (partially) unlinked stereo to MS, and very often in MS I prefer it fairly heavily linked, say ~70%. So you’re just doing a trace more in the middle, if need be, but not just making it weak.
The thing with the MS compression fad is that I’m hearing lots of what I’m calling ‘horseshoe mixes’. All central bass, then moving up into all wide stereo content with a recessed, flat, boring centre from say, 200Hz up. Initially impressive but ultimately, BORING.
I’ve said it a zillion times: ear candy is great and everything, but don’t forget the meat and potatoes…