🚀 go-pugleaf

It is a self-evident truth (whether Katz likes it or not) that if a re-released 16/44.1k track remastered by whatever means to 24/96k sounds different on his equipment, then something has been added or taken away, so you are not talking about the same product. Or failing that, his reproduction chain is in some way faulty, which of course he'd hotly deny. I think that Arny's quote from Maaenburg is probably a much more realistic take on this situation, though.

But when Katz says:

      Therefore, produce your master at the highest resolution, and at the end
      (the production master), use a single process to reduce the wordlength
      or sample rate. Multiple processes deteriorate quality more than a single
      reduction at the end.




You can't say that he's wrong, at all. He isn't referring to the present situation, but the one where you have a choice about the mixdown format you use, and it simply reiterates what I said before - avoid unnecessary bitrate changes. That still applies in the compilation situation - just because somebody else has already done the reduction, there's no point in compounding a felony, is there?