Re:Vinyl set to outsell CDs for first time since 1986
by monocasa on Sunday September 08, 2019 @01:00PM
(#20911701)
AFAIC, dynamic range compression is very nearly as bad as clipping. It just curves out the signal so you don't go straight off of a cliff, but you're still trying to fit 10 lbs of potatoes into a 5 pound bag.
Re:Vinyl set to outsell CDs for first time since 1986
by stjohnswarts on Sunday September 08, 2019 @01:58PM
(#20912138)
You don't understand it then especially with modern 24 bit and up sampling available in flacs and other standards. clipping is almost always going to be worse than dynamic compression except under the worst edge case conditions in 16 bit and above (standard CD limit)
Re:Vinyl set to outsell CDs for first time since 1986
by monocasa on Sunday September 08, 2019 @02:55PM
(#20912572)
I'm literally a engineer who does DSP, so yeah I do understand it.
24-bit doesn't tend to matter, because somewhere in the mastering pipeline will inevitably be 16-bit nine times out of ten.
And the compression they use in 16-bit to try and have their cake and eat it too has more noticeable than the issues with vinyl IMO. Yes, there's technically plenty of room in there for the signal without compression or clipping, but like I showed above there's systemic reasons why they don't use the bandwidth effectively.
You mention that "it's getting worse", but it's unlikely that you'd be able to find a release mastered/remastered with noticable digital clipping within the last decade.
As one example of how basic a concern this is for digital mastering, Apple's "Apple Digital Masters" (formerly "Mastered for iTunes") program strictly limits it, recommendeds testing for the existence of clipping both pre- and post-encode, and provides tools (afclip, AURoundTripAAC Audio Unit) for helping to do that.
Generally, "loudness wars" refers primarily to dynamic range compression, which can easily destroy a track (in the musical sense) even if there is zero clipping.