upvote
There are tools to mitigate clipping artifacts, and tools to generate new transients for overly compressed files, but they're not a silver-bullet and the new material that is generated is more of a best guess than a true replacement for not over-compressing a mix in the first place.

These tools are most useful when used earlier in the process. Like when you just tracked an amazing vocal take, but the gain was too hot on one or two notes. The tools can mitigate some of the distortion artifacts to make it more usable. Applying these tools to complex material like a full mix will have some improvements, but at that stage there's less guarantee for convincing restoration of the record.

What I think non-professionals don't understand is that a record that is characterized by heavy compression is not something that happened at the very end with the mastering stage. It is an aesthetic choice that was made dozens of times along the way while recording, arranging, and mixing. Heavy compression is not necessarily a bad thing. Lots of amazing-sounding records harness it well. It's an art AND a craft. It takes audio engineers and producers years to do it well and with taste.

reply
No, you need the original mix to remaster it yourself.

If you just amplify the whole track until its max amplitude reaches the medium's maximum, yes you could undo that.

But the loudness war aims to make the whole track even louder than that, by quietening those max peaks so they don't clip, then that gives you room to amplify the rest of the track even further. The dynamic range of the recording is permanently reduced.

reply
people said "it's impossible to separate tracks (voice, bass, ...) after they are mixed". true in theory, but neural-networks can separate them in practice

same here, but there is no real market for somebody to bother yet

reply
> people said "it's impossible to separate tracks (voice, bass, ...) after they are mixed". true in theory,

and true in practice too. what you get out of an ai demuxer isn't an exact match for what went into the mix. it's a plausible approximation, much like ai upscalers, upsamplers, colorizers, and such.

reply
> but neural-networks can separate them in practice

That's a massive stretch.

They are able separate them sort of, but they are nowhere close the original quality of the individual tracks.

reply
"Assuming no clipping" is the biggest problem there, because the loudness wars resulted in a ton of very lossy clipping and similar artifacts. Arguably that sort of distortion became part of the expected sound, though, so just because it isn't reversible doesn't necessarily mean it is a problem.

In the open metadata world there is ReplayGain which analyzes music peaks and tries to create a negative gain to equalize the dynamic range to a standard volume at both the individual track and full album level.

Apple Music, Spotify, and others have proprietary but similar systems.

(As someone who deeply loves to shuffle an entire library, having a music player that supports ReplayGain has long been a personal requirement.)

reply
ReplayGain sounds pretty cool. Does it pre analyse your library ?
reply
ReplayGain is nice - but note it doesn’t ’fix the compression’. Compression and dynamic range is about loud/quiet _within_ the track. ReplayGain just turns the volume up and down for the entire track, the point being so all your tracks play back at about the same level. It saves a preset on the volume knob for you essentially.

If you remember making a playlist where one song is suddenly much louder than the last, and you’re riding the volume knob on every other song, you’ll see why this is nice!

reply
Yeah, you run an analyzer on your library and it creates MP3 or Ogg Tags that the player. Often you can leave the rest of the file as it was originally, just the new metadata tags.

On Windows I've always liked Foobar 2000 for its strong ReplayGain support, both automating the analyzing pass and respecting the metadata in the tags once saved. On Unix I was using Banshee for playback and automating analyzing with a pair of CLI tools I've forgotten the names of, one was MP3 specific and the other Ogg/FLAC-specific, as I recall.

reply
Even without clipping (which loud mixes almost invariably have), you lose resolution; for digital it should be obvious that if you start with M distinct values and remap them to N distinct values, you can't reverse it if M>N (which it will be for compression).

For analog there are similar limitations, but it's limited by other factors like noise.

reply
Short answer, no not really. It won’t ever be as good as a proper uncompressed mastering
reply
you can use an expander or something more advanced like Ozone 12's Unlimiter. you still lose signal when you compress even if you're not clipping so it won't be perfect
reply
No, it's a destructive process. It's like trying to get back the original texture of wood after you've sanded it down. The best you're going to get is an approximation of what it might have been like.
reply