How far can dynamic compression be standardized?

I've been pondering the use of compressors on playback devices - and in particular, if it is feasible to standardize compression technology across all players and genres of music.

I already know that they are fairly well known for movie playback, but they don't exist on virtually all music devices. This is because there hasn't been a good use for an additional stage of compression for CDs - everything is compressed already! OTOH, movie soundtracks have well-defined (and large) dynamic range standards, so that a compressor is useful for all sorts of situations (playing very quietly, at bedtime, in the car, etc).

We already know that music is mastered loud because consumers demand it to be loud. We also know that compression is approaching a "brick wall" of maximum loudness at 0dbFS. And from what I understand of modern digital mastering, most mastering compression is only done with a few different tools that vary mostly in the details.

Here's what I'm getting at. If consumers can obtain complete control over the loudness of the music they listen to, then the Loudness War would essentially be over, and CDs can be mastered arbitrarily dynamically. Everybody would be appeased. The final mastering result may still have per-track compression as the instrument requires, but it could be convertible to a wide range of mastering loudnesses.

Obviously this is a fairly pie-in-the-sky idea and there would be numerous fundamental technical, commercial and societal hurdles to overcome before anything like that would happen. And I'm definitely not the first guy to think of this. Nevertheless I think there are some nuts-and-bolts questions I'd like to ask to the audio engineers here, just to appease my curiousity about the technical issues.

• How equivalent are professional compressor plugins nowadays, in terms of post-mixdown compression? Are the main differences a matter of preference? Could an average listener care about a change in the compressor used, given roughly equivalent settings between two different compressor implementations?
• How prevalent is classic multiband compression for mastering, without the use of proprietary algorithms?
• How genre- and album-dependent are the detailed mastering choices, such as attack/decay rates and band settings? I'm not fully convinced that they have much to do with the music itself, and more to do with the listener. ie, mastering a classical record like a hip-hop record is probably going to make it sound better to a hip-hop fan who doesn't like classical.
• How content are most listeners with the usual FM brickwall limiter practices?
• How much does this differ from HDCD, which as I understand is trying to solve a pretty similar problem?

 

Standardize? It's a great idea, but how do you deal with the band that wants his CD to play louder than the next guy's CD, and standardized compression in the playback gear won't allow it? There goes the standard...

 

Again, since the loudness is in the hands of the listener, said rocker could just turn the knob as far as he/she wants. Perhaps there could be some sort of "hinting" in place for releases, for suggested mastering settings. Just like fonts are hinted for proper legibility and style at all sizes.

This does bring up the question of if it's even legitimate to think of a loudness control as a defined range, when some people will want to crank it further than that range. Perhaps it should have an unbounded upper range, limited only by how much distortion is added.

EDIT: And again, mastering loudness is expected to plateau eventually, so rockers will be SOL anyway.

 

What Grant said.

 

It might be true that consumers like it, but they certainly aren't clamoring for it. Mostly people are oblivious to the issue.

 

Most high-end mastering is still done with analog and digital outboard equipment. The most recent album I mixed was mastered with Apogee D/A through Manley and other analog gear, with Apogee A/D feeding a Pyramix workstation; there, native plugins were used for EQ, and I believe final Sample Rate Conversion and dithering were done in Pyramix. Other high-end MEs use anything from complete analog chains to ProTools with plugins all the way. Plugin mastering limiters such as the Waves L2 or L3, or the UAD-1 Precision Limiter, sound different from one another at similar settings, but both serve the cause of limiting the dynamic range. As with anything in the chain, the average listener might not know why he or she likes one mastered mix over the next (comparing two of equal squashed loudness), but he or she might hear that one sounds harsh and the other sounds smooth, for example. As with anything audio, its a matter of choosing the right tool for the job.

Multiband compression (where one device treats different frequency bands with different levels of compression - a good idea on paper, at least) is actually used less than you might think. Brickwall limiter abuse is much more common.

IMO a good mastering job fits the music, not the audience.

We all grew up with FM limiting practices, and the dynamic range and power available for FM transmission hasn't changed in decades. Some current processers actually sound better that previous generations of processers. The current trend toward hyper-limiting of pop music makes many hyper-limited recordings sound just awful when they are treated by the FM station. If an easy, intelligently designed limiter was available on each car stereo and music was no longer hyper-limited, I think consumers would eventually find as many uses for their on-board limiter as they do for their tone controls.

---

When I listen to XM radio, I find myself turning the radio up and down a little. I love this. Some older recordings need to be cranked a little, but sound great because they were not hyper-limited originally or at re-mastering.