I've read about proposals to create "black lists" of music producers and mastering engineers who harmfully compress the records they create on the Wikipedia article on Loudness War.

What about it? Anyone interested in setting this up? In a Wiki+Forum combination maybe? Of course backed up with HA.org like research standards: music samples for verification, screenshots of waveforms, amplitude stats, replaygain figures, etc...

I'm not so sure it's the producers/engineers that should get blacklisted. I'm sure the labels are dictating how loud their albums sound when they end production. Regardless of an album being too loud or not, how do you know they didn't squash everything during the mixing process?

There are more overcompressed albums and engineers, then properly mastered albums.

Therefore, it would make more sense to use a whitelist (automatically blacklist everyone, unless he's on the whitelist).

This would also solve chang's argument. If the albums mastered by an engineer, tend to constantly be non-compressed, then for that to happen, BOTH would need to be proper: the mixdown and the mastering. Thus, any non-overcompressed album automatically means proper mastering (regarding overcompression).

- Lyx

@ChangFest: Well, don't worry. Of course, this includes the mixdown process, too. That's why I said "music producers and mastering engineers", naming both explicitly thus pointing out that overcompression can either happen during the recording and mixing process or the mastering or both. AFAIK the producser and the mixing engineer is often the same person, or both are working together, whereas the mastering is often done under a seperate supervision, out of reach from the producer and band... Yeah, it's not that way with all bands... all three stages, recording, mixing and mastering can happen at the same place or at three or even more studios, supervised by one set of people or more...

You cannot generalise this from start and I don't intent to (like saying "let's focus on mastering engineers!"), I think all this detail might become interesting when dealing with a specific album or artist and needs to be evaluated and researched from case to case.

@Lyx: Haha, maybe that might be a better idea, indeed.

It would be far more productive to whitelist/blacklist reviewers instead of mastering engineers. Quite a few engineers know full well the issues with overcompression but their decisions typically are overridden by producers or label execs. The only real way to change this situation is to change the perception of the public, if not at least the industry, and the most direct way to do that is to promote influential people who advocate this sort of thing.

Hm, well I'm aware that it's not always the engineers or the musicians who are responsible. Instead this list should contain the name of the person or label that is responsible. It's wrong to say that it's always the label, because it wouldn't explain why so many indie records are badly mastered, where there is virtually no pressure from the label itself onto the engineers to make the record sound "hot". Those ndie bands who have full controll over their record either do it on purpose or they (or someone they hired) screwed up because they have no clue about mixing and mastering.

A simple approach from a user's point of view would be a database that connects all these variables: artist, album, date of release, format (CD/LP/DVD-A), producer (+engineers), label, etc. Now when you look up a producer, you will see a list of records he was involved in: some are red (sound bad), some are green (sound good), plus all the other data (artist, year of release, label, other engineers involved).... that way it's easy to see patterns. For example producer A did do some bad records, but only when he was working for an artist from label B, when he was doing work for an artist from label C, he made good records... now you look up label B and see a lot of bad releases, then you look up label C and also see lots of bad releases, except a few artists, then you look what's up with those artists that seem to care about good sound and who they usually work with... and so on. That way one can see more clearly: who in the industry is likely to spoil a record and who's less likely to do that and who's strictly against this kind of practive. Top lists will point out the individuals or labels with the highest/lowest percentage of clippressed records. If you want, these Top Ten/100 lists are the "black-" and "white lists" then.

Now who decides whether a record (or release) sounds bad? Registered users, who own the release can submit a vote, plus a set of information, like short samples, screenshot of waveforms, amplitude statistics, etc, in order to somewhat backup their decision. It will never be totally objective since noone's actually been there when the engineer pressed the buttons or turned the knobs of the limiter unit. But hopefully many contributions will make it somewhat reliable at least for the more popular releases/labels/producers/artists. And I think there are many ways to make identifying clippression and distortion but also mildly compressed records easier for inexperienced contributors, too.

I think that's the case most of the time. It largely depends on the style of music too. Jazz and Classical recordings are going to have an excellent dynamic range and aren't going to be clipped, where as rock and pop recordings will always suffer. The real challenge is when a CD is "remastered". A lot of remasters tend to be garbage these days, although there a few exception cases. You know what would be even better? a standard like with Bob Katz K system. That's what ReplayGain set out to achieve, although if a CD is clipped it's clipped.

QUOTE(Fandango @ Aug 2 2006, 23:04)

Now who decides whether a record (or release) sounds bad? Registered users, who own the release can submit a vote, plus a set of information, like short samples, screenshot of waveforms, amplitude statistics, etc, in order to somewhat backup their decision. It will never be totally objective since noone's actually been there when the engineer pressed the buttons or turned the knobs of the limiter unit. But hopefully many contributions will make it somewhat reliable at least for the more popular releases/labels/producers/artists. And I think there are many ways to make identifying clippression and distortion but also mildly compressed records easier for inexperienced contributors, too.

I think the black list will give some beneficial effect. But a user *has to* contact a record label or executive engineer about the issue, along with providing waveforms etc., and then submit their answer too, such as "we know, we intentionally spoiled the sound" or "really? I did not noticed anything on my equipment" .

But you'd still have to get over the hurdle of overcompressed meaning different things to different people... some people can't stand anything mastered louder than an 80's CD, others don't mind so long as the sound remains natural and doesn't clip everywhere. Some people think all clipping is an abomination, others don't really care so long as it's not blatantly audible. Obviously everyone is going to say that Californication sounds terrible and any given Autechre CD sounds great, but the middle ground would be tricky to deal with consistently--I can name more than a few CDs I think sound good enough but that others would have issues with. And you have to take into account certain albums sounding intentionally terrible or abnormal. Like Ulver's uber kvlt blakk metvl Nattens Madrigal, which basically never pops below -7RMS except for the acoustic bit in the first song.

Yes, I think besides the huge amount of work such a project will cost, the subjectivity that goes along with every submission are the biggest issues.

The subjectivity might be overcome by a clever rating system. As I mentioned above the web interface will show you a percentage of how many records an artist/producer/label put out are being considered "bad". Now the actual releases being reviewed need an even more sophisticated rating system. On a scale of 5 reviewers can rate (by guessing but also by analysis) different aspects of the record like the amount of compression, clipping and distortion. Besides analytical tools will provide objective statistics: peak sample/number of clipped samples, RMS, and so on.

There's a lot of subjectivity involved, but when you look at review websites like imdb.com, you can see that they somewhat managed this problem.

So I think it is prossible to get quite useful results, but it's probably new grounds because I don't think anyone has ever done this in music science and it's tricky because it involves limited objective data, as well as subjective judgement and sparse first hand information. While for some albums there might be info about the recording and mastering available (which would give the most accurate info), for most it is not. So there are many types of inputs that have to be transformed into a useful rating somehow including a probability of accurateness. That's actually a scientific task, and seriously should only be carried out by individuals who have scientific experience.

My "Anyone interested in setting this up? In a Wiki+Forum combination maybe?" was actually a bit of a naive and precipitant idea.

Rating things as "bad" or "good" produces way more heat than light. Wouldn't a table listing mastering engineers/labels/producers and the average replaygain score of their albums achieve the same thing and allow people to make their own decisions?

For example, on cheap boom boxes highly compressed CDs sound better, not worse. The fact is that a large percentage of consumers of music use just such devices to reproduce their favourite tracks.

But RG is only useful as a way of telling if something might be sonically atrocious... Arcturus' Radical Cut clocks in at about -12db but doesn't even hit fullscale on every drumbeat, for instance. It isn't even a brick if you zoom all the way out. Nightwish's Once, on the other hand (...I didn't get into metal with the manliest of bands) only gets about -7 or so but sounds flat and dull. So while you could just give people RG values and let them make decisions based on that, they wouldn't necessarily be good decisions.

true, RG is not a proper measure, it has to be subjective and it has to be supported by more than 1 reviewer, also instead of making a people blacklist it would be wiser to make albums blacklist (including year, label, ect.)

RP values should only be one bit of information that will count into the final rating of a release. Practically it's nothing more than a clue.

smok3, as I said somewhere above, there could be several dynamically generated "hitlists" of albums/producers/bands/labels/release years/whatever showing an average rating of quality (in case of albums) or percentage of good/bad releases. These hitlists, top/worst 10/50/100 lists or score lists (...whatever) are much better than an actual white- or blacklist that contain nothing else but names and no ratings/scores.