My first attempt at any kind of wiki article. Comments? Corrections? Karl
Jan. 3, minor changes, added note about lack of emotional expressiveness caused by dynamic flattening, independent of distortion. Karl
someone should mention the idea of headroom (which is basically the left-over space on the volume meter when the track reaches its highest level). songs with digital clipping have no headroom as there is no space between the peaks and the maximum possible levels of the track. songs without clipping may have headroom, but brick-wall limiting can still result in restrained dynamics. some argue that among "loudness war" records, those with ample headroom are more listenable than those without any, even if they are "squashed" just the same. lack of headroom in general means an improper representation of transients, so a cd that clips but is not so compressed may sound much worse than a cd that is heavily compressed/limited but has headroom to keep from clipping.
Can we get some examples of albums released during the years of the "loudness war" that dont subscribe to the loudness philosophy? For instance, I believe most of Brendan O'Brien's releases (Pearl Jam, Stone Temple Pilots, Soundgarden's "Superunknown") don't really have any clipping. There's a totally audible difference between Audioslave's first album (not engineered by O'Brien), which has over-the-top digital clipping, and their second album (engineered by O'Brien) which seems to substitute tape clipping for digital (correct me if I'm wrong but I definitely think I hear tape on that album). Even so, the dude clearly does use compressors and limiters extensively.
Tool has a few albums, such as Lateralus, which have tons of headroom.
This article needs some work. It doesn't seem to be in a neutral POV and parts of it are wrong (or at least misleading.) There's a big difference between distortion caused by clipping and distortion caused by "companding," for example. It's unlikely that commercially produced CDs will actually have clipped waveforms on them; clipping is more likely to come from the playback hardware. People just assume that heavily expanded audio must be clipped but this is usually incorrect. Also, many of the concepts in this article do not apply to digitally produced (not recorded) music, especially purely synthesized music, where "louder" can actually improve the quality of the sound, at least in theory. Better general explanations of the concepts involved and fewer references to external sources would improve this article and make it more neutral. As it stands, it seems more like a magazine article to me.
I don't want to discourage you, but you did ask for comments... Please try to improve this article as it has some important concepts.
I have to agree that this article needs some serious work. It does not have a neutral POV and is rife with weseal words. I'm going to go ahead and add the Weasel template to the page. Personally, I'm an audiophile and I absolutely can't stand the "loudness war," but it really needs to be NPOV. Its an encyclopedia article, not a rant. --Tom 07:36, 26 July 2006 (UTC)[reply]
Yes, although I am not professional, my understanding of the subject is that maybe masterers would occasionally allow clipping of maybe half a decibel, but more often, dynamic devices are used with brick-wall limiters on the end. Could someone in the industry please verify, but this information sounds like a description of mastering ten years ago. Nowadays mastering is often done completely digitally and the engineer is not guessing at the peak levels at all. Engineers rarely treat digital full-scale the same as analog saturation point. More often, they simply compress or limit the mix aggressively which causes the whole thing to pump with the kick drum. One reason why engineers are able to achieve high dynamic ranges is because with digital technology, the clip point is accurately known and not guessed at all. Peak metres are readily available, and an engineer can tell exactly when clipping will occur and is able to position a limiter so that it limits exactly at the threshold of clipping as opposed to the analog days when the line between clean and saturated was more blurry. I have generally noticed more compression during mastering in recent times, and less distortion. (there was some serious saturation in early Led Zeppelin albums).
Compression and limiting still decreases the quality in music though, and I do believe that there is little understanding of audio quality by consumers. It seems that people are quite happy to listen to low quality MP3 music. My opinion is that it is worthwhile sacrificing some RMS level and not always using all 16 bits of CD quality audio, however it may be a good idea to represent other people's ideas in the article. I think it would be advisable to place more emphasis on the fact that through decreasing the dynamic range of the audio, the 16 bit quality of CD is better put to use for more of a song. It is worthy of note that by digitally encoding a passage of audio six decibels louder, there are twice as many different levels used by the samples making up the waveform.
This is my understanding, but I do not pretend to be an actual masterer, just a hobbiest, so I think that we should wait for someone in the industry.
I came across this page while trying to bring together all Sound recording and reproduction topics under the Wikipedia:Root page concept. This is a topic that causes me much grief and I think it is important enough to stand alone as a branch of the Root page, where it will be easily found by those looking at any audio page. It is of equal interest to the listener as to the recording engineer, and so I have not placed it under Post production. --Lindosland 16:45, 29 March 2006 (UTC)[reply]
Merge request has been removed per majority opinion after reviewing the discussion pages both here and at Talk:audio mastering. -- Dept of Alchemy 21:57, 10 May 2006 (UTC)[reply]
I was thinking that this article was about the military use of loud music as a weapon. One example was the US Army attack by playing loud rock and roll against Manuel Noriega when he was refuged in the Nuncio residence in Panama City.
Many have suggested boycotting recordings that they feel showcase the phenomenon to the point of significantly lessened satisfaction with the product (often to the point of lessened, or even nonexistent, listening as opposed to otherwise) to communicate the existence of disdain for the practice to the offending parties, though it is often stated that such an attempt would be interpreted by the music industry as wanton piracy.
This bit doesn't make sense to me. Is it a result of a section being chopped out? As I read it, the implication is that boycotting a recording is interpreted as wonton piracy. That is nonsensical. -- Jon Dowland 20:30, 25 May 2006 (UTC)[reply]
Actually it makes perfect sense. The music industry is so arrogant nowadays that it simply refuses to believe that people won't buy CDs just because they sound awful. Instead they will (incorrectly) assume that just because sales are down "people must be downloading this music illegally". Totally unable of course to realise that because it sounds horrible and gives you a headache on the original CD, it's still going to sound horrible and give you a headache when ripped to MP3. -- Squirrel 09:32, 29 June 2006 (UTC)[reply]
Fair enough. Doesn't change the fact that music ruined by loudness wars is fatiguing to listen to for any length of time (even after volume levelling) due to the amount of compression. And all this treatment does is to make all recordings sound the same, flat and lifeless, and means nobody wants to spend their hard-earned cash on them. Squirrel 10:40, 23 July 2006 (UTC)[reply]
These quotes are taken from the Dolby paper linked-to after the quotations:
"Despite the lack of explicit calibration of loudness level to a reference signal level, loudness on media designed for the consumer has hitherto ended up with some degree of consistency. Consistent loudness has been achieved, by and large, by one of two methods: either by conventional taste (e.g. a lot of pop music sounds wrong if it has a large peak-to-average value, so it ends up with just around 6 dB between the average and peak levels: not in itself a bad thing); or because the medium itself imposes performance limitations on the programme that dictate an optimum use of the medium."
"Consumers use programme loudness - typically dictated by speech levels - to set their volume level. But clearly there will be a relationship between the acceptable range of average speech loudness and the reference levels / peak levels in consumer audio systems. There are some overriding factors that a consumer expects from the audio system in the home. Broadly speaking, the basics are:
1) outputs do not distort; 2) the volume levels match between different audio sources, e.g. between TV channels and between analogue TV and anything else that plays through the same system.
Fig. 3 shows the basic requirements for a consumer TV audio system. (see linked-to paper)
The three outputs shown are designed to feed three different applications. The Line Output drives a hi-fi system; the RF Output is tailored specifically to match the audio levels that result from typical RF-demodulated TV audio. The Digital Out socket supplies the undecoded bitstream to a home cinema system. To satisfy condition (1) above, the peak audio output from the broadcaster should always be ≤ +8 dBu. To satisfy (2), the broadcast signal must always have the same properties as analogue FM audio to match the RF output requirements (because the reference level does not measure loudness, and signals cannot exceed +8 dBu on a consumer receiver). This leads us to the unpleasant conclusion that, unless the broadcaster and receiver have some new way of dealing with the relationship between audio programme levels and programme loudness, the broadcast signal must have the same properties as analogue FM. Most disturbing is that this applies also to digital media even when using MPEG stereo audio on digital TV, and MPEG or PCM audio on digital packaged media. This rather cuts across the broadly-held belief that digital media offer the promise of a new level of audio quality."
- Loudness and in broadcast audio Dynamic Range — the Dolby solution, Tony Spath, Dolby Laboratories, Inc. Available freely from the EBU: http://www.ebu.ch/trev_293-spath.pdf
It's possible that the harshness of modern pop recordings is because of the difference in playback volume and equipment between the studio and the home. In the studio, mastering will be done at a high volume, probably over 100 dB, and on speakers that have a good deal of bass extension. As the loudness contours show in the article, at higher volumes the tonal quality of the sound changes, the with the mid-range and treble more disguised (or masked) by bass energy. Volume levels in the home will most likely be lower than in the studio and the speakers will likely have less bass extension. At high volumes the bass output will not be as high as in the studio because of the limitations of the typical domestic loudspeaker and amplifier. This will affect the tonal quality of the music.
First attempt was perhaps not neutral, but the removal of the first paragraph also removed the whole point of the remasters section.
Some remasters do indeed sound better (eg very early CDs), but more recent remasters are just "oh, bit of smiley EQ and 4:1 above -10dB... and let's make it louder!" The ABBA examples mentioned are one of my favourite songs of all time, and it's tragic to see it destroyed in this way. So forgive me if my POV is a little less neutral than it could be on this particular track.
Squirrel 10:45, 18 July 2006 (UTC)[reply]
I was under the belief that the first audio CDs were not released until 1982 or 1983. It seems baffling how one could rip a recording from a CD which hadn't been invented in the 'one of us' section. Perhaps someone could clarify.tommylommykins 07:15, 21 July 2006 (UTC)
ABBA's final recordings (The Visitors album and the 1982 sessions) were all digitally recorded. The Visitors was one of the first CDs released when the format was launched in early 1983. Digital recording has been around since the late 70s but was still expensive, so most studios didn't have digital multitracks back then.
The CD I have here says on the spine "DIGITAL STEREO", and in the booklet "Digitally recorded and mixed at Polar Music Studios, Stockholm". The CD format was under development in 1981 when this album was released, but it looks like they had CD in mind when it was being recorded. Either way, it says on the back cover and on the disc itself "© 1981 Polar Music International". So it's fair to assume that the original Polydor release of the CD was taken straight from the 2-track digital master.
My friend has the 1997 remaster of this album (which is also heavily compressed), this says © 1997 on the back cover.
Hope that clears things up. Squirrel 10:50, 23 July 2006 (UTC)[reply]
Can we think of some other 'warning tags' to throw up here...? Thus far we have an article that BOLDLY! claims to have "compromised" "neutrality" by "weasel words", is "in need of attention from an expert on the subject" AND has some information which "has not been verified" and "might not be reliable"...
It isn't as if Wikipedia doesn't have enough of these $%*#@! tags all over the place (more so recently); as this article is regarding the "Loudness War", I really do not think it warrants all this special attention.
Where do you think the folks over at Britannica 'draw the line'? Nothing is perfect; this article is informative, and covers a 'pop-cultural' term/idea. Do we really need to warn the reader of possible bias in this article? Aren't they old enough to 'handle it'?