RANT #7: Music Into Noise: The Destructive Use Of Dynamic Range Compression (by Wes Lindstrom)
Note: This rant was completely written after all the comments on this page were added.
A while back I realized that the current article that I contributed was, in fact, really fucking terrible. Poorly written, with tons of inaccuracies and stupid humor. Since then I’ve made a brand new article for my upcoming website, loudnessrace.org, which should be up about the time the sun explodes. Anyway, here is the new article which should suck a bit less than the old one and will hopefully replace it:
Why Today’s CDs Sound So Terrible: An Explanation of Overcompression, Hot Mastering and the Loudness Race
It's often been said that "what you don't know can't hurt you". But sometimes there's a difference between sensing and knowing, and it's those times when you're being affected by something you can't identify that you are in for a world of hurt. Like when that masked armed robber that breaks into your house, mutilates your cat and steals your belongings. You call the cops on the wrongdoer in hopes to get some retribution, but with no ID they're not really going to have an easy time making out the current threat. Which brings us to the recording industry.
With the advent of digitized music and file-sharing services, the general public has finally its first glimpse at the ludicrous and downright immoral ideas that run through the minds of the Recording Industry Association of America (RIAA) through their masses of lawsuits and paralegal tactics in an attempt to shut down any means of transferring music that isn't officially sanctioned by them and them only, which includes the current cesspools of bubblegum trash known as MTV and commercial radio conglomerate Clear Channel. And through the publication of knowledge given to industry experts such as Steve Albini, the "price fixing" policies of the RIAA, gradually lifting CD prices to ludicrous levels and paying the artists who made the very material on recordings possible a mere fraction of what is given to the labels and advertisers, has been brought to light as well.
But there is still one major issue left to deal with in regards to the industry. And as stated in the opening paragraph, it's the one thing that has been nagging you for years, but you just can't put your finger on what it is. And it manifests in issue of people's opinions of new music. So many people switch the radio station whenever the latest commercially-saturated pop song comes on the radio, and often long for "the good old days". Maybe it is the music, or perhaps it is something else. Something residing deep within the public's subconscious that they can't identify.
This article is dedicated to educating the people about just what that phenomenon is.
You've probably heard the term "compression" numerous times in regards to the subject of music and audio. The advent of converting verbose CD audio into a space small enough to send through a modem. But words often have more than one meaning, and can often refer to things that differ drastically. The most commonly known definition of compression in audio refers to eliminating barely or unnoticeable details in sound to fit audible data in a much smaller space. However, audio compression has existed in a different form long before the advent of digital music; this kind works on audio levels. Rather than squeezing audio into a less information-intensive format, this kind reduces them to a narrower range of volume.
What is it for, you ask? Compression has long been a tool to manipulate sound because it works on a microacoustic level. Think of someone fiddling with the volume knob during a song, and reduce that to mere milliseconds. That's how dynamic compression works. It's actually a very versatile thing; it can be used to shape the impact of an instrument when used lightly or alter the overall dynamics entirely. Think of it as the "glue" that holds everything in the recording together. Compression can make details in the music more readily audible, and can bring different sections of a piece closer together in level. This is used all the time in radio stations to get music heard consistently over the constant background noise of the environment.
The purpose of compression is to fit a wide array of details in sound into a smaller space, which brings us to the issue that has brought a lot of chaos into the industry: the limits of the recording medium. Every format has a limit for how loud the sound can be pressed on it, and compression is involved when it comes to getting as much sound against these limits as possible. Since the human ear responds to the average levels of music rather than what the sound peaks fully at, compression can be used to make recordings sound louder by attenuating small portions of sound to make more room for what the ear will perceive.
The recording industry knew this back during the '50s when jukeboxes packed with 45" RPM singles were introduced. Record labels had suddenly developed the notion that if they cut their records louder than a competing label's record they would be noticed more by the listeners and therefore outsell the competition. Each label kept cutting their single more "hot" than the last to extreme levels. Eventually, bands such as the Beatles popularized the concept of 35 RPM LPs, better known as albums, designed for home listening, leading 45"s to no longer become the industry's main selling point and thus quelling the "jukebox scene".
From then on, producers began to start looking into use of compression techniques for effects besides overall loudness. George Martin was one of the first to make significant use of them, and it's been said that his work with the Beatles was ahead of its time because of techniques like such. Phil Spector was another pioneer in this respect, well known for his heavily compressed "wall of sound" style. By the '70s, use of compression had gone beyond simply being a safety device and bands such as Pink Floyd and Steely Dan continued to take it in new directions and push it even further. But it was the '80s when compression really began to be the end-all-be-all of recording. Now that the age of music video had taken hold, music became more of a commercial venture than ever. And with the advent of comptuers, producers, engineers and artists were clamoring over the wealth of things that could now be done in the realm of recording. Suddenly all sorts of production "gimmicks" had come into play: canned reverb, pitch correction, take splicing, synthesizers, multitracking, the works. And compression was definitely high on the list. The explosion of the art of recording as both an artistic and commercial venture led to producers and artists getting a sound that people would notice. Something that would be in your face. No longer was the wall of sound a calling card for certain recording engineer prodigies, it was now the standard.
Back to the recording medium. Analog mediums such as vinyl, cassette and 8-track have an amplitude limit called the saturation point. Everything under this point is referenced by a negative number, which means that a portion of the music that is recorded at -12db will be played back at 12 decibels lower than what your stereo is set at. Likewise, sounds that are recorded above this point are measured in positive numbers. The reason the saturation point is referred to as a limit is because anything over the limit goes beyond the indented physical limitations of the medium and becomes distorted. As long as very small portions of the material go over this limit the sound can remain relatively transparent, making it a good reference point for the recording's transient peaks. In fact, a lot of artists and producers have a certain fondness for the warm, fuzzy sound that tape saturation makes, and often recorded their music extremely hot intentionally just for this purpose. The Rolling Stones were famous for this, with records that sound like they were run through a Bigmuff distortion pedal.
CDs, on the other hand, are different. Rather than being a physical medium, they are digital, which means they consist entirely of binary information. Unlike analog, digital audio cannot be recorded at levels over 0db at all, and rather than saturating or distorting, all sound that goes over the mediums limit will be stuck at full scale, resulting in a straight line. This is what is known as clipping. Whereas saturated sound produces a warm, fuzzy distortion, the sound clipped waves make are anything but; rather it results in a harsh, ear-splitting, unnatural noise that is very noticeable if enough of a peak is squared off. That's why digital recording software and digital mixers strongly suggest recording at a soft level so as not to go over 0db.
Anyway, in 1982 a new medium of recording was released to the public: the compact disc. Promising portable sound that never degraded with play or over time, it seemed to be the future of music. At first it was more or less a niche market due to the extremely high prices of early CD player models. Labels only mastered CDs out of necessity, rather than demand, often putting very little effort into the CD mastering process. This combined with digital/analog technology still in its infancy led to some very botched and terrible-sounding CD masters, often tracked at ludicrously soft levels due to engineers cutting straight from the original analog mixes. In the late '80s, however, CD players had reached consumer-level prices and exploded onto the market. Suddenly, labels began to take this new medium very seriously, and digital mastering became a legitimate profession, with mastering houses taking great care to cut as high a quality CD master as possible, and using compression to get more overall volume out of the medium.
Even then, the current technology didn't allow engineers to get the kinds of levels analog allowed without making subtle alterations to the dynamics. Enter the early '90s and the digital limiter. Finally, engineers could get their CDs as loud as vinyl and consistently peak to full scale without risk of excessive clipping or audible compression. They could now take full advantage of the upper regions of the CD's dynamic range, and for the first half of the decade or so, it was good.
But then something went horribly wrong.
Without an enforced standard for loudness similar to the one used by the film industry, engineers were wildly inconsistent with how they were to go about the new technology. How much do you limit? How much clipping is too much? Engineers began to get it into their heads that what they had were tools to make their CDs as loud as they possibly could, and eventually the artists began to pick up on it. Frustrated with the limitations of CDs, producers and artists began to resort to heavy amounts of compression and readily audible levels of clipping to try and get a "hotter" record than the competition. The loudness race of the 50's had come back.
As more and more volume was being squeezed out of the CD, the problem began to get exponentially worse. What was considered unthinkable one year became the standard the next. More and more CDs fell prey to the hot CD disease, leaving the situation we have today where the market is saturated with them, and overcompressed, heavily clipped CDs have now become the norm rather than the exception. And the damage done is a lot worse than you might think.
Dynamics is one of the key elements of music. Any music professor will tell you that. The light and shade of a quiet section bursting into a fortissimo passage, a vocalist raising her/his voice, the sudden rush of a drum fill bring excitement to the music, and when those things are neutered the excitement is lost. And it's not just dynamics that overcompression hinders. Transients, the peaks and valleys of a waveform, are necessary for music to have room to breathe. When compressors are cranked to excessive levels these transients are turned from sharp, punchy attacks to dull nubs, and much of the detail is lost as the harmonics are mashed together in a murky mess. The end result is sound that is dull, unnatural, distorted, and painful to listen to.
Now you know why you have a hard time listening to modern music.
The loudness race is one of the most tragic things to ever happen to the music industry. Listening to today's square waves, then going back to a CD released 8 years ago and marveling at the wonderful detail, depth and dynamics of those older CDs makes one long for those days when people took good care of their music. We've been denied the full potential of the CD medium for over half a decade. Our music, good and bad, has been turned to sonic mush, robbed of its excitement and aural splendor, murdering great music with noise-ridden mockeries of itself. It's time we took a stand and told the people responsible for this that we won't take any more, that we want those first 15 years of the CD back, that we want a safe industry loudness standard. No more contests to see who can release the hottest CD leading to mangled sonic bricks.
Knowledge is power. Through getting the word out to the public and the artists, producers, engineers and record executives we can stop the loudness race once and for all.
P.S. For more information on dynamic range and the loudness war, check out the following websites:
George Graham: Whatever Happened to Dynamic Range on Compact Discs?
Digital Domain: Compression in Mastering
Woodpecker Records: Compression: Tool or Sin?
Pro Audio Rx: What happened to Dynamic Range?
Stereophile Magazine: Dynamics and Dynamic Range
John Vestman Mastering: Hot CD Disease!
homerecording.about.com: Compression Master Class
Electronics Projects Online: Dynamic Range Processor
http://users.belgacom.net/gc247244/myths.htm
Post your comments about this rant
COMMENTS
Your rant against overcompression and reduced dynamic range in todays CDs is right on target. I had long noticed that CDs have gotten louder, but I wasn't aware of the negatives until I recently bought a new Rush album. Rush happens to be one of my favorite bands but their latest album Vapor Trails is a glaring example of how bad this problem has gotten. The music on the CD is great but the sound and production is absolutely terrible. The levels have been pushed so high that clipping in the form of audible crackling is clear and loud (and extremely annoying) on a number of the tracks. It borders on the unprofessional. Godspeed on your quest to stop this asinine trend in music.
The loss of dynamic range in pop music has been incredibly sad, and as noted above, the damage done to Rush's Vapor Trails album by this stupidity are nearly fatal. Alex has admitted he'd have to go back into the studio and re-record most of his tracks on the album to fix the problem. Where were the professionals while this was happening? What the hell were the engineer and producer doing? I've mailed letters of complaint to both Atlantic and Rush...fat lot of good it will do, however.
Check out this quote from Alex:
"Even as he was hitting the links, Lifeson was on the phone four or five times a day with Lee, who was forced to deal with unexpected glitches that didn't emerge until late in the recording process.
"We found problems that we didn't hear in mixing that were apparent in mastering. To get the kind of levels (we wanted), we had digital distortion. We remixed a couple of songs half-way through the mastering, through the remix, back to mastering," says Lifeson.
"The poor guy (Lee) was doing this on his own. It really shook him up."
[email protected] (Teejay Riedl)
I agree with Wes’ technical assessment of the technologies at work here, but I’m not entirely convinced regarding his “black helicopter” justification for applying excessive compression. While his points are valid, the matter of individual perception still comes into play.
Here’s an example: way back in the early 80’s, I borrowed a copy of Mozart’s clarinet concerto in D from the public library, on an LP. I took the LP home, and recorded a cassette tape of it. That tape was placed into yonder car, as it happens to be one of my favorite pieces of classical music, and for about 10 years I had it in the glove box, ready when needed.
Because of the “limited dynamic range” which the LP mastering engineers had applied to keep the needle from jumping out of the record grooves (caused by the excursion brought on when the lower-pitched instruments started to rumble - on an LP, low bass requires a large amount of "hills and valleys" along the groove wall), the recording was perfect for the automobile listening environment – the quiet passages were quiet, but still audible, and the loud passages were sufficiently high in volume to enjoy without reaching to turn down the system.
Fast-forward a decade… I bought another version of the piece on CD, recorded in the new, glorified D-D-D methodology: that is, digitally recorded, digitally mixed, and digitally mastered. Using the enhanced dynamic range afforded by the CD, I found I could not listen to the CD in the car. When I turned up the volume in order to hear the quiet passages, the loud sections were so loud they made me jump.
Fast forward another decade, to another version, again in DDD format on CD, but this time with the advances in multiband compression used in the mixing/mastering process… and finally, I can enjoy the Concerto again.
The prescient reader has probably already begun the response to flame me, yelling “Hey! Wes is talking about POP music, not classical!”… but the fact remains that the same thing was happening across all genres. When I first heard Nirvana’s “Smells Like Teen Spirit” on the radio, the dynamic range was being compressed. That was back in, what, 1992?
Even though both the producer and the radio station applied their respective compression, it still wasn't enough to tame the disparity between loud and quiet enough for the automobile listening environment.
It all boils down to a matter of taste – music production fads come and go. The “reverse/gated reverb” drum sound of the 1980’s came and went, and even though the current trend is to have vocalists with no musical range groaning into a dry mic as if they had old whiskey-soaked gym socks stuffed down their larynx (aka Creed and associated clones), I’m confident that’ll pass too.
Extend the paradigm of the points Wes raised, and you’ll realize that if Cool Edit, Sound Forge, Audio Cleaning Lab, etc. are so inexpensive (as are multiband compressors – you can by a Behringer 9024 Ultradyne 6-band compressor for about $175), even if the record companies stopped squashing the sound, anyone who needed enhanced legibility in a public space (as Wes correctly asserts) now has the technology to squish the sound themselves.
Rather than tilt at THAT windmill, I’d prefer to topple the Corporate Recording Industry on the principle of their stylistic stranglehold on the genres of music they arbitrate the radio stations play, and getting rid of Creed and associated clones… but undoubtedly, whatever popped up to replace them would be even worse.
Oh, bloody hell – I’m geezing!
Great piece of writing... Thought you might be interested in this:
mailto:http://www.informatik.fh-hamburg.de/_windle_c/Logologie/MP3-Gefahr/MP3-risk.html
Hi I couldn't find the email address for Wes on your site (author of rant #7) so I decided to email you. Basically I completely agree with what he said and I just want to say that if you think music lovers have it bad, you should be an audiophile! For years audiophiles have been moaning about the over-use of compression in music cds and the disasterous effect it can have on the dynamic range in music- affecting musicality, impact, naturalness and transparency. I actually have tried to find a solution to this problem and though I'm not sure, I think I've found it. It's the first (and only?) hardware de-compressor on the market by spl called the dynamaxx. It seems like the 'holy grail' for audiophiles, this device looks like it could turn poorly mastered albums into something closer to the original sound. I was wondering whether you or Wes knew of it or had used it?