Search CE Pro






Print  |  Email  |  Comments (23)  |  Share  |  News  |  Follow on Twitter, Facebook, Google+ or RSS

Dynamic Range Day: Inside the Loudness War

Happy Dynamic Range Day! Founder Ian Shepherd, a U.K.-based mastering engineer, says the movement is gaining momentum as more people grow tired of listening to compressed music.


Dynamic Range Day

Dynamic Range Day was founded by British mastering engineer Ian Shepherd to educate the public about the “Loudness Wars,” which are harming the quality of today’s commercially released music.

Within the world of audio, both on the consumer and pro sides, there has been no greater point of contention than the increasing levels of compression used to produce music.

The phenomenon grew to the point where it was given the name “loudness wars,” because it reached a level where one engineer would try to outdo another engineer by making his recording louder. The problem is that it employing compression destroyed the dynamic range of many of today’s recordings.

Over the past few years, Ian Shepherd, a U.K.-based mastering engineer, Blu-ray and DVD author and owner of Mastering Media, Ltd., has advocated through his popular Production Advice website an end to the loudness wars.

Shepherd’s most visible action has been his founding of Dynamic Range Day. This year the movement will be marked on Friday, March 16, with awards for the most dynamic recordings, and a variety of prizes from manufacturers in the consumer and professional audio markets.

To help celebrate Dynamic Range Day, CE Pro asked Shepherd to explain what forces are driving the loudness wars and what can be done to stop them.

What does a mastering engineer do, and how is it different than a recording/mixing engineer?
Mastering comes after recording and mixing - in a nutshell the idea is to make the music sound as good as it can possibly be. When you are mixing, you’re balancing elements of the mix to make a song - guitars against vocals for example - whereas in mastering you’re balancing song against song to make an album.

Photos: 9 Albums That Maintain Dynamic Range

Another way to look at it is, it’s like Photoshop for audio. You get the best camera, you take the best photo you can and when you load it up on your computer you need to adjust the contrast, color, brightness - maybe you want to crop it, and that's what mastering is for audio. Adjusting contrast and brightness is like using compression and limiting to manage dynamic range and loudness; color is equivalent to EQ; taking out clicks and hums is a bit like red-eye reduction! Those are really close parallels.

What type of training do mastering engineers typically receive - are you a musician, did you go to a music school, did you start as a FOH (front of house) engineer, mixing engineer?
imageDynamic Range Day founder Ian Shepherd
I’m not sure there’s a typical route, but a lot of mastering engineers have come to the point where they were successful recording engineers first. I was the opposite, I did a degree in music and physics, and at the same time I was working on the live sound crew of the university sound crew. I was doing FOH and I learned to mic up guitars, amp cabinets and drums.

I went straight out of that to working at a studio for free. I worked there for several months and then SRT [Sound Recording Technology], a company based out of Cambridge took me on as a tape operator, initially. They did a real variety of work - lots of indie rock and pop - and they had a contract to record the Royal Philharmonic Orchestra, which at the time I think it was the biggest classical contract to be signed. They trained me as a mastering engineer from scratch when I was still young, I don’t think that’s typical. I worked there for 15 years, in the end.

I’m a musician - I play trombone, although once I started work I couldn’t find time to practice. Over the years I’ve done a lot of recording and mixing as well as mastering - and Enhanced CDs, which is a process of putting videos on CDs. From there we went into DVD, and when I left SRT it was because I wanted to move onto Blu ray - now all of that happens at Mastering Media.

Is it important for a mastering engineer to work with a mixing engineer and record producer during the mastering process?
I really like it whenever the artist or engineer is involved. We call that an attended session, but it doesn’t always happen. I can ask questions and get feedback, so it’s a really interactive process, but often you don’t have that opportunity. Sometimes something was recorded 10 or 20 years ago, and so you don’t have the opportunity to ask those questions.

I like it, but it’s not crucial - although there’s often an approval stage. If you’re mastering a catalog album, you’ll mail them a CD master and get their feedback that way.

Briefly, what is the background behind the term "Loudness Wars"?
It’s the name given to the trend for higher-than-average levels to be put on CDs. The idea is that if you play one CD and then another, your CD is louder. It’s based on the idea that “louder is better” - if you play two identical CDs, but one at a fractionally higher volume, people will normally say they prefer [to hear] the louder one.

That only works when you have plenty of headroom, though - a volume control on an amp, say. On CD, there’s a hard limit to how high the signal level can go, so you’re pushing the music harder and harder against this “brick wall,” so there’s less and less room to make it louder, and so it has to be “crushed” in to fit.

This problem happened before on vinyl, it was famous on the Motown label, for example. But by the 1980s, when I started listening to music, had it faded away - as I think the current loudness war will fade away, eventually.







Subscribe to the CE Pro Newsletter

Article Topics

News · Audio · Video · Digital Media · Apple · Itunes · Dynamic Range Day · Ian Shepherd · Loudness Wars · All topics

About the Author

Robert Archer, Senior Editor, CE Pro
Bob is an audio enthusiast who has written about consumer electronics for various publications within Massachusetts before joining the staff of CE Pro in 2000. Bob is THX Level I certified, and he's also taken classes from the Imaging Science Foundation (ISF) and Home Acoustics Alliance (HAA). In addition, he's studied guitar and music theory at Sarrin Music Studios in Wakefield, Mass.

23 Comments (displayed in order by date/time)

Posted by Shawn  on  03/16  at  08:55 AM

I wish everyone a Happy Dynamic Range Day and want more dynamic range in my CD’s. Case in point the new Van Halen album, I scanned every song on the album and they have max dB.

Posted by Chris  on  03/16  at  09:20 AM

Interesting article/interview. Thanks for that, Robert. I look forward to more articles on this very important topic.

Posted by Carl  on  03/16  at  09:33 AM

I agree more than 100%, if that’s possible.  i was involved in the recording industry in the late 60’s through the late 70’s.  When I was learning the trade, I was taught that 45rpm master cuttings were purposely over-compressed to compensate for cheap record players’ poor amplifier circuits.  33rpm cuttings were much “gentler”.  Even master tapes were sometimes recorded with the console’s VU meters pegged.  That was typically done with music aimed at nondiscriminating listeners.

Now I’ve seen the same thing, not only with a lot of modern music, but even with older music that was previously released without overcompression.  It is not only unfortunate, I believe it is downright stupid!

Posted by andysummers  on  03/16  at  09:42 AM

Is there an advanced search engine that has all the ever produced CD that have had compression add to pop rock film scores jazz hip hop and so on and so and so on…

So we can find out if one of many of our CD that we have brought has been compressed and when uncompressed version will be released sigh.

compression shouldn’t have been used on CD leave it to use to add in limiters or compressors to soften it down on our sound system if we feel its too loud or what ever don’t use it on the recording leave the recording as it is.

Posted by Gene DellaSala  on  03/16  at  10:41 AM

Good article Bob and glad you wrote it to help bring this very important topic to light.

I wrote something similar a few years back as you can see below:

http://www.audioholics.com/news/editorials/the-dumbing-down-of-audio

Posted by Robert Archer  on  03/16  at  10:50 AM

Hi Gene, thanks for reading and the link. It’s amazing the actually quality of satellite radio versus what it’s advertised to be.

Posted by Joe Finn  on  03/16  at  11:43 AM

Thank you for the great article, Bob.  Who cares?  Music lovers do, that’s who!

Posted by Mike  on  03/16  at  06:18 PM

I only want good dynamic audio when I’m sitting in my home.  More often than not, I’m stuck with headphones or listening in the car and office.  I pretty much need some compression to hear the whole song.  Isn’t it fun listening to classical music in headphones?  One moment it is too loud, the next it is practically silent.  Perhaps there is an opportunity here to provide both mixes in the file/disc/whatever or implement it in the players so the listener can select.

Posted by Darren R. Starr  on  03/16  at  08:07 PM

This was an interesting article. It seems the first time I’ve seen an interview with an “audio engineer” who appears to actually understand most of what he’s talking about. Unfortunately, he probably lost a huge number of readers when he started referring to dynamic compression vs. digital compression.

Digital audio is actually a pretty hard topic for many people, most often a much harder topic for people who are artistic first, technical second. If you don’t have a clear understanding of how the audio is stored, performing a task like mastering is a black art.

Here’s the thing… if you studio record audio using high bit depth (higher is better, floating point is better for range, but worse for precision), you end up with the tracks to edit. Then, when you edit, you should try to keep the “loudness” so that the greatest peek doesn’t get clipped or “dynamically compressed”. This will retain the highest level of precision. However, a proper audio editing program should render internally using 64-bit signed integers or larger for the mastering render. Up to that point, there should be no rendering except for monitoring and preview.

When the final to be “mastered” is output by the editors, it should have the peeks maximized.

Here’s where loudness becomes an issue. dB of loudness in digital is based on a “standard” from AES which suggests that a tone played with peeks between this level and its inverse level (the wave in the editor is between these two lines), is a decibel. That allows digital to analog converters lacking power amplifiers to output a pre-amp signal within a specified voltage/current range where this numerical value equals that voltage/current. Therefore, if all music is mastered with a similar set of average levels, then whether you’re playing Chopin or Metallica, the volume dial shouldn’t have to be moved much during transitions between songs.

The real problem is, by using those levels in lossless audio, it means you’re losing a huge amount…. yes, you’re actually losing a huge amount of detail because you’re effectively encoding most of your audio in 10-13 bits instead of 16-bits as the remaining bits create values outside of the range of the specified decibel level and if used, would force the listener to adjust their volume dial.

This is where MP3 and AAC can actually outperform CD for the same bit rate (or in some cases less). MP3 and AAC provide fields which define relative levels. This means that you an encode music using the full 16-bits and then provide reference levels for the loudness of the bit values. Alternatively, whether on CD, AAC or MP3, you can choose to ignore those values altogether and simply identify the average level per track on ingest to your player (you don’t actually use CD do you?).

AAC and MP3 are both lossy, I would need ten pages to list the issues with MP3, though AAC is much simpler… AAC is actually multiple different types of compression and the difference between a good encoder and a bad one is the psycho-acoustic model used to determine the ideal path for a given set of audio units. The original AAC specification defined different types of compression, I think HE-AAC2 defines quite a few more.

MP3 is a very simple, very static audio compression. Talking about things like dynamic compression is almost a misnomer in this context, in fact, MP3 is much dumber as it simply clips. This is what causes the crackling that even non-“Lame ass, I’m a listening tester wanna-bes” can hear sometimes.

But the loudness war is in fact a good thing… binary values of samples should define a predefined decibel level. Instead, the audio frames should define a relative value so that high quality D/A converters can adjust the levels at a higher bit detail level. The problem with the loudness wars is, some mastering guys insist on making the audio so loud, they’re forced to “dynamically compress” the peeks so their average will adjust to insanely high levels. Effectively clipping out their actual music.

On the other hand, the artists who make that kind of music arguably don’t make music but noise, and therefore it seems appropriate they would actually want to make their noise even noisier smile

Posted by Ironic that he mentions "The Bends"?  on  03/16  at  09:47 PM

Given that the album is named for a syndrome known as decompression sickness?

Posted by Slartibartfast  on  03/17  at  01:27 AM

@andysummers: Yes, not every release, but you can contribute too: http://www.dr.loudness-war.info/

Posted by Robin K.  on  03/17  at  03:48 AM

Excellent article. Also it’s the reason why most “remastered” albums sound horrible, especially if you’re a fan of the original recording. It’s painful to listen to an album you grew up on being re-released and sound like someone rode a sonic steamroller all over it.

Posted by Cletis  on  03/17  at  07:01 AM

The two fundamental theses of this article are that (a) bad engineers abuse compression in order to make tracks louder throughout, including the quietest parts, and (b) using fixed-rate 128-bit sampling when encoding an MP3 can exacerbate, or even synergistically worsen, that effect. Well, the first point is obviously completely outside the direct control of the consumer, but since all common digitizing tools that output MP3 default to using VBR that ranges up to 320 bits/sec, the whole point of this article is pretty much meaningless outside the world of professional sound engineers.

Posted by Jordan  on  03/17  at  02:55 PM

I wonder what he thinks about vorbis vs. MP3…?

Posted by James Martin  on  03/17  at  07:12 PM

I understand Mr.Shepard’s contention, and, yes, MP3s are massively compressed, thats where their appeal came from - because at 56.6kbps you could get a song downloaded in about 20 mins if it was encoded at 128kbps. Yet, reading this article, the moment I read that this douche worked on VHS to DVD mastering, he fell out of favour: DVD audio mastering is atrocious. At some stage, idiots like Ian Shepard decided to go with realism over convenience. Heres my case; I don’t bother renting DVDs or Blu-Ray discs. Its not that I don’t want to support an industry, local retailers, or artists. Its simply that I don’t want to piss off all of my neighbors! Lets take a random example off the shelf like District 9. What I want is to be able to hear the actors deliver their lines so I understand whats going on. What I get is earth-shattering explosions and tearing car-tyre sounds disrupted by ear-pulverising machine-gun fire, with a whisper of dialogue in the background. Fundamentally, all this massive bit-rate/audio level headspace has provided is a medium for morons like Ian to masturbate over irrelevant audio detailing and totally ignore the “normalise” feature. I would argue that audio engineers that worked on projects like “Them” (B&W ant sci-fi VHS film) had a better understanding of what audiences wanted. Actual recording engineers can see when their tracks are clipping, and actually do look at waveforms, so they do get to choose what they think is the appropriate level of compression for their application. Some performers simply care more about their recording quality than others. Its that simple. This guy is rubbish.

Page 1 of 2 comment pages  1 2 >
Post a comment
Name:
Email:
Choose smileys | View comment guidelines
Remember my personal information
Notify me of follow-up comments?

Sponsored Links

  About Us Customer Service Privacy Policy Contact Us Advertise With Us Dealer Services Subscribe Reprints ©2013 CE Pro
  EH Network: Electronic House CE Ideas Store Commercial Integrator ChannelPro ProSoundWeb Church Production Worship Facilities Electronic House Expo Worship Facilities Expo