This is about compression. Not the sort of compression that makes file sizes smaller, though the two types are often confused. Audio compression reduces dynamic range (the difference between loud and soft) in order to make recordings sound ‘punchier’ or – at least perceptually – louder.
One of the advantages of a louder sounding recording is to make albums more ‘impressive’. This is often considered particularly important on smaller speakers, through radio, or via mp3 where most users are not listening in ideal conditions (ie: on little headphones, on the bus).
But audiophiles – and, increasingly, just people who like music – are complaining that too much compression and not enough dynamic range is killing the quality of recorded music. And they kind of have a point.
There’s a video that demonstrates this nicely:
Typically, the compression has been added in CD mastering (a large part of the reason that vinyl aficionados prefer the sound of records) and it happens a great deal with radio. In fact, FM stations notoriously compress the hell out of everything all over again – ostensibly to protect their transmitters from ‘clipping’, but mostly to just sound louder than the competition.
The problem with compressed audio (apart from just the sound that results from the lack of dynamic range) is that it’s fatiguing to listen to. People can listen to music with ‘air’ in it for far longer than something that is sonically relentless.
Now, I raise this because there is a tension between hypercompression (the kids are listening out of the speakers of their cellphones, for goodness sake!) and what I call dynamic puritanism.
I’d argue that compression isn’t of itself a bad thing, and as a former sound engineer and producer, I’ve been known to put compression on the sorts of things that purists would be horrified at (I compressed a grand piano on a jazz recording at 3:1 once and drove it fairly hard too).
But compressing a whole mix so that there’s no dynamic range whatsoever kills the sound of a recording (and here comes my point) far more than data compression does. A lot of contemporary recordings have been degraded far more because of audio compression than because of file compression.
Taking out the dynamic range on a rock or a hip hop record makes it sound worse than converting it from CD quality to an mp3 (as long as we’re talking decent bitrates).
So before we start getting sniffy about what mp3 is doing to audio fidelity, let’s sort out the more important problem first. Just ease back on the compression when you’re mastering your masterpieces, would you?
Download Andrew Dubber's new book Music in the Digital Age - or, if you already have and you've been enjoying it or finding it useful, please consider paying for it here.
Tagged: audio, compression, mastering, mixing
4 Trackbacks
You can leave a trackback using this URL: http://newmusicstrategies.com/2008/08/11/whats-the-loudness-war/trackback/
[...] a blog-entry that caught my attention because i thought about this for quite some time now. read it here. it’s about the loudness race/loudness war. this term describes the fact that producers [...]
[...] the post, entitled “What’s the Loudness War?” Andrew Dubber points out the difference in compressed file size and audio compression, which [...]
[...] usually would be at the expense of the normal dynamics of the music. Andrew Dubber blogged about it here with a video that demonstrates the [...]
[...] The Loudness War – A great article about whether or not audio compression is a good or bad thing. [...]
27 Comments
Hey there Andrew,
Big Al Wagner here from the Project Studio Network Podcast.
Thanks for this post. We have an initiative on this very front going on. Please check out our non-profit foundation to help turn this trend around at http://TurnMeUp.org.
This was founded by Grammy-winner Charles Dye, recording artist John Ralston and me. It reflects our real-world experience with consumers artists, labels and mixing and mastering engineers.
Hope to have you on the show again for an update on your endeavors.
Keep up the great work!
I agree completely. As a producer and musician, the compression war has affected the way mastering engineers work. I often have to nearly BEG my mastering guy to back off the compression to let the mixes breathe a little because the poor guy is so used to compressing everything to death.
On the other hand, a well compressed mix is great for in the car when there is lot of ambient noise blocking the quieter moments in the mix. I can’t tell you how many times I reach for the volume knob when listening to Damien Rice’s “O” album so I can actually hear it during the quiet parts.
I think it would be nice to have a button to push to select whether a mix is compressed or not when you are driving.
Hello Andrew!
We’ve been there too. Until the late 90ties we were suffering from the fact that our CDs were just not as loud as major label products. Then we could afford the gear and were so proud that our tracks hit the meters as hard as the stuff from prince and other highly compressed recordings
It took us a while to realize that this was dead end street.
These days with Blue Star we are using decent compression on single tracks and mastering. Compression is a wonderful tool, it can highlight the essence of a track and we use it wherever it helps the music.
When we master production music that will be part of compilation CDs, we have too check that our tracks don’t sound weak compared to the other material but also not too compressed.
All in all the loudness war is a fading….. except for commercials.
But what about the loudness war that’s happening at live concerts? I have the feeling that many live mixers have severe hearing damage and I know many people who don’t go to concerts anymore because the sound levels are too much.
Thanks for posting that. I’m on my own personal crusade against over-compression.Nothing turns a good song into crap like too much compression. That video is extra cool. Had to share that with my friends.
I guess it’s the same with bad design, when form beats function. Dynamics are, and have always been, one tool of expression for artists (at least the better ones) – so it’s a loss for our art. But what can one do if he/she wants to be on radio or played by a DJ or something like that. Not joining the race makes the song sound wimpy and weak compared to the others, compressing the hell out of it sacrifices an important part of the music itself…
On the other hand, does the consumer even care about that? Chances are, he/she doesn’t know about this matter and the non-audiophile-guy might not even notice the difference. Listening through low-grade gear often makes it not possible even for experienced listeners…
And, if the consumer would know about it, wouldn’t he rather choose the over-compressed version because he could listen to it everywhere without adjusting the volume all the time?
This is an interesting question. Would the ‘standard-listener’ appreciate a recording with full dynamic range (which, if I remember correctly, is already compressed down to 90dB on a CD, right?) or would he rather think of it as thin and weak – and therefore of minor value – because the competitors (including commercials) use compression to ridiculous lenghts and he got used to that? Which is followed by the question: could the artist afford to go against the customer’s wishes or would he have to adapt? Even if it means to reduce the artistic value?
Ha. By the way, my home stereo has a button to compress while listening. It’s called ‘night mode’ and is meant to reduce the dynamic range so you can watch an action movie without disturbing the neighbours – pretty cool feature.
So where does this lead us? Two versions of the same song? Equipment with such buttons? Unlikely, the audience does not care enough to make this a major issue. But Peter Blue might be right about the fading of the war. A few years ago every musician panicked because real instruments wouldn’t be needed anymore in the future – computers were to take over everything. But did it happen? Today it feels more like a renaissance of playing instead of only programmed stuff all around – and I guess there will come a time (again) when dynamics do matter….
Great discussion by the way :-)
Andrew
I am beyond stoked to see you writing about this topic. The loudness war is a vicious cycle of loudness chasing loudness for all the wrong reasons. When we launched our blog it was the first topic we wrote about…as I believe it to be among the top issues in the culture of modern music. Our post is called “Welcome to the land of loudness” and there is a mention of Allen Wagner’s TurnMeUp project….you can read or in depth analysis of the loudness wars at http://www.studiomanifesto.ca.
@Audiot
“Thin and weak” is actually the result of compression abuse. When the dynamic range is substantially reduced, you are left with something that is one dimensional in perceptible dynamics. In other words instead of a range of dynamics (quiet parts to louder parts) music gets reduced to one dynamic….loud.
If there is no quiet there can be no loud. When a recording has good dynamic range, the loud parts punch out of the mix. This leaves the listener with the choice of turning up the volume on their stereo and hearing beefy punchy music. Overly compressed material is artificially embedded loudness…it aims to be loud even when your volume is turned down. Which is exactly what a listener naturally does when assaulted with a wall of relentless loudness.
I have a theory which many people will not except. It is that the fatiguing, painful quality of audio compression/limiting abuse is a key reason for the decline of sales in music. Not to undermine the other reasons often discussed on this and other blogs. When music becomes painful to be subjected to, its no surprise that people will loose interest in it.
Good luck with TurnMeUp Allen. our studio, Euphonic Sound, will be contacting you shortly to find out how we can become more involved.
“If there is no quiet there can be no loud”
That in itself is a great philosophy to live by. I had no idea about this type of audio compression and that video was a true mind melter. It also makes sense when I think about how “lifeless” some modern recordings sound.
I’ve played in a heavy rock band for 8 years now and I realised the importance of that first sentence maybe 2/3 years ago, i.e “if there is no quiet there can be no loud” in the context of songwriting and live performing. I learnt the all important aspect of dynamics….that to have even two moments of pin drop followed by a huge wall of sound was always going to more effective than the wall of sound all the time…The same goes for tempo, slowing down adds more breadth to a song or piece of music..but I drift again….excellent piece as always Dubber!
There’s a certain degree of “horses for courses” about this issue – some tracks just suit a lot more compression and limiting than others.
While squashing the hell out of everything just for the sake of loudness is an ultimately futile exersize and I’m delighted that so many people are waking up to this, it’s equally worth remembering that some tracks sound great being brickwalled with no dynamics left in them.
It’s not a good policy to be too dogmatic and try to apply the same principle to every master you do – that is after all how we ended up with this problem.
Hey Andrew,
As a musician/recording engineer, I am in total agreement. Many people give the cd a bad rap, blaming the format on something that was completely driven by human desire, not the cd itself. CDs that were produced in the 80′s/early 90′s, especially reissues, were pretty much like listening to vinyl in the sense that the dynamic range wasn’t boosted like the cd’s of today.
When digital recording came along, not only did the frequency response expand, but the dynamic range did as well. Vinyl has a much more limited dynamic range compared to digital (cd), so with increased bandwith capabilities and more headroom, bands/mastering engineers/producers/labels got drunk on volume power.
And many people don’t see the big deal with loudness. I didn’t myself until recently. I took out some old cd’s from the early 90′s and listened, turning my stereo dial up to 5 rather than the usual 2 (where it used to be when I was a kid, ironically), and Wow! What an experience. I heard music again!
The dumbest thing about this whole matter is all based around club music, strictly club music. That’s pretty much the only place that it counts. The radio already compresses the hell out of the music, so it really just makes your music sound worse if you’ve already smashed the hell out of it. Any other genre that cares about the quality of their music should NOT be concerned with making it loud for the radio and all that BS.
It’s not even a complex issue either. All they have to do, is just not turn it up so loud in the mastering process. Simple fix, less effort.
Don’t blame the CD.
Blame the folks behind the wheel!
@darryl – one clarification: it’s the other way around. More dynamic range means a greater difference between quiet and loud. Dynamic range is reduced with greater compression.
I received my first CD player in 1988. In 1989, I had my first disappointing experience – and it was the opposite problem. The original CD release of Simple Minds’ New Gold Dream was way too quiet, and it just didn’t have any punch to it at all. A later remastered edition was infinitely superior.
But overall point here is that effects of any kind – particularly compression – should be added like you add salt to food. You can make a dish tastier by adding salt. But that doesn’t mean it gets tastier the more salt you add.
This discussion is getting better and better…
@James Pew: You are right of course, that’s exactly what I meant. Sorry, sometimes my English is just not clear enough. Certainly it’s the overly compressed songs that lack something or maybe better: lost something. But the problem for people like us is, that our music – if we’re not participating in the loudness race – sounds weak compared to the massive loudness of other songs, because it is just not that loud…. Not necessarily in the case of home-stereo-listening, I was more thinking about radio and dj-nights in clubs, events where people hear music that’s new to them. If the consumer already owns your CD and listens to it at home, the loudness war is irrelevant. IMO, it has only an effect when it’s about the first impression of something new. I hope it’s clearer now. If not, just believe me: we’re on the same side ;-)
And about your theory: I think this is NOT too far fetched and I’m absolutely willing to agree this is one major reason but there are at least 3-4 other ones as well… Come to think of it, this could be another interesting discussion – I guess it already was and is in countless blogs… oh well…
I thought I should mention The Bob Katz Honor Roll – http://www.digido.com/misc-content/honor-roll.html This is a list of recordings that mastering engineer Bob Katz considers to have good dynamic range. What’s interesting about this is that included on the honor roll is the debut record form Rage Against The Machine. This offers proof that heavy stuff can be mixed and mastered with punchy dynamics and still stand up against other recordings in the genre.
@Audiot
“But the problem for people like us is, that our music – if we’re not participating in the loudness race – sounds weak compared to the massive loudness of other songs, because it is just not that loud”
This is only the case if the volume control stays in the same place. If you increase the volume on the full dynamic range recording it ends up sounding bigger and punchier than the squashed recording. How to deal with this problem from a consumers point of view….well we need to start with awareness. This is why I’m a big fan of the TurnMeUp label included on recordings that break away from the loudness regime and attempt to give back the dynamic range to consumers.
On the other hand there is nothing more annoying than listening to a compilation or mix CD where the levels don’t stay consistent. I don’t like the idea of limiting circuits built into stereos (or other playback devices) – A more sophisticated approach would be to have stereos analyze the average loudness and make an automated volume adjustment prior to playing the song. Can someone please invent that?
Andrew,
Thanks for correcting me, came out wrong. I meant to say that the dynamic range is cut (compressed), and the overall volume level is boosted.
I just wanted to say that there are albums that have been completely ruined for me because of the loudness wars. I just wanted to second the Bob Katz reference from James Pew, Bob has alot of information about good mastering techniques and such. Another fear of mine is I’m seeing older albums being re-released that have been “Digitally Remastered”. I have never bought one of these re-mastered cd’s, largely out of fear that they have probably been squashed to death to fit in with modern records!
I certainly agree with this post – the stuff has got to breathe! But interestingly, the compression issue that has irritated me the most is the lack of compression in classical music. There’s nothing like uncompressed symphonic music on hi-fi speakers at home, so I really am not suggesting that should change. But in the car, which is where I have the most time to listen, Beethoven’s 9th is inaudible at parts and blaringly loud at others at without adjusting the volume.
I have taken to compressing some of these tracks myself so that I can enjoy them without constantly manning the volume knob (and risking getting blasted by a sudden crescendo). But this is very time consuming, especially if you’re talking about the 5 disc set of all of Beethoven’s symphonies, and especially if you take the time to do quality compression. I would love to see classical labels offer a compressed version of their songs for listening in less than ideal environments.
the problem i see with my own music is the thought that some kid randomising his ipod might hear one of my tunes, and because its not as loud as everything else decide they dont like it. so i feel that i have to keep up with the competition in the volume stakes.
i have a decent stereo that does not need to be cranked up to be loud and clear but a lot of people (the majority, i suspect) listen to music on cheap systems that are already distorting being driven harder than they should. surely if compression/limiting is used well then these people may turn their stereos down and hear the music a little cleaner.
there are two sides to this argument, i am not on either. besides which, i use plenty of distortion creatively so i dont have a great deal of dynamics to lose anyway
Totally agree. I have worked on a few recordings where the focus in the end, for the client has been to get it louder and prouder, and to my ears we’ve been getting into dangerous territory. Compression and limiting are dangerous processes that should be seriously contemplated and evaluated before getting slapped on mixes for the sake of loudness. I’ve been really hoping that the burbling up of this issue over the last few years would begin to gain momentum, and we’d see more and more artists’ mixes demonstrating restraint in the loudness war. Thanks for writing about this, Andrew!
Just a note that audio compression typically screws up data compression. Why? Because MP3 et al are *perceptual* audio codings: they eliminate what’s quiet, or encode it at lower quality relative to the louder parts. If you compress the dynamic range, things that *should* be too quiet to hear get retained and the codec runs out of bits to encode what you actually wanted to be heard.
I think Mobile Fidelity Sound Lab (mofi dot com) really cares about their re-issues – no compression whatsoever. They make it a point to get all the dynamic range in there. It really sounds amazing on a good system – and even on my ipod (in lossless format of course).
My CD album HereAfter was mixed and mastered by the old school genius John Wood (John Martyn, Nick Drake, Richard Thompson) ..who flatly refused to use any “program” at any stage of the process.
Instead we opted for a great Studer machine to mix on and some wonderful Valve outboard gear and great transparent speakers to master on .. and yes it breathes a lot more than stuff I’ve just had put through a Waves type program.
When I asked him why, he replied “if you want it louder .. turn it up”
Great discussion here. I went through this issue recently with my own indie release, when I had it mastered by Sterling Sound. These guys are one of the top mastering studios in the world and do all the popular rap and rock releases, which are as hot (compressed) as you can imagine. But I let the mastering engineer use his judgment instead of asking for a blaringly loud master, and it turned out much less compressed than other albums he’d done for major labels. The result IS that it’s not as loud overall as most major label releases. Does it sound better? I think so.
There’s something seductive about the huge levels of compression these days, where everything is in your face from the beginning, but after listening to the several songs that never change dynamically, it starts to sound pretty lifeless and fake.
It’s ironic that indie artists that no one cares are still free to have their masters done right, but that being on a major label or getting major airplay means having the life squeezed out of your dynamic range.
That was pretty funny reading the complaints about too much dynamic range in classical music. Manning the volume knob every time I want to listen to classical recordings drives me crazy!
James Pew is absolutely right in saying that the loudness war is a major cause of the decline in record sales. Obviously the industry thought, or assumed, it would actually boost sales, but in practice it’s having the OPPOSITE affect. It’s not only audiophiles who are being repulsed by it, but also millions and millions of regular music fans, who want to be able to listen to, focus on, appreciate and experience the music – and find themselves physically UNABLE to do so, because the non-musical square waves of these “bricked” CDs induce fatigue, pain and sometimes even nausea. Not only that, but what with the sonic-details in the music being lost through clipping and over-processing, the listeners are being short-changed/ripped-off, because they are literally NOT GETTING ALL THE MUSIC – what they are getting is a maimed, bastardised version of the music, with parts chopped off!
It is through alienating this LARGE group of regular music fans (ie. buyers/potential-buyers) that the industry has run itself into the ground. For proof, check out the following links I found on amazon.com and a music-trading site. These are music CUSTOMERS, ticked off by the loudness war, sending out messages to other customers, warning people NOT to waste their money on particular CD releases/remasters, or even to bother listening to them:
http://www.amazon.com/tag/victim%20of%20the%20loudness%20war/products
http://www.amazon.com/gp/richpub/listmania/fullview/R120N2FPI04EV4/ref=tag_tdp_rb_elst
http://www.amazon.com/review/product/B000002AP1/ref=dp_top_cm_cr_acr_pop_hist_1?%5Fencoding=UTF8&filterBy=addOneStar
http://www.amazon.co.uk/review/product/B0000251W3?filterBy=addOneStar
http://www.sharoma.com/trading/loudness.htm
Years before i did even know about the Loudness War, I listened alot to Red hot chili peppers “Californication”., I remember that i turned down the volume alot because my ears hurt during longer playtime. I was just a regular teenager and didnt know anything about HIFI and Dynamics. Like Chris wrote on the post before:
“It’s not only audiophiles who are being repulsed by it, but also millions and millions of regular music fans”,
Cant even enjoy a live Rock Concert anymore due to the “Loudness War Effect” Im telling you! Its everywhere, spreading like a virus!
I guess i have to use “Soft Clipping” for home listening and Earplugs for live concerts in the future.
I’ve posted a video based on a presentation I gave on this topic at a recent convention of the Audio Engineering Society: http://www.sfxmachine.com/docs/loudnesswar/ . The presentation discusses the audible effects of hypercompression and looks at the loudness war in terms of game theory. After presenting evidence questioning the idea that louder recordings sell better, it suggests some possible ways to de-escalate the loudness war.
As much as I agree with this post, I don’t agree.
The loudness war is quite stupid, I know, I have mixed and mastered many recordings, and trying to take a recording sound “up to par” with the competition is very difficult without touching the volume knob. It takes intense equalizing, separate instrument equalizing, exciting, compression, etc.
But, the loudness war is not as bad as you took it to be.
Dynamic compression is NOT THAT bad. It take the loud parts and trims the peaks with a limiter. (NOT CUTTING, AS IN THE VIDEO) When you do this, the quite parts become louder.
There are many advantages to participating in the loudness war when creating a mix.
When you push the volume on the limiter, (With the right limiter setting of course) makes the recording flat (as in frequency response). When you push them, like say, Metallica’s Death Magnetic album, you get a very flat, and lush frequency response across the spectrum. If you push the recording the wrong way, you get no dynamics, no punch, and tons of distortion, just like the album I mentioned above and in the video that was used as a example.
I find that tweaking a recording to be flat across the spectrum with a 4-band compressor and a limiter works like magic rather than a equalizer and exciter, like most people use. Plus, when used right, you don’t even need to equalize the separate instruments. In my opinion, the loudness war is as good as it is bad. Bad if done horribly wrong, like the AC/DC remasters and most everything else that comes from Sterling Sound.
I have mastered and mixed many CD’s using a very interesting technique. I recorded the instruments, leveled the volume, put them all together, and ran a limiter, and a 4-band compressor on top of it all. All the instruments share the space in the spectrum, rather than eqing them to have only dominate certain spots on the spectrum. (Some people say that this causes frequencies to collide, but with a limiter, that rids that problem) With the limiter, pushing the volume as loud as it will go, while making sure absolutely no distortion occurs in each band, and a perfect limiter setting creates a VERY flat response, and sounds great on any system, while keeping the punch and clearity of the mix. This is how a mix should be done if you want to participate in the loudness war. No distortion filled noise. Distortion is what makes your head hurt from listening, not the volume. Punchy dynamics are created from the 4-band compressor.
You can still have excellent dynamics and clarity in a mix while participating in the loudness war. In the video above, the guy simply cut the waves, which created insane amounts of distortion, along with many other errors in the example.
Thanks!
-Kris
@Kristoffer McElhaney:
But like I said before, you have skills if you can make a recording with very small limiting, sound better than a recording that participated in the loudness war. I know of very few that do that and you can tell sound better when compairing without touching the volume knob. I have been starting to create more of those mixes as I get more experienced.
I myself am a music arranger, and the loudness war makes it almost impossible to understand guitar parts!! Many times with rhythm guitars, I have to pay attention to the bass line, because the distortion is so high, and the complete head-room of all the frequencies don’t come through well. I’m also tired of getting headaches from listening to music!!