Dynamic Range Compression: Are The Loudness Wars Over?
, November 8th, 2013 08:44
Nick Southall speaks to American mastering engineer Bob Katz and Hookworms about how, with the launch of iTunes Radio, the tide is finally turning against full dynamic range compression
The other day I bought a copy of Nathan Fake’s rather lovely Drowning In A Sea Of Love album. Nothing unusual about that, except that I’d first bought a copy in 2006 when it came out, and then sold it again, because it was too loud as far as I was concerned.
Seven years ago I was at my most militantly anti-loudness, and I culled plenty of records from my collection that I deemed to be offensive to my dynamic sensibilities, denying myself the potential pleasure of listening to plenty of really, really good music that just happened to be loud. I’m more pragmatic these days, more concerned with how things are mixed and the integrity of frequency ranges and timbres than with pure dynamic range, or lack thereof, but my tastes have definitely changed over recent years: I listen to practically zero mainstream rock or pop these days unless it’s exceptional or dynamic or both.
But it seems as though my struggles to reconcile my sonic and aesthetic tastes might be about to close, because, according to renowned American mastering engineer Bob Katz, the Loudness War - as the quest to be the most attention-grabbing music on the radio (or anywhere else) has become known - might be about to end.
A couple of weeks ago Bob released an announcement on his website stating that the war had been won, and that we – listeners, fans, musicians, engineers etc who give a damn about sound – were the victors. And all because of Apple.
It comes down, in Bob Katz’s thesis, to iTunes Radio, a streaming service which has yet to reach UK shores but which has been building an impressive 20 million plus user base in the US over the last few months.
If you’ve ever played back music using an Apple device – iPod, iPhone, iPad, iEtc – or else out of any computer using iTunes, then you might have come across ‘Sound Check’, a little tickbox in the preferences, which promises to “automatically adjust song playback volume to the same level”. It comes ticked by default on new iGadgets, and makes sure the transition from an old Byrds song to something super-modern by Battles isn’t too jarring. You can untick it, and your old, quietly-mastered music will sound tiny next to the obese monoliths of unsubtle sound we’ve become used to, but most people probably don’t.
With iTunes Radio, though, Sound Check isn’t a tickbox; it’s a standard algorithm or whatchamacallit or gizmo or thingamajig or process that all the audio is run through on its way from Apple’s gargantuan servers to whatever headphones or speakers you’re listening on, whether you like it or not. And Bob Katz likes it a lot, especially after spending several hours carrying our loudness measurements, and finding that current super-loud pop tracks were being attenuated by approximately seven dB in order to make their loudness equal to more sensitively-mastered material. In other words, rather than turning quiet songs up, Sound Check turns loud songs down.
I emailed Bob and asked him to tell me more.
Bob Katz: Everyone should turn on Sound Check in your iTunes players. Right now. Leave it there at least for a month and tell us whether you like it better. First of all that will illustrate what iTunes Radio will sound like before it comes to the UK. I also think it does a nice job of dealing with the differences between hot and quiet CDs, the ‘Pink versus Pink Floyd’ issue.
So, what does Sound Check have to do with eliminating the loudness war? First of all, there's the obvious; it brings down the overcompressed, super-loud hot stuff which is common today to allow the more open-sounding stuff from yesterday a chance to sit on an even footing. I switched between a 70s station and a ‘pure pop’ station on iTunes Radio the other day and the transition was seamless.
Does this mean I’m going to have to re-buy albums from the last 10 years as they get remastered to be quieter?
BK: How many different copies of The Wall have I purchased over the years? Or Thriller. We get saturated by this business of remastering! And many remasters sound worse than the originals.
I think there will be some remastering, but comparatively little. I just see a gradual opening up of sound quality with new releases, for the better, as soon as producers become aware that iTunes Radio is ‘the great leveller’. It took us 33 years after the invention of the Compact Disc to arrive at a very fucked up point. (I use that word for emphasis). I think it will only take us five years to return back to sanity, now that iTunes Radio points the way of the future.
Why are Apple controlling the user experience by doing this?
BK: Apple did this because it’s a better listener experience, and because they had to. Radio has always needed to do some form of regulation from song to song. Sound Check is just a better way to accomplish this. Sound Check sounds much better than any smashed version of FM radio I've ever heard. It also reproduces the original material more authentically, without altering the dynamics or the impact. Finally artists will hear their record on the radio as they intended it to be heard.
So I look at Apple's wish to ‘control the user experience’ from the positive side. It's going to be a better experience for the consumer. Put on Pink Floyd followed by Pink or Green Day. The latter will blow up your loudspeakers. Commercial radio has been trying to control the user experience, with disastrous sounding results, but at least it's consistent loudness-wise. So why can't Apple do it better?
We could pick nits and talk about audiophile quality versus 256 kbps AAC quality, but the fact is all those differences pale compared to the smashing and squashing that traditional radio does to an artist's work. European radio is already moving in the same direction - just levelling, no processing. UK radio may also be moving in that direction. The U.S. will be a holdout, but I predict in five years when internet radio and iTunes Radio starts killing terrestrial radio in sound quality and number of listeners, that terrestrial U.S. radio will come around.
What's the difference between this and 'Mastered for iTunes', and has 'Mastered for iTunes' been a good or bad thing, generally?
BK: Mastered for iTunes is an entirely different initiative. It’s simply an initiative to get higher quality conversions to AAC than ever before, by supplying 24-bit masters to Apple for encoding. This actually has raised the quality of the AAC releases so it is a good thing, generally. 256 kbps AAC sounds as good as or better than 320 kbps mp3, by the way. And 320 kbps mp3 is just a bit below CD quality. However, when the 256 kbps are made from the 24 bit master, everything turns around, and many AAC masters sound better than the CD because the AAC preserves more of the depth and space from the 24-bit original than the 16-bit CD. I still prefer the sound of the 24-bit master of course, and the 2496 master sounds even better than that. But for commercial streaming radio, iTunes Radio already stands out above the rest in sound quality to my ears.
Does any of this make much difference if people are going to be listening in less-than optimum settings, at their desk in an open-plan office, in a car, via headphones on a train, etc?
BK: Of course the worse the playback environment, the worse the sound issues. Soft passages in noisy places is not a problem currently addressed by iTunes Radio or Sound Check. Some of the other streaming services employ limiting and compression, and I do not subscribe to lowering sound quality to the lowest common denominator during transmission. The solution has to happen in the player: in the car, while jogging, in the gym, etc. And it will happen. In a noisy place I would turn on compression in the player so you don't have to ride the volume control during soft passages. Apple has the smarts to, for example, use the microphone that's built into the iPhone as a sensor for outside noise.
Is there a danger that the Loudness War now just manifests itself as a battle between streaming and online radio services?
BK: Yes, I do have that concern. But I believe that Apple is the 200lb gorilla and hopefully its standard will win. Let's hope I'm right. I don't want to see a Loudness War between internet stations either.
It’s a fallacy to judge good sound by loudness range. Rudolph Ortner, an Austrian Sound engineer, wrote a Masters Thesis called The Evolution of Loud. For it he measured the sound characteristics of over 10,000 chart-topping songs from 1950 through 2011. This is a big, statistically-valid survey because of the large number of samples. He demonstrated that loudness range has not decreased one bit over all those years! In fact, in the last couple of years of his survey, loudness range slightly increased by about a dB. So engineers were taking some advantage of soft passages and making more soft passages.
The real smoking gun of sound quality loss during the loudness race is NOT loudness range, but rather "PLR" (Peak to Loudness ratio). Let me explain in layman's terms. Take a recording you know well, like a good Steely Dan [song]. Steely Dan is known for their sound quality and the clarity of their drums and percussion. However, Steely Dan records are very "radio ready" as these records typically have very low loudness range. So, what distinguishes a Steely Dan from, say, a Miley Cyrus ‘Wrecking Ball’? What distinguishes them is that they are not compressed from the top down. The transients (drums, percussion, cymbals) are not deteriorated.
One thing that a record like Tool's Aenima has in common with Miley Cyrus is that Aenima is highly compressed from the top down. Percussion instruments have lost all their snap and punch. But that's the style. And remarkably, Aenima has far more loudness range than Steely Dan. One could argue (fairly) that the producers of the Tool record chose to squash the transients as a stylistic choice. I'm not going to deny that.
But one could also argue that this record, or, more appropriately, Miley Cyrus' ‘Wrecking Ball’, are real victims of the loudness war because they have absolutely no transients left, no life, due to the reduction of transients and the introduction of severe distortion in their place. That's the biggest sign of the loudness war, compression of transients and distortion. You will still find loudness range in many of these hypercompressed pieces, therefore loudness range is obviously not a measure of sound quality.
And PLR (peak to loudness ratio) is how Sound Check is going to improve sound quality. Artists who wish to use less compression and obtain better and clearer sound and more impact will no longer be afraid to produce masters that are lower than the rest. Because they will know that iTunes will bring the other songs down. It's already happening. I've already convinced some of my artists and clients that they don't have to compress as much as before in this new world of loudness-normalized media. This is very good news. I've been waiting for it for 33 years.
[I also spoke to MJ from Hookworms, and asked what he thinks vis a vis the Loudness War in 2013.]
MJ: We cut the last Hookworms 7" with Alex Wharton at Abbey Road almost entirely flat. There was no compression and if I remember right, a half dB shelf of EQ for some presence on top of the original mix I took there and that was it. That song was playlisted on 6music for over a month and anytime it came on and I happened to be listening I didn't think, "Hell, that sounds weak." The BBC smash stuff through limiters so hard that it ends up being the open and dynamic sounding track that retains an excitement. I was grateful to see that first hand to reinforce my opinions that loudness is a race to the bottom.
In a way I think that, ironically, to have a quiet or dynamic record now is to engage with the listener's attention far more than any homogenised hyper-compressed pop music, if only because it stands out and they have to reach for their own volume control. And wasn't that the aim in the first place?