Maximum Mona. Whether we’re talking painting or audio files, consuming every iota of available space is rarely attractive, as symbolized by my Limited Mona Lisa portrait (Image 1). We need the audio equivalent of what visual artists call “negative space”—unpopulated portions of the piece whose emptiness/quietness makes the busy/loud parts feel exciting. But countless rock and pop recordings of the last 25 years are more like Maximum Mona than Leonardo’s original. Image 2 shows what happens to a waveform when it receives such treatment. There’s more total energy, but at the expense of drama and variety.
Image 2
Few audio pros dig this trend, so why do they do it? Partly it’s because our ears like loudness, at least to a degree. Our hearing isn’t linear. When music gets quiet, we perceive less low end relative to everything else. (For a scientific explanation of this, google “Fletcher-Munson curve.”)
But the abuse of loudness is about competition and fear. This attitude has long been prevalent in radio, where loudness is a genuine advantage as listeners scroll through the dial. But in the early ’90s, producers and mastering engineers launched a volume arms race. Even though many technicians knew this is a bad idea, they feared that their mixes would sound wimpy alongside their competitors’ work. Thus began what the recording industry calls “the loudness wars.”
Extra-Limp Bizkit. I first realized how extreme this was in 2001, when I was doing some in-house work at Interscope Records. One day in the studio I encountered mastered files for a Limp Bizkit track. The waveforms in Pro Tools had no peaks or valleys whatsoever. It was just a fat sausage of nasty-sounding loudness.
Here’s a quick guide to loudness abuse. In Clip 1, you hear the original mix, already with a fair amount of compression. In Clip 2 through Clip 11, I edge up the limiter setting, starting with 2 dB’s worth all the way up to a massive 20 dB reduction.
I’ve used one of the limiter plug-ins from Ozone 8, iZotope’s suite of mastering tools. It’s damn good software. It’s amazing how decent the morbidly over-limited tracks sound! But even though I monitored at low volume, a headache descended at around the time I crossed the -10 dB threshold. (Try toggling between Clip 1 and Clip 11. The contrast isn’t subtle.)
Loudness war correspondent. There’s a long chapter on this topic in Greg Milner’s Perfecting Sound Forever, a brilliant book on recording technology. Milner covers the entire history of recorded sound, with an emphasis on the ways evolving technology changes music, rather than the other way around. (Historians would call this viewpoint “technological determinism.”) He dissects such notoriously over-limited albums as Red Hot Chili Peppers’ Californication, Rush’s Vapor Trails, and Oasis’s (What’s the Story) Morning Glory?
Since Milner’s book was written, there’s been some much-needed backlash to the loudness wars. Still, at the recent NAMM show in Anaheim, one of mainstream rock’s most successful mixing engineers displayed some of his tracks on the big screen. They all looked like that 2001 Limp Bizkit track: total sausage party.
Only you can determine the amount of limiting that brings your mix to life and the amount that squashes it dead. As you experiment, consider this: If you want something to sound really loud, juxtapose maximum and minimum levels. When you hear an oldies jukebox, Joan Jett’s “I Love Rock ’n’ Roll” sounds louder than anything else because there’s silence between every big beat. Much hip-hop and EDM works the same way. The artful use of negative space can make your music loud and dramatic.
Round off, chop off, or expand? Limiting is simply an extreme form of compression. If compression rounds off the loudest portion of a recording, limiting amputates it. In other words, limiting is high-ratio compression. (Yes, you can usually use a hardware or software compressor as a limiter.) Most DAWs include both compression and limiting plug-ins, and there are scads of third-party options, often modeled after popular analog devices.
Maximizing has no precise definition. Some say it’s merely a sexy marketing term for limiting. No single company owns the term, and many use it. But generally speaking, it implies a “smarter,” more musically sensitive version of limiting.
Venus de Maximizer. Consider the three variations of Venus de Milo (Image 1). The first is an unaltered photo depicting the statue’s actual proportions—think of that as a recording with no limiting applied. In this Greek statue metaphor, size equals volume. So if Venus were a recording, we’d size her up so that the top of her head just reaches maximum loudness before distortion. In digital, that’s 0 dB, represented by the red bar. (That process is called normalization.)
The second image depicts Venus through a limiter. If we scrunch down the top of her figure (representing the loudest sections of the recording), we can increase her size/volume. But ouch! She’s grotesque and hard to look at, just as a crudely limited recording sounds grotesque and hard to listen to. It’s volume at the cost of naturalism.
The third image is the maximized Venus. Here, we’re not simply compacting the top portion—the compression is distributed across the figure. Yes, this distorts the natural proportions. But she doesn’t appear as disfigured as in the second image. If anything, she’s closer to average human proportions. (That is, with a body height 7.5 times as great as the height of the head alone.) Sometimes artfully applied maximizing sounds more natural than natural.
Pump it up. Okay, enough with the sculpture metaphor. Let’s hear what these processes might do to a recording. Clip 1 is a brief solo guitar recording. I’d applied some compression in the mix, but added no limiting after the fact.
Clip 2 features a limiter setting of -2 dB. That means only the loudest note attacks have been squished. It’s a fairly subtle setting that sounds reasonably naturalistic.
Clip 3 features the same limiter plug-in at a far more aggressive -15 dB setting. It’s an unpleasant effect, with ugly popping on the note attacks and weird distortion.
For the next two clips, I’ve substituted the Maximizer plug-in from Ozone 8, iZotope’s suite of mastering plug-ins. Clip 4 features a maximizer setting of -3 dB.
It’s louder and punchier than Clip 1, but it still sounds like a real guitar. But the overly heavy -10 dB setting in Clip 5 introduces all sorts of nasty artifacts.
Check out these screenshots from the Ozone Maximizer plug-in in Image 2.
The upper screenshot shows the light -3 dB setting. The portion of the waveform in blue depicts the degree of amplitude reduction. Only the loudest note attacks are affected, and those only slightly. Meanwhile, the heavy setting alters the audio to an extreme degree.
Strike up the band. The next five clips follow the same procedures, but with a full-band track. Clip 6 has no limiting.
Clip 7 uses light limiting, while Clip 8 is heavily limited.
There’s a light maximizer setting in Clip 9.
And an extreme one in Clip 10.
Use with moderation. The moral is pretty obvious: A little limiting or maximizing can make a track feel more engaging and energetic, but too much makes it flat and ugly. But even if you know that, it’s easy to lose perspective while applying these techniques—especially it you’re performing homespun mastering right after a long mixing session. Plus, your ego is on the line: Of course you want your music to be as loud and proud as possible! It’s often wise to make several mixes, applying settings of varying strengths, and revisit them later with fresh ears. It might also be a good idea to compare your work to commercial recordings.
Or maybe that’s not such a good idea. When listeners complain about how crappy many modern recordings sound, it often has to do with excessive limiting. And that’s a topic we’ll take up very soon.
