By Craig Anderton
If there’s one thing in the world a mastering engineer doesn’t want to see, it’s a mix that looks more like a sausage than audio (Fig. 1).
Fig. 1: Mixes like this are no fun to master, and they can’t be mastered to reach their full potential.
This is usually due to someone who straps a maximizer-type dynamics processor across the master mix bus for a “loud” sound, without realizing that it ties the mastering engineer’s hands (who likely has better tools for making audio loud anyway). However, lately I’ve been getting something more disturbing: mixes that look a lot like Fig. 1, but upon closer examination, have clipping issues.
Fig. 2: The peaks circled in red are clipped.
In Fig. 2, you can see that the waveforms are “flat-topped,” which causes clipping distortion. One reason this happens is because the mix engineer doesn’t realize that with digital, “going into the red” almost invariably generates clipping, so they don’t get too bothered when the overload light goes on. But another issue is personal taste: Some people like the sound of digital distortion, and figure a little clipping won’t hurt. I don’t agree, but hey, there’s no accounting for taste.
In either case when this file goes to the mastering engineer, remember that mastering puts a sort of magnifying glass up to the audio. Once digital distortion is “baked into” a mix, there’s almost nothing the mastering engineer can do to remove it. The end result is a sort of fuzzy, harsh quality that robs definition and causes ear fatigue. If you really want digital distortion, it’s better to do a mix without it, then tell the mastering engineer that you’d like the sound pushed somewhat into distortion. The mastering engineer will most likely hold his or her nose, but “the customer is always right,” and they’ll do their best to give you want you want.
Fig. 3: The peaks circled in red have been severely compressed, but still sound better than being clipped.
Fig. 3 also shows a mix that has virtually no headroom, but it’s due to excessive amounts of compression and limiting, not clipping. The waveform peaks aren’t flat-topped, but simply reduced in level. Although this file is still far from ideal from a mastering standpoint, and won’t let mastering reach its full potential, at least it’s better than clipping distortion.
The solution is simple: Don’t use any processors across the stereo bus (like maximizers, compressors, EQ, and the like—remember, the mastering engineer probably has better tools than you do for the task). Having effects on individual tracks is fine, of course; you just don’t want to apply effects to the entire mix.
Also, set levels so there’s plenty of headroom when mixing—I generally don’t let peaks go much above –10 to –6dB max in my mixes. Mastering can always make it loud, but it can’t get rid of distortion or most effects that you added to a mix.
Some engineers balk at leaving this much headroom, because they like to push the output, and say that a squashed effect is part of their “sound.” In that case, sure, throw a maximizer across the stereo out and mix away—but bypass it before exporting the final mix. Then include a note to the mastering engineer saying you want a really loud, squashed mix; they’ll add the maximizer during the mastering stage, which will lead to better results. If it doesn’t, then it’s easy to just take the mix and add more or less dynamics processing. If that dynamics processing is already part of the mix, there’s nothing that can be done to “undo” it.
When you’re mixing, get the mix and the balance right—but don’t mix too hot and introduce distortion, or you’ll just end up with a frustrated mastering engineer,. What’s worse, the recording will never be able to reach its full sonic potential.
Craig Anderton is Editor Emeritus of Harmony Central and Executive Editor of Electronic Musician magazine. He has played on, mixed, or produced over 20 major label releases (as well as mastered over a hundred tracks for various musicians), and written over a thousand articles for magazines like Guitar Player, Keyboard, Sound on Sound (UK), and Sound + Recording (Germany). He has also lectured on technology and the arts in 38 states, 10 countries, and three languages.