In a previous article, we wound up speaking to the use of limiting on the master bus—how it’s used, specifically, to maintain desired levels without going into the realm of clipping. All this begs the question:
What is clipping in audio, anyway?
Audio clipping is what happens when you push the gain of a signal past the capabilities of the gear handling that signal. Clipping can happen in digital and analog realms alike, but digital clipping does sound different from analog clipping.
Analog clipping happens when you push a signal higher than the peak voltage of a given piece of gear can handle. This can be any piece of signal processing—say a compressor or an equalizer. Digital clipping occurs when you push a signal past 0 dBFS in any digital medium with fixed-point processing, such as the typical A/D or D/A converter. In these cases, 0 dBFS is the absolute highest value your computer can handle above this level information is discarded resulting in slicing off the top of the waveform.
The audible effect of clipping in both cases is distortion, but the nature of the distortion is different depending on the process. Analog clipping adds complex harmonic distortions that sound different dependent on the analog processor. In other words, distorting tape will sound different from distorting tubes, solid-state amplifiers, and so on.
On digital systems, however, there is one kind of clipping distortion: a squared off sound wave as the signal hits the digital ceiling. It looks quite like a square wave (if you’re into synthesis, you’ll recognize the look), and indeed, it has the hard, gritty characteristics of said wave. Digital clipping does not have the benefit of being impacted by a lovely piece of analog gear, where it will take on the unique characteristics of that circuitry (though some circuits try to mimic the distorted quality of analog, these are often approximations of pleasurable distortions more familiar to us throughout recorded history. So, when we speak of digital clipping here, we refer not to digital recreations of analog clipping, but to the squared-off sound described above).
It all boils down to one immutable fact: clipping adds distortion to the signal, and this distortion is by and large unwarranted.
…But what if it isn’t?
Audio clipping as a creative tool
Despite being labeled technically as a “fault” or an “error,” the sound of clipping can be desirable for certain styles of music and specific musical instruments. Thus, in creative pursuits, clipping is worth investigating.
Let’s take the drums, for instance. Analog clipping on a drum track can add a rounded, warm, hyper-saturated distortion that might lend itself well to hip hop, both modern and vintage. Take a drum loop through a digital clipping process, however, and you might be on your way to some of EDM’s more gritty, grimey sub-genres. Given the right context, both sounds can be acceptable in a production or a mix.
Clipping as an error
Sometimes, as an engineer, an audio file will come to you with audible clipping that you don’t want. You’ll know these by the distorted quality you wish to remove. In my practice, this is particularly odious in post-production scenarios, where an actor clips a microphone by yelling suddenly.
I do my best to mitigate clipping with tools like RX 6’s De-clip module, De-crackle, and Deconstruct to fix these issues. Often times it’s quite a simple process—hitting the “suggest” button on De-clip and letting the algorithm handle the best way to mitigate distortion. In fact, this very process came in handy while de-clipping Eugene Mirman’s vocal on an episode of Startalk Live that I mixed last year.
Sometimes I’ve noticed that De-crackle works better on material that’s distorting, but not necessary clipping. For a more in-depth look on how to use these modules, check out this link.
Clipping in mastering
Yes, some mastering engineers do employ audio clipping—either distorting a piece of analog gear for harmonic saturation, or actually clipping the A/D converter itself. Mostly this is done to achieve the levels clients expect, though some engineers do pursue clipping techniques for a specific sound.
Still, I wouldn’t jump into this process right away. Do your homework on the sound of clipping before seeing if it’s right for your mastering process—and still limit afterwards so as not to create distortions that you didn’t intend (at least, that’s what I would do).
Indeed, if you find that the sound you’re looking for isn’t coming naturally with basic limiting, try some creative limiting possibilities. Here are a couple of ideas:
Try different limiting algorithms
As we noted in a previous article, many limiters offer different limiting algorithm—some of them even providing clipping or soft clipping; Ozone 8 is no exception. In O8, there’s no fewer than 10 different options for limiting styles, with each one sounding a bit different than the others (indeed, there’s even a “Clipping” algorithm in IRC III). Spend some time listening to what these algorithms do to the music when driven a bit harder than you would in the real world (this is for practice, after all), and you’ll begin to hear which algorithm might suit you best. For a more thorough dive, try investigating algorithms using the delta test I mentioned previously.