Audio Distortion

When it comes to video issues that can ruin a shot, it can sometimes feel like an endless minefield of problems: overexposure, underexposure, dirt on the sensor, a smudge on the lens, an incorrect shutter angle, the neutral density filter at the wrong setting, too much digital noise, rolling shutter issues; the list really does go on.

However, the number of audio issues, outside of external factors like a noisy tractor or the hum of an old fridge, is much slimmer. In fact, it really comes down to two main issues. One: the microphone is too far away, making the audio too quiet, and two: increasing the volume in the recording produces digital noise, or the gain is set too high, resulting in a distorted audio signal.

In this article, we’ll focus on the issues that arise with point two: audio distortion. 

What is audio distortion?

Quite simply, distortion occurs when a signal overloads. Often, this is caused by a combination of high gain and a microphone being too close to a loud source, which makes the signal too strong for the recording device or format to handle. 

Within the audio sphere, gain refers to how much the user wants to amplify the audio signal before it is recorded. We shouldn’t refer to gain as ‘volume’ because it technically isn’t. However, it works similarly. On a field recorder, you can turn the gain knob left or right to decrease or increase the amplification of the audio signal. However, unlike volume, gain affects the strength of the input signal before it is processed or recorded, whereas the volume quietens or loudens the already mixed audio. 

Also, somewhat confusingly, while 0 on for volume would indicate no sound or “mute” for gain, 0 is way too loud. You want to capture audio at -12 dB and -6 dB, where the gain shows clear peaks and valleys that display the dynamics of the sound. But when the gain is set too high, say 0 dB, the audio signal can push beyond what both the device and format can capture. 

As noted in our pre-production checklist, we referenced a term called dynamic range. And this refers to how much detail can be captured between the brightest highlights and the darkest shadows. Well, audio has a similar element with a dynamic range between the quietest and loudest sounds. When you push the amplification too high, and the signal moves beyond that range, you will receive…yep, distortion. 

Unlike video issues such as digital noise, which can be rescued with noise reduction software, or out-of-focus shots, which can sometimes be salvaged with sharpening techniques, distorted audio is beyond saving.

Once an audio file is clipped and distorted, it cannot be fully recovered, and of course, we never want to be in that position. The keyword here is clipped. 

Understanding clipping 

There are different types of audio distortion (some put to creative use, which we will talk about later on), but the most common audio distortion issue related to video is clipping. Clipping is the audio equivalent of overexposure in video; just as turning up camera settings too high results in blown-out highlights, clipping means the audio is pushed past its limits, losing all clarity. 

Different audio formats can handle different amplitude levels before distortion occurs. Some offer greater headroom to prevent unwanted clipping. However, these formats usually appear in more expensive audio recording tools. 

Visually, clipped audio appears as a solid block of waveforms rather than having natural peaks and valleys. 

In the example above, track 1 shows a normal audio waveform with natural peaks, while track 2 contains a distorted file with clipped peaks. And audibly, instead of producing clear sound, it often creates a muffled yet painfully loud, walkie-talkie-like distortion. 

As noted earlier, with clipped audio, you can’t simply lower the volume of that clip; that information has been destroyed. 

So, how do we stop this from happening?

Preventing audio distortion 

Audio distortion can result from improper level settings in editing, but it’s primarily a pre-production issue, and we know that the main causes are improper mic placement or gain levels set too high. With this, there are two things to look for. 

While a boom operator should position the mic as close as possible for a cleaner signal, loud sounds, such as an actor screaming or a car engine revving, will require more distance. Finding the optimal mic distance beforehand prevents clipping.

Image via Unsplash

Conversely, the gain levels must be adjusted appropriately if the microphone position cannot be moved. If a dialogue scene has a moment where the voices rise, a signal increase will result if the gain is too high, causing peaks to clip. Adjusting gain in real-time for louder sections is key here. If you’re unsure how the script will play out, it’s always advisable to be part of the blocking rehearsal to see how the scene plays out. 

Image via Lewis McGregor

Depending on your device, you can also activate a limiter. A limiter is a safety net that will stop the loudest parts of the signal from clipping. But in this instance, it’s always advisable to acknowledge that the limiter is saving you and not something to be relied on, as other issues may arise. 

In following these tips, you should be in a position where your captured audio is not distorted. However, even then, you can still distort audio in post-production by incorrectly setting the levels of the audio clips. 

In practice, keep these audio elements within the following ranges:

• Dialogue: -15dB to -12dB
• SFX: -20dB to -8dB
• Music: -22dB to -15dB

Be sure to check out our extensive guide on setting the correct audio levels.

Fixing audio distortion

Is there a way to recover this level of distortion?

Some specialist audio tools, such as Adobe Audition’s DeClipper, can attempt to repair clipped audio. A de-clipper is essentially an audio effect that reconstructs missing data in clipped sections by analyzing the surrounding audio and filling in the gaps.

It’s not perfect, but a declipper can improve the original distorted audio if re-recording isn't possible. You can see the declipper in action in this quick two-minute video from YouTube audio expert Mike Russell.

Of course, this is a case-by-case tool. Much like blown-out highlights in video, once the audio is clipped beyond recognition, the original data is permanently lost. If your recording sounds like a microphone placed directly next to a loudspeaker, no amount of post-processing will fully restore it.

Creative use

Okay, while audio distortion is inherently harmful to captured dialogue or clean recordings, intentionally distorting an adequately recorded signal is a time-tested technique for creating unique and powerful sounds.

In the world of music production, nearly every rock track features an electric guitar that has been run through some form of distortion, whether it’s an overdrive pedal, a fuzz pedal, or a digital interface preset. In fact, distortion is a defining characteristic of rock and metal music; without it, these genres simply wouldn’t exist in the way we know them today.

But what about sound effects and dialogue? Well, there are different types of distortion, each defined by the algorithm used to create its effect. Saturation, for example, is a form of distortion but differs in the harmonic content added. Regular distortion adds random frequencies and noise, creating a harsh effect that fills the spectrum.

Saturation is more organized, and the extra harmonies have some relation to the original sound, resulting in a fuller sound. This effect can be used on vocals to make them sound thicker and stand out, but also on any sound that needs a boost.

From Marvel’s “Avengers: Age of Ultron,” Ultron is a good example of saturated distortion. 

Another type of distortion is a crusher. A bit crusher reduces the sample rate or resolution, degrading the sound quality. This can create an 8-bit vibe like an old-school Mario game.

When distortion is used as a tool, it’s often associated with harsher effects, like synth bass in the cyberpunk genre. The noisy artifacts can symbolize rebellion and danger, and the hard-hitting synth bass in cyberpunk or dark synth music adds an underground feel. In these instances, it’s worth noting that audio distortion is not always an issue; it exists on a spectrum. 

However, two things to note about these creative applications are that you need to ensure the audio has been captured correctly. The best use of distortion is when we convert a clean signal into a distorted one. Second, you will need advanced audio editing software to implement these effects.

Final thoughts 

So, what’s the gist of it? Distortion is a simple issue; you just need to control mic placement and gain levels. Do that, and you’ll capture clean audio. If not, you’ll get distortion, and there’s no fixing it.