From subtle warmth to screaming walls of noise, distortion is one of the most versatile effects in the studio engineer's armoury.
The term 'distortion' describes any change made to a waveform: even a simple change to the amplitude is technically a form of distortion, as is EQ, which is, in essence, an amplitude change to specific frequencies. More commonly, though, we intend it to mean something that happens when audio passes through a non‑linear device, like a saturating tube amp or a clipping preamp. In both cases, the main function is to change the amplitude, but the distortion process introduces new harmonics that are musically related to the original signal in some way. Obvious distortion is something to be avoided if you want to make accurate recordings of acoustic instruments, but it has many creative uses in popular music that extend far beyond making the electric guitar sing the blues...
As we explored in our recent feature on analogue warmth (SOS February 2010), the harmonic content of a signal gives a sound its timbre (it's what makes a flute sound different from a clarinet). Harmonic distortion is the introduction of extra harmonics that are musically related to those already present, resulting in a change in timbre. Even‑order harmonic distortion tends to sound musically sympathetic, smooth, and bright in a constructive way. Many simple valve‑based circuits (including most using triodes) tend to generate mostly even‑order harmonic distortion. Odd‑order harmonic distortion (the predominant type produced by analogue tape, for example) tends to sound rough or harsh, gritty or edgy, and is often associated with added 'richness' and 'depth'.
The higher harmonic components created by analogue distortion tend to sound gritty and fairly unmusical — so rolling off the high end can produce more ear‑friendly results. In guitar amps this happens naturally, as their loudspeakers tend to roll off quite quickly above 4kHz. Without the speaker, the top end of a guitar through a fuzzbox sounds like a wasp in a paper cup!
No prizes, then, for guessing that one way to add very organic‑sounding distortion to a track is to feed the recorded sound out to a guitar amplifier and then mic it up, returning the miked signal to the mix. This technique is often used for 're-amping' guitar and bass parts that were initially recorded via a DI — but it can also produce interesting results on drums, synths, and even the voice.
Whether it's a plug‑in, or a hardware device such as a Line 6 Pod, an emulation of a guitar speaker can work just as well as a real speaker. Try using one in place of a low‑pass filter to tame high frequencies. As with re‑amping, you're not limited to guitar signals: try it with distorted synths, organs or electric pianos. In fact, this approach can be more flexible than re‑amping, as most modelling plug‑ins give you a choice of speaker cabinet types — so try them all and pick what sounds best to you for your part.
When recording, err on the side of caution: don't add more distortion than you will eventually need. You can always add more during mixing if you need to, but there are very few practical ways to reduce distortion in a recorded piece of audio. If you're uncertain, it pays to play safe and use a splitter or A/B box to record an undistorted, 'dry' version of the input signal alongside the distorted one.
Having said that, one way to reduce the apparent distortion of guitar sounds at the mix stage is to double them as precisely as possible with a clean‑sounding synth part. Even if you mix this synth low enough so that its synthetic nature isn't identifiable in its own right, you can still achieve what appears to be a noticeable increase in the guitar's note definition.
Try to maintain contrast. Unless you're producing death metal, it's probably a good idea to balance distorted sounds with clean ones. For example, if your mix includes two filthy-sounding guitars, they may stand out more if the drums and bass are kept relatively clean.
The general perception of distortion is that it makes audio sound less clear. However, if we confine the distortion to a specific part of the frequency spectrum, it can actually enhance the sense of clarity. The Aphex Aural Exciter works on this principle: it takes a high‑pass‑filtered version of the input, distorts it via a non‑linear amplification circuit, and then mixes the result back with the clean signal. This process synthesizes new upper harmonics that may never have been present in the original signal, creating an impression of brightness and detail. By contrast, EQ can only enhance what already exists in the signal in the first place.
In general, harmonic enhancers should be used sparingly, to avoid making the end result over‑bright or fatiguing, but they can be used more aggressively to brighten snare drums and other percussive sounds. Even if you're not using a harmonic enhancer, adding a distorted signal to the unprocessed original can yield musically interesting results, and this is easy to try by copying your original audio to two tracks and then adding a distortion device or plug‑in to only one of them.
One word of warning, though: the ear is easily fooled into liking louder and brighter sounds, particularly as it gets tired from too much listening — so always review aurally excited or otherwise 'enhanced' audio in the cold light of the next day, to see whether it really is an improvement and not just louder, brighter and nastier!
Tubes are often used in dedicated distortion processors, but don't expect a good tube mic to sound obviously distorted. When the classic capacitor studio mics were designed in the middle of the last century, valves or tubes were the only game in town, and the designers did their best to make them behave as linearly as possible. The remaining distortion artifacts that create a significant part of the tonal character of those mics are, therefore, quite subtle. These are the characteristics that we now tend to describe as warmth or smoothness. Modern recreations that deliberately add more tube distortion than the bare minimum generally sound less musical than correctly designed tube mics.
Different analogue gear reacts to being overdriven in different ways, so some experimentation is necessary. If you have enough audio I/O on your system, try routing tracks through external analogue gear or guitar pedals, as true analogue distortion almost always tends to sound more natural than its digital emulations.
Compare different bits of gear, and pick which one works best. Equipment with transformer‑coupled inputs or outputs can usually be persuaded to add some nice distortion character as the magnetic cores of the transformers saturate. But be aware that you'll need to introduce a lot of gain at the front end to provoke saturation or clipping, and will therefore require a lot of attenuation at the output to regain sensible signal levels!
If you're more into programming than playing acoustic or electric instruments, try taking something as simple as a sine wave and playing it through a number of different distortion pedals. Something like the Boss MT2 guitar pedal (pictured above), with its swept mid control, will add texture and can yield some very interesting sounds.
While heavy distortion — and clipping in particular — should generally be avoided (other than perhaps tube‑style distortion on guitars or super‑fat synth-bass sounds), it can be used creatively on short, bright sounds such as snare drums. The aggressive snare sounds on some rock records were created by deliberately overdriving the input stage of the analogue mixing console to provoke clipping. Equally, a little analogue distortion can save a limp‑sounding snare, filling out the decay portion of the sound.
Analogue tape adds subtle distortions, especially when driven close to its maximum level. Some mastering houses create an analogue feel for the track by copying it onto analogue tape, then converting it back to a digital format — something that's particularly popular on classic rock styles of music. Of course, mastering engineers use high-quality tape machines running at high speed to keep added noise to a minimum, but you can try to achieve the similar results with any analogue tape recorder, even a cassette. Though there are plug‑ins that aim to emulate the sound of analogue tape, the real thing often works best!
All forms of distortion make the original signal more harmonically complex. This means that where more obvious amounts of distortion are being added, it can be more difficult to give each sound its own space in the mix. Artificially band‑limiting a distorted signal, using both high‑ and low‑pass filters, can help it sit more comfortably in a mix, by preventing its spectrum from spreading over too wide an area.
If you're mixing together distorted and undistorted versions of the same track (by duplicating tracks, using distortion as a send effect, or re-amping), be aware that you need to check the phase relationship between the two signals. Otherwise, comb filtering can potentially suck the life out of the combined sound. A good method of dealing with this is to use a dedicated phase‑rotation device on one of the tracks to adjust the way the mixed sound meshes. Fortunately, these aren't difficult to find in software form these days: Betabugs' freeware Phasebug offers a simple solution, while Voxengo's PHA979 and the Universal Audio plug-in version of Little Labs' IBP give you more flexibility. Remember also that most (but not linear‑phase) equaliser plug‑ins also change the phase response of the audio that they process (ie. they shift the phase of some frequencies relative to others), so you'll need to re-check your phase match if you add or adjust any EQ.
One of the classic mixing applications of distortion is to increase the density of upper harmonics in bass sounds, so that they can still be clearly heard on small speakers. This has two advantages over trying to EQ to the same end: it adds harmonic information that's not present in the original signal, usually allowing a greater range of effect than EQ; and it affects different pitches in the part more evenly. A popular distortion for this application is provided by Tech 21's SansAmp pedals, which have received Sound On Sound name‑checks in this role from both Spike Stent (U2, Depeche Mode, Oasis, Björk) and Mark Endert (Savage Garden, Rihanna, Maroon 5).
One of the challenges of mixing hard‑edged rock music is that when the heavy guitars enter the equation, they mask all the presence frequencies of the lead vocals, leaving them sounding muddy — even if the raw vocal tracks are as clear as day. Some distortion on the vocals can make a big difference here, and many mix engineers add the necessary edge by overdriving compressors or mixing in dedicated distortion send effects.
While using distortion on drum loops can often help to create a much more characterful and aggressive sound, it can also break up the bass drum's low‑end attack. For this reason, it can make sense to split the drum loop into several frequency bands to separate out the low end, and apply less drive in the lower band. Craig Anderton's old Cubase Quadrafuzz plug‑in, Ohm Force's Ohmicide or MDA's freeware Bandisto will do this for you automatically, but you can also set things up from scratch by using complementary high‑ and low‑pass filters on duplicate drum‑loop tracks to split the different spectral regions for independent processing.
Differentiating the sound of your delay repeats from that of your dry signals is one good way to keep an upfront‑sounding mix, and distortion is one fruitful avenue to explore for this purpose. Indeed, the appeal of many of the old analogue delay devices can be traced partly to the subtle (or not so subtle!) distortion characteristics they added to delay repeats
As with any effect, think where you want to introduce distortion in the signal chain — particularly in relation to any other effects. In a guitar effects pedal setup, for example, you'll get a very different result putting distortion before a wah‑wah pedal than after; similarly, a delay pedal might be better placed in an effects loop, after preamp distortion, than in the chain before the preamp stage. Try different configurations and you'll become more familiar with how your effects interact with one another.
No discussion of distortion would be complete without a few words about digital distortion. Obviously, digital systems can be used to model analogue distortion, but you can also experiment with reducing the word length or sample rate of an audio file. Reducing the word length (bit depth) without using dither (a process usually referred to as 'bit crushing') creates quantisation distortions in the digital audio. Reducing the sample rate can bring about aliasing distortion at frequencies within the audible range.
Aliasing distortion has the interesting property that the distortion frequency components can appear at frequencies below the fundamentals of the source sounds (because they're related to the mathematical difference between source frequency and sample rate). That doesn't happen with most analogue distortion processes, which produce harmonic distortions above and based upon the fundamental frequency of the source. Of course, there's also the harsh sound of simple digital clipping: this is rarely desirable, though it can certainly add an edge to a snare drum!
Digital distortion might not generally sound as musically pleasing as analogue distortion, but it does have useful creative potential, nevertheless: it can be reminiscent of the lo‑fi sound heard in early video games, for example, or may be used to replicate the sound of 'vintage' digital gear such as early samplers. .