Most producers are familiar with the most common audio processors: EQs, compressors, reverb, and delay. These can be (and often are) used in every project, being the main tools in a producer’s arsenal.
In a mixing context, these are the most important plug-ins you’ll be dealing with. In a creative or sound design focus, however, we can take advantage of more specialized processors to create interesting sounds.
While there are lots of unique-sounding audio effects for sound design, there are a few you’re likely to come across time and time again: chorus, flangers, and phasers. These processors are all related, and are often confused with each other, but they have notable differences that are worth pointing out.
It’s possible to use these processors without knowing exactly how they work, but being more informed on what’s actually happening will allow you to use them more intentionally and more effectively.
But before we get into the effects themselves, let’s brush up on a bit of audio physics:
The basics of phase
As you may already know, audio comprises sound waves. These waves are the result of vibrating (moving) air molecules. When something emits a sound (your vocal chords, a passing train, a loudspeaker, etc.), air molecules are moved, resulting in changes of air pressure in space and time—sound waves.
Our eardrums, in turn, vibrate according to these waves. With this information, our brains interpret the sound waves entering our ears as audio.
As with any other type of wave, audio waves have points of maximum positive amplitude and maximum negative amplitude. These are called crests and troughs.
When two waves interact with each other, the result depends on how the waves overlap. When crests overlap with crests (or troughs overlap with troughs), constructive interference occurs. The amplitudes of the sound waves will sum, resulting in a louder signal at the point of overlap.
When a crest and a trough overlap, the waves will cancel each other out. This is known as destructive interference.
When two waves of the same frequency interact in a way that the crests and troughs align perfectly, we say that the two waves are “in phase” with each other. If pure destructive interference occurs, where the crests of one wave align perfectly with the troughs of the other, we say the two waves are “180° out of phase” with each other.
In everyday life, it would be somewhat rare for two sound waves to align so perfectly with each other. Room acoustics can create phasing due to sound waves reflecting off surfaces and interacting with each other, but this requires specific room dimensions and audio content. Additionally, as the walls of a room are fixed, it is difficult to adjust the phasing that occurs here.
However, in the world of music production, we have the ability to manually and intentionally create exact duplicates of audio signals. Therefore, we can experiment with the concept of phase to produce the effects found in chorus, flangers, and phasers.
When signal enters a chorus plug-in, a copy (or several copies) of the signal will be made. This signal is exactly the same as the original signal in amplitude, frequency, and phase position.
The copied signal is then slightly delayed. This creates a difference in the phase of the two signals, the original and the copy.
Next, the amount of delay time is modulated with an LFO. This causes gradual changes in the frequency (wavelength) of the resulting signal. Like the “Doppler Effect,” this will have the effect of varying the copy’s pitch. This delayed, modulated copy is then mixed with the original signal.
The result is that the copy (or copies, depending on the chorus unit) slightly varies from the original in pitch and time. This is used to imitate the effect of multiple instruments or singers, who would never be perfectly aligned with each other in the real world.
Note that, because the pitch is modulated in the copy, it and the original signal are rarely at the exact same frequencies. This minimizes constructive and deconstructive interference, while still taking using phase shifting techniques.
In the chorus plug-in below, two copies of the original signal are made. Each copy (Delay 1 and Delay 2) have adjustable delay times. The plug-in also includes separate filtering of the first delay and the choice to modulate the second delay’s delay time, leave it fixed, or turn the delay itself off.
In the Modulation section, we can see the LFO parameters. We can set the amount of modulation (in ms) of the delay time and the rate of modulation (i.e. the LFO’s amplitude and rate). Note that this plug-in has a feedback parameter, though not all chorus plug-ins will.
In the example below, the first and third guitar arpeggios are dry. The second and fourth have been processed with chorus.
Here are a couple classic examples of the chorus effect in recordings:
Dire Straits - “Sultans of Swing” (Mark Knopfler’s guitar):
The Police - “Walking on the Moon” (guitar in intro):
Of the three processors we’re discussing today, chorus is generally the most gentle. Chorus is best used to wash a sound out and make it more ambient.
The effect can be overdone for some highly characteristic sounds, but doing so can cause them to lose presence in the mix. Therefore, chorus can be a great addition to supporting layers.
Flangers are similar to chorus effects, both in their design and sonic properties. Like in chorus, flangers create a copy of the original signal and delay it. Flangers require much shorter delay times.
An LFO is then applied to modulate the delay time. Like chorus, we now have a copy of the original signal with a modulated delay time. This copy is then mixed with the original signal.
Because the copies of audio are identical, interference will begin to occur. Resonances will be created at some frequencies, but more noticeably, a series of notches (notch filters) will also be created across the frequency spectrum. Due to its shape, this is called a comb filter.
This type of filtering occurs in both flangers and chorus. Because flangers have shorter delay times, notches will be created in mostly high frequencies. Chorus units will create notches in lower frequencies due to their longer delay times.
Because the notches in flangers are generated by frequency relationships within the signal itself, they will occur at harmonically-related intervals (based on the frequency content of the original signal).
By using an LFO to modulate the copy’s delay time, these points of interference will change. The notches on the comb filter will appear to “move,” resulting in a searing resonance that sweeps to the rate of the LFO.
Flangers also make great use of feedback, sending the output back to the input to receive more processing. This accentuates the notches and resonances, resulting in the harsh, metallic timbre characteristic of flangers.
In the flanger plug-in below, we can see that the delay time between the original signal and the copy is set below the XY pad. This is where it’s also possible to increase the feedback.
The LFO’s rate and the “amount” that it modulates the delay time can be set on the right side of the plug-in, with various waveshapes available for the LFO. This plug-in offers an additional feature, as the LFO for the left and right channel can be separated and at different phases.
In the example below, the first and third guitar arpeggios are dry. The second and fourth have been processed with a flanger.
Here are a couple classic examples of the flanger effect in recordings:
Van Halen - “Unchained” (especially in the intro guitar):
Rush - “Spirit of Radio” (again, intro guitar):
The sound design capabilities of flangers are obvious, especially when parameters are automated in an interesting way. However, the constantly oscillating comb filter can become tiring to listeners quickly.
Therefore, flangers can be great as transitional effects and momentary ear candy to keep the listener’s attention.
Phasers / Phase shifters
It shouldn’t be a surprise that phase shifters (also known as phasers) operate on a similar idea to chorus and flanging, apart from one key difference.
Once again, a copy of the original signal is made. Instead of delaying the signal, however, it is passed through a circuit called an “all-pass filter.” This type of filter does not affect the level of frequency content in the signal, but does introduce a phase shift around a set frequency.
Connecting another all-pass filter afterwards would create a single notch (one tooth of a comb filter). Phasers work by stringing several all-pass filters together in series to create a series of non-harmonically related notch filters.
An LFO can then be used to modulate these notch filters, similar to the motion in a flanger. Due to the inharmonic relationship between the phaser’s notch filters, the effect sounds more gentle—between that of the chorus and flanger.
In the phaser plug-in below, the number of notches introduced into the frequency spectrum is assigned by the “Poles” parameter. The cutoff frequency of the all-pass filters is set using “Frequency”. This phaser also has a feedback option, giving it capabilities similar to that of a flanger.
The native Ableton phaser repeats the LFO parameters for the native flanger, with LFO amount, shape, rate, and phase all being adjustable.
In the example below, the first and third guitar arpeggios are dry. The second and fourth have been processed with a phaser.
Here are a couple classic examples of the phaser effect in recordings:
Led Zeppelin - “Kashmir” (phaser on the drum for whole song):
Billy Joel - “Just the Way You Are” (phaser on the rhodes piano in the intro):
The phaser and flanger create very similar sounds, and can be used more or less interchangeably. As mentioned, the flanger can sound more extreme than a phaser, so phasers can be used when more nuance is needed.
Often confused due to their similarity, these processors all incorporate phase shifting in different ways. Chorus combines it with pitch modulation, flangers use it to cause harmonic-based comb filtering, and phasers employ all-pass filters to phase shift without the use of delays.
While chorus, flangers, and phase shifters may slightly differ in their functions and application, all can be used effectively to create interesting timbres in sound design. Their ability to produce ambience, locationality, and modulating timbres (in ways that the standard processors can’t) should secure their place in any producer or sound designers tool kit.