12. Chorus, Flanger & Phaser
In the last video we talked about the delay effect and a lot of different things we can do with it. In today's episode will follow the same direction as we will explore 3 effects that use a very similar process. We'll talk about flanger, chorus and phaser effects.
They are called modulation effects because they usually have one or several LFO inducing movement in the sound.
When a signal goes through a flanger, it is duplicated into 2 copies. One is played back as is and the second one is delayed slightly. It's generally a very small delay time, below 15 ms late compared to the first signal.
So up to now it is really like a delay effect.
These 2 signal are then layered and this create some phase cancellation and amplification.
At some points, for some frequencies, the 2 waveforms will both have positive values or both negative values, so there values will be added and make the signal more powerful.
At other points, for other frequencies, one will be positive while the other will be negative, so they'll cancel each other out.
As a result some frequencies in the sound will be amplified, and other will be cancelled or attenuated.
We can then trace a graph to show which frequencies are boosted and which one are cancelled, and if we do so, we get something like this.
This is the same shape that a comb filter have, and that's why some comb filters have the name flanger attached to them.
And because the second signal is uniformly delayed, these frequency cuts are uniformly spread out along the frequency spectrum.
It follows a harmonic series (the second cut should be at twice the frequency of the first one, the third cut should be at 3 times the frequency of the first cut, etc...).
Then if we move the delay time (of the second signal) we move this harmonic series, just like if we'd move the cutoff of a comb filter.
On many flanger effect you should also have a feedback parameter, that sends the result of the effect back into it, that basically accentuate the effect of the filter. And on some flangers you also have the option to choose the polarity, to create a feedback or a feedforward comb filter. You can check the episode about filters for more info on that.
Now the flanger effect is often associated with a sweeping sound. That's because usually, the delay time is controlled by an LFO, that makes it move up and down. So the LFO will move the equivalent of this filter up and down, creating the sweeping effect
So to sum it up, on a flanger effect you often have a delay time and a feedback parameter to basically set the position and the strength of the comb filter.
Then you have a rate and a depth knob which define the speed of the LFO and how far it will move the delay time.
A chorus effect is very similar to a flanger effect.
When a signal goes through a chorus effect, it is also duplicated. The first one is also played back as is, and the second one is also delayed, but generally it's delayed by a higher amount, between 20 and 50 ms.
Then a chorus effect also generally an LFO that modulate the delay time the second signal, just like a flanger. But, unlike a flanger, modulating the delay time will also alter the pitch of the signal. Just like with a delay effect when we right click it to put it in repitch mode, shortening the delay time will make the pitch higher, and making the delay time longer will make the pitch lower.
So with the LFO, the pitch of the second signal will go up and down, just like a vibrato. So mixed with the dry signal, it makes a dissonance, and as any dissonance, that creates kind of a beating In the sound. If the effect is very drastic it can make a straight up wobbly dissonant sound, but with more subtle values, it is often perceived more as a rich, shimmering tone.
In term of frequency response, when we mix the delayed signal with the dry one, it also makes the effect of some kind of a comb filter, but different than a flanger.
It is also possible to have more than one copy of the signal, to have several delays with different settings, to accentuate the effect.
Because usually, there is no feedback control on a chorus. I mean, some of them do have a feedback control, but it's not always there.
It's an interesting effect that gives the impression that a sound is played by more than one source, hence the name chorus. And it is often used on vocals, pianos or guitars to add thickness or interest to the sound.
So to sum it up, on a chorus effect, you would have a delay time parameter just like on a flanger. Then you'd have a rate and an amount knob to control the speed of the LFO and how much it moves the pitch of the second voice.
Also, some chorus effects are called stereo chorus. This stereo effect is achieved by delaying the left and the right channel differently, with 2 different LFOs that have different speeds. So it would pitch shift the left and the right channel differently as well
A phaser effect is also very similar to a flanger effect, but is a bit more complex.
The signal that enter a phaser effect is also duplicated, with one signal being played back as is and the other being processed.
The processed signal is kind of delayed as well, but it's delayed differently that in a flanger effect.
The signal goes through one or several filters that are called all pass filters.
And an all-pass filter doesn't really cut any frequency. What it does is phase shifting the signal in a non-linear way.
Basically what it means is that it delays some frequencies in the sound differently than others. So you can have the bass frequencies delayed more than the higher frequencies for example.
So when you mix this signal with the dry signal, you still get some phase cancelling, you still get kind of a comb filter effect, but compared to a flanger effect, a phaser would have less cuts. Somtimes, even just one.
The frequency cuts do not necessarily follow a harmonic series.
As a result, a phaser effect would sound less harsh that a flanger effect. Because the effect is not "tuned" to a particular frequency, it will sound less drastic.
Actually, the less all pass filter in your phaser, the softer the effect. And the more all pass filters used in your phaser, the more frequency cuts you add.
Some phaser even let you activate or deactivate these filters, they often refer to it as "poles" or "stages".
So on a phaser effect you generally have a frequency knob and feedback knob, to move the series of cuts and their strength. And you'd also have an LFO with a rate and a depth parameter, and that would modulate the frequency, so that can create a sweeping effect close to a flanger but less harsh. Then you may have a pole or stage parameter to use more or less filters in your phaser.
So the thing to remember about the flanger, chorus and phaser is that they all use a short delay time to alter the sound and they all have an LFO to give it some movement. And it's this LFO that gives them their status of modulation effects.
Personally, in a sound design phase, I like to use them as static effect, without using the LFO, so I would turn its amount to 0. I like the texture it gives.
But on the opposite, you can use a very fast LFO. Past a certain rate, you can't really hear the oscillations of the LFO and it becomes some kind of distortion. Just like in FM synthesis, where you oscillate the pitch of an oscillator so fast that it becomes a distortion.
Or you can use a normal or slow rate for the LFO to simply have motion in you sound.
Another thing to keep is mind is that the result of these effects are equivalent to adding comb filters, which means subtractive synthesis. That means that these effect would work better on sounds with a rich harmonic content. Or at least you'll hear the effect more on sounds that have a lot of harmonics. So it can work well on sounds generated with FM synthesis or that has distortion on them for example.