|
Ok here we go.
All dynamics processors (compressors, gates, ...) have what is called a sidechain. This is the signal path in the device that contains the detecting circuit, in other words the part of the processor that controls it.
Under "normal" use, the sidechain receives the signal you present at the dynamics processor (or in the cheaper models/plugins, you don't have access to the sidechain yourself). For example, you compress a kick. The kick signal also passes in the sidechain path. Thus it is the kick that is detected, and that will determine (according to the threshold/ratio/attack/release) how your kick will be compressed.
Now, on more extensive models, the sidechain is accessible on it's own. This means you can present one signal at the processors input, and ANOTHER signal at the sidechain. What does this mean. Well your second signal (at the sidechain) will "control" the processing. Processing that is applied to the first signal (that was presented at the processor input). At the output you only hear the signal that passes in the compressor, not the signal that passes in the sidechain.
What uses are there for that you might ask. There are plenty. I'll state the most used ones, but you are free to experiment.
- A trick that is commonly called kick over bass (sometimes bassducking). You use a compressor on a bassline, but present the KICKDRUM at the sidechain. As I explained, the kick will "control" the compressor, but the bass will be compressed. With kick over bass you can either make that famous pumping effect, or just use it to mix the bass in higher.
The aim of this "trick" is the bass will be compressed down just as the kick goes, and revert back to it's normal level inbetween kicks. A little bit of tinkering with the attack and release times and you'll see you'll come to a result pretty quick.
- Ever heard about the talkover function on a dj / radio mixer? Music is playing, when the dj starts to talk, the music lowers in level. That's done in the same manner, with sidechaining. There are actually two method to achieve that.
With a compressor (music in compressor, vocal splitted to the sidechain) : vocal is speaking, music gets squashed down.
With a gate, in ducking mode (which means it actually closes down when you go beyond the treshold. Ducking is kinda the opposite of gating). Same setup, almost same effect.
Usually the ducker (so the gate) is prefered. With a compressor, the harder the guy shouts in the mic, the more the music gets compressed. With the ducking, it's a fix attenuation that is applied.
- A popular effect nowadays is rhythmical gating. I believe some people also refer to it as the trancegate, as this has been used a lot in trance music. Put a synth in a gate. In the sidechain, you put a rhytmical component (for example a kick). In this example we'll play a steady tone on the synth.
Now you have two options. You use the gate as a gate. You play on the synth, you won't hear much. Now the kick plays a rhythmical pattern. If you set your gate settings right, the gate will open on each kick. Your steady synth tone will be "demuted" on the rhythm of the kick...
If you want the opposite (and that's more like the trancegate) you use the gate in ducker mode. Your synth tone will play, and be cut rythmically with the kick.
- A deesser. Nowadays you find it as apart plugins, but in practice it's nothing but a compressor. For example you have a singer that has sybilance problems (harsh s'es). If you would eq it out, you'll lose a lot of other "useful" high frequencies.
Better is the deesser. You split the vocal. The "unprocessed" vocal goes in the compressor. Then you take the second one, remove the low frequencies, and emphasize those S sounds, with an EQ. This EQ'ed vocal goes into the sidechain. Now you just have to play with the compressor parameters until you find the right balance.
What happens? Well the compressor is triggered by the EQ'ed signal, that contains mostly the bad S sound. So the compressor will not move in normal circumstances, but only when the singer makes an S sound. So the compressor will only reduce S sounds... It will de-ess. Of course you can try any frequencies for that matter for funny effects.
- Finally, a less popular trick for amateur people, but still widely used (consider this as a bonus knowledge). Suppose you want to put lots of reverb on a vocal (or lots of mic ambience on a miked drumkit, works also very well). But when you put the amount of reverb you want, the sound gets all washed out and undefined. What you do is you put a compressor (works smoother than using a ducker) on the reverb (that's why usually effects are returned to channels, so you can EQ and process the effect separately again), and you split the vocal into the sidechain. The singer sings, the reverb gets pushed down. So you got the definition. As soon as the singer stops singing, the compressor stops working and the reverb gets back to the level you want...
As you see, sidechaining can produce much useful results, and this is the reason why much people still complain Cubase doesn't support this natively yet, and right they are.
Hope it got clear now,
Thy
|
b?
