Compression is one of the harder concepts to grasp when it comes to music production. When I was first learning, I found it an overwhelming to go through so many tutorials that either didn’t show practical use of compression or didn’t explain why you’d do it. So we’re going to start from the very begining and keep moving onward from there.
Before we learn about compression, it’s important to understand how the your DAWS repesents sounds that you hear. Here’s a quiet sound and a loud sound.
And here’s a way too loud sound:
(sound cloud actually minimized the distortion by limiting the sound, but you can still here, it sounds bad)
0dB is basically “loud.” Its as loud as a speaker can play a sound without distorting it. A quiet sound like the one shown below peaks at -10dB and a loud sound is close to 0dB. The distorted sound goes above 0 which is into the ear blasting zone. We’ll come back to all this in a little bit. Now lets look at a real situation.
Compression to Minimize Volume Spikes
Here’s where we start. This is a synth arp that has a very strong start and then a soft fade out point. Let’s say you want to smooth out this first spike in the wave. You could just turn the volume of the track down:
This worked, the peak at the start is now less powerful, but the quieter part is now WAY quieter! Let’s say we don’t want to make our entire sound quieter but we just want to smooth out that initial bump. This is where compression comes into play.
Here’s a quick look at the controls:
It’s important to understand what a compressor does before looking at the controls. Basically, when a sound plays, the compressor looks for the volume to go above a certain level (threshold #1). When it does, it essentially lowers the volume level until the volume goes back below that level. The attack (#2) release (#3) and ratio (#4) effect how much and how fast the imaginary volume knob is turned down. Lastly, the gain (#6) control is used to bring the volume back up after the spikes are smoothed out by the compressor.
Don’t worry if you don’t get all this yet. The samples below will help you figure it out.
1. Compressor threshold is what decides “when” the trigger starts working, when a volume goes over a certain level, it starts turning the volume down from that point. Lowering this makes the compressor more sensitive to spikes in your sound. Raising it makes the compressor less sensitive.
2. Attack is the speed at which the compressor starts working when it is triggered.
3. Release is how quickly the compressor “let’s go” when its done being triggered.
4. Ratio is “how much” the volume is turned down when the compressor is triggered.
5. Knee effects how the volume changes when it beigns.
6. Gain is basically just a volume knob that applies to the compressed sound. We’ll look at more about what this does later. For now leave it alone at 0db.
7. Gain Reduction – This gives you a visual representation of how the compressor is working. If you don’t see anything happening here, your compressor is not working.
* The limiter is actually another control, it cuts off any stray bits of sound that get too loud after the compression is applied.
The best way to understand what is does is to look at an actual sample.
Here’s the raw wave again with the spikey loud part at the front.
Here it is compressed:
Here’s the two waves overlapped:
See what the compresor did here? The original sound (turquoise) had a large spike at the begining and that has now been minimized (green is the compressed sound). There is still a little bit of a spike – the nature of the sound has not been drastically modified – but it has been minimized by a few dB by the compressor. As you can see the quiet parts of the sound are mostly unaffected. Only the louder moments have been “turned down” slightly.
Compression To Raise volume
Now let’s look at a new wave that has a different problem. This synth note has a strong attack and then a sustain that is a little weak (I did it on purpose for the sake of the tutorial)
Let’s say we want to bring up the level of the trail on this note. When I was first learning compression at this point I said, why not just turn the track’s volume up? We could do that but look what happens:
Now let’s apply a compressor to this same sound. We’re doing the same thing as the previous one, but with a new purpose. We are going to first even out the spike.
Now we’ve evened out the spike. This gives us a new situation:
Since we lowered the spike, there is now this new empty space up here. You migth be saying, “didn’t we want to make the sound louder? It’s quieter now! What is your deal Beatscribe!?”
Well, now we can turn up the Gain on the compressor and look what happened:
As you can see, the overall sound is now louder. It hasn’t changed too much in form, but we have brought up the lower trailing part without going into the red. Now, let’s say we really want to post this even higher, just turn on the compressor’s limiter. Put the threshold around -.01db and crank up the gain. Here’s the new sound with the gain cranked:
See what happened? We just “added loudness” to the wave without distorting it or drastically modifying it. The quiet trail is now much louder, but the spike at the beginning never goes over 0db either. Of course, using the limiter with the compressor does modify your sound in a bigger way, but overall your sound will still sound the same without distorting.
Next time we’ll look at the multipressor, which is sort of like a compressor combined with an EQ. Than after that, we’ll look into sidechaining and some other cool stuff you can do with a compressor.