Introduction
Compression - you've heard about it, you've seen the plugins, you may have even used it yourself...
But do you understand it? Scrolling down this page, it might looks daunting - I promise, it's not!
Compressors are incredibly versatile tools that can be used in a pretty vast array of ways, but I notice that many times that they're used in recordings (or other audio applications), they're used in a way that does more harm than good. I thought I'd do a little bite-sized walkthrough on the basics of compression here, for anything who isn't too comfortable using them.
Firstly,
For voice work, there is no 'correct' setting to put on a compressor. Everyone's voice is different, every recording has different peaks. So let's get that out of the way first - there is no magic setting I'm going to give you here. Instead, I'm going to help you understand it, so you can make sense of it and create your own settings.
To my knowledge, every workstation has a built in compressor (including Audacity). So, you won't need to spend any money to get your hands on an effective compressor. I use Cubase and I can speak for it that the default Steinberg Compressor is a great little tool - no worse than the paid ones I also have.
What are all these buttons?
Depending on what compressor you use, you'll see a different array of buttons. There's a few images below of different compressors I float between, which you'll very quickly notice are all very different - look closer, though, and you'll notice some striking similarities in the buttons.
(in case you're curious, in order, these are - CLA-76 by Waves, Pro-C 2 by FabFilter, Steinberg Default Compressor, Audacity Compressor)
So - these all look very different, but are functionally the same thing.
There are a few settings you'll see on every single compressor, regardless of which ones you use - thankfully, these are the only ones you'll really need to be in-the-know about.
Threshold, Ratio, Attack Time and Release Time. There's also make-up/gain on some compressors, which I'll briefly touch on, but it's fairly self explanatory.
Threshold
Simply put, the threshold slider is a level you set to trigger the compressor. Once the signal going into the compressor hits the threshold you've selected, the ratio that you've selected will kick into effect. There's really not much more to a threshold - it's easily the simplest setting on here!
For voice work, I generally have a threshold that just catches the peaks of an audio. Having a threshold much lower will flatten the dynamics of the audio - something that's desirable in some situation, such as music production (say, for drums/parallel compression) but is less so when mixing vocal parts.
Using a compressor that has a level meter inside of it can be handy to see where your peaks are hitting. This makes it a bit easier to find a comfortable level for your threshold to sit at.
Ratio
Ratio is a bit more confusing to explain. The first thing I want to do here is share an image that does a great job of showing how a ratio affects a signal (pulled from the iZotope website, thanks iZotope!):
In this image, the threshold is 40dB. Once the signal hits that level of 40, that's when the signal begins to be attenuated by the compressor.
As you can see, a 1:1 ratio is no attenuation, with the level decreasing the higher the ratio gets - hitting the top ratio (or inf on many compressors) is just limiting. This means no signal gets past the threshold level.
A good way to think of the ratios is like this:
For a 5:1 ratio, for every 5dB of signal that exceeds the threshold, only 1dB will pass.
Similarly, a 20:1 ratio, for every 20dB of signal that exceeds the threshold, only 1dB will pass.
Effectively, the compressor is quashing (or, compressing) the signal going past that threshold. Nice and simple.
For VO work, again, a high ratio will flatten dynamics if the threshold is too low and isn't generally too desirable in voice work. You tend to hear fairly compressed voices on radio shows or podcasts, but not in commercials/acting etc. (at least, not often).
Find a good middle ground that just gently attenuates the peaks that are creeping past your threshold (something relatively low, like 3:1). You're only trying to control these peaks, not flatten your voice.
Attack Time
Attack time is certainly one which less people have much knowledge on, but it's nice and simple. Put simply, attack time is a set amount of time (usually measured in ms) that the compressor 'waits' before triggering it's effect.
(super handy infographic from MasteringTheMix)
So, if you have a 200ms attack time on a compressor, once a signal exceeds your threshold, the compressor will only take full effect once it's waited the 200ms (or 0.2 second) delay.
Why is this useful? Attack time is usually something that's toyed with to toy with the transient of a sound. This is the 'hit' at the start of the sound, if you will - imagine a drum being hit; the initial impact of the stick on the drum and the 'crack' it produces is the transient.
Audio engineers/producers will use a higher attack time to accentuate the transient of a sound, whilst keeping the rest of the sound under control, creating a more impactful hit. For voice work, however, we generally want to keep this (very) low. If a sound is getting past our treshold, we want to grab it and deal with it as quickly as possible. Transients in voice work don't sound too great.
Release Time
The other end of the compression chain, release time is another set amount of time that the compressor takes to 'release' the signal.
Super simple and easy to understand. For our voice work which is likely to be only lightly compressing, a fast release can sound nice and natural, keeping our dynamics in check. Having the release time a bit too high is going to keep the compressor's effect active for a little too long - remember, we're here just to catch the peaks and go.
There's not really much more to say about release time!
Make-up Gain & Knee
Another really short one - make-up really is just a gain stage for your output signal.
In absolute laymans terms, this is just a volume control for the final, processed signal on the other end of your compressor. Use this if you need a slightly boost in the signal from where the compressor has reduced it.
I'll package knee in here too, incase your compressor has one - this is a 'smoothing' option for the transition of your signal between the pre and post compressor phase. If you'd like a smoother transition that's perhaps a little less jarring, you'd choose a soft knee. A hard knee would be chosen for a near-instant compressor effect.
Conclusion
And that's really all there is to it!
Compression is daunting, I get it. It will take time to get used to.
My top advice, however, is to just be really light with it. Overcompressing is really jarring to me and I hear a hell of a lot of it out in the voiceover sphere. The way I view compression in my general work is that it's just a tool and not FX - if it's changing the sound of my voiceover too much, I'm doing it wrong!
Hope all of this helps and makes sense.
Peace out. Be kind.
Elliot
Comments