WHAT exactly are DYNAMICS anyway?
Dynamics processing deals with adjusting the dynamic range of a piece of audio. That means deciding how loud or soft it will be, moment to moment.
For our purposes, this is a relative term – things are loud or soft in relation to something else.
Why? It’s because we are “cooking” with audio ingredients in the final analysis, and we want our final meal, whether “candy” or “rare steak”, to tantalize your senses.
This means managing the dynamics of the song on both the macro level and the micro level, from how loud the ‘t’ on the end of the word “bright” sounds, to how the entire pre-chorus song section is imbued with a heavy feeling of intense pressure, sustained until the chorus entry explodes and unleashes all that pent-up energy.
You can compress whole stereo mixes of songs, or you can turn down an annoying click or pop for an instant, long enough to hide it’s unfortunate presence in your wonderful, emotionally raw vocal take.
Sounds are loud or soft, either relative to other sounds in your mix or to the technical upper limit of system gain, namely the headroom available in your mix. Headroom may be the unknown word to you here.
Headroom is the amount of gain available to you before your system clips (distorts) the signal.
In digital audio systems like a DAW, the maximum level of the system is called 0dBFS.
This means zero decibels is agreed as a calibrated scale mark present on your fader scale markings showing the point of Full Scale audio along that fader’s travel.
Full Scale means the maximum level permissible with digital audio before clipping occurs. Stay below FS all the time, no exceptions.
Digital clipping will occur way above the point at which the analog pre-amp will clip in front of the A/D converters of your DAW interface inputs. The interface is, of course, where analog signals you plug in are converted to digital audio.
All audio levels in a DAW environment will have a negative dBFS value, with -6dBFS being an example where the audio signal is a full 6 dB quieter than the Full Scale level of 0dB.
At -6dBFS, t is already exactly half as loud as it would have been if it had been set at Full Scale, the 0dBFS mark.
The 0dBFS level is always clearly marked, and appears near the top of your onscreen faders in your DAW mixer page.
Always leave your main output fader set at the Full Scale 0dBFS mark. Change your other faders, but leave the Master Fader.
MAXIMIZERS – DO NOT USE
People use maximizers mostly across the main stereo output of their mixes. I don’t recommend this until you have been recording for a decade or so. On the other hand, experimenting and living with the mixes is easy to do and will teach you a lot. Try out a few maximizers on free trial periods and see what they are about. In five minutes, you’ll see for yourself, because they only have a few controls.
Naturally, you can use them on individual tracks to fatten them up, as well as on whole mixes.
I strongly recommend that you do not do this on music you intend to release.
Leave mastering to a mastering professional. Once over-processing a master is done, it cannot be undone – the master is effectively dead and you have wasted your efforts.
With these things, you lower the threshold to hear the effect, but the more you lower it, the more impact it has. It’s easy to overdo this and ruin the sound without noticing immediately (it can be subtle at first), so be very, very careful not to squash the life out of your music or get weird distorted warbling sounds and terrible grotty bass as side effects. Danger, Will Robinson!
In mastering audio for clients, I usually set the maximum level that can be present in my final masters to -0.15dB, which is just less than Full Scale, and allows for a little leeway against inter-sample peaks. This is done with a plug-in called a maximizer, such as Waves L2 or L3 and lots of great alternatives by McDSP or UAD or Slate Audio or Sonnox and many others.
Be extra careful with using these things across your stereo master output – you don’t want to set the threshold any lower than where you get about 5dB or 6dB of gain reduction. That’s also going to be way too much effect for jazz or natural styles, but for pop and dance and rock you’re probably at around the right amount of squishing and re-inflating.
Please get a mastering engineer to do this stuff, because it takes considerable years of experience and some fancy equipment, especially full-range monitor speakers, and finely honed critical listening skills to master music well.
You know it makes sense and they don’t cost much per song. Why not take that last step seriously and get a really great result? I’ve had new clients say they didn’t know it was possible to sound that much better quite a few times now, so find a great mastering engineer and invest in your music.
Clipping digital audio sounds appalling and nobody will want to hear it. Seriously. We have countless distortion tools available and that is not one of them.
TYPES OF DYNAMICS PROCESSING
First up, don’t process tracks for no reason. That said, you will want to use dynamics processing a lot as you get the hang of the types available. It can make your vocals really in-your-face and present, or it can smooth out dramatic level changes in an unruly bass guitar track. You’ll love it.
I tend to refrain from using any dynamics processing on the input signal to a recorder or DAW. It is so easy to do later, and there is no significant noise penalty in waiting for playback to choose compression settings. These days, systems are so quiet it is pointless to compress a sound as it arrives in 99% of cases.
Having said that, if you have purely temporary access to a great sounding compressor or limiter in hardware form, by all means colour the sound with it as you record. Take care though – compression is extremely hard to remove, and in fact it’s usually impossible. Once squished, it’s usually squished in character forever more.
We have compression and expansion, two sides of a coin. We have gating,an extreme form of gain reduction. We have limiting, which is compression with extreme ratios. They have their uses. We also have level maximizers of course, and multiband dynamics processors that act on specific frequency ranges.
We have various parameters to figure out like input level, output level, threshold, attack, release, ratio, range, hysteresis, decay, hold time, and so on.
A lot of these exotic controls interact with each other, so when you move one thing it affects the way something you already set behaves, so you have to go back and readjust the initial parameter.
It’s like tasting and adding salt, or not. You get used to a little interplay with your compression to get it where you want it.
It can take quite a while to get good at this stuff, not only because it’s techie, but also because it takes a while to recognize unwanted artifacts of the processing.
It can be really hard to hear the wanted or unwanted changes sometimes, but all in all, it’s like anything else – practice makes perfect. The trick is to set the devices up to overdo the effect quite a bit and then back off to what you really wanted. That way you know what you are hearing and can adjust to taste quickly.
I’m going to start the guided tour tomorrow in my next blogpost “PROCESSING- DYNAMICS Part 2″. We’ll begin with a look at the types of compressor you might use and how and when to use them.
See you then!