As any pro engineer or producer will tell you, dynamics are one of the
most important considerations with any mix. Proper use of dynamics processing
is crucial to finishing off a mix and making it sound more professional,
and without it you'll have a hard time getting that high quality finish
to your music. How often have you been watching TV and noticed that the
adverts seem to be much louder than the program you're watching, even
though you haven't touched the volume control? It's all down to the clever
use of dynamics by advertising companies, because they know that the
louder something is, the more likely it is to grab your attention. It's
the same with music: if one track appears louder than another, it'll
grab the listener's attention more. Dynamics processors play a huge part
in this, making music appear louder, more 'in ya face', and generally
much more appealing to an audience.
Dynamics processors take in a range of machines, including compressors,
noise gates, Finalizers and de-essers, so letÃ¢â‚¬â„¢s have a look at
what each can do for your music.
The most important processor to have around for increasing the loudness
of a mix is a good compressor. This is designed to limit the dynamic
range of a track, allowing you to turn the overall volume of a mix up
louder before finally mastering it to CD. The best way of explaining
how compression works is to give an example, so let's say you've just
mixed a track involving drums, bass, lead guitar, vocals and a synth
playing a chord over the top.
Obviously you'll need to record it at the highest possible volume, but
when you turn the volume of your mix up, in certain places throughout
the track the vocals, guitar or chord parts can push the signal indicators
into the red, which results in the sound distorting, so you have to turn
the volume of the mix down to prevent it. Of course, you could always
drop the volume of the offending instrument or vocal, but if you do,
it could sound too quiet in the parts of the track where it wasn't clipping.
In the old days before all this newfangled technology was available,
the only way to counteract this was by employing an engineer with lightning
reflexes and a good ear who listened for the peaks in a track and dropped
the fader at the offending part before pushing it up again for the rest
of the mix - essentially a forerunner to mix automation.
Things were certainly made easier by the introduction of compressors,
simply because they allow you to set a volume threshold to control when
the compression kicks in. In our example this would be set just below
the offending part's peak. If the offending signal tries to go into the
red during playback, the compressor immediately reduces the signal, preventing
the sound from distorting and allowing you to push the overall volume
of the track up more, making your mix louder and more impressive.
However, as tempting as it may be to keep compressing and pushing the
volume up louder and louder, there are limits. Too much compression can
force your mix to lose its dynamic edge, resulting in a rather flat sound.
So it's a case of compressing as hard as you can while also listening
out for the dynamic range of your mix, and getting the right balance
between these two considerations.
Looking at the controls on a compressor, you'll notice that there are
probably several others in addition to a threshold dial. They also include
ratio, attack, release, gain and sometimes a setting for hard or soft
The ratio is the amount of compression applied to a signal when it exceeds
the threshold you've set. This is listed, not surprisingly, in values
such as 2:1, 4:1, 6:1 and so on. So, for example, a ratio of 2:1 would
compress the output signal to half that of the incoming signal, and the
higher the ratio, the more the level is reduced once it exceeds the threshold.
The attack and release (the latter sometimes known as 'decay' on compressors)
are exactly the same as they are with a synth envelope, controlling how
the compressor will behave when a signal exceeds the threshold. Attack
defines how quickly the compression starts, and release controls at what
stage compression stops after the signal level is lower than the threshold
again; a long or slow attack will not activate the compression until
the signalÃ¢â‚¬â„¢s level has been over the threshold for that length
of time, whilst a slow release will mean that the compressor returns
to its resting position fairly quickly. Obviously, this also has limitations
and should be set carefully depending on what signal you're compressing.
Using a long attack on a kick drum will have no effect at all, as the
signal only exceeds the threshold for a short amount of time and normally
decays fairly rapidly. Also, if you set a fast attack time but a slow
release on a kick drum accompanying some other lower level tracks, the
compression (triggered by the kick) would be held for too long and so
would audibly punch holes in the signal.
While using the envelopes, you may also need to take the knee setting
(if offered) into account. This can be considered as the shape of the
compression curve between the compressed and uncompressed signal (either
side of the threshold). A hard knee results in more aggressive and sudden
compression, whereas a soft knee creates a smoother or more transparent
transition from uncompressed to compressed.
So, if you've set the threshold, ratio, attack and decay correctly,
the compressor should compress effectively and reduce the dynamics in
a sound, but you also need to compensate for the overall reduction in
volume of the signal. This is where the gain control comes in: you can
use this to bring the signal back up to its original level (minus the
peaks that have been compressed, of course).
When finalising a track prior to burning it to CD, it's usual to compress
the whole mix, helping you to increase the volume even further, and this
is when the most common mistakes are made. The threshold has to be set
very carefully; if it's set to react to the kick drum, for example, it
will compress/uncompress on every hit, resulting in a mix that pumps
and 'breathes', rising and falling in loudness.
There are no rules as to what settings to use on a compressor, so it's
only possible to offer guidelines (see p51). In the end it will depend
on your ears and the mix you're working with. Apart from reducing the
dynamic range of an instrument, compressors can also be used in a creative
context, as any dance musician will tell you. Drums in dance music, particularly
the kick and snare, are very heavily compressed; so much so that you
can actually hear the loss of dynamics in the sound. The main reason
many dance tracks have such a huge beat is that massive compression has
been applied. As we mentioned at the beginning, however, there's more
to dynamics than just compressors...
Close the gate
Noise gates can be just as important as compressors in creating a powerful
mix. When recording a guitar or vocals - in fact, anything using a microphone
- unless you're in a professionally sound-treated room, you're guaranteed
to capture some background noise. While the vocalist is singing or the
guitarist is playing his piece it isn't as obvious, as the sound being
recorded is louder than the background noise; but when they come to rest,
the noise will become more prominent.
Noise gates work much like compressors, except in reverse: any sounds
that exceed a threshold setting are let through unprocessed, while sounds
below are simply cut off altogether, resulting in total silence. As with
compressors, they also feature attack and decay settings that react as
you would expect. Using a fast attack will result in the sound being
cut immediately, while a slower setting would slowly gate the signal.
A typical noise gate will also feature a couple of extra controls: hold
and range. Hold controls how long the gate will be open or how long the
level below the threshold is cut/reduced, while range adjusts the amount
of signal that can be heard while the gate is closed or how much the
signal below the threshold is reduced by.
To better explain the range control on a noise gate, let's introduce
the extra input that some gates (and compressors) feature: the sidechain.
A signal, usually from a microphone, is connected to the sidechain, and
instead of the threshold functioning as normal, the sidechain takes control
of it. As you speak, the noise gate drops anything else playing through
it thatÃ¢â‚¬â„¢s below the threshold to the volume or "range",
set by the range control. It's the same with a compressor; as you speak
(or sing), the 'background' music going through the main inputs and outputs
of the unit will be compressed to the level determined by the ratio control.
There are plenty of uses for a sidechain with both compressors and noise
gates. With compressors, the most common use is to 'duck' instruments
to make room for vocals. For example, a guitar can swamp the bandwidth
that it shares with the vocals in a mix, so by sidechaining the vocals
into a compressor that has a guitar signal as the input, when the vocalist
sings, the guitars get compressed, allowing the vocals to sit in the
mix better. Noise gates are popular with DJs, as by sidechaining their
voice and setting a low range, when they speak, the music drops in level,
allowing you to hear them more clearly.
As with every other processor and effect in today's studios, compressors
and noise gates have been used and abused way beyond what they were originally
designed to do. The most memorable use of a noise gate in recent years
was on Olive's You're Not Alone, but it's a trick that's been employed
again and again - even Blondie found a use for it (the pad in Heart of
Glass). This sort of sound is achieved by feeding a simple hi-hat pattern
into the sidechain input of a noise gate, then playing a chorded pad
or string sound through the same gate, so that every hi-hat hit closes
and opens the gate, causing a stuttering effect on the pad.
Another trick is to put plenty of reverb on a kick drum and then set
the noise gate to cut off the reverb's tail, giving a louder and more
prominent kick. There's more to dynamic processors than just noise gates
and compressors though; there are also de-essers and finalizers.
De-essers can be employed on vocals to reduce sibilance. This is an
effect caused by a singer over-stressing the 't' and 'sÃ¢â‚¬â„¢ sounds,
resulting in the recorded vocals sounding like a dripping tap, a steam
engine or, in the worst cases, both. Now, while most vocalists shouldn't
need any specific treatment - and it's always best to avoid applying
any if possible - using a bright capacitor mic, adding compression or
processing with a bright reverb will bring up the levels of sibilance,
which is where the de-esser comes in. They work by literally targeting
a narrow frequency band, just the Ã¢â‚¬ËœsÃ¢â‚¬â„¢ or Ã¢â‚¬ËœtÃ¢â‚¬â„¢,
and compressing it. These devices are designed to reduce the effect of
the over-stressed 's' and 't', although if at all possible it's best
to try and capture the performance right in the first place, as de-essing
can have a detrimental effect on vocals, robbing them of their live feel.
Finalizers are used to sprinkle a little of that oft-mentioned 'fairy
dust' over a track. When you've finished mixing a piece of music, sometimes
it can still sound a bit plain, so feeding it into a finalizer can add
an extra sparkle gained from boosting the lower and upper harmonics,
along with adding small amounts of reverb to brighten up the image. Appealing
as this may sound, many engineers and producers avoid finalizers like
the plague, as they can add too much polish at the expense of the energy
and Ã¢â‚¬ËœgrittinessÃ¢â‚¬â„¢ that you might have originally wanted,
resulting in an over-produced sound. There's also the problem of users
attempting to use them to make a bad mix sound better, which simply doesn't
work. Instead, it merely emphasises the worst parts.
So, now we have reached a good point from which you can begin using
dynamics processors to liven up your mix and make it louder, but it's
important to remember that there are no hard and fast rules. It's down
to your own ears and, as with any effects processors, compressors, gates
and their ilk, definitely worth experimenting with as wildly as possible
to come up with new sounds of your own.
Setting up a compressor
Table of typical compression settings:
|Fastest possible or
|-5 to -9||2:1 to 3:1||Soft|
|Drums||5ms||10ms or Auto||-15||5:1 to 8:1||Hard|
|4ms to 10ms||10ms||-4 to -8||Ã‚Â 4:1||Hard|
|4ms to 10ms||10ms||-2 to -10||Ã‚Â 8:1||Hard|
|Slowest possible or
|-3 to -8||4:1 to 12:1||Soft|
|BrassÃ‚Â||Fastest possible||Fastest possible or Auto||-10 to -14||2:1 to
|Hard or soft|
|Fastest possible or
|-10 to -14||8:1 to 14:1||Hard or
All these values are meant as a rough guide, the exact levels of threshold
and ratio are very much dependant on the signal being compressed, not
to mention the character of the sound you want to create. This threshold
guide could also be misleading in certain situations where a different
scale is provided; on the VoiceMaster Pro for example, a dBFS scale is
provided so a conversion will be necessary (assume 0dBFS to be around
20dBu, so Ã¢â‚¬â€œ5dBu to be around Ã¢â‚¬â€œ25dBFS). Perhaps you could
begin with these settings and play around with one parameter at a time
to decide what gets the best audible results.
On the other hand, if you want to set up the compressor without the
use of the table, then once you've got your sound fed into a
compressor, the first control you need to adjust is the threshold, as
this sets when the compressor will start to Ã¢â‚¬ËœsquashÃ¢â‚¬â„¢ the
signal. In most cases you adjust the threshold so the compressor works
on the loudest parts of your signal. Next, you need to set the ratio.
This setting will depend entirely on how you want your music to sound,
so if you're writing dance music, you should set a fairly high compression
ratio. Set the attack and release according to the nature of the waveform,
if there are lots of transients (fast peaks) or not and whether they
want to be compressed or left alone (read the earlier passages in this
dynamics tutorial for more on attack and release times). Compressing
a signal will lower its output volume, so once you have compressed the
signal sufficiently, you'll need to raise the volume from the compressor
and not from the desk, as this can unbalance a mix.