The basics

Compression basically explained: the reason people ask about compression more
than anything is because they find it the hardest concept to understand or hear. A
basic explanation I heard when I first started was thinking of compression like an
automatic volume control, when the audio is loud it gets turned down and when it's
soft it gets turned up. This means sharp signals are now curved and fading signals
are now picked up and last longer. It also means smoother sounds and fatter notes.

Soft knee, hard knee

Unless you have a software compressor or a really high-end model, you won't be
able to choose the setting. You simply have to decide whether to get a soft knee
compressor or a hard knee one. Try to buy a soft knee compressor as it will be
useful on practically everything without crushing the sound. Mostly used on vocals
and mixes, it means a larger amount of compression can be applied, while hard
knee compressors, which can be heard working, will be typically used on bass. It is
more of an audible effect than soft knee.

Yours is a Urei

Kate Bush was rumoured to have two compressors across her vocals: the infamous
studio compressor Urei 1176 one on its flat-out setting (all ratio buttons pushed in)
and another added just in case any peaks got through. It sounds crazy, but the
1176 is famous because of its super-soft character on vocals, and on this high a
setting she would have had every breath picked up by the mic and every peak
squashed. This gave her a unique sound.

Vocals

Vocals are one of the hardest and most dynamic sounds you may come across. My
advice would be to try and catch the peaks in the song. Use soft knee, set the ratio
around 2:1 (but maybe as high as 6:1 for voiceovers and spoken word), attack to
0.09ms, release to 100ms then adjust the threshold to catch the loudest parts of the
song, so you get about 8dB of reduction.

Get what you pay for

Software compressors are fantastic now and the built-in compressor in Emagic's Logic has done the job for me on
many vocals now. However, I still went out and paid for a classic compressor like the hand-wired perfection of the
all-valve Chiswick Reach. This put across the outputs of Pro Tools is amazing. A lot of money compared to
software, but the sound is worth every penny.

The beat goes on

Drums can be transformed by compression in a mix. On a snare try a soft knee, use a ratio of 4:1, a long attack
and a little longer release, then adjust the threshold to just grab the first couple of dB of reduction. Now try
adjusting the attack shorter, and the threshold higher to adjust the sound to fit the track, the R&B type of snap or
the pop type of slap.

Stereo low

Old-school engineers often use the trick of sub grouping the drums to a stereo pair then applying a stereo
compressor to achieve a pumping sound.

...and for everything

Pretty much everything will sound better with a little compression, the whole sonic from bass drums to flutes.

Take care on tape

Always remember that, if you're committing tot ape, then err on the side of caution. Remember you can always
add more if you want to, but it's impossible to remove.

Try and try again

Don't be afraid to experiment and, as with all things audio, 'use your ears!' If it sounds good to you, then it m buest
right, regardless of what the manual says.

The basic rule

Yes, yes, it's been said before, by many people, but you can never have too many compressors!

Double your money

Instead of putting a whole sound through a compressor, a neat trick is to split it to two channels, heavily compress
one of them and mix that with the uncompressed channel. This works particularly well on drum sounds and can
be applied to, say, an individual snare drum or a stereo submix of the whole kit (or some of its constituent parts).
The compressed version of the sound can be tweaked to make it pump by setting an appropriately short release
time and can then be added to the uncompressed version to get a more exciting and dynamic rhythm bed.

Multiband

When working with a sound source which covers a full (or at least large) frequency spectrum, such as a complete
mix, normal compressors tend to introduce a 'pumping' effect. This is because the lower frequencies which tend
to trigger the compressor will normally be doing something quite different to the higher frequencies, yet the
compressor will attenuate the entire output by the same amount. Multiband compression, as the name suggests,
uses a crossover to split the full-bandwidth input sound into smaller bandwidths which are then compressed
separately. The results are then mixed back together, the result being a much louder,tighter mix which doesn't
pump or sound squashed.

23 Compression tips

Sidechain

The sidechain, or key, is the signal within a compressor which is used to control the
output level. It is when this signal exceeds the threshold that compression is applied
to the main signal running through the unit. A lot of hardware compressors (and
some of the better plug-ins) have external sidechain - or key - inputs so you can use
the characteristics of one sound source to compress another. In this way you can,
for example, use a kick drum track to make a synth pad pump in time with the
music.

De-essing

Used mainly on vocals, this is a technique for reducing the level of sibilant sounds
('s' or 'sh', etc) which are significaonutldye lr than the other sounds associated with
speech. The problem areas of these sibilant sounds tend to be above 7-8kHz,so
reducing a microphone's volume when these frequencies are encountered is the
order of the day. Take an auxiliary send from the vocal channel, feed it through an
equaliser and then into your compressor's sidechain input. Insert the compressor
into the vocal channel. Then use the equaliser on the sidechain to boost the sibilant
frequencies. You'll need to be able to monitor the sidechainn i some way, as, as
with all things, using the old shell-likes is the only way to go. Listen to the main
vocal channel to determine the best settings for the threshold and ratio, but use a
fast attack and release time to make the effect as transparent as possible.

Guitars

Overdriven and distorted guitar sounds rarely need compression as the process of
overdriving them introduces a fair amount of compression anyway. Clean and
acoustic guitars are a different matter. For that classic, funky, clean sound, use a
fairly low threshold, a ratio of at least 3:1, fast attack and quite a slow release. Add
a touch of chorus or flanging to add the extra icing on the cake. Strummed parts,
electric or acoustic, are better with a higher threshold, the aim being to reduce the
volume of the transient parts of the sound (the strums themselves) while keeping a
steady overall level. As ever, use your ears!

Bass

One of the best uses for a compressor is on bass sounds - get a solid, steady
foundation to your track, and the rest will stand up beautifully. Probably. Optimum
settings vary here, as much depends on the type of bass sound - synth, electric,
acoustic, etc - but as a general rule, use fast attack and medium release times, a
medium threshold and a ratio of between 3:1 and 5:1.

Have an idea

I realise all those dials and buttons are just calling you over to start fiddling as soon
as you plug the compressor in, but have some sort of purpose first. Are you trying to
make the sound more punchy, more smooth or just keep it under control? Listen
and think before you leap.

On the attack

I try to have the attack up (ie, not at its quickest setting) a little on most things if
possible; guitars, for example, really benefit from extra 'front'. Many recommended
compressor settings for parts such as vocals suggest the quickest attack time, but
try it up and see what you think. You can add a little attack on the most unlikely
things to help them stand out more.

Using bypass

Having a stereo compressor over the whole mix is generally a good thing, but
monitoring through it when mixing is hard work. While it can be useful to know what
the final compression will do to the mix, you'll probably just be fighting it the whole
time. Leave it on bypass until the mix is almost finished.

For effects

There's lots of fun to be had compressing effect returns, if you have the free mixer
channels to set it up. If your delay doesn't quite fade away as you'd like, then
compressing the return may give you the control you're after.

Pre- or post-EQ

The difference between having the compressor before the EQ section or after it can
be drastic. When learning about compression, try both options and hear the way the
compressor affects your EQing, and vice versa.

                         

Equalization is the most-used, most-mis-used, most-over-used and most-under-used signal processing device. It is also the most powerful. By definition an equalizer is a gain control that raises or lowers gain at a specific set of frequencies without affecting the gain at other frequency ranges. I learned a lot about equalization by sitting down with a graphic equalizer and lots of records. After about 30 hours of listening to the effect of different bands of equalization on the records, I began to "hear" the effect of certain frequencies on overall mixes. Since I had a job in mastering, this was a direct application as to how I was going to use equalizers. Doing an exercise like this is the very beginning of training in using equalization. There are commercial CD packages which attempt to duplicate this type of test. They are no real substitute for switching the equalization on and off as the CD (or music source) plays. The next step in the training is to apply specific equalization to specific instruments. Recommended or "key" frequencies are used and the equalization is switched on and off while soloing the instrument and then while listening to the instrument in the mix. The point of this "ear" training is to gain the ability to hear what frequencies would be needed to bring up or down in a specific mixdown. Until one obtains the ability to "hear" frequency, one has only limited ability to use equalization.

                                         Frequency Ranges:

A key to understanding equalization is to gain an understanding of the effect of different frequency ranges on music and instrument sounds.

                                                   The "First Octave"

The first usable octave for most recording is the 40 - 80 Hz range, with equalization settings centered around 50 Hz. This range of frequencies is often referred to as "Low Bass" There is sound between 20 Hz and 40 Hz but little or no sound from instruments. The lowest pipes of a pipe organ will get into this range but more "ordinary" instruments like Bass Guitar, Upright Bass and Foot Drums do not. The lowest pitch on a bass guitar or string bass is at 41 Hz. Thunder, earthquakes and rumble from the building shaking extend below 40 Hz. While mixing, watch out for objectionable sounds below 40 Hz caused by building shifts and mic stands moving with heavy footsteps. If there is objectionable sounds in this range, the range can usually be taken entirely out with a filter. The first octave that we deal with (40 - 80Hz) gives more of a "feeling" and sense of "power" to the sound. This range is way down or non-existent in smaller stereo systems. This range is difficult to hear at all at medium and low volume levels because of the Fletcher Munson Effect. To properly set the amount of low bass in your mix or in your instrument sound, you must listen both loud and soft. You also may want to listen to the mix or instrument on large and small speaker systems. Too much energy in this range will make the mix sound muddy on large speakers played loud and still sound good on small speakers played at a medium volume. You want the mix or instrument to sound larger and more powerful over large speakers without sounding muddy. Rap, Hip Hop and "Dance" music (under various names) often have extra energy in the low-bass range. This is what causes cars equipped with sub-woofers to shake. Usually, however, it is not the entire mix that is boosted below 80 Hz, but just, for example, the foot drum. By boosting the energy on only one or two instruments, "clarity" can be achieved without "mud."

                                            The Bass Range

Covering about 1.5 octaves, from 80 Hz to 250 Hz, this range of frequencies determines the "fatness" and "fullness" of the instrument's sound. Equalization is usually applied centered around two frequencies, 100 Hz and 200 Hz. For guitars and bass, the 100 Hz range tends to add body and fullness. Excessive energy in this range tends to make these instruments sound "boomy.," This range of frequencies is still greatly affected by the Fletcher-Muson Effect; this means you will need to listen to the mix and instrument both loud and soft. Similar to how the 50 Hz range affects the bass and foot, the guitars should sound fatter when played loud, not boomy. Reducing the 100 Hz energy on the guitar will usually cause distinction between the bass and guitar parts. The lowest fundamental frequency on a guitar is around 80 Hz. For vocals the 200 Hz range determines the fullness of the vocal. This range can often be reduced to increase distinction on the vocal. If, however, boosting in higher frequencies on the vocal makes the sound "thin" or "small" a boost of 200 Hz. will restore fullness. When 100 Hz is reduced on a guitar or bass to reduce "boom," at small boost at 200 Hz can be helpful to keep the instrument from sounding "lumpy" (certain notes hard to hear and others standing out). The guitar and bass have almost equal energy at their fundamental and 2nd harmonic frequencies. Thus if a range of notes becomes hard to hear because of a at lot of 100 Hz, reducing energy at 100Hz and adding energy at 200 Hz will help the notes be heard again.

                The Bass Presence / Lower Mid Range Covering

about one octaves from 250 Hz to 500 Hz, this range accents ambience of studio and adds clarity to the bass and lower-string instruments (Chello and Upright Bass). Too much boost can make higher-frequency instruments muffled sounding and low-frequency drums (foot and toms) have a cardboard box quality. Equalization in this range is applied at many frequencies but most often between 300 Hz and 400 Hz. The lower part of this range (250 Hz to 350 Hz) is sometimes referred to as "Upper Bass" and is used to increase distinction and fullness on the vocal, especially on female singers. The Lower Mid Range in general can be viewed as the "Bass Presence Range" Increasing this range gives clarity to the bass line and the lower-register of pianos and organs. Clarity and distinction can be obtained between the foot drum and bass guitar by both reducing the foot and increasing the bass guitar in this range, at the same frequency. This range is often reduced for overhead drum and cymbal microphones to increase clarity and presence on these instruments' and reduced on lower drums (foot and toms) to reduce boxiness.

                                                       The Mid Range

The Mid Range band of frequencies covers two octaves from 500 Hz to 2 kHz. This range can give a horn-like quality to instruments (500 Hz to 1 kHz) and a "tinny" sound (1 kHz to 2 kHz) or a telephone-like quality (all of the range). Equalization usually centers around 800 Hz and 1,.5 kHz. The mid-range also tends to accent the presence (800 Hz) and attack (1.5 kHz) of the bass guitar. The lower pitches of a rhythm guitar can be given more attack by a boost at 1.5 kHz. For your Mid Range Instruments (vocals, guitars and piano) this range is most-often reduced rather than accented. Reducing 500 - 800 Hz on an acoustic guitar can remove the "cheep" sound and make it sound more "silvery." Reducing 800 Hz on a vocal makes it sound less nasal and have more body and presence. For snare drums, a reduction of 800 Hz can take the tinny, cheep sound out of the drum and make the snares have more sizzle rather than rattle.

                                      The Upper Mid Range

Covering about one octave, this range of frequencies is responsible for the attack on percussive and rhythm instruments and the "projection" of mid range instruments. Equalization can be applied at any frequency in this range but still somewhat centers around 3 kHz. On the foot drum, boosting 2.5 kHz or 4 kHz increases the attack. 2.5 kHz sounds more like a felt beater and 4 kHz sounds more like a hard-wood beater. These frequencies can also be used to increase the attack or "hit" sound on toms and snare drums. Guitar lines often get more attack and distinction with equalization added at this range. A small boost (1-3 dB) for the vocal will increase projection. Adding too much energy, in this range, makes it hard to distinguish the syllables of the vocal and can cause listening fatigue. This range of frequencies is often reduced on background vocal to give them a more "airy" and "transparent" sound.

                                        The Presence Range

Although this range covers a mere half-octave of 4 kHz to 6 kHz, it is an often-used band of frequencies. This range makes most vocals and melody instruments sound closer and more distinct. Over-boosting causes a irritating and harsh sound. Equalization centers around 5 kHz.

                                          The Treble Range

Covering approximately that last two octaves of sound (6 kHz to 20 kHz), this band of frequencies is responsible for the brilliance and clarity on instruments. Equalization centers around 7 kHz, 10 kHz and 15 kHz. The vocal "S" sounds are at about 7 kHz, making this a frequency that is avoided for vocals. Care must be exercised in reducing 7 kHz on vocals, however, because the vocal will sound dull very fast. The breath sound of the vocal is at 15 kHz and above, giving a breath quality without much accent on the "S": sound of the vocal. The 7 kHz frequency is also the "metallic attack" frequency on drums The "sizzle" of cymbals is at 15 kHz. When equalizing, 10 kHz and above is often used as a general "brilliance" frequency band.