Snoman R. Dance Music Manual: Tools, Toys, and Techniques

Подождите немного. Документ загружается.

PART 3

Mixing & Promotion

374

THE VERTICAL PERSPECTIVE

The ﬁ nal perspective of a mix is the top to bottom of the soundstage, with

higher frequency elements sitting towards the top of the stage and lower fre-

quencies towards the bottom. Much of this positioning will already be dictated

by the timbres used in the mix. For example, basses will tend to sit towards

the bottom of the mix and hi-hats will naturally sit towards the top. However,

it should be noted that these, and any other timbres, will contain frequencies

that do not necessarily contribute to the sound when placed into a mix.

A typical case in point is that the basses used in dance music do not consist

entirely of low-frequency elements but also some mid-range elements and

possibly some high frequencies too. If you were to tonally adjust this type of

sound by removing all of the mid-range frequencies, the bass would move

more towards the bottom of the soundstage, making some space for instru-

ments to sit in the mid-range. This ‘corrective ’ adjustment is one of the most

fundamental aspects of mixing and is accomplished with the EQ on a mixing

desks channel.

EQ is a frequency volume control that allows you to reduce or increase the gain

of a band of frequencies contained within a sound, but it can be used for much

more than simply making space in a mix . Applied cautiously it can be used to

make insipid timbres much more interesting, add more deﬁ nition to the attack

portion of a sound to pull it to the front of a mix or prevent the bass from

overpowering and ‘muddying’ the entire lower frequency range of the mix.

However, if you apply it without knowing exactly what affect it will have on a

sound and those surrounding it, a mix can fall apart within minutes and you

end up with nothing more than a muddy, ill-deﬁ ned mess that doesn’t trans-

late at all well on a typical club PA system.

To understand the concept behind using EQ, we need to revisit the ﬁ rst chapter

which explained that all sounds are made up of a number of frequencies that

are a direct result of the fundamental and its associated harmonics at differing

amplitudes. Also, that it’s this predetermined mix of harmonics that helps us

establish the type of sound we perceive, whether it’s an earth-shaking bass, a

screaming lead or a drum kick.

We only require all these frequencies present if the sound is played in

isolation. If it’s mixed in with a number of other instruments, we only need

to hear a few basic frequencies because our hearing/perception will persuade

itself that the others are present – they’re just masked behind the other

instruments.

This natural occurrence is one of the main keys to mixing. If frequencies that

we don’t necessarily need are removed, well have more room for the important

frequencies of other instruments that we do need to hear. This means that we

need a way of pinpointing a speciﬁ c group of frequencies within a sound that

we can remove or enhance. In ‘professional’ desks this is accomplished with a

parametric EQ.

Mixing

CHAPTER 19

375

Fundamentally, a parametric EQ con-

sists of three controls: a frequency

pot, a bandwidth pot and a gain pot.

The frequency pot is used to select

the centre frequency that you want to

enhance or remove, while the band-

width pot determines how much

of the frequencies either side of the

centre frequency should be affected.

Finally, the gain pot allows you to

increase or reduce the volume of the

chosen frequencies.

Suppose you had a lead sound that

contained frequencies ranging from

600 Hz to 9 kHz but the frequen-

cies between 1 and 7 kHz were not

required. Using the frequency pot, you

could home in on the centre of these

(4 kHz) and then set the bandwidth

wide enough to affect 3 kHz either

side of the centre frequency. If you

then reduce the gain of this centre fre-

quency, the frequencies 3 kHz either

side would be attenuated, too.

The size of this bandwidth is often

referred to as the Q (quality). The

smaller the Q number, the larger the

‘ width’ of the bandwidth ( Figure 19.3 ).

Such a precise bandwidth as shown in

the ﬁ gure is entirely theoretical. In the

real world the two extremes of the band-

width are not attenuated at right angles.

This is because, similar to synthesizer’s

ﬁ lters, EQ has a transition period or ‘response curve ’ that allows it to appear

more natural. Typically, this is around 3 dB per octave but can sometimes be

6 dB per octave ( Figure 19.4 ).

The problem with this approach is that if you apply a particularly heavy cut

or boost and the Q does not remain constant, the bandwidth will increase in

size as you boost or cut further. Typically, the width of the Q is measured at

3 dB from the ﬂ oor, so if the bandwidth were set to remove 2 kHz either side

of the centre frequency at an EQ boost of 3 dB and you were to then boost by

another 3 dB, the bandwidth would affect more of the frequencies either side

of the centre frequency ( Figure 19.5 ).

(Centre frequency)

1 kHz 7 kHz4 kHz

Gain

cut

FIGURE 19.3

Theoretical EQ

(Centre frequency)

1 kHz 7 kHz4 kHz

Gain

cut

FIGURE 19.4

Practical EQ

Becomes wider

3 dB boost

6 dB boost

FIGURE 19.5

Non-constant Q

PART 3

Mixing & Promotion

376

This action is known as a ‘non-constant

Q ’ because the Q changes as you increase

or decrease the gain at the centre fre-

quency. This creates an unnatural sound,

so many professional desks and external

EQ units will vary the response curve

as the gain is increased or decreased to

maintain a constant bandwidth. This

produces a sound that appears much

more natural to our ears and is the sign

of a well-designed EQ unit ( Figure 19.6 ).

Q should only be used when discussing the theory behind EQ, and I wouldn’t

advise taking it any further than this book. In a studio situation you’re com-

municating with musicians, not techno freaks, so it’s usual to talk about the Q

in octaves. Otherwise you could ﬁ nd yourself laughed out of the studios. For

those with an unhealthy academic interest into how the Q value can be trans-

formed into octaves, you can use the following math theorem:

()4

215

×=

−

⎛

⎝

⎜

⎞

⎠

⎟

log Octave

But for those who simply can’t be bothered, it’s worth memorizing this table as

it covers the most commonly used Q settings while mixing:

Q Setting Octave Range

0.7 2 Octaves

1.0

Octaves

1.4 1 Octave

2.9 ½ Octave

5.6 ¼ Octave

Generally speaking, a bandwidth of 1/4th of an octave or less is used for

removing problematic frequencies, while a width of a couple of octaves is

suitable for shaping a large area of frequencies. For most work, though, and

especially if you’re new to mixing, you should look at always maintaining a

non-constant Q of 1.0. This is sometimes referred to by many engineers as the

‘ magic Q ’ since it commonly avoids any masking problems (we’ll look at this

in a moment) and sounds much more natural to the ears.

As of yet, we’ve only concerned ourselves with one form of EQ – parametric –

but there are shelving EQs too. These will, or at least should, be available on

Remains constant

3 dB boost

6 dB boost

FIGURE 19.6

Constant EQ

Mixing

CHAPTER 19

377

most semi-professional and all professional desks and consist of low- and high-

pass ﬁ lters along with low- and high-shelf ﬁ lters. These two different forms of

ﬁ lters are often confused with one another due to their similar nature, but it is

important to differentiate between them.

Both low-pass and high-pass EQs are based on exactly the same principle as

the low- and high-pass ﬁ lters used in synthesizers. The low-pass (sometimes

referred to as high-cut) will attenuate the higher frequencies that are above the

cut-off point. A high-pass (sometimes referred to as low-cut) will attenuate all

the lower frequencies that are below the cut-off point. Conversely, a low-shelf

ﬁ lter will boost (or cut) any frequencies that are below the cut-off point, while

a high-shelf will boost (or cut) any frequencies above the cut-off point.

Of course, this does mean that if only low- and high-pass ﬁ lters are available

on the desk, it is possible to emulate the action of a low- and high-shelf and

this is the reason why the two types are often confused. For instance, if you

need to emulate the action of a high-shelf, you could use the high-pass ﬁ lter

and increase the EQ overall output level.

The speciﬁ cations of the mixer will depend on how much control you have over

the ﬁ lters and shelves but, generally, a shelving ﬁ lter will have two controls: one to

select the frequency (often called the knee frequency) and one to adjust the gain

of the frequencies above (or below) the shelf. Some of the more expensive desks

may also offer the opportunity to adjust the transition slope of all these ﬁ lters,

and this can sometimes be important when we need to ‘roll off ’ frequencies.

Since many speaker systems are incapable of producing frequencies as low as

40 Hz, they should be removed altogether and the attenuation of frequencies

above or below the cut-off point is often termed as ‘rolling off ’. Thus, if you

were asked to roll off all frequencies below 40 Hz, you would employ a high-

cut ﬁ lter and position the cut-off point at 40 Hz. Or at least you would if the

transition slope of the EQ ﬁ lter were more or less vertical. As we’ve seen, EQ

units will use a transition slope to make the effect appear more natural to our

ears, but we don’t necessarily need this when rolling off frequencies that are

not required; so it is useful to reduce this if it happens to be particularly long

(6 dB per octave and over) ( Figure 19.7 ).

Naturally, all this control requires dedicated pots on a mixer and additional

circuitry to implement it; so many of the cheaper mixing desks will not employ

shelving or low/high-pass ﬁ lters and in some instances may not even offer any

parametric EQ. Indeed, it’s fair to say that the cheaper the desk, the less EQ

control you’ll be offered. EQ can broadly be categorized into three areas: ﬁ xed

frequency, semi-parametric and graphic.

Fixed Frequency

Fixed-frequency EQ is possibly the worst type to have on any mixing desk as

it’s the least ﬂ exible, but as it is the cheapest to implement, many cheap con-

sumer desks will employ it. Fixed-frequency EQ consists of two, sometimes

PART 3

Mixing & Promotion

378

three, rotary pots that control the low, mid and high frequencies. To accom-

plish this, they commonly use low/high-pass ﬁ lters with a ﬁ xed cut-off fre-

quency at 12 kHz and 75 Hz for the high and low band, respectively. If there

is a third pot, then this controls the mid-band and is generally a notch ﬁ lter

with a centre frequency that’s ﬁ xed at around 2 kHz and has a ﬁ xed bandwidth

of approximately two octaves (see the chart for the respective Q!). This type of

EQ makes precise, frequency-speciﬁ c shaping impossible, and unless you have

an external parametric EQ unit it is far from suitable for creating a good mix.

Semi-Parametric

Semi-parametric EQs (sometimes known as sweep EQs) utilize a similar design

as ﬁ xed-frequency EQs with a ﬁ xed low/high-pass ﬁ lter at 75 Hz and 12 kHz,

respectively. Rather than having a ﬁ xed mid-band, only the bandwidth is ﬁ xed

and you are free to sweep the centre frequency. This type of control is typi-

cal of mid-priced consumer mixing desks, but while offering more freedom

to select frequencies, the ﬁ xed bandwidth means that it isn’t versatile enough

for creative EQ. It can be particularly difﬁ cult to produce a good mix with this

limitation.

Graphic

Graphic EQs do not appear on mixing desks but are sometimes used by mas-

tering engineers to set the overall tonal colour of a full mix. They’re also pos-

sibly the most instantly recognizable of all EQ units. Most home stereos now

have them ﬁ tted as standard, appearing as a number of faders, each of which

offers control over a speciﬁ c frequency band.

The size of the bandwidth covered by each fader obviously depends on how

many faders it has. So the more faders available, the better the EQ, since each

bandwidth will be smaller. The reason these are referred to as graphic EQs is

(Harmonics/frequency)

High shelving

(Harmonics/frequency)

Low shelving

Fundamental

Filtered harmonics

Unfiltered harmonics

FIGURE 19.7

The action of shelving

ﬁ lters

Mixing

CHAPTER 19

379

because you can often tell what type of music people listen to by the way the

equalizer is conﬁ gured. Hence it’s rather graphic in nature.

With the intervention of computers and audio sequencers, this graphic nature

is not just limited to graphic EQs anymore, as many software mixing desks

and some of the latest digital desks will mix a parametric EQ with the visu-

als of a graphic EQ to produce a paragraphic EQ. This gives all the freedom

of a parametric EQ with the visual ease of using a graphic EQ. The immediate

beneﬁ t of this is obvious. Rather than having to determine the current EQ set-

tings through a series of often cryptic pot positions, you can actually see the

frequency range being affected.

That said, keep in mind that your potential audience does not see EQ settings,

they only hear them and so you should do the same. One of the biggest prob-

lems with mixing on any ‘visually enhanced ’ software desk is that you have the

tendency to use your eyes more than your ears and this can be disastrous. If it

looks right, it may not sound right; if it sounds right, it most probably is.

PRACTICAL MIXING

Armed with the theory of mixing, we can approach a mix on a more practical

sense, and it should go without saying that the ﬁ rst step to creating a good mix

is to ensure that you can actually hear what you’re doing reliably. If you don’t

hear the frequencies within a mix accurately, then you certainly can’t EQ or

affect them accurately. This means that you must use loudspeaker studio moni-

tors to scrutinize a mix rather than placing your trust in everyday hi-ﬁ speakers.

No matter how ‘hyped ’ these are by manufacturers for having an excellent

frequency response, all hi-ﬁ speakers deliberately employ frequency-speciﬁ c

boosts and cuts throughout the sonic range to make the sound produced by

them appear full and rounded. As a result, if you rely entirely on these while

mixing, you’ll produce mixes that will not translate well on any other hi-ﬁ

system. For instance, if your particular choice of hi-ﬁ speakers features a bass

boost at 70 Hz and a cut in the mid-range, you’re obviously going to produce

a mix based on this. If your mix is then played on a system that has a ﬂ atter

response there will be less bass and an increased mid-range.

It’s impossible to recommend any particular make or model of studio monitor

because we all interpret sound differently, but generally you should look for a

model that makes any professionally produced music you like sound moder-

ately lifeless. If you can make your mixes sound good on these, then it’s likely to

sound great on any other system. Then again, you shouldn’t aim for a totally ﬂ at

response, and it’s usual to look for monitors that are a happy medium between

coloured and ﬂ at. Indeed, most dance musicians will deliberately use monitors

with a vaguely coloured response, since 99.9% of people who are going to listen

to it will do so in a club, in a car or on a typical home stereo system.

Even with a good monitoring system, it’s important to also note that the

actual positioning of them will have an effect on the frequencies they produce.

PART 3

Mixing & Promotion

380

Placing any monitor close to a corner or wall will increase the perceived bass

by 3, 6 or 9 dB due to the boundary effect. This is similar in some respects to

the proximity effect that’s exhibited when singing too close to a microphone –

there’s an artiﬁ cial rise in the bass frequencies.

With monitors this can be attributed to the low-frequency waveforms emanat-

ing from the rear of the monitor onto the wall or corner directly behind it.

This low-frequency energy combines with the energy produced by the monitor

itself and doubles the perceived bass. As a result, the general rule of thumb is

that unless they’re speciﬁ cally manufactured to be used close to a wall (i.e. they

have a limited bass response and use the wall to produce the low-frequency

energy), they should be placed at least 3 feet away from any wall. Of course,

this may not always be possible, and if it isn’t then you’ll have to learn to live

with the increased response by listening to as many commercial mixes you can

and training your ears to accept the response they produce.

On that note, before you even approach any mix you should take time out to

listen to some of your favourite commercial mixes. We hear music everyday

but we very rarely actually listen to it, and if you want to become competent at

mixing, you need to differentiate between hearing and listening. This means

that you need to sit down and play back your favourite commercial mixes at a

moderate volume (just above conversation level) and try to pick out the action

of the bass, the drums, vocals, leads and any sound effects. Listen for how the

melodies have been programmed and ask questions such as:

■ Is there any pitch bend in the notes?

■ Does the drum loop stay the same throughout the track or does it

change in any way?

■ What pattern are the hi-hats playing?

■ Do they remain constant throughout?

■ What effects have been used?

■ Where is each instrument placed in the mix?

Although many of these may not seem related to mixing, listening this closely

for small nuances in drum loops, hi-hat patterns and basses, etc., will train

your ears into listening closely to any music and will help you to not only

begin identifying frequencies but also increase your awareness of arrangement

and programming techniques.

Most important of all, though, despite the huge array of audio tools available

today, the most signiﬁ cant ones are those that are stuck to the side of your head.

Continually monitoring your mixes at excessive volume or listening to any music

loud will damage the ears ’ sensitivity to certain frequencies. A recent survey con-

ducted showed that those in their late teens and early twenties who constantly lis-

tened to loud music had more hearing irregularities than people twice their age.

Also, keep in mind that dabbling in pharmaceuticals for recreational use leads

to a heightened sensory perception. Although they may have played a large role

in the development of the dance scene, they don’t help in the studios. Apart

from losing hours from ﬁ ts of laughter about terms such as ring modulation,

Mixing

CHAPTER 19

381

the heightened sensory perception makes (a) everything sound great or (b)

everything sound wrong, so you spend the next 10 h twiddling EQ on some-

thing that doesn’t need it.

VOLUME ADJUSTMENTS

The ﬁ rst step in creating a mix is to begin by setting the relative volume lev-

els of each track, and this means taking the musical style into account. You

need to identify the deﬁ ning elements of the genre and make sure that these

are placed at the very front of the mix. Typically for all dance music this means

that the drums and bass should sit at the front of the mix with everything else

that must be located centrally sitting behind them. This includes the vocals as

more often than not they will not be the truly deﬁ ning part of the music.

It’s generally best to begin mixing by starting with these deﬁ ning points, so

it’s prudent to mute every channel bar the drum tracks and commence mixing

these. As the kick drum is the most prominent part of a drum loop, set this so

it’s at unity gain and then introduce the snare, claps, hi-hats, cymbals and other

percussion instruments at their relative levels to the kick.

At this point don’t concern yourself with panning, EQ or effects but try to ﬁ nd

a happy medium so that the drum levels are close to how you imagine the ﬁ n-

ished product to sound. Follow this by adding the bass, then the vocals (or the

next important part of the mix) and gradually add each audio track in order

of priority, adjusting the volume levels until all the instruments are playing

together in the mix. It’s vital that you’re not concerned with masking, EQ or

effects at this stage. You need to make the mix work quickly before you start

jumping in with compressors, EQ or effects as these can quickly send you off

on a tangent and you can lose the direction of the mix. If you’re experiencing

troubles setting the volume levels, then a quick trick is to switch the mix to

mono as this allows you to hear the relative volume levels more accurately.

This form of ‘additive ’ mixing is only one way to approach a mix, and some

users feel more at ease using the ‘subtractive ’ method. This means that all the

faders are set at unity gain, and each is reduced to obtain the appropriate levels

required. Which to use is entirely up to your own digression and depends on

which you feel most comfortable with.

Whichever method you feel at ease with, an approach I often recommend is

that when adjusting any volume fader, do not listen to the instrument you

are adjusting the volume of, rather listen to the whole mix and how adjusting

the volume is affecting the mix. For instance, as you increase the volume of the

strings (if they are present), do not listen to the strings, rather listen to how the

other instruments are being affected by the strings ’ increase in gain.

PANNING

With all the relative volume levels set, you can then begin to pan the instru-

ments to their appropriate positions. This is where your own creativity and

PART 3

Mixing & Promotion

382

scrutinizing of previous mixes of the same genre will come into play, as there

are no ‘strict’ rules for creating a good soundstage. Of course, this isn’t going

to be much help to those new to mixing, so what follows is a very generalized

guide to where instruments are usually positioned in most dance genres.

Instrument Pan Position/Description

Kick Positioned central so that both speakers share

the energy

Snare Positioned from 1 o’clock to 2 o’clock or central

Hi-hats Positioned at the far left of the soundstage with a

delayed version in the right

Cymbals Positioned central or from 1 to 2 o’clock

Percussion Positioned so that the different timbres are in the

left and/or right of the soundstage

Bass Positioned central so that both speakers share

the energy

Vocals Positioned central since you always expect the

vocalist to be centre stage

Backing vocals Occasionally stereo so they’re spread from 2 to

4 o’clock

Synth leads (trance) A stereo ﬁ le positioned fully left and right with

perhaps a mono version sat central

Synthesized strings Positioned at 4 o’clock or stereo spread at 9 and

3 o’clock

Guitars Positioned at 3 o’clock

Pianos Commonly stereo with the high notes in the left

speaker and low notes in the right

Wind instruments Positioned in the left or right speaker (or both) at

3 and 9 o’clock

Sound effects Wherever there is space left in the mix!

Apart from placing the kick, bass and vocals in the centre of the soundstage,

note that this is a very general guide and the panning of instruments will deﬁ ne

your own particular style of mixing. Indeed, the best solution to panning

instruments is to have a damn good reason as to why you’re placing a timbre

there in the ﬁ rst place. Always have a plan and don’t just throw sounds around

the image regardless.

Mixing

CHAPTER 19

383

In direct contradiction, in many dance mixes the positioning of instruments

is rarely natural, so you should feel free to experiment. In fact, in many cases

it may be important to exaggerate the positions of each instrument to make a

mix appear clearer and more deﬁ ned. Nevertheless, you shouldn’t position a

sound in a different area of the soundstage just to avoid any small frequency

clashes with other instruments. In these circumstances you should try to EQ

the sound beforehand; if it still doesn’t ﬁ t, consider panning it out of the way.

After the instruments have been panned it’s highly likely that the relative vol-

umes will have adjusted. You’ll need to go back and readjust these so that the

instruments are all at their appropriate volumes. Again, while panning the

instruments, listen to how the rest of the instruments are being affected, rather

than listening to where the instrument being panned is moving.

E Q

Although there are no rules on how to use EQ when it comes to mixing, some

generalizations can be made. Firstly, any EQ is essentially acting like a ﬁ lter

and this means that like any other type of ﬁ lter EQ introduces resonance at the

cut-off point and phase shifting. In other words, it applies small amounts of

comb ﬁ ltering into the audio, which sounds like small amounts of distortion.

This is where the quality of the EQ unit being used plays a major role: less

competent/cheaper EQ units will introduce more comb ﬁ ltering than the more

expensive units.

As a result, you need to exercise care while EQ’ing sounds so that it doesn’t

become too apparent; this is why many engineers will advise that you should

look to cutting frequencies rather than boosting them. This is because boosting

any frequency will make the phase distortion much more noticeable. To sup-

port this theory further, in the ‘real’ world our ears are more used to hearing

reductions in frequencies than boosts since frequencies are generally reduced

by walls, objects and materials.

Yet another reason to cut rather than boost lies with the recording of the audio

itself. To increase the signal-to-noise ratio, any audio should be captured as

‘ hot’ as possible. If an audio ﬁ le is already close to clipping, EQ boosts may

push the audio to distortion.

As mentioned, these are only generalizations and so in direct contradiction, on

occasion, boosting the frequencies that are lower in volume than the funda-

mental can be used as a creative sound design tool. Indeed, one of the most

common misconceptions about EQ is that you should never really have to use

it if the sound is recorded correctly in the ﬁ rst place.

The viewpoint stems from when EQ was ﬁ rst used as a means of correcting

the frequency response of a recording to prevent it from compromising the

recorded sound. This was because recording in the early days was carried out

with poor-quality gear when compared with today’s standards and, therefore,

frequencies had to be adjusted to reproduce the original recording.