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

Подождите немного. Документ загружается.

PART 3

Mixing & Promotion

364

for anything less than the very best – it may be an overused analogy but you

wouldn’t expect to make a great tasting cake if some of the ingredients were

out of date!

Perhaps, most importantly, before you approach the mix you need to ask your-

self one ﬁ nal yet vitally signiﬁ cant question about your track. Close your eyes,

listen back to the music and ask yourself ‘can you feel it? ’

Above the arrangements, sounds, processing and effects this is the ultimate

question and the answer should be a resounding (and honest!) yes. Dance

music is ultimately about ‘groove ’, ‘vibe’ and ‘feel’ and without these funda-

mental elements it isn’t dance music. A track may be beautifully programmed

and arranged but if it has no ‘feel’ then the whole endeavour is pointless.

MIXING THEORY

At the outset, it’s important to understand that mixing is a creative art: there

are no right or wrong ways to go about it. All engineers will approach a mix

in their own distinctive way – it’s their music and they know how they want it

to sound. A unique style of mixing, provided that it sounds transparent, will

deﬁ ne your creative style as much as the sounds and arrangement do. Despite

this, there are some practices that you can employ that will put you on the right

track to producing transparent and detailed mixes.

The ﬁ rst step to creating any great dance mix is to understand the theory

behind mixing and how it can be adapted in a more practical sense to suit your

own particular style. This means that you need to comprehend how the sound-

stage of a mix can be created and adapted, how frequency ranges are grouped

together and how to come to terms with our natural hearing limitations. We’ll

look at each of these in detail.

HEARING LIMITATIONS

Our hearing is far from perfect because not only do we perceive different frequen-

cies to be at different volumes but the overall volume at which we listen to a mix

also determines the dominant frequencies. This inaccurate response can be traced

to a time when we lived in caves. Since talking (or more likely grunting) to each

other in a cave resulted in vast amounts of reverberation, our hearing adapted

itself to concentrate on the frequencies where speech is most evident and deci-

pherable in this situation – approximately 3 –4 kHz. This means that at conversa-

tion level, our ears are most sensitive to sounds occupying the mid-range and any

frequencies higher or lower than this must be physically louder in order for us to

perceive them to be at the same volume.

At normal conversation levels it’s sixty-four times more difﬁ cult to hear bass frequencies

than it is to hear the mid-range and eighteen times more difﬁ cult to perceive the high range.

Mixing

CHAPTER 19

365

If, however, the volume is increased beyond normal conversation level, the

lower and higher frequencies gradually become (albeit perceivably) louder than

the mid-range. This again is related to when we lived in caves and it played a

part in our survival. It allowed us to understand when something was shouted

rather than spoken – not many people are in the habit of casually telling you

there’s a hungry-looking bear behind you.

In the 1930s, two researchers, Fletcher and Munson from Bell Laboratories,

were the ﬁ rst to accurately measure the uneven response of the ear and so we

refer to the Fletcher Munson contour control curve .

It’s absolutely vital that you keep this contour control in mind while mixing

since it means that the bass energy produced will depend entirely on the mix

volume. For example, if you balance the bass elements at a low monitoring

level, there will be a huge bass increase at higher volumes. Conversely, if you

mix at high levels, there will be too little bass at lower volumes.

Mixing appropriately for all three volumes is something of a trade-off as you’ll

never ﬁ nd the perfect balance for all listening levels, but whenever mixing dance

music you should check that it sounds okay at low and medium volumes but

great at high volume. That is, of course, assuming that clubs are your market –

they don’t play music at low or medium levels ( Figure 19.1 ).

FREQUENCY BANDS

The second step is to become familiar with how frequencies are generally

banded together as these bands play a large role whenever we speak about sub-

bass, bass, mid-range, etc. This isn’t as difﬁ cult as it may initially sound: we’ve

100

20 50 100 500 1000 5000

110

100

10 000 20 000

Intensity (dB)

Frequency (Hz)

FIGURE 19.1

The Fletcher Munson

contour control curve

PART 3

Mixing & Promotion

366

been subjected to these frequencies since we were born; you just need to under-

stand how they are grouped together. To help you decipher this, what follows is

a list of how frequencies are generally banded and described.

Sub-Bass: Under 50 Hz

At frequencies this low, it’s impossible to determine pitch; nonetheless this

range is commonly occupied by the very lowest area of a kick drum and bass

instruments. Notably, most loudspeaker and near-ﬁ eld loudspeaker monitors

cannot reproduce frequencies this low reliably, and in most genres of dance

music all signals this low will be rolled off to prevent loss of volume in hi-ﬁ

systems or to prevent damaging the bass response in club PA systems.

Bass: 50 – 250 Hz

This range is typically adjusted when the bass boost is applied on most home

stereos, and where most of the bass is contained in all dance music mixes. EQ

cuts (or boosts) around this area can add deﬁ nition and presence to the bass

and kick drum.

Mid-Range Muddiness Area: 200 – 800 Hz

This frequency range is the main culprit for mixes that are described as sound-

ing muddy or ill-deﬁ ned. It is often the reason why a mix is fatiguing to the ear.

If too many sounds are dominating in this area, a track can quickly become

tiring and irritating.

True Mid-Range: 800 – 5000 Hz

As previously touched upon, the true mid-range is where loudspeakers produce

most of their energy; human hearing is particularly sensitive to these frequen-

cies. This is because the human voice is centred here, as are TVs and radios;

thus, even small boosts of a decibel will be perceived to be the same as boost-

ing 10 dB at any other frequency. Subsequently, you should exercise extreme

caution when adjusting frequencies in this area.

High Range: 5000 – 8000 Hz

This range is typically adjusted when the treble boost is applied on most home

stereos and is where the ‘body’ of hi-hats and cymbals often reside. This area is

sometimes boosted by a couple of decibels to make sounds artiﬁ cially brighter.

Hi-High Range: 8000 – 20 000 Hz

This ﬁ nal frequency area often contains the higher frequency elements of cym-

bals and hi-hats. Some engineers also apply a small shelving boost at 12 000 Hz

to make the music appear more hi-ﬁ delity. This often adds extra detail and

Mixing

CHAPTER 19

367

sheen without introducing any aural fatigue. It does, however, require plenty of

care as boosting anywhere in this region can intensify any background hiss or

high-frequency noise.

There’s much more to applying EQ to a mix than understanding the separate

frequency bands and to become truly competent at engineering. It’s also vital

to train your hearing so that you can identify timbres and the corresponding

frequency in hertz. Unfortunately, this isn’t something that you can pick up

just from reading about it but will only come with practical experience and

careful listening ( Figure 19.2 ).

Bass

CDEFGAB

CDEFGABCDEFGAB

CDEFGABCDEFGABCDEFGABCDEFGAB CDEFGABCDEFGABCDEFGABCDEFGAB

Kick

Snare

Hi-hat

General percussion

Piano

Strings

Leads

Acoustic guitar

Electric guitar

Woodwind

Organs

Pads

Bass

Mid-range

Treble

Middle C

220 Hz 440 Hz55 Hz27.5 Hz 880 Hz 1.76 KHz 3.52 KHz 7.04 KHz 14.08 KHz

FIGURE 19.2

A chart deﬁ ning the

general ranges taken

up by instruments in

dance music

PART 3

Mixing & Promotion

368

CREATING A SOUNDSTAGE

Finally, you need to be able to envisage the ‘stage’ on which the instruments

will be placed. When approaching any mix it’s usually best to imagine a three-

dimensional room – the soundstage – on which you can place the various

instruments.

Sounds placed on this stage can be positioned anywhere between the left or

right ‘walls ’ using a pan pot; they can be positioned at the front or back, or

anywhere in between, using volume; and the frequency content of the sound

will determine whether it sits at the top of the stage (high frequencies), the

middle (mid-range frequencies) or the bottom (low frequencies).

The concept behind mixing is to ensure that each sound occupies its own

unique space within this room so that it can not only be heard but also ﬁ ts in

well with everything else. To do this we can break the soundstage into three

distinct areas: front to back, horizontal and vertical. We’ll begin by examining

the front to back area.

FRONT–TO–BACK PERSPECTIVE

One of the primary auditory clues we receive about how far away we are from

a sound source is through the intensity of air pressure that reaches our ear

drums. As we touched upon in Chapter 1, sound waves spread spherically out-

wards in all directions from any sound source. The further these have to travel

the less intense the sound becomes. In other words, the further we are from a

source of sound, the more the sound waves will have dissipated, resulting in a

drop of volume.

The intensity of sound is governed by the inverse law which states that:

‘ Sound pressure decreases proportionally to the square of the distance

from the source ’

Roughly translated, this means that each time the distance from the sound

doubles it’ll become roughly 6 dB quieter.

If we interpret this in the context of a mix, we can say that the louder an instru-

ment is the more ‘in your face ’ it will appear to be. However, although many

dance mixes appear to have everything right in your face and at the front of the

soundstage, this is far from the case. Indeed, depth perception is the ﬁ rst aspect

to take into consideration when producing a good mix.

If every instrument were placed at the forefront, all the volumes would be at

equal gain; this would produce a cluttered image as every instrument ﬁ ghts

to be at the front. What’s more, the mix would appear just two-dimensional

because when listening to music our minds always work on the basis of com-

parison. That is, for us to gain some perception of depth there must be some

sounds in the background so we can determine that some sounds are at the

front of the mix and vice versa.

Mixing

CHAPTER 19

369

This means that you need to determine which sounds should be up front and

which should be progressively further back in the mix but this isn’t as difﬁ -

cult as it may sound. Dance music, by its very nature, is based on rhythm and

groove because that’s what we dance to. Thus, in all genres of dance it makes

sense that both the drums and bass should take soundstage priority over every

other instrument.

The most obvious way to provide this depth perspective is with volume adjust-

ment, as it’s natural for us to believe that the louder a signal is the closer it

must be. This only applies to sounds that we are familiar with, though, and

for dance music it’s quite usual to use unnatural or unrealistic timbres. We can

overcome this problem by taking advantage of the fact that how our ears and

mind perceive a sound depends on the acoustic properties it projects.

As touched upon in the very ﬁ rst chapter of this book, the higher the frequency

of a sound the shorter the wavelength becomes; therefore, we can assume that

if a high-frequency sound has to travel over a long distance, many of these

frequencies will dissipate along the way. This means that if a high-frequency

sound is playing a good distance away from a listener, some of the high-

frequency content will be reduced.

This effect can be particularly evident with passing cars whose stereo systems

take up more space than the engine and shift enough air to blow out a candle

at 50 paces. You hear the low frequencies while the car is at a distance. As it

approaches, the higher frequencies become more pronounced until it passes

by, whereby the higher frequencies begin to decay as it moves off into the dis-

tance. We can therefore emulate this effect with some creative EQ. Cutting

just a few decibels at the higher frequency ranges of a sound makes the sound

appear more distant when it is heard in association with other sounds with

low-frequency content. Alternatively, increasing the higher frequencies using

enhancers or exciters can make a timbre appear much more up front.

Notably, this frequency-dependent effect is an important aspect to consider

when working with a compressor since if a fast attack is used the transient will be

clamped down, thus reducing some of the high-frequency content and pushing

it to the rear of the mix. This is often the cause of numerous problems, as most

usually you would like to compress to prevent clipping or to introduce ‘punch’

into a full mix. While you could always increase the gain to bring the timbre

to the front of the mix, it would not work as well as if they all had the high-

frequency content intact. The only suitable way to avoid this is to use a long attack

on the compressor or to employ a multi-band compressor and set it to compress

only the lower frequency content, leaving the higher frequencies unaffected.

Another characteristic of the front-to-back perspective is derived from the

amount of reverberation the signal has associated with it. All sound has some

natural reverberation as the reﬂ ections emanate from surrounding surfaces,

but the amount and stereo width of these reﬂ ections depends on how far away

the source of the sound is. If a sound is far away, then the stereo width of the

PART 3

Mixing & Promotion

370

reverberations will dissipate as they travel through the air, and they will be sub-

jected to more reverberation. This is important to bear in mind since many artists

wash a sound in stereo reverb to push it into the background and then wonder

why it doesn’t sound quite ‘right’ in context with the rest of the mix.

From this we can also determine that if you need to place an instrument at

the rear of a mix, it’s wise to use a mono reverb signal with a long tail and to

remove some of the higher frequencies. This will emulate the natural response

we’ve come to expect from the real world, even if that particular timbre doesn’t

occur in the real world. Most good reverb units will allow you to remove higher

frequencies with an on-board ﬁ lter. If not, by returning the reverb effect into

a normal mixing channel you can use the channels EQ to remove the high

frequencies this way.

Of course, sounds that are up close or at the front of a mix will have little or

no reverb associated with them but in some instances you may wish to apply

reverb to thicken the timbre. On these occasions you should apply it in stereo;

you should also aim to keep the stereo width of the reverb contained to pre-

vent it from occupying too much of the left and right perspective. Also, use a

short tail and a pre-delay of approximately 50 –90 ms to separate the timbre

from the effect to prevent it from washing over the attack stage; otherwise it

may force the instrument to the back of the mix.

Always keep in mind that applying effects too heavily can make sounds very

difﬁ cult to localize and it’s absolutely vital that each instrument be pinpointed

to a speciﬁ c area within a mix. Otherwise it will appear indistinct and muddy.

THE HORIZONTAL PERSPECTIVE

The next consideration in a mix is the horizontal plane – the distance and

positioning of sounds between the left and right walls of the virtual room. The

major audio clue that helps us derive the impression of panning and stereo can

be determined by (a) the volume intensity between sounds or (b) the timing

between sounds .

The principle behind altering the volume of sounds to produce a stereo image

was ﬁ rst realized by Alan Blumlein in the early 1930s. An inventor at EMI’s

Central Research Laboratories, he researched the various ways in which the ears

detect the direction from which a sound comes. Along with deriving the tech-

nique for creating a stereo image in gramophone records, he also ﬁ gured that

to maintain realism in a ﬁ lm, the sound should follow the moving image.

This technique was ﬁ rst employed in Walt Disney’s Fantasia, when sound engi-

neers asked Harvey Fletcher (of the same Fletcher Munson curve) if he could

create the impression of sound moving from left to right for the movie. Drawing

on Alan Blumlein’s previous work, Fletcher came to the conclusion that if

a sound source is gradually faded in volume in one speaker and increased

in the other it will gradually move from one to the other. The engineers at

Disney put this idea into practice by using a potentiometer to vary the volume

Mixing

CHAPTER 19

371

between two speakers and labelled the process ‘panoramic potentiometer ’,

hence the term pan pot.

Although this volume-intensity difference between two speakers is still the

most commonly used method for panning a sound around the image, we

can also receive directional clues from the timing between sounds. Otherwise

known as the directional cues, Precedence or Haas effect, this process takes

advantage of the Law of the First Wavefront, which states that:

‘ If two coherent sound waves are separated in time by intervals of less

than 30 ms, the ﬁ rst signal to reach our ears will provide the directional

information ’ .

In layman’s terms, if a direct sound reaches our ears anywhere up to 30 ms

before the subsequent reﬂ ections, we can determine the position of a sound.

For example, if you were facing the central position of the mix, any sound leav-

ing the left speaker would be delayed in reaching the right ear and vice versa

for the left ear, an effect known as interaural time delay (ITD). Considering that

sound travels at approximately 340 m/s, this effect can be emulated by setting

up a delay unit on a mono signal and delaying it by a couple of milliseconds,

producing the impression that the sound has been panned.

For this effect to work accurately, we also need to consider that our ears are

on the side of our head and therefore our head gets in the way of the frequen-

cies from opposite speakers (an effect known as head-related transfer func-

tion). This means that, provided we are facing the centre of the stereo image,

some of the higher frequencies emanating from the left speaker will be reduced

before they reach the right ear, simply because sound doesn’t travel through

our heads, it has to travel around it. This effect can be simulated by cutting a

decibel at around 8 kHz from the delayed signal.

Naturally, to accomplish these types of effects you need to use mono sounds

within the mix. Here is where many practicing musicians make one of their

biggest mistakes. With all of today’s tone modules, keyboards and samplers fea-

turing stereo outputs, it’s quite easy to fall into the trap of using stereo sounds

throughout an arrangement, but this only leads to a mix that lacks any real

deﬁ nition.

A stereo image is formed by spreading any sound to both the left and right

speaker using one of the two methods: effects or layering. The results are then

panned left and right so that they play across both speakers. While this always

makes them sound much more interesting in isolation, it’s only to persuade

you to part with your money for the synthesizer.

If you made a mix completely from stereo ﬁ les, they would collate in the mix

and all occupy the same area of the soundstage. Of course, you could narrow

the stereo width of each ﬁ le with the pan pots so that they don’t all occupy the

same position, but this approach is not particularly suitable for making a

transparent mix.

PART 3

Mixing & Promotion

372

The soundstage for any mix should be transparent enough that you can pic-

ture the mix in three dimensions with your mind’s eye and pinpoint the exact

position of each instrument. Many dance mixes are quite busy, encompassing

anything from 8 to 12 different elements playing at once. If these were all ste-

reo it would be difﬁ cult, if not impossible, to ﬁ nd a pan placement for each.

This often leads to an inexperienced engineer resorting to unnecessary EQ in a

futile attempt to carve out some space for the instruments or start to ‘creep the

faders’.

Creeping faders is typical of most inexperienced engineers and is the result

of gradually increasing the volume of each track so that it can be heard above

other instruments. For instance, they may increase the gain of vocals so that

they can be heard above the bass, and then increase the drums so they can

be heard above the vocals, and then increase the bass so it can be heard with

the drums and then move back to the vocals and, well, you get the picture.

Eventually, the mixer is pushed to its maximum headroom and the mix turns

into loud, incomprehensible rubbish.

Unnecessary EQ and volume creep can be avoided by utilizing mono sound

sources, since if two instruments share the same frequency range they can be

separated by panning one sound to the left and the other to the right. There’s

little need to spatially separate them by panning them as far as possible in each

direction and, in many cases, a panning space of approximately ‘two hours ’

will allow both instruments to be heard clearly. By ‘hours’ we mean the physi-

cal position of a pan pot on a mixing (i.e. 1 o’clock, 2 o’clock, etc.) Thus, for

a 2-hour panning the right pan would read 1 o’clock and the left would read

11 o’clock.

This physical pan pot measurement is only mentioned in a theoretical sense.

You should always pan with your ears and not your eyes. All mixing desks are

not equal and where one may place the sound at a certain position at a certain

setting, the same setting on a different mixer may place the sound in a different

position.

There is yet another reason why you should avoid using stereo ﬁ les throughout

and this is based upon our perception of stereo. Just as different volumes aid

us in gaining some perception of depth within a mix, the same is true of stereo.

It’s in our nature to turn and face the loudest, most energetic part of any mix,

which in many mixes is the kick drum. If this happened to be a stereo ﬁ le,

it would be spread between the left and right speakers and the main energy

would be spread across the soundstage. If, however, mono samples were used

and the kick drums were placed centrally, we would perceive the position of

other sounds much more clearly, which would increase our overall sense of

space within the mix.

On top of this, when working with mono ﬁ les on a stereo system we also per-

ceive the volume of a sound by its position in the soundstage. This means that

if a signal is placed dead centre, the energy is shared by both speakers and it

Mixing

CHAPTER 19

373

will appear louder than if the same signal, at the same gain, were placed in the

left or right speaker alone.

Typically, this perceived volume difference can be as much as 3 –6 dB, so many

professional and semi-professional mixing desks (and some audio sequencers)

will implement the panning law. According to the panning law, any sounds

that are placed centrally are subjected to either 3 or 6 dB of attenuation, which

can be set by the user. It should be noted that not all mixing desks will imple-

ment the panning law, so after panning the instruments to their respective

positions you may need to readjust the respective volumes again.

Perhaps, most important of all, if you wish to pan instruments faithfully, your

speaker system must be conﬁ gured correctly in relation to your monitoring posi-

tion. Ideally, you should be positioned at one point of an equilateral triangle with

the loudspeakers positioned at the other two points. This means that speakers

should be positioned an equal distance apart from one another and your listening

position to ensure that the signal from each reaches your ears at the same time.

If there is just a few inches difference between these three points, the sound

from one speaker could be delayed in reaching your ears by a couple of mil-

liseconds which results in the stereo image moving to the left or right. What’s

more, you also need to ensure that the volume from each speaker is the same

when it reaches your current listening position – even differences as small as 1

or 2 dB can shift the image considerably to one side.

A possible solution to producing an accurate soundstage would be to mix

through headphones, but this should be avoided at all costs. While they are suit-

able for listening to a mix without disturbing the neighbours, they overstate the

left and right perspective because they’re positioned on either side of your head.

Keep in mind that when listening to a mix from loudspeakers, sound from the

left speaker reaches your right ear and vice versa for the left speaker. If you mix

using headphones alone, you can easily over- or understate the stereo ﬁ eld.

Finally, any stereo mix must also work well in mono. Most mixing desks will

offer a mono button that, when used, sums the channels together into a mono

signal. This prevents any phasing that, while not immediately evident in ste-

reo playback, can result in a comb ﬁ ltering effect. While this may seem worth-

less when all of today’s hi-ﬁ systems are stereo, most TV and radio stations still

broadcast in mono, most clock/transistor radios are also mono, and so are

many club PA systems.

This means that if you’ve relied entirely on stereo while mixing, then parts of

the mix can disappear altogether. This doesn’t mean you have to destroy the

stereo mix. When you switch the mix to mono, the image will simply be col-

lated into the centre. When in mono, the general tone of the instrument and

their volumes should remain constant; if not, you will need to re-examine the

mix and check for what is causing mono compatibility problems. In the major-

ity of cases this is caused by using stereo ﬁ les or the application of too many

stereo effects.