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

Подождите немного. Документ загружается.

PART 2

Dance Genres

334

curves outwards as it falls, creating a ‘boomier’ sound. Additionally, it’s often

worth sending the kick to a reverb unit set to a medium pre-delay and short

tail to help lengthen the timbre and prevent it from dropping off too sharply.

If you decide to use a kick sample or a GM module and there is too much of

an attack on the sound, a compressor set so that it clamps down on the tran-

sient may produce the sound you require. That said, in the long run a preferred

option would be to use a transient designer such as featured in the Waves

Transform Bundle plug-in or SPL’s hardware transient designer. These can be used

to remove the front of the kick to produce a more relaxed-sounding timbre.

In direct contrast to the kicks, the snares often use a sharp transient stage, as

these dictate the rhythmic ﬂ ow of the loop and need to be apparent in the loop –

more so if it’s a particularly busy loop packed with auxiliary instruments. The

typical snare can be created by using a triangle oscillator mixed with pink

noise. A high-pass ﬁ lter across these generally produces the most common chill

out snares, but it’s also worth trying a notch or band-pass ﬁ lter to see which

you prefer. The amps attack should be set fast while the decay can be set any-

where from medium to long depending on the sounds that are already in the

mix and how slurry you want the snare to appear.

Alternatively, if it’s possible with the chosen synthesizer, it’s prudent to employ

a different amp EG for both the noise and the triangle wave. The triangle wave

can be kept quite short and swift by using a fast decay, while the noise can be

made to ring a little further by increasing its decay parameter. The further this

rings out, the more slurred and ‘ambient’ it will become. If the snares are too

bright, brightness can be removed with some EQ or, preferably, try replacing

the triangle wave with a pulse and experiment with the pulse width.

The snares also often beneﬁ t from small amounts of reverb by sending, not

inserting, them to a reverb unit. A typical snare room preset that’s available on

nearly every reverb unit around will usually sufﬁ ce, but depending on the loop

it may be worth reducing the decay slightly. Again, if samples are used in place

of synthesis, the Jazz snares on most GM modules produce a fairly good rendi-

tion which may be more suited to your track. Using a transient designer along

with sending the snare to a reverb unit will often help you achieve the typical

‘ chilled’ feel.

Some chill out will beneﬁ t from a snare with a less prominent attack phase

which can be accomplished by using a transient designer to remove the attack

phase. On top of this, it may also be worthwhile introducing a small amount

of pink noise over the top of the snare to give the impression of a brush stick

used to play. Alternatively, you could just use a brush kit from the GM standard

or sample one from a Jazz record. For obvious legal reasons I can’t condone

this latter approach but that’s not to say that a proportionate amount of dance

artists don’t do it.

The hi-hats can be created in two ways: with noise or with ring modulation.

Typically for chill out, ring modulation produces better results. This is accomplished

Ambient/Chill Out

CHAPTER 17

335

by modulating a high-pitched triangle wave with a lower pitched triangle to

produce a high-frequency noise. As a starting point, set the amp envelope to

a zero attack, sustain and release with a short-to-medium decay. If there isn’t

enough of a noise present, it can be augmented with a white noise waveform

using the same envelope. Once this basic timbre is constructed, shortening the

decay creates a closed hi-hat, while lengthening it will produce an open hat.

As touched upon, not all chill out will use MIDI programmed loops and some

artists will borrow loops or at least elements of loops from previous records.

For obvious legal reasons I can’t condone this behaviour but nevertheless it is a

popular technique. These sampled loops are often imported into a sample slic-

ing programme and messed around until it sounds nothing like the original.

Of course, you shouldn’t just settle with rearranging parts. After they’re remod-

elled, it’s worthwhile importing them into a wave editor and using a transient

designer to model the transients and releases of the drum hits. Alternatively,

another approach is to use two sampled loops and mix them together in a

sequencer. Time-stretch the loops so they share the same tempo, and play the

loops simultaneously, and any conﬂ icting elements can be cut-out from one of

the loops ( Figure 17.3 ).

This technique is useful if you wish to create particular poly-rhythms that are

difﬁ cult to accomplish any other way. On top of this, it’s worth experimenting

by moving one of the loops back in time to make some of the elements of the

loop occur in different places.

FIGURE 17.3

Lining up a series of

loops and removing

element to create a

basic rhythm

PART 2

Dance Genres

336

With the loop laid down, you can then look towards effecting and processing

the rhythm, although roughly speaking, chill-out loops are not compressed for

numerous reasons, including:

■ They’ve already been sampled from a record and are already compressed.

■ They are from a synth and are already compressed.

■ They don’t need to ‘pump’ musically.

■ Second-order harmonic distortion from valve compression isn’t usually

required.

However, if the loop has been composed of numerous elements from differ-

ent sources, compression may help to bring some of the levels under control.

Typically, as a starting point, set a relatively slow attack with a medium release

and a ratio of about 4:1. Then adjust the threshold so that the loop registers

approximately 4 or 5 dB on the gain reduction meter. It’s usually wise to ensure

that the loop doesn’t pump, so always experiment with the compressor’s

release to make its action as transparent as possible.

Although in many genres it’s usually worth avoiding applying any effects to

drum timbres, chill out is often an exception to the rule. While it’s unwise to

wash the entire loop in cavernous reverb, if it’s lightly applied to the snares and

kicks, it can often present a more laid-back feel to the timbres. That said, any

reverb or any effect for that matter must be applied cautiously. The drums form

the backbone of the record, and if an effect is applied too heavily it will often

destroy the transients, reducing the bottom-end groove to mush.

MELODIC LEADS

Although it isn’t unusual for a chill out track to be based entirely around syn-

thetic instruments, most do tend to contain real instruments. These can range

from acoustic guitars playing Spanish rhythms, Classical strings, pianos, wind

or a combination thereof. Roughly speaking, these are best laid down before

the bass, as much of the attention of the track is commonly directed towards

the lead and the bass simply acts as the underpinning. Since most real instru-

ment leads are commonly sampled from sample CDs or other records, it’s

much easier to form the bass around their progression, rather than to try to

re-programme a lead to sit around a bass. Although it is possible to cut up and

rearrange a lead in a sample splicing programme to suit a pre-programmed

bass, this can often result in the lead behaving unnaturally, so it’s prudent to

rearrange the bass instead.

Samples of real instruments do not necessarily have to be used (although in

the majority they do sound better when used for leads), and provided that

you have a good tone module it’s possible to programme realistic instruments

through MIDI. If you take this approach, however, it’s important to take note

of the key and the range you write the song in because every acoustic instru-

ment has a limit to how high and low it can play. For instance, with an acoustic

guitar the (standard) tuning is E, A, D, G, B, E with this latter E a major third

Ambient/Chill Out

CHAPTER 17

337

above middle C. This is incredibly important to keep in mind while program-

ming since if you exceed the limitations of an acoustic instrument the human

ear will instinctively know that it’s been programmed rather than played.

ACOUSTIC GUITARS

The key to programming acoustic guitars is to take note of how they’re physi-

cally played and then (provided the tone module produces an authentic

sound) emulate this with MIDI. Firstly, you need to take into account the way

that a guitarist’s hand moves while playing. If the style is Hispanic, then the

strings are commonly plucked rather than strummed; so the velocity of each

note will be relatively high throughout since this will be mapped to control the

cut-off (i.e. the harder a string is plucked, the brighter the timbre becomes).

Not all of these velocity plucks will be the same, however, since the performer

will want to play in time with the track, and if a note occurs directly after the

ﬁ rst, it will take a ﬁ nite amount of time to move the ﬁ ngers and pluck the

next string. This often results in the string not being accentuated as much as

the preceding one due to ‘time restraints ’. Conversely, if there is a larger dis-

tance between the notes, then there is higher likelihood that the string will be

plucked at a higher velocity.

In many instances, particularly if the string has been plucked hard, the reso-

nance may still be dying away as the next note starts, so this needs to be taken

into account when programming the MIDI. Similarly, it’s also worth bearing

in mind that a typical acoustic guitar will not bend more than an octave, so it’s

prudent to set the receiving MIDI device to an octave so you can use the pitch

bend wheel to create slides. In between notes that are very close together, it

may also be worth adding the occasional fret squeal for added realism.

If the guitar is strummed, then you need to take the action of the hand into

account. Commonly, a guitarist will begin by strumming downwards rather

than upwards, and if the rhythm is quite fast on an upwards stroke it’s rare that

all the strings will be hit. Indeed, it’s quite unusual for the bottom string to be

struck, as this tends to make the guitar sound too ‘thick’, so this will need to be

emulated while programming.

Additionally, all the strings will not be struck at exactly the same time due to

the strumming action. Obviously, this means that each note on message will

occur a few ticks later than the previous, which will depend on whether it’s

strummed upwards or downwards and the speed of the rhythm. To keep time,

guitarists also tend to continue moving their hands upwards and downwards

when not playing; so this ‘rhythm ’ will need to be employed if there is a break

in the strumming to allow it to return at the ‘right’ position in the arrangement.

Finally, and somewhat strangely, if guitarists start on a downward stroke they

tend to come in a little earlier than the beat, while if they start on an upward

stroke they tend to start on the beat. The reason behind this is too complex to

explain plus it would also involve me knowing why.

PART 2

Dance Genres

338

If you want to go even further and replicate trills or tremolo, it’s worthwhile

recording pitch bend movements onto a separate MIDI track and imposing it

onto the longer notes of the original guitar track by setting it to the same MIDI

channel. These pitch bend movements are recorded onto a separate channel.

Since once it’s imposed onto the notes it’s likely that the pitch bend informa-

tion will need further editing.

WIND INSTRUMENTS

Although fundamentally there are two types of wind instruments – brass and

reed—in the concept of MIDI programming they’re both programmed follow-

ing similar principles since they both rely on variations in air pressure to pro-

duce the timbre. As a result, if you’re planning on making music that contains

wind instruments, it’s prudent to invest in a breath controller. These are small

devices that connect directly to the MIDI IN port and act similar to any wind

instrument, albeit they do not produce any sound of their own. Rather they

measure changes in air pressure as you blow into them. They convert this into

CC2 (breath controller) messages that can then be used to control the con-

nected MIDI synthesizer or sampler. These can be expensive, though, so if the

use of wind instruments is only occasional it’s worth programming the typical

nuances by hand into the sequencer.

Firstly, the volume and brightness of the notes are proportional to the amount

of air pressure in the instrument. All good MIDI instruments will set the veloc-

ity to control the ﬁ lter cut-off of the instrument, so the brightness can be con-

trolled with judicious use of MIDI velocity. The volume of the note, however,

is a little more complicated to emulate since many reed instrumentalists will

deliberately begin most notes by blowing softly before increasing the pressure

to force the instrument to become louder. Essentially, this means that the notes

begin softly before quickly rising in volume, and occasionally pitch. The best

way to emulate this feel is to carefully programme a series of breath controller

(CC2) messages while using a mix of expression controllers (CC11) to control

the volume adjustments. Alternatively, brass instruments will often start below

the required pitch and slide up to it; this is best emulated by recording live

pitch bend movements into a sequencer and editing them to suit later.

On top of this, many wind instruments also introduce vibrato if the note is

held for a prolonged period of time due to the variations in air pressure

through the instrument. While it is possible to emulate this response with

expression (CC11) controllers, generally speaking, introducing small pitch

spikes at the later stages of the sustain can produce better results. It should be

noted that these pitch ‘spikes’ only appear in the later stages of the notes sus-

tain, though, and should not be introduced at the beginning of a note. Possibly

the best way to determine where these spikes should appear is to physically

emulate playing a wind instrument by beginning to blow at the start of the

MIDI note on and when you begin to draw short of breath, and insert some

pitch bend messages. Alternatively, if the synth allows you to fade in the LFO,

Ambient/Chill Out

CHAPTER 17

339

you can use this to modulate the volume by setting it to a sine wave on a slow

rate, modulating the volume lightly. As long as the LFO fade-in time is set quite

long, it will only begin to appear towards the end of a note.

On the subject of breathing, bear in mind that all musicians are human and

as such need to breathe occasionally. In the context of MIDI this means that

you should avoid playing a single note over a large number of bars. Similarly,

if a series of notes are played consecutively, remember that the musician needs

enough time to take a deep breath for the next note (wind instruments are not

polyphonic!). If there isn’t enough time, the next note will generally be played

softer due to less air velocity from a short breath, but if there is too short a

space the instrument will sound emulated rather than real.

Finally, you need to consider how the notes will end. Neither reed nor brass

instruments will simply stop at the end of the note; instead they will fade

down in volume while also lowering in pitch as the air velocity reduces

(an effect known musically as diminuendo). This can be emulated with a series

of expression (CC11) messages and some pitch bend.

As touched upon, some tracks do not employ real instruments and rely solely

on synthetic. If this is the case it’s often worth avoiding any sharp aggressive

synthesizer patches such as distorted saw-based leads since these give the

impression of a ‘cutting’ track, whereas softer sounds will tend to sound more

laid back. This is an important aspect to bear in mind, especially when master-

ing the track after it’s complete. Slow, relaxed songs will invariably have the

mid-range cut to emphasize the low and high end, while more aggressive songs

with have the bass and high hats cut to produce more mid-range and make it

appear insistent.

In MIDI programming, if synthetic timbres are used, then many tracks tend to

refrain from using notes less than 1/4th in length since shorter notes will make

the track appear faster. Additionally, with longer notes many of the tones can

utilize a long attack, which will invariably create a perception of more relaxed

music. In fact, long attack times can form an important aspect of the music.

It’s good to experiment with all the timbres used in the creation of this genre

by lengthening the amps/ﬁ lters attack and release times and taking note of the

effect it has on the track.

Vocal chants are sometimes used in the place of leads as these can be used as

instruments in themselves. Typically, in many popular tracks these are sampled

from other records or sample CDs, but it’s sometimes worthwhile recording your

own. Characteristically, condenser microphones are the preferred choice over

dynamic as these produce more accurate results, but the diaphragm size will

depend on the effect you wish to achieve. As touched upon in the chapter on

recording real instruments, many producers will use a large diaphragm for vocals,

but if you’re after ‘ghostly’ chants, a small diaphragm mic such as the Rode NT1

will often produce better results due to the more precise frequency response.

Of course, this is simply the conventional approach and, if at all possible,

PART 2

Dance Genres

340

it’s worthwhile experimenting with different microphones to see which pro-

duces the best results.

Vocals will also beneﬁ t from compression during the recording stage to prevent

any clipping in the recording, but this must be applied lightly. Unlike most

other genres where the vocals are often squashed to death to suit the com-

pressed nature of the music, chill out relies on a live feel with the high frequen-

cies intact. A good starting point is to set a threshold of ⫺ 9 dB with a 2:1 ratio

and a fast attack and moderately fast release. Then, once the vocalist has started

to practice, reduce the threshold so that the reduction meters are only lit on the

strongest part of the performance.

Similar to the drum rhythms, effects can also play a large part in attaining the

right lead sound and you should feel free to experiment. Reverb is the main

suspect here, but phasers and ﬂ angers can also work especially well if they are

lightly applied. A good (if often overused) technique to produce ghostly vocal

chants is to reverse the vocal sample, apply a large cavernous reverb and then

reverse them back the right way around again. This way, the reverb tail-off will

act as a build-up to the vocal chant. The key is, as always, to experiment.

BASS

With the lead laid down, the bass is often programmed next to form the basic

groove of the song. Generally speaking, this tends to be quite minimal, consist-

ing mostly of notes over 1/8th in length to prevent the groove from appear-

ing too fast. These are often derived from the principles encapsulated in the

dub scene, staying quite minimal to allow space for the lead to breathe without

having to compete with the bass for prominence. Indeed, it’s incredibly impor-

tant that the bass is not too busy either rhythmically or melodically and many

chill out tracks borrow heavily from the pentatonic scale used in dub and R

‘ n’ B. That is, they use no more than a ﬁ ve-note scale playing a single harmony

restricting the movements to a maximum ﬁ ve-semitone shift.

Naturally, the bass should interact with the lead and, as touched upon in the

chapter on music theory, this can be accomplished through using parallel,

oblique or contrary motions. In many cases, an oblique motion is the preferred

option since a majority of this music is driven by the melody and you don’t

want the bass to detract from it. Having said that, you should feel free to try

out both contrary and parallel motions to see which produces the best results.

What is important, however, is that the bass coincides with the drum rhythm.

Bear in mind that if the drums have been affected with reverb, the subsequent

tail-off between each hit will leave less room for the bass, so it will need to be less

melodic or rhythmic. Conversely, if the drums are left dry, it’s possible to employ

a more rhythmic or melodic bass without creating any conﬂ icts ( Figure 17.4 ).

Typically, most chill out basses are synthesized rather than real since the timbre

is often quite deep and doesn’t want to attract as much attention as the lead.

Analogue synthesizers usually provide the best results due to the uncontrollable

Ambient/Chill Out

CHAPTER 17

341

yet pleasing phasing of the oscillators. Nonetheless, the beginnings of a typical

timbre can be programmed on most synthesizers. A good starting point is to

use a sine wave with a sawtooth transposed up from it by 5 or 7 cents. Unlike

most of the timbres used throughout the production of chill out, the attack

should be well deﬁ ned to prevent the bottom-end groove turning to mush, but

there is usually very little decay, if any, to prevent it from becoming too plucky.

Indeed, in much of the music of this genre, the bass tends to hum rather than

pluck to prevent it from drawing attention away from the chilled feeling. The

best way to accomplish this type of sound is to use a simple on/off envelope

for the amp envelope. This is basically an envelope with no attack, decay or

release but a quite high sustain. This forces the sound to jump directly into

the sustain stage which produces a constant bass tone for as long as the key

is depressed. If this results in a bass with little or no sonic deﬁ nition, a small

pluck can be added by using a low-pass ﬁ lter with a low cut-off and high res-

onance that’s controlled with an envelope using a zero attack, sustain and

release but a medium decay. By adjusting the depth of the envelope modulat-

ing the ﬁ lters or increasing/decreasing the decay stage, more or less of a pluck

can be applied. If you use this process, however, it’s prudent to employ ﬁ lter

key tracking so that the ﬁ lter action follows the pitch.

At this stage, it’s also prudent to experiment with the release parameters on the

ﬁ lter and amp envelope to help the bass sit comfortably in the drum loop. In

fact, it’s essential to accomplish this rhythmic and tonal interaction between

the drum loop and the bass by playing around with both the amp and ﬁ lters

A/R EG before moving on. This is the underpinning of the entire track and it

needs to be right for the style of music.

Since these waveforms are produced from synthesizers, there is little need to

compress the results – they’re already compressed at the source. Also, it’s unwise

to compress the drums against the bass to produce the archetypal ‘dance music ’

pumping effect, as this will only accentuate the drum rhythm and can make

FIGURE 17.4

A typical chill out bass

PART 2

Dance Genres

342

the music appear too ‘tight’ rather than relaxed. Similarly, effects are usually

avoided on the bass timbre because they can not only spread the timbre across

the stereo spectrum but they can also bring too much attention to the bass.

Of course, that said, the entire track may be based around the rhythmic move-

ments of the bass. If this is the case you should feel free to experiment with

every effect you can lay your hands on!

CHORDS/PADS

Slow-evolving pads can play a vital role in a proportionate amount of chill out.

Since many of the tracks are thematically simple, a pad can be used to ﬁ ll the

gap between the groove of the record and the lead and/or vocals. What’s more,

slow-evolving strings often add to the overall atmosphere of the music and can

often be used to dictate the drive behind the track.

We’ve already covered the principles behind creating chord structures in the

chapter on compression, processing and effects, but as a quick refresher, the

chords here will act as a harmony to the bass and lead. This means that they

should ﬁ t in between the rhythmic interplay of the instruments so far with-

out actually drawing too much attention (rather like the genre as a whole). For

this, they need to be very closely related to the key of the track and not use a

progression that is particularly dissonant. Generally, this can be accomplished

by forming a chord progression from any notes used in the bass and experi-

menting with a progression that works. For instance, if the bass is in E, then C

major would produce a good harmony because this contains an E ( C–E–G ).

The real solution is to experiment with different chords and progressions until

you come up with something that works.

Once the progression is down, it can then be used to add some drive to the track.

Although many chill out tracks will meander along like a Sunday stroll, if you

place a chord that plays consistently over a large number of bars, all of the feel

can quickly vanish. This can sometimes be avoided by moving the chords back in

time by a small amount so that they occur a little later than the bar. Alternatively,

if they follow a faster progression, they can be moved forward in time to add some

push to the music. It should be noted here, however, that if the pads employ a

long attack stage, they may not actually become evident until much later in the bar,

which can destroy the feel of the music. In this instance, you will need to counter

this effect by moving the chords so that they occur much earlier ( Figure 17.5 ).

The instruments used to create these are more often than not analogue in

nature due to the constant phase discrepancies of the oscillators, which cre-

ates additional movement. Thus, analogue or analogue emulations will invari-

ably produce much better results than an FM synth or one based around S &S.

What’s more, it’s of vital importance that you take into account the current fre-

quencies used in the mix so far.

With the drums, bass, lead and possibly guide vocals playing along, there will

be a limited frequency range where you can ﬁ t the chords in. If the timbre used

Ambient/Chill Out

CHAPTER 17

343

is too harmonically rich it will be incredibly difﬁ cult to ﬁ t them into the mix

without having to resort to aggressive EQ cuts. This will change the entire char-

acter of the chords and in some instances may make it inappropriate for the

music; so it’s much better to programme the chords timbre to ﬁ t in now, rather

than when approaching the mix. Ideally, you should look towards producing a

‘ mix’ that sounds right during this construction stage and not rely on the mix-

ing desks EQ later to cure a host of problems. Keep in mind that the EQ on

a mixing desk is very subtle and usually kept aside for subtle tonal changes.

This obviously means that there are two programming possibilities for the

chords. They either need to be quite lush to ﬁ ll out any noticeable gaps in the

mix or they need to be quite thin so they do not collide with the rest of the

instrumentation.

If the mix is particularly busy in terms of frequencies, then it’s prudent to build

a relatively thin pad that can then be ‘thickened’ out later, if required, with

phasers, ﬂ angers, and reverb or chorus effects. This is accomplished by using

pulse waves, as they have less harmonic content than saws and triangles and do

not have the ‘weight ’ of a sine wave (plus, of course, they sound more interest-

ing!). Generally only one pulse wave is required with the amp envelope set to

a medium attack, sustain and release but a fast decay. If this timbre is to sit in

the upper midrange of the music, then it’s best to use a 12 dB high-pass ﬁ lter to

remove the bottom end of the pulse; otherwise use a low-pass ﬁ lter to remove

the top end and then experiment with the resonance until it produces a general

static tone that suits the track. Following this, set an LFO to positive modu-

lation with a slow rate on the pulse width of the oscillator and the ﬁ lter to

add some movement. Which waveform you use for the LFO will depend on the

track itself, but by ﬁ rst sculpting the static tone to ﬁ t into the track you’ll have a

FIGURE 17.5

Chill out chord

structure