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

Подождите немного. Документ загружается.

PART 1

Technology and Theory

184

All notes that are played are recorded direct onto the piano roll editor, whereby

they can then be edited within the sequencer. For instance, badly played notes

or poor timing can be corrected and then transmitted back to the synthesizer to

recreate a perfect performance. Remember, however, that the quantize value is

set too low, everything you record will be moved onto the next pulse and there-

fore you could end up losing the human feel you’re trying to inject!

For many music projects one synthesizer will rarely be enough and you may

wish to add samplers, drum machines and further sound-generating devices to

the sequencer. This can be accomplished in one of two ways, utilize an instru-

ments THRU port or purchase a multi-MIDI interface.

In addition to the IN and OUT MIDI ports, some of the more substantial

synthesizer also features a THRU port. With this, information received at the

IN port of the synthesizer can be transmitted back out (i.e. repeated) at the

device’s THRU port to the IN connector of the next device. Connecting syn-

thesizer in this manner is often referred to as a ‘daisy chain ’ and allows much

more elaborate set-ups ( Figure 9.4 ).

Using this method it is possible to connect, for example, a synthesizer, a drum

machine and a sampler and have all these instruments playing together to rec-

reate a performance. There are, however, a few problems with daisy chaining a

number of synthesizers together.

Firstly, any message transmitted from a sequencer takes a ﬁ nite amount of time

to travel down the MIDI cable to the synthesizer. While this is certainly not

noticeable on the ﬁ rst few synthesizers, if the information from the sequencer

is destined for the ninth synthesizer down the daisy chain, it has to travel

through the eight previous synthesizer ﬁ rst which can result in the information

arriving 8 ms late resulting in a delay, an effect otherwise known as latency .

While 8 ms may not appear to be too drastic on reﬂ ection, the human ear is

capable of detecting incredibly subtle rhythmic variations in music and even

slight timing inconsistencies can severely change the entire feel of the music.

Indeed, a timing difference of only 8 ms can destroy the strict rhythms that

Program Select

Group - Pads

14 - Breeze

15 - Killer Waves

16 - True Blue

17 - The Sunami

18 - Ever lasting

Windows has experienced a

serious error and needs to

shut down.

SHUT DOWN

MIDI OUT

MIDI IN

MIDI OUT

1 a101 Trance Lead

OS3.1Loading...

MIDI THRU

FIGURE 9.4

A more elaborate MIDI

set-up

Sequencers

CHAPTER 9

185

are required in some dance music genres and becomes especially evident with

complex break beat or drum and bass loops.

The second problem arises when we consider how does each synthesizer know

which information is destined for it. Simply sending an instruction such as a

pitch and note-on/off message from the sequencer would result in every syn-

thesizer in the chain playing that particular note. This latter problem can be

solved through the use of MIDI channels.

All of today’s sequencers can transmit MIDI information on different channels.

Much like the different channels on a television, MIDI signals can be transmit-

ted to a synthesizer on a number of different MIDI channels. The foremost rea-

son for this is because many synthesizers are multitimbral, in other words, they

can play back a number of individual instruments simultaneously. Typically,

most of today’s synthesizers are 16-part multitimbral but some of the more

recent are 32-part multitimbral. This allows the sequencer to transmit, say, a

piano part on channel 1 and a bass part on channel 2 and – provided the syn-

thesizer was set up correctly – it would play back the piano part in a piano

sound and the bass part in a bass sound.

This channel information, however, can also be used to communicate with spe-

ciﬁ c synthesizer within a daisy chain. By specifying in the synthesizer to ignore

(or pass through) channel 5, for example, any signal reaching that synthesizer

on channel 5 would be passed to the MIDI THRU port and into the next syn-

thesizer in the chain.

While this does provide a solution for connecting numerous MIDI devices

together, it does come at the expense of losing a number of channels in each

device so that the information can pass through it. It also doesn’t solve the

problem of possible latency further down the daisy chain. The only solution to

circumvent these problems is to employ a multi-MIDI output device.

Multi-MIDI interfaces are most commonly external hardware interfaces con-

nected to the sequencer, and offer a number of separate MIDI IN and OUT

ports. Using these you can use different MIDI outputs to feed each different

device rather than having to daisy chain them together. Notably, though, when

looking to purchase a multi-MIDI interface, you should ensure that it utilizes

multiple busses and does not operate on a single buss.

A single-buss interface will offer a number of MIDI outputs but they are all

connected to one MIDI buss. This means that the 16 channels will be divided

between the available outputs on the MIDI interface. In other words, these

basically act as a substitute for devices that do not feature a MIDI THRU port.

Conversely, a multi-buss MIDI interface will provide 16 MIDI channels per

MIDI output on the interface. Therefore, if a multi-buss interface offers four

MIDI OUT channels, you’ll have 64 MIDI channels to transmit over. Many of

these multi-buss interfaces also feature a number of MIDI inputs too, saving

you from having to change over cables if you want to record information from

different synthesizers ( Figure 9.5 ).

PART 1

Technology and Theory

186

FIGURE 9.5

A multi-MIDI set-up

Sequencers

CHAPTER 9

187

Of course, having multiple MIDI inputs begs the questions as to why would

you want to record into a sequencer from numerous synthesizer. After all,

if the sequencer only records simple data such as note-on/off and not the

actual sounds, and the MIDI data received from all synthesizer keyboards is

essentially the same, why not just one synthesizer as the main keyboard for

recording?

If MIDI was only capable of simple pitch, note-on and note-off messages,

anything performed via MIDI would sound incredibly dull and metronomic.

While some dance music protagonists will argue that there is no human ele-

ment behind any electronic dance music – it all sounds as though a machine

has created them – this couldn’t be further from the truth. As already touched

upon, under close scrutiny you’ll ﬁ nd that even the most metronomic sound-

ing techno is crammed full of tiny sonic nuances to prevent it from becoming

tedious. Indeed, this is the very reason why some songs appear to ooze energy

and soul while others seem motionless and drab.

To give any music energy, drive and that indeﬁ nable character it requires the

human element. Every ‘real’ musician will perform with expression by con-

sciously or subconsciously using a series of different playing styles. For

instance, a pianist may hit some notes harder than others; a guitarist may slide

their ﬁ ngers up their guitar’s fret board bending the sound, or may sustain

some notes while cutting others short. All of these playing nuances contribute

to the human element behind the music and without it music would sound

dull and lifeless.

Indeed, much of the energy and ﬂ ow of any music can be attributed to slight

or aggressive development of the sounds throughout the length of an arrange-

ment. While any ‘real’ musician playing the piece naturally injects this into

a performance, capturing this feel over MIDI requires you to be able to send

more than simple on/off messages through MIDI.

With MIDI, sonic nuances can be either programmed or recorded live using a

series of MIDI CC messages. As an example, when you play a note and move

the pitch bend wheel, the wheel sends a continual stream of messages to the

synthesizer’s engine, which in turn bends the pitch either upwards or down-

wards. This message consists of a number informing the synthesizer what

parameter is being adjusted, followed by another number informing it by how

much it is being adjusted. This stream of messages can also be transmitted to

the MIDI OUT and recorded into the sequencer as a series of MIDI control

change messages (abbreviated to CC).

Of course, unless you happen to have three hands – two to play the instrument

and one to move a controller on the facia of the instrument – it is much easier

to record a performance and then create another MIDI track to then record any

movements of the controller. While you can always record the movements onto

the same track as the MIDI notes, it is a better practice to record them on a dif-

ferent track since it allows you to mute the CC movements without having to

mute the MIDI notes too.

PART 1

Technology and Theory

188

Once recorded, these CC messages can be edited within the sequencer to

perfect the performance if required. This form of ‘automation’ can easily be

accomplished in most sequencers using the sequencers toolbox. You can draw

over the CC messages with a pencil, or straighten the movements into a gentle

slope using line tools. Generally speaking, all sequencers will offer a graphical

interface that permits you to do this.

CC MESSAGES

All synthesizer will (or should!) feature a pitch bend wheel but what if you

want to control an element of the synthesizer that has no real-time control on

its facia? There are two solutions to this. The ﬁ rst is to record pitch bend move-

ments into a sequencer and then re-assign the pitch bend to control another

aspect of the synthesizer, or you could just draw in automation using the

sequencers toolbox.

Re-assigning CC controllers is dependent on the sequencer in question, but

most will offer a method of doing so, typically, in the form of a drop-down

box. You click on the drop-down box to choose a different CC number and the

recorded (or drawn) automation will automatically be re-assigned to the new

CC number.

However, it is important to note that not all synthesizers will understand the

same CC messages. For example, you may re-assign pitch bend to control pan-

ning but if the synthesizer does not understand the CC number for panning

it won’t do anything. Alternatively, the CC number that equates to panning

on one synthesizer may be set to control the resonance on another, so you

wouldn’t receive the desired effect.

In an effort to avoid this and encompass some kind of uniformity between dif-

ferent synthesizer, the general MIDI (GM) standard was developed. This is a

generalized list of requirements that any synthesizer must adhere to for it to

feature the GM symbol. Apart from determining what sounds are assigned to

which numbers and how percussion sounds should be mapped across a key-

board, it also deals with the type of MIDI messages that should be recognized.

In total, there are 128 possible CC messages on any MIDI capable device but

since sequencers all rely on computer chips in one form or another, 0 is classed

as a number; therefore, they are numbered from 0 to 127. In the GM standard,

many of these controllers are hard-wired to control particular functions on the

synthesizer, so provided that the CC number is encompassed by the GM stan-

dard, your guaranteed that it will control the correct parameter no matter what

synthesizer the messages are transmitted to, provided that the synthesizer is

GM compatible.

Appendix D Features the Full List of CC Messages

As touched upon previously, each of these controllers must have variable

parameters associated with it to permit you to set how much the speciﬁ ed

Sequencers

CHAPTER 9

189

controller should be moved by. Indeed, every CC controller message must be

followed by a second variable that informs the receiving device how much

that particular CC controller is to be adjusted. Some CC controllers will offer

up to 128 variables (remember that’s 0 –127), while others will feature only

two variables (0 or 1) – essentially on/off switches. It is also important to note

that some CC messages will have both positive and negative values. A typical

example of this is the pan controller, which takes the value 64 to indicate cen-

tral position. Hence, when panning a sound, any values lower than this will

send the sound to the left of the spectrum while values greater than this will

send the sound to the right of the stereo spectrum.

While the GM standard is certainly useful in allowing synthesizer and sequenc-

ers to communicate effectively, it is not perfect since many synthesizer have

outgrown the original speciﬁ cation of MIDI and offer many more user-accessi-

ble parameters. Consequently, two major synthesizer manufacturers developed

their own MIDI standards that act as an extension to the GM standard. The ﬁ rst

is Roland’s GS system.

Contrary to popular belief, GS does not stand for general standard but is sim-

ply the name of the microprocessor used to power the engine (only Roland

seems to know what it stands for). Nevertheless, this protocol is compatible

with the GM standard while also adding variations to the original sound set,

access to two drum kits simultaneously and a collection of new CC messages.

The second is Yamaha’s XG system, which is also fully compatible with GM but

offers more advanced functions than both the GM and GS formats. This system

allows you to use three drum kits simultaneously, along with three simultane-

ous effects and is also partly compatible with the Roland’s GS system. This is

only partly compatible since Roland constantly updates the GS system with the

release of new synthesizer.

As if these formats were not enough, the developers of the GM standard

released the GM2 standard. This is yet again fully compatible with the origi-

nal GM standard, but also adds a series of extensions to the original, including

MIDI tuning and even more controllers.

Nevertheless, even with these new speciﬁ cations and additional controllers,

synthesizers are continually released which go beyond the standards set by the

manufacturers and there will be instances where you need to access or control

parameters that are not covered by these standards, therefore you will have to

use System Exclusive messages.

Universal System Exclusive Messages

System Exclusive (often abbreviated to SysEx) is overlooked by an incredible

amount of musicians since it can initially appear to be incredibly complicated.

After all, musicians rarely want to start working with hexadecimal numbers

they just want to make music.

PART 1

Technology and Theory

190

Regardless, it is an art well worth learning since it can be used to great effect on

every MIDI compatible unit. If you’ve programmed a synthesizer with your own

set of personalized sounds, you could dump them all to one easy-to-manage

ﬁ le on the sequencer, allowing you to re-install then whenever required.

Alternatively it can be used to control or adjust parameters on a MIDI unit that

has very few or no controls on its facia. It can even be used to adjust CC’s that

are listed on the MIDI speciﬁ cation but are not recognized by the MIDI unit.

Furthermore, by inserting a string of SysEx messages at the beginning of a MIDI

arrangement, you could take your own MIDI ﬁ le to a friend’s studio and, pro-

vided that they have the same synthesizer, the SysEx at the header of the ﬁ le

would not only set the sounds for each channel but also reprogramme the syn-

thesizer to all the sounds you used.

To understand SysEx ﬁ rst requires an understanding of hexadecimal, or at the

very least, how to count in hexadecimal. The normal way in which we count

is known as Base 10, based around the logic that we have 10 ﬁ ngers, therefore

every time we count above 9, we have to move over one decimal point.

0 …1…2…3…4…5…6…7…8…9…

Then we move over one decimal point and start again:

1 ⫻ 10 ⫹ 0 …1…2…3…4…5…6…7…8…9…

Take the number 1650 as an example. We can break this down into the follow-

ing calculation:

1 ⫻ 1000 ⫹ 6 ⫻ 100 ⫹ 5 ⫻ 10 ⫽ 1650

This could also be viewed as:

1000 100 10 0

1650

With the number 1650 there are four decimal places; in other words, every dec-

imal place in a number to the left of the decimal point is ten greater than the

previous one.

Hexadecimal works at Base 16, not Base 10, we cannot move over 1 decimal

place yet, so we need to continue by using letters in the place of 10, 11, 12, 13,

14 and 15. These are:

A ⫽ 10

B ⫽ 11

C ⫽ 12

D ⫽ 13

E ⫽ 14

F ⫽ 15

Sequencers

CHAPTER 9

191

Therefore we count like this:

0 …1…2…3…4…5…6…7…8…9…A…B…C…D…E…F

When we count to 16, we then move over one decimal place which would

equate to 10, in other words F has moved over a decimal place to 16. Following

this, 17 would equate to 11 (16 ⫹ 1), 18 would equate to 12 (16 ⫹ 2), 19

would equate to 13 (16 ⫹ 3) and so forth. When we reach 26, we then have to

start using our letters again which would equal 1A.

That’s 1 ⫻ F (16) ⫹ 1 ⫻ A (10) ⫽ 26 (in other words 16 ⫹ 10 ⫽ 26).

In a more practical situation, we’ll say that part way through an arrangement

you want to change the waveform of a LFO. This is obviously not covered in

the MIDI CC list but can be enabled using SysEx, usually listed in the ﬁ nal

pages of a synthesizer manual.

To programme a certain type of behaviour using SysEx messages, you need to

know what SysEx message to send to the synthesizer. SysEx messages typically

consist of a hexadecimal address, which pinpoints the function within the syn-

thesizer that you want to adjust, along with a hexadecimal value, telling the

function how you want to adjust it.

Thus, by sending a SysEx message to the synthesizer reading 00 00 11 04

(derived in this case from the values shown in Table 9.2 ), the LFO rate would

be set to waveform number 4.

It’s all a little more difﬁ cult than this, however, because you can’t just send the

address and SysEx information. You also need to provide basic ‘identiﬁ cation

information’ about the synthesizer that you are transmitting the message to,

even if there is only one synthesizer in the entire set. If this additional informa-

tion is not also sent, the message will simply be ignored. Thus, all synthesizers

require that the full SysEx message is either eight or nine strings long, combin-

ing all the elements described, to complete a message. A typical message is F0,

41H, 10H, 42H, 12H, 00 00 11H, 04H, 10H, F7. To better understand what this

means, the message can be divided up into nine distinct parts, as shown in

Table 9.3 .

Looking at Figure 9.3 , the SysEx message we want to send, comprising the

Address and the SysEx value, can be recognized in parts 6 (00 00 11) and 7

(04). The other seven parts of the message provide the required header and

synthesizer identiﬁ cation information.

SysEx Message Structure

Address Parameter SysEx Value Adjustable Amount

00 00 11 LFO waveform 00–04 0–4

Table 9.2

PART 1

Technology and Theory

192

Parts 1, 2 and 9 are deﬁ ned by the MIDI speciﬁ cation and are required in all

SysEx messages. The ﬁ rst part simply informs the synthesizer that a SysEx mes-

sage is on its way and sends the value F0, while the last part, part 9, informs

the synthesizer that it is the end of the message with the value F7.

The second part of the message is speciﬁ c to the manufacturer of the synthesizer.

Each manufacturer employs a unique number, which is quoted in the back of

the synthesizer’s manual. In this example, 41H is the identiﬁ cation tag used by

Roland synthesizers; thus, only a Roland synthesizer would prepare to receive

this particular message and any other manufacturer’s synthesizers will ignore it.

The third part of the message is used to identify a particular synthesizer even if

it is from the same manufacturer. This can be changed in the parameter pages

of the synthesizer in case another device in the set-up shares the same number,

so that you can send SysEx messages to the one that you want to control rather

than to all of them.

The fourth part of the message contains the manufacturer’s model identiﬁ ca-

tion code for that particular synthesizer to ensure that the message is received

and processed only by synthesizers with this particular model identiﬁ er (ID).

The ﬁ fth part can be one of two variables – 12H or 11H – used to specify whether

the message is sending (12H) or requesting (11H) information. If a synthesizer

receives a SysEx message it recognizes, it will look at this part to determine whether

it needs to change an internal setting or reply with its own SysEx message.

An 11H message is usually employed to dump the entire synthesizer’s patch

settings into a connected sequencer. By sending this, along with the appropri-

ate follow-on messages and pressing record on a sequencer, it’s possible to save

any user-constructed patches. This is useful if the synthesizer has no user patch

memories of its own.

As already observed, sixth part contains the address of the function on which

the SysEx message is to act, which in this case is the LFO rate. Most synthesizer

manuals provide an address map of the various functions and the correspond-

ing addresses to use.

The seventh part can serve two different functions. If the ﬁ fth part contains the

value 12H (indicating that the message is being sent), then part 7 will con-

tain the data being sent, as shown in the example we’re using (where the SysEx

value 04 is sent to set the LFO rate). If the ﬁ fth part contains the value 11H,

indicating that information is being requested, the value of this part indicates

the number of bytes you want the synthesizer to return in its reply message.

Example SysEx Message

12 3 4 5 6 7 8 9

F0 41H 10H 42H 12H 00 00 11 7F 10H F7

Table 9.3

Sequencers

CHAPTER 9

193

Whether eighth part is included will depend on the type of synthesizer you use.

Some manufacturers employ this ‘extra’ byte as a checksum, which is used to

validate the message and to ensure that the receiving synthesizer does only what

the hexadecimal code asks it to. Although MIDI is a fairly stable platform, there

will be the odd occasion when notes stick and messages become corrupted dur-

ing transmission. If this happens, it’s possible that the original message could

end up asking the synthesizer to do something entirely different, such as eras-

ing all user presets. Errors like this are avoided with a checksum because if the

checksum and message do not match, the SysEx message will be ignored. This is

why you need to understand how to calculate and count in SysEx.

A checksum is calculated using the following formula, taking the SysEx mes-

sage shown in Figure 9.3 as the basis for the calculation.

For example, convert the address and SysEx hex values from parts 6 and 7 of

the message to decimal values and add them together to give the value H .

Based on the values from Figure 9.3 :

Part 6 _ 00 00 11 (hex) and converts to a decimal value of 17.

Part 7 _ 04 (hex) and converts to a decimal value 4. Therefore; H _ 4 _ 17 _ 22.

If the value of H is greater than 127, then we need to minus 128 ( H _ 128)

from it to produce the value X. In this case H _ 22, which is less than 127, so X

is also 22.

H _ 22, which is less than 128; therefore X _ 22.

Finally, we convert the value X to hexadecimal to derive the checksum value for

the message. The decimal value 22 in hexadecimal is 16. Therefore, the check-

sum value is 16.

Although creating your own messages can be boring and time-consuming, it

does give you complete control over every aspect of a synthesizer, allowing you

access to parameters that would otherwise remain dormant.

It should, however, be noted that only one SysEx message can be transmitted

at any one time and additional messages cannot be received while the synthe-

sizer is using that particular function. Furthermore, any synthesizer will pause

for a few microseconds while it receives and processes the command, so any

two messages must not be sent too close together so as to give the synthesizer

time to process the previous command. If a second message is transmitted

while the current command is being processed, the synthesizer may ‘lock up ’

and refuse to respond and a ‘hardware boot ’ (switch off then on again) must

be performed to reset it. Consequently, the timing between each SysEx message

should be considered when transmitting any messages.

These timing problems make SysEx pretty useless if you want to continually

increase or decrease a parameter while it is playing back. For instance, if you wanted

to emulate the archetypal dance music ‘ﬁ lter sweep ’ where the sound becomes

gradually brighter, you would need to send a continual stream of messages