BBC Micro Music Masterclass
Until fairly recently, if you hoped to play music of reasonable standard, you had to resign yourself to years of arduous practise on a saxophone, guitar or, worse still, violin. Ah, you may argue, surely this wasn’t true of the new wave bands? Yes, even during the punk phase of pop music bands still had to get together for hours on end, even if they were simply deciding in what order to play the songs at their next gig. Computers have completely changed this state of affairs. Modern musicians now find themselves able to create technically sophisticated music in their own bedrooms by using a bewildering array of drum machines, synthesisers and sequencers that would put the BBC Radiophonic Workshop to shame.
All the above mentioned machines are basically microcomputers which are dedicated to the particular task in hand, such as producing drum beats or digitally synthesising sounds. We can learn a great deal about the best ways of programming sound on the BBC Micro from examining these devices, so in this chapter I would like to describe how some of them operate.
Computers are ideally suited to act as substitute drummers. They are accurate, unimaginative, stupid and do not mind undertaking boring and repetitive tasks for hours on end. It is not surprising then, that when computers began to infiltrate music one of the first tasks they were set was to act as a drum machine.
The original drum boxes that came attached to home organs contained a small selection of fairly standard and uninteresting pre-programmed rhythms and sounds. This was obviously an unsatisfactory state of affairs and everyone soon realised that it would be much better if a rhythm was programmed by the musician to fit the particular tune being played at any given time. The problem with this scenario was to make the programming simple enough for machine-shy musicians, a not inconsiderable percentage of the music community both then and now. One of the first drum boxes devised to solve this problem was called the Roland Dr Rhythm.
This machine was cheap and straightforward enough for even the most recalcitrant musician to feel at home with and approach optimistically.
The sounds were produced by pre-set noise producing circuitry, but the small on-board computer allowed almost limitless variations of rhythm to be tapped in by the user. On the Dr Rhythm each bar of music is simply divided into sixteen subdivisions for 4/4 time or twelve subdivisions for 3/4 time. The rhythm for each drum (bass drum, snare, sidestick and accent) is tapped in rather like sending morse code, using one button for a drum hit and another for a rest. Once a bar of rhythm has been completed for a particular drum it can then be checked against the inbuilt eighths or sixteenths hi-hat before proceeding to another drum. The diagram below shows how we would produce a pattern of on beat brass drum and off beat snare drum against hi-hat eigths:
This method of breaking down a bar into subdivisions is fundamental to all drum machine programming. After the introduction and success of Dr Rhythm, drum boxes which incorporated larger memories and more facilities soon became generally available. The Roland Drumatix and TR808 machines were amongst the most popular and had a wider selection of drum sounds to their credit, as well as the ability to chain together bar length patterns. This chaining technique allows complex drum parts to be programmed, complete with fills, tempo and time changes.
However, looked at objectively, one sees that these machines are only superficially more sophisticated than the Dr Rhythm. To gain a substantial improvement the number of subdivisions per bar must increase. In the Dr Rhythm and its Soundmaster counterpart semiquavers are the shortest playable notes. In 4/4 time it would thus not be possible to play triplets (16/3). This limitation could be overcome by increasing the number of subdivisions to forty-eight per bar. A machine which could subdivide bars this finely could not only play semiquavers (since a semiquaver is 1/16 of a bar, and therefore would equal 48/16=3 divisions) but also quaver triplets (quaver triplets occupy 1/12 of a bar and therefore would occupy 48/12=4 divisions). The first machine to reach and make available this level of sophistication was Roger Linn’s LM-1.
To call a Linn Drum (or the Oberheim, MXR, etc., counterpart) a drum box would be rather like referring to a Stradivarius violin as a fiddle. We are in a completely new ball game here, for these machines are very far removed from the more primitive devices we have been looking at to date. They differ in two main areas. Firstly let us look at the sounds. Linn sounds are digital recordings of real drums which are captured in Rows (Read Only Memory). D/A (Digital to Analogue) converters are then used to turn this digital information back into noises.
Secondly, in the programming area, all these machines are capable of far more subdivisions per bar than, say, the Tp808. In Linn terminology we talk about correcting to the nearest 1/16, 1/32 or 1/48th of a bar or recording in real time. This is because all these computers have abandoned the morse code method of the Dr Rhythm, for tap buttons which allow you to play the drum rhythms into the machine. Needless to say, this gives a much more spontaneous feel to the music produced by this new and exciting generation of machines.
In addition to pattern entry we are also able to choose a song mode which allows easy construction of complex arrangements. As the sophistication and size of memory increases, so editing facilities also have to be refined accordingly. All Digital Drum computers allow the deletion, insertion and copying of both overall and individual drum patterns. At least half of the pop records which are currently being aired on the radio are made by using computer drums of one kind or another. For a simple application of some of these principles, see the ‘Drum Machine’ program in Chapter Nine.
Drum machines are by no means the only musical area into which computers are insinuating their technology, merits and effects. In ten short years, for example synthesisers have changed from being gangling giants which took up half a room to the sleek, keyboard size instruments which we now find in virtually every well-equipped studio.
The early synthesiser developed by Robert Moog and used by Walter Carlos to produce his famous Bach recordings looks quite unlike the instruments we see Depeche Mode and other contemporary bands play today. This is because the original Moogs were built as a series of modules, connected together in different ways to produce the various effects. These modules had the following functions:
VCO: the Voltage Controlled Oscillator produced the actual sound in the form of a sine, square (like the BBC) or sawtooth wave.
VCF: the Voltage Controlled Filter alters the nature of the basic sound by filtering out certain frequencies.
ENVELOPE: which acts in a similar way to the BBC’s volume envelope.
KEYBOARD: the keyboard in these early instrument had to provide two sets of voltages. The control voltage (CV) controls the pitch of the note played by supplying set voltages to the VCO. A standard of one volt per octave is now widely used such that a CV of one volt gives a pitch of C. The gate voltage controls the duration for which the note is played. This voltage was normally zero for no note, and 2.5 volts or more for a keypress.
Of course, various other modules were often included, for instance to introduce vibrato into the sound using a low frequency oscillator (LFO) output or to generate white noise, but the above VCO, VCF and ENVELOPE generator are all that is needed, when amplified, to produce the familiar synthesiser sound we have come to know and love.
Voltage Voltage Voltage Controlled Controlled Controlled Oscillator Filter Amplifier OUTPUT
A modern synthesiser still includes the traditional components but instead of connecting together these elements with patch leads, a microprocessor is used to achieve the same effect. This computer’s function varies according to the machine. Thus, it can merely be used to connect the keyboard to the various modules, as in the case of monophonic synths, such as the Sequent-ial Circuits PRO ONE or the Roland SH-101, or it can be utilised to memorise entire sets of control settings, as with the polyphonic Oberheim OB8 and Roland Jupiter 8 machines.
In Britain, a cheap but impressive-sounding synth called the Wasp was the first computer-controlled instrument to become widely used in the music business. Like the more modern SH-101 and the Pro One, this synthesiser employs a microprocessor to interface the piano keyboard with the analogue oscillators and filters. Opposite, courtesy of the manufacturers, is a block diagram of the Roland SH – 101 elements to illustrate the current state of the art.
A polyphonic synth is simply the amalgam of a number of monophonic synths which are put in the same box and controlled by the same keyboard (five in the case of Prophet, eight in the case of Jupiter). These added voices create extra work for the computer, however, in more ways than one!
First of all, the computer has to convert up to eight keypresses into control codes in order to be able to define pitches for the VCOs. Then it has to decide which VCO will generate which note. Finally, just as before, it has to send gate information to the relevant envelope generator. This can prove quite a problem when the player could be anyone from Semprini to my tone-deaf cat running up and down the keys! If you turn to look at the organ program in the previous chapter you will find a scaled down solution to this problem.
In the process of developing better and faster synthesisers it soon occurred to the manufacturers that since a computer was needed on board anyway, it might as well be used to its full capacity. Soon afterwards the programmable polysynth was born. One immediate benefit was that instead of having to keep drawings of control knob settings to hand to reproduce your favourite sound, it now became available at the Hick of a switch. This is because, after a sound is set up on a programmable synth, the knob settings are stored in the computer memory. The sound is also assigned a number, and whenever the number is typed in, the sound is called up.
This development changed the face of music. Previously, in order to fully exploit the possibilities introduced by the synthesiser musicians also had to have a boffin like mentality. The patch leads and bewildering array of knobs and switches put many a potential synthesist off the whole idea of tackling this promising but daunting hardware. Now that sounds are available simply by pressing a button, all that twiddling and tweaking can be done in the privacy of your own home, thus, saving hours of valuable studio time (and a lot of red faces at concerts).
One logical extension of using a computer to create rhythms is to get it to play melodies. A machine that does precisely this job is known as a sequencer. A sequencer, therefore, is any device which is used for the automatic control of an external sound synthesiser.
Sequencers originally functioned entirely without the use of computer technology, for they were simply a series of modules which could be set to particular voltages and were then triggered sequentially. These voltages caused the external synthesiser to play a sequence of notes. Since such a repetitive task is ideally suited to computers, this old-fashioned analoguemethod of sequencing was quickly abandoned in favour of digital technology.
Simple sequencers capable of recalling up to one hundred steps are now built into modern synthesisers as a standard facility. The Roland SH 101 and the PRO ONE can both be interfaced with a drum box, which makes it possible for the sequence stepping to be synchronised with the drum machine’s tempo. If you hunt out some of their records and listen to them bearing this in mind you will soon realise that this type of mechanical sequencing is used extensively by Georgio Moroder and Kraftwerk.
More sophisticated sequencers such as the Roland MC-4 and the Oberheim DSX allow many channels of music to be recorded along with phrasing and dynamics. These machines have developed from the single channel one hundred note sequencers in the same way that the Linn Drum developed out of the early rhythm boxes. By recording the sequence using increasingly smaller and smaller time intervals, you arrive at an effect which precisely mirrors real time playing.
The MC-4 Microcomposer, for example, is a music-dedicated 48k microcomputer. Musical information is entered into the computer using a combination of the function keys and a numeric keypad. To input a melody three types of information must be entered:
1. Pitch: A 125 note chromatic scale is numbered as follows:
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 C C# D D# E F F# G G# A A# B C C# D D#
To enter pitches the edit mode is selected and the chromatic pitch values (CV’s) typed in sequence. A bar structure can be set up at this stage by pressing the <<>MEASURE END<>> key whenever a barline is required.
2.Step: The step time is defined as the number of steps from the start of one note to the start of the next in the sequence. Suppose, for example, we decided to define a bar of 4/4 as 192 steps:
Note Value Step Time semibreve 1 192 minim 1/2 96 crotchet 1/4 48 quaver 1/8 24 semiquaver 1/16 12
This high number of subdivisions per bar is chosen because the larger this figure the more subtle the phrasing that results when we input the final set of information.
3. Gate: The gate value defines the phrasing of the notes and can be any number from zero to the step time. For example it the gate value equals the step then legato phrasing results. If the gate value is less than the step then stacatto phrasing results. A zero value for gate produces a rest. The following musical example should make the functions of Pitch, Step and Gate clearer. CV’s for Pitch are entered as above. The Step time is the entire length from the start of one note to the start of the next, and the two quavers hence have a quaver rest added (24+24=48). The Gate time input then gives the phrasing of the music:
Time Base = 48
Pitch (CV): 21 23 24 26 21 Step: 72 24 48 48 192 Gate: 70 20 24 24 144 Measure End
The MC-4 has four channels each with two pitches, one step and one gate. As a result, it becomes possible for the musician to compose harmonically complex music. In practise, this would be next to impossible, were it not for the sophisticated editing commands. Any part of a piece can be deleted, copied, transposed or inserted by means of a fairly straight-forward command syntax. For instance, if we wished to copy bars 1 to 16 at the end of a 32 bar tune the following syntax would be used:
COPY 1,16,1 <>>;ENTER<>>;
The cursor is set at bar 33 (the end of bar 32), the COPY key is pressed followed by 1(bar 1 start), 16(bar16 finish), 1 (one time), and
This refined form of sequencing is a long way from the haphazard analogue approach and would be totally impossible to achieve without a micro - and a very clever piece of software.
This latest giant leap in synth technology has seen the computer take over the production of certain forms of musical composition almost entirely. Not content with just controlling and storing sounds, some ambitious microchip decided it should get in on the act and generate the sounds as well! Instruments such as the Yamaha DX7 and the PPG Wave 2.2 are simply microcomputers disguised as keyboards.
Wave shapes are generated digitally rather than using simple sine or square wave oscillators. These waves are processed and shaped inside the computer and finally appear as analogue sounds, courtesy of a D/A converter. The most sophisticated electronic instrument available to date belongs to this class of digital synthesisers and is called the Fairlight CMI. Unlike the instruments described so far, the Fairlight makes no pretense of being anything other than a computer. The alphanumeric keyboard, twin disc drives and VDU are enough to send the average musician screaming from the studio. Music makers such as Vince from the Assembly (ne Yazoo), Depeche Mode and the world famous producer Steve Levine, on the other hand, have taken to the revolutionary new instrument like the proverbial duck to water.
The Fairlight can best be described as a complete music production system in one box. It can either generate sounds itself, from digitally produced sine and square waves, or sample noises from the real world, create a scale of similar sounds and play them back. Once in the computer, the sound waveforms can also be displayed on the VDU and directly modified to create startling new effects. Once the sound has been chosen it can be played manually on a piano keyboard (with eight note polyphony) or programmmed, using a variety of compositional languages. Using any one of these languages, the Fairlight can play up to eight different sounds at the same time to create complex arrangements which were previously only made possible by using a large group of highly trained musicians.
Because the Fairlight is a software-based instrument it is much less prone to obsolescence than its more traditional counterpart such as the Wave 2.2. Since its inception many modifications have been made in order to improve its sampling and refine the programming languages.
For those of you who are as yet unfamiliar with the concept of sampling I should insert a short explanation of the term at this point in the text. Sound sampling involves digitally recording a natural sound into a computer memory. Basically, the amplitude of the wave shape of the sound is measured against time and then this information is recorded (as a series of numbers) by the computer.
AMPLITUDE AMPLITUDE VALUES SOUNDWAVE TIME SAMPLING TIME SEQUENCE
The quality of the sample is dependent on the frequency with which the sampling takes place. The Fairlight samples at a maximum frequency of 30000Hz (30000 samples per second). Once in the computer, the sound can be displayed and changed using a light pen or numerical input.
Vibrato, filtering and tuning changes can all be made simply by using the correct commands. Since the facilities of the CMI are so wide ranging each area is designated as a separate ‘page’. These are as follows:
Page No. Title Functions 1 INDEX Self explanatory 2 DISK CONTROL Loading/saving files and programs 3 KEYBOARD CONTROL Tuning, polyphony, scaling, instrument files 4 HARMONIC Graphs, looping ENVELOPES 5 WAVEFORM Additive synthesis GENERATION 6 WAVEFORM Light pen, merging, display DRAWING 7 CONTROL Key velocity, portamento, vibrato PARAMETERS 8 SAMPLING Rates, filtering 9 SEQUENCER Basic sequencing D WAVEFORM DISPLAY Display L LIBRARY Disk library, files C MCL Music Composition language R PAGER Rhythm sequencer
From a brief survey of the above; you have doubtlessly realised that it is beyond the scope of this book to go into detail about the Fairlight or any of the other instruments mentioned in this chapter. However they have provided the inspiration for many of the programs contained in this book and, hopefully, knowing something about them will provide you with a few ideas of your own for further experimentation.