24 tracks? Pah! How about a whopping 24 notes? Steve Howell concludes his series on the modular systems of yore by taking a look at the somewhat capacitorially‑challenged analogue sequencer... This is the last article in a three‑part series.
It's hard to remember sometimes, when I sit down at my 99‑track, 64‑channel MIDI sequencer with 100,000 note storage, that it wasn't so long ago that sequencers were somewhat more moderate. By moderate, I mean 24 notes! And that was top of the range! I seem to remember that ARP claim that they invented the sequencer, around the early 1970s, although I am sure I have seen them on earlier Moog systems. Of course, other manufacturers included analogue sequencers, for example Roland, in their System 100, 100M and 700 synths.
Early analogue sequencers have a row of rotary controls which are 'tuned' to shunt out different voltages. When these voltages are applied to the CV inputs of an oscillator, they create a melody, or, more accurately, given the limited number of notes, a riff. The sequencer has a step counter which cycles through the controls at a rate dictated by a clock (typically a square wave generator with a variable rate and pulse width). This clock is also used to turn the envelope generators on and off (see explanatory side panel 'Stepping Out', elsewhere in this article, for a more detailed explanation). This setup results in a painstaking process of melody creation, requiring the individual tuning of each step in the riff, although later sequencers (the imaginatively named 'ARP Sequencer' to give but one example) have quantised controls, and tuning is incremented in semitones. Pressing Play (i.e. engaging the clock) starts the sequencer, switching in the counter to move through the steps sequentially. There are also Stop and Continue keys, whose function is fairly self‑evident. When running, these old sequencers look the business, as LED step indicators switch on and off as the steps play.
Whilst some analogue sequencers can remember maybe eight or 16 steps of monophonic riffery, some offer as many as a staggering 24 notes! However, other functions exist, so that instead of one 24‑note monophonic sequencer, you can allocate the 24 notes over three channels of eight notes each. Clocked from a common source, this allows three‑part harmonies to be created. Alternatively, one channel can be used to pitch the oscillators, whilst the other two are used to control the filter(s) and amplifier(s) respectively, for control over dynamics. Other functions allow you to set the number of steps in the sequence, and selectively switch steps on or off. A Note Length function allows you to vary the length of the steps from legato to staccato, and a Direction switch sets — you guessed it — the direction of the sequence. The fact that all this can be accomplished while the sequencer is running means that the analogue sequencer is a versatile, interactive, creative tool for improvisation that modern sequencers cannot begin to emulate. And don't forget, as well as adjusting the sequencer's parameters, you can be adjusting the sound's parameters as well — I think this is what's meant by 'playing the synthesizer'!!
But stop thinking of the sequencer as a means of playing sequences! You're forgiven for thinking this way, because that's what modern MIDI sequencers do. The analogue sequencer, by contrast, is actually just another controller to be used in the arsenal of goodies on offer in a modular synth. Playing pitched sequences is just one of its tasks.
Whilst LFOs, sample and hold units, envelope generators, and so on offer a degree of control over their voltage outputs, an analogue sequencer is a totally open, programmable and triggerable controller. By setting the controls, you may create your own control shapes, and apply them to any control input on the synth. You can even use the sequencer to control itself! We saw last month how the random output of the sample and hold module can be used to control its own rate. Similarly, we can route the sequencer's output to its own clock modulation input, or step length modulation input (to control the duration of each step — a high voltage will generate a fast step, a low voltage will cause a long step). Not spectacularly useful with a simple monophonic sequencer, but if you can split the sequencer into two or more channels, one channel can be reserved to control note length.
More simply, perhaps, by overriding the internal clock with the keyboard's Gate output, you can use an analogue sequencer to provide preset, pre‑determined sequences of control voltage for controlling many aspects of a sound. For example, plug the Gate output of the keyboard into the External Clock Input of the sequencer, and set the controls for a variety of levels. Now, as you play each note, you will step through the sequence, advancing one step for every note you play. If you apply the output of the sequencer to the filter cutoff, you can have a different pre‑programmed tone for each note. With careful settings, you can create some interesting vocal articulations, literally making your filter 'speak'. Tomita pioneered this technique, using it on many of his albums to create a most effective 'singing synth', although he was prone to overuse it!
By clocking the sequencer from the keyboard in this way, it is also possible to impart dynamics. In the early days, of course, keyboards had no velocity sensitivity. By routing the sequencer output to the VCA's control input, you can set pre‑programmed velocity values for the melody you want to play. Of course, the limited note capacity of a typical analogue sequencer means that phrases have to be kept short. When Walter (now Wendy) Carlos, Tomita and others were creating the early synth albums, this wasn't a big problem, as much of their music was recorded phrase by phrase, concentrating on only a few notes at a time. In this way, they could set the ideal sound for a phrase, record it, make a few subtle changes to the next phrase's sound, and so on, building up the music bit by bit. Laborious? Yes, but the results were glorious!
Yet another use for an analogue sequencer is in One‑Shot mode. Instead of churning out an endless repetitive sequence, triggered either from the Play key or from a keyboard, the sequencer will run through its sequence once and stop. With a fast, arpeggiating sequence that can be triggered by every note played, it is possible to create some very exotic textures indeed, especially when layering sounds with binloads of effects.
Of course, a variation on the sequencer is the arpeggiator. Few modular synths had this, but one that did was the Roland System 100M's 4‑voice controller keyboard (sold separately from the synth itself). The arpeggiator works much as you would expect — it runs either up, down, or up and down any series of notes you care to play (for example, the composite notes of a chord), at a rate determined by its clock — or an external clock. Roland would also sometimes include the fascinating Random setting, which takes the notes you are playing, but runs through them in a random order. It really is a wonderful effect, and when that and a multi‑channel sequence are running in sync, with just a hint of synchronised delay, the musical and rhythmic effects are equally relevant and musically interesting today.
At the dawn of synth history, the keyboard was an optional extra — you had to order them separately with early ARP and EMS gear — but with those synths manufactured by Serge, they were almost forbidden. These were intended for the creation of 'real' or 'serious' synth music, and, to this end, they featured what were called 'function generators'. I must admit to having little experience of these synths, but, from what I understand, the function generators were truly open‑ended, voltage‑controlled, triggerable controllers that could act as envelope generators, as LFOs with a totally programmable waveform output, or as more or less anything in between, and were capable of running at extreme rates, from one cycle every few days to low audio. By setting various rise and fall times, you could create repetitive, interesting control waveshapes, or, alternatively, you could trigger them much like standard ADSR envelope generators. The EMS VCS3 and AKS synths had something similar, in that the Envelope Shaper (as it was called) offered attack/hold/decay/off times, and could be programmed to cycle, by setting the Off control fully clockwise. The decay parameter was also voltage‑controllable, allowing you to create bizarre 'trapezoid' (as EMS called them) control shapes.
Of course, other controllers, such as joysticks (EMS), ribbon controllers (Moog), and touch keyboards (Buchla, Serge, EMS) abounded. Some esoteric keyboards (often referred to as Scale Programmers) had a knob for each key, so that you could tune the keyboard to odd scales.
The real beauty of the modular synth is not the individual modules, but the myriad ways they can be combined and routed, and the interplay between them. Sometimes I wonder if the invention of the MiniMoog was the best thing that happened to synthesis or the worst! Certainly, it made synths more affordable and manageable, but it did limit the true possibilities, and set a trend that was never reversed, as other manufacturers copied it to the point where it became an industry standard layout for all synths (and samplers) that were to follow.
Analogue sequencers and arpeggiators are not only the preserve of those few with enormous modular systems; they are also included on more lightweight synths. A (far from definitive) list might include the following:
SYNTHS FEATURING ARPEGGIATORS
- Roland SH101.
- Roland Juno 6 & Juno 60.
- Roland Jupiter 4, Jupiter 6, & Jupiter 8.
- Oberheim OB8.
- Korg Mono/Poly.
- Sequential Circuits Pro One.
The Yamaha CS30L synth features a 10‑step tunable analogue sequencer similar to the ones described on modulars in this article. In addition, stand‑alone analogue sequencers were manufactured, such as the 24‑step, 12‑channel Korg SQ10, which, much like MIDI sequencers, generates no sound of its own, and has to be linked to a Korg MS10 synth. Finally, to bring us right up to date, there is the Doepfer MAQ 16/3 MIDI/analogue sequencer, designed in consultation with Florian Schneider of Kraftwerk, which was reviewed in the July 1993 edition of SOS. This is a 48‑step sequencer (also usable in a 3‑track, 16‑step configuration). Not only is the pitch of every note tunable, but so is MIDI note information (pitchbend, aftertouch, velocity, and so on), bringing analogue sequencing firmly into the MIDI era.
Analogue sequencers are as relevant today as they were in their heyday — assuming you have the synths to connect them to! Using them with MIDI sequencers can add new dimensions to your music; you can have the MIDI gear provide the basis for the backing track whilst you twiddle over the top in real‑time with the analogue sequence(s).
Despite what you might think, getting them to sync up is no problem. Kenton Electronics' Pro2 and Pro4 MIDI‑to‑CV converters both provide clock outputs for a solid sync from a MIDI clock.
If you own a Roland MPU101 MIDI‑to‑CV converter, you can also provide a clock output, but you lose the ability to use the unit as a MIDI‑to‑CV converter at the same time. Probably the best way to overcome this is to set aside a track on a separate channel of your MIDI sequencer, and fill it with 8th or 16th notes (at any pitch). The MPU101 will convert this into a gate signal, which can be output from one of the MPU101's Gate outputs. You can then use this signal to drive the analogue sequencer via its Clock input. Both these methods are equally effective for syncing arpeggiators to MIDI.
Festooned with controls, the analogue sequencer is a knob twiddler's delight! There are rows of dials which you 'tune', and these are stepped through at a rate set by an internal clock. The setup can be summarised graphically, as in Figure 1.
When the clock is running, each pulse it puts out advances the step counter one increment, until it reaches the end, whereupon it repeats. The clock also feeds the envelope generators to trigger them. A close up of a typical panel might look something like Figure 2.
The Length control sets the duration of each step. This is actually just a pulse width control, which sets the length of the On time by varying the symmetry of the square wave clock signal being sent to the envelope generators through the Clock Out. The longer the On time, the longer the step. Sometimes, the pulse width can be governed by an external controller, such as an LFO, or even the sequencer itself! The Rate control sets the speed of the sequencer, and the External Clock input overrides the internal clock, allowing the sequencer to synchronise to other clock sources, for example other sequencers, arpeggiators, or early drum machines. You can also override the internal clock with the Gate output of a keyboard, and step through the sequence manually.
The Steps dial controls the number of steps in the sequence, whilst Direction sets the sequence to play either forwards, backwards, forwards and backwards or, if you're lucky, randomly. The One Shot switch determines whether the sequencer will repeat its sequence endlessly or just play it once through, and the 1x16/2x8 control allows you to choose whether the sequencer behaves monophonically or has two tracks with a capacity of eight notes each.
Play, Stop and Cont(inue) control playback of the sequencer, and the Trigger input allows you to start the sequencer remotely. The controls for each step are identical, and the exact number of steps varies according to the sequencer you are using. Typically, each step is set up using the Set button, in conjunction with the Steps rotary control (to select which step you wish to set). Pressing the Step button plays each step in turn as you adjust these controls. LED indicators light up to show the current step being adjusted, and also which steps are playing as the sequence runs (looks great!). Underneath the Step control is a three‑way switch that selects whether that step will Play, Skip (i.e. miss out completely with no gap) or Rest (i.e. skip the step, but leave a gap). Occasionally, analogue sequencers have an individual CV output for each step, which can be routed anywhere you like for auxilliary control purposes. The main CV jacks (not shown here) output the entire sequence.
Of course, the example given here is not a particular analogue sequencer, but an example of a typical one. Most looked something like this, although the ARP Sequencer used sliders.