Guide to Electronic Music Techniques
Dave Smith, the designer of the Prophet 5, was once asked why said instrument didn't feature any form of polyphonic portamento — it would have cost an almost negligible amount to provide such a facility, and as the instrument already had a 'glide' control (for monophonic use), no cosmetic alterations would have been necessary to the main control panel. He was of the opinion that because such a facility was beyond the control of the player, it was musically invalid.
Let's see what he meant by that. Portamento, as we discovered last time, is an analogue effect caused by the slewing of the control voltage routed to the VCOs. Thus the pitch slides from note to note as opposed to stepping cleanly when portamento is removed. In effect there will always be some degree of portamento between notes produced by the same oscillator, but the slew rate is so fast as to be aurally undetectable. Figure 1 shows a simplified analogue portamento circuit.
Consider a five voice analogue synth which offers poly-portamento; here we'd have five voices each with a similar circuit in the CV to VCO line. To determine the slew rate (octaves per second) a control voltage is applied (Cs) and this obviously will be the same for all voices. Say we played a five note chord, then another as shown in Figure 2a. Since all the notes have a different interval to travel they will all reach their destinations at a different instant in time; 2a shows a simple 1:1 relationship where the voice assignment leads the lowest note played to be routed to voice 1 etc. but more often than not this assignment won't be applicable and the situation will be even more confused as in Figure 2b.
This is the problem to which Dave Smith was referring — and there's no simple way out. If you think about it, polyphonic portamento/glissando is not too common a phenomenon in the world of acoustic music. Most acoustic instruments are monophonic anyway, and of those polyphonic ones the effect isn't that often used. Generally polyphonic glissando effects from keyboard instruments can be obtained by running one's hands up and down the keyboard — and you can perform exactly the same operation using the synthesiser's keyboard. Bottle neck, pedal steel and Hawaiian guitars are examples of polyphonic portamento that are 'under control' but again clever use of the pitchbender on your synth can get you pretty close to the desired effect. It is when you come to orchestral or ensemble synthesis, where you are simulating groups of instruments, that you may find yourself caught short without an accurate system of polyphonic portamento.
Dave Smith may have decried the musical validity of polyphonic portamento in the past, but I note that the Pro 600 has now got it. And let's face it, it is a pretty dramatic sound when five or six voices are swooping all over the place, moving from one chord to another, even though they might not 'get there' all at the same time. If you use just a hint of poly-portamento the gap between the notes is usually so small as to be virtually negligible. So even a crude polyphonic portamento effect is worth something.
The Prophet could have incorporated this if it had been felt desirable, but some poly-synths, by their very design criterion, cannot accommodate this effect. These are the top octave tone generation models. All the pitches are derived either from a single high frequency VCO and divided down, or from 12 top octave generators which produce all the pitches for every note. Only voice assignable synthesisers can provide this facility.
Digitally Controlled Oscillators (DCOs) have thrown somewhat of a spanner in the works. These devices derive the pitch of the oscillators from the processor which puts out a pulse to a waveshaper that does the same job as a VCO, only it's cheaper and more efficient (no tuning problems etc.). However, portamento now becomes a bit of a problem as there is no voltage line in which to stick a slew generator. So if the synthesiser is going to offer this facility it will need to process the pitch data with much greater accuracy, and output a continually varying pulse train to the waveshapers. Some DCO based machines do provide poly-portamento facilities (eg. Kawai with their SX-210), others feel that the extra memory and processing room required doesn't warrant the effect, and thus don't feature it on their machines. Check this one out if you feel you want this effect on your machine.
Now perhaps we can start to solve our initial problem — the machine gun effect of all the notes reaching their destination at different times. If the machine has powerful processing facilities it is a (relatively) simple matter to build in a system whereby the user defines the slew time (T) rather than the slew rate. As the second chord is played the processor looks at these Destination notes (D1, D2,etc) in turn and then from the Origin chord (notes O1, O2 etc.) is calculated the distance each voice has to travel (D1-O1, D2-O2, etc.) and thus the slew rate (S) for each voice can be determined (S1 = (D1-O1)/T, S2 = (D2-O2)/T, etc.). This information can then be used internally by the processor to vary the pulse train accordingly, or, in an analogue system with separate slew limiters for each, to define the slew control voltage (Cs in Figure 1).
This is an extremely versatile system once initiated as it facilitates all manner of accurate sliding effects. For example, it would be possible to specify a start point either at the top or bottom end of the keyboard (say), so that any chord could be assigned a perfect positive or negative Glide effect, which is invaluable for certain imitative synthesis effects.
When using polyphonic portamento, even in its basic form, one must remember the importance of voice assignment. Systems vary from synth to synth, but knowing the order of note assignment, and when the next note played is going to be 'robbed', gives you so much more control over the output of your instrument — there's a lot more to synthesiser playing technique than is apparent to the piano convertee.