We look at how electronics, synths and a PET computer have influenced this versatile British composer.
Probably best known for his contribution to 'War of the Worlds' LP, we find out how electronics, synths and a PET have influenced his music.
Ken Freeman's success as a keyboard player and writer is largely due to his combined talents as a musician and a technician. At the age of sixteen his band, 'Second City Sound', had already made a single and were managing to make a living playing pubs and clubs in and around Birmingham. Six years later they were to earn the dubious distinction of winning Opportunity Knocks six times in succession. "From our success on Opportunity Knocks we were offered a summer season in Brighton with Ronnie Corbett," recalled Ken, "it wasn't exactly inspiring, and soon after that I left the band."
During this period Ken had been teaching himself electronics through the pages of hobby mags, and having left the band he was able to concentrate on completing a string synth project that he had been toying with for some years. "In those days there weren't any synths around with a really natural string sound. It occurred to me that when a string section was playing together, each player's vibrato was slightly out of phase with that of his fellows. I therefore designed what was actually an organ rather than a synthesiser, with three oscillators and three LFOs per note. The front panel provided controls for adjusting the pitch of each of the three groups of oscillators and each of the three groups of LFOs, and by slightly detuning the oscillators relative to each other and putting the LFOs out of phase, I was able to get a very rich and authentic string sound."
In partnership with a Leslie speaker manufacturer in Harrow, Ken built a second, more refined prototype, and during 1971, exhibited it at the 'Musical Instruments Fair' at the Russell Hotel, London and received overwhelming interest. A company called Ling Dynamic Systems wanted to manufacture it and financed the construction of a pre-production model to be exhibited in Frankfurt the following February. Although Ken and his partners once again experienced a great deal of interest, they were also subject to what Ken loosely describes as 'industrial espionage'. In the same period they were forced to pull out of their contract with Ling Dynamic Systems due to a financial disagreement, and so they were left with a pile of orders and no manufacturing facility. Whilst they were fighting Ling for release from the contract, other manufacturers were bringing on to the market various adaptions of Ken's idea, and so the marketing impact of the Freeman String Synth had been somewhat pre-empted. Ken is keen to pass on the advice to any would-be inventors, that if you get a good idea, don't put it on general display until you have the final product ready to sell, and then flood the market.
The cost of negating the contract with Ling had left Ken completely broke, and he was therefore forced to take a job servicing electronic instruments. However, he still had the second prototype of the synth, and with it he managed to get some sessions. At that time the natural quality of its sound was without equal, and even though he had no other keyboards, Ken soon found himself doing around ten sessions a week. Even today he still prefers it to modern synths for full chordal string sounds.
Since that time Ken has become more and more in demand in the studio, working with many acts including Elkie Brooks, Jon Anderson, Mike Oldfield, Justin Hayward and in 1975 with David Essex on the single 'I'm Gonna Make You A Star', which was produced by Jeff Wayne. In the same year Ken was part of the band that accompanied Essex on his American tour, with Jeff Wayne as MD. The synth sound on the record had been produced in the studio by quadruple tracking, and so to recreate it on tour Ken had to play a duophonic ARP Odyssey and the monophonic ARP 2600, while Jeff played a second 2600. As the tour promoters were going to have to hire the synths anyway, Ken decided to buy an Odyssey and offset the cost of its purchase by hiring it to them for the duration of the tour. "The Odyssey was very well designed, and although the sound quality isn't that good, the facilities for internal routing without the need for patching and so on are great. If it had a more powerful VCF it would still be very useful — as it is I don't really use it anymore."
The Odyssey provided Ken with his first experience of interfacing synths with computers. He built a bi-directional A to D, D to A interface between it and his PET to allow lines to be played on the synth, stored in the PET's memory and then subsequently played back through the synth. He could build a whole piece up line by line, storing each one on disc, and by using his Korg Rhythm 55 drum machine as a trigger pulse, he was able to play the lines in slowly and then have them played back at a faster speed, all perfectly in time. Of course, only one line could be replayed at any one time and so it was necessary to have a multi-track facility to get a complete piece recorded. When the first line was recorded, the trigger pulse was also recorded on a separate track, and thereafter was used as the trigger for subsequent lines. In this way sync between the individual lines was ensured. This method had several advantages: Ken could save time by writing the piece at home using his own four and, later on, eight-track facility, and then go in to the studio with the sure knowledge that a piece, which might under normal circumstances be technically very difficult to play, could be played back from the PET perfectly in time and perfectly in sync at any tempo required. With studio time costing what it does, that kind of reliability is very valuable. Ken wrote and recorded the current LBC station ident using this method.
After the David Essex tour, Jeff Wayne started to use Ken on a lot of his jingle sessions, and when he started work on the recording of 'War of the Worlds' in the summer of 1976, he hired Ken to play all the synth parts. The project took three years to complete and during the first two of those years Ken had little time to do anything else. "The equipment was relatively primitive in those days and things that would be easy to do now took a long time. We were using a 48-track recording facility, and including all the bouncing down, we must have actually used over one hundred tracks to get the sounds right. It was very good experience."
In 1977 Ken bought what was one of the first CS80's in the country, and armed with his experience with the design of the string synth, set about modifying it. In its normal state the CS80 is capable of 8-note polyphony, with each note using two VCAs, VCFs and VCOs. If, however, a single note is played, all eight pairs of VCOs are tuned to that note (that is, until another key is depressed) but only two of them are actually triggered. Ken's mod simply allowed a single key to trigger all eight pairs at once. He then put the LFOs through a phase shift network with eight, 45° stages, each of which were then used to modulate each of the eight pairs of VCOs, thereby giving the rich chorus-like effect that had been the basis of the string machine design. With the mod switched in, it is of course reduced to monophony, and Ken tends to use it to get a big sound for single line parts, preferring the old string synth for the full chordal sounds.
In January 1980 he signed a contract with Jeff Wayne Music to write and produce jingles, and to allow himself greater freedom to build up complete pieces on his own at home, he updated his TEAC 4-track to a TEAC Tascam 8-track. His complete equipment list now reads as follows: a TEAC Tascam 8-track, a Revox A77 and B77, a Studiomaster 12:2 mixer, a Korg Rhythm 55, a Roland Space Echo, a JVC portable VHS video, a PET 32K plus dual floppy disc drive, 3040 series, two Commodore 5¼in. floppy disc drives, a JCL high speed assembler, a TCM heat sensitive printer, the old and faithful string machine, a Yamaha CS80 and his latest acquisition, a Synclavier 32 voice.
When Ken bought his Synclavier in May 1980 it was, once again, one of the first in the country. He saw an advert for it in the American magazine, 'Computer Music Journal', and flew out to Norwich in Vermont to take a look. "I was very impressed with the bowing effect on the string sounds, especially on the cello; for rhythmical work it's unbeatable. Each sound has a digitally transparent clarity, and again because it's digital there is an absolute precision about everything it does. I bought a 32-voice model because of the extra flexibility it allows; for example, a long glissando can be more effective because you can have thirty-two notes sounding before the first one cuts off, whereas with the CS80 you are limited to eight notes at any one time. It has a vast wealth of unusual sounds, with the software continually being updated. The facility to store data on disc is also invaluable, both in the studio and live. In the past it has often been the case that you'll get a sound that a producer likes during one session, and then at a subsequent session, having been asked to set the same sound up again, the producer will claim that it isn't quite the same as the original one, even though the settings are identical. With the Synclavier the exact sound can be stored on disc in seconds, and recalled in seconds any time it's required in the future, and there's no argument. Each disc can store up to eight banks of eight different sounds, giving sixty-four different sounds per disc. It takes less than a couple of seconds to recall a bank, and then recalling each of the eight individual sounds is practically instantaneous. Working live, this gives a very wide variety of sounds which can be accessed at the touch of a button, or at the most, the changing of a disc."
Ken continues to do a lot of work for Jeff Wayne and has therefore had a good opportunity to use Jeff's Fairlight: "The real sound facility is great, it makes it possible to make a tune out of almost any noise you want. It's limited though by its short sample length, which is only one-and-a-half seconds; after that it has to loop back and read the same data over again. This inevitably causes a glitch as it loops back, and although you have complete control over how much it reads and where it loops back to, you can spend ages trying to get a smooth, reasonably glitch-free loop, and still end up with something unusable; it's really too time consuming for normal studio work. Each overall sample period is split up into 128 samples, and if you can get each cycle of the sound to fit one sample, it gives much smoother loops, and so it is important to try and get the sample rate and the pitch of the sound right.
"The sample frequency also determines the cut-off frequency of the digital recording, and to get an overall sample period of one-and-a-half seconds, the high frequency response is brought down to about 6kHz which tends to make all the sounds a bit woolly. It's also a bit noisy for normal use: There's an odd hissing noise which accompanies each note as it's played. I don't know what it is and I haven't been able to get rid of it. Being able to process real sounds though, really does provide new scope: for one jingle we used the sound of a paper bag being screwed up, played back at a different pitch, backwards. Only the Fairlight could do that."
Since signing the contract with Jeff Wayne Music, a major part of Ken's work has been writing music to go with television commercials. When working with a visual medium, it is often necessary to have certain sounds or effects take place in synchronisation with specific points in the film or video. When working with film, there are so many different units to relate to a position in it, that complications and confusion often arise. The speed at which each individual frame of picture is moved past the projector's gate can be either twenty-four or twenty-five frames per second, and the units in common use for film length include frames, feet and frames, seconds, minutes and frames, and minutes and seconds. Ken has written a program for his PET to simplify the writing process. Film cue data can be entered into the PET in any of the above forms, and it will then list it in any of the other forms as selected. The most common method that Ken employs is to take the computer along to the viewing of the film, and referring to the countdown at the beginning, start its clock at the first frame of picture. As cue points arise in the film, he enters them by depressing another key on the keyboard, and they are then listed in order on the PET's screen. This timing sequence can then be 'played back' again, starting it at the beginning of the film. The PET will give out a blip at each cue point so that the producer can check that they are all in the right place. Any that are found to be a little out of sync can be individually adjusted in either direction. Having got all the cues in the correct positions, comments relating to each cue can be added to the list of timings on the PET screen, e.g. man turns corner, woman opens bonnet etc., and of course the data originally entered can be displayed in any of the previously mentioned units. When the tempo is entered for the piece of music, all the cue points are also printed on the screen as numbers of beats, and if the tempo is subsequently changed, the beat count is also changed accordingly. The cue points can then be used to trigger a synth or any other electronic instrument via a synchronous pulse from the user port, to give an effect at exactly the right point in the film. More often though Ken simply takes a VHS cassette of the picture home with him, and uses the PET to verify that he is writing the jingle as it was discussed at the run through, and that he isn't making any mistakes concerning the various cue points. A 4k7 ohm resistor connected in series to the line input of his mixer allows the PET to give an audible beep at each one.
Ken's style of writing often lends itself to the use of a rhythm machine, and so he has written another program for his PET which allows him to program a complete piece at home, and then if necessary, let the PET run it through in the studio, for the master recording. His rhythm machine is a Korg Rhythm 55, and although there is a difference in logic levels, the PET's 5V is enough to trigger it. The program is written half in BASIC and half in machine code using a JCL high speed assembler. The BASIC makes it operationally simple enough for an inexperienced user, and the machine code allows it to be fast enough for the demands of the program. Being an 8-bit micro, the PET is limited to the control of eight functions, which Ken finds to be sufficient for normal use. The eight available are: bass drum, snare, tom one, tom two, claves, cowbell, hi-hat and finally a tom-tom accent control which gives the toms a long or short decay time. A dot matrix is printed on the PET's screen eight dots high and up to thirty-six wide, with each line denoting a different function and each dot being a beat or rest. The eight lines are programmed individually, with an asterix replacing a dot wherever a sound is to be played and a dot remaining to denote a rest.
In this way a graphic description of the rhythm is built up on the screen as you write it, which is very useful, especially for non-drummers, forgetting a rhythm that's in your head into the PET accurately. At the beginning of each thirty-six by eight dot frame, information in the PET regarding the length of the next frame, its tempo and certain other functions that we'll look at later, is updated. If the rhythm pattern is to stay the same for perhaps eight or sixteen bars or so, you can use the full width of the frame as there will be nothing to update. However, if you want changes in tempo to take place within one or two bars, it will be necessary to have the frame updated within that period, so that the new data can be acted upon. These accelerando and rallentando functions allow the tempo to be speeded up or slowed down to a new tempo over any given number of beats. If at any point you want the rate of change or the direction of change to alter, the frame will have to be updated to allow the new information to be read in. The PET's 32K memory has a capacity for a maximum of 255 full frames, and the program has the following facilities:
FILL. Having entered a bar, it can be duplicated as many times as you like, saving you the trouble of entering the same pattern more than once.
COPY. You can copy any single bar into any other existing bar — if you come across a pattern that is better than an existing one, you can substitute it in as many bars as you like quickly.
MOD. To modify the pattern in an existing bar.
DELETE. To delete an existing bar.
INSERT. To insert an extra bar between two existing bars.
TEMPO. To adjust the rate at which each time division is clocked through.
ACCEL. To effect an accelerando.
RALL. To effect a rallentando.
PLAY. Will play from the bar entered through to the end of the piece or until it is stopped by depressing the 'space' key.
SAVE. Will dump sequential data from RAM on to disc.
LOAD. Will dump sequential data from disc into RAM.
There is a second version of the program which omits the tempo, rallentando and accelerando functions and allows an external sync pulse to be used instead.
The fact that the data update part of the program is written in machine code makes it fast enough to access the new data in between beats, and therefore without interrupting the rhythm.
Interview by James Betteridge