Produced in the Studio for Electronic Music of the WDR, Cologne in 1976, premiered in Darmstadt 1976

Listen to 5 German Dances on Sound Cloud ►

Upon receiving the commission, my first thoughts for this composition were heavily influenced by the technical environment in which I would be realising it: the Electronic Studio of the WDR. This is the same studio - only slightly enlarged and modernised over the years - in which Stockhausen had made Gesang der Juenglinge, Kontakte and Hymnen. The kit which he used or had had built by technicians of the radio was still standing around hooked up, most of it broadcasting or test equipment from the 50's, such as huge W49 variable band-pass filters, enormous super-stable K&B sine wave generators or massive impulse generators, not to mention the two big reverberation plates in wooden crates which had, fortunately by that time, been moved to the cellar. And besides that, huge German broadcasting quality mixing desks, one 8-track tape deck, six 1-inch 4-track tape decks and numerous stereo 1/4-inch stereo and mono decks, some of which could run at 76cm/sec. All technically perfectly maintained, very solid, but pretty old stuff for someone like me who had played around with a Moog and already done several live performances with a VCS3.

As a result of all this big heavy stuff for making, treating and recording sound I began planning to make a piece in which I would treat sound if it were a physical object with mass, density and dimension. I intended to subject it to as many basic physical manipulations - compression, expansion, tension, torsion, compacting it, carving bits out of it, exploding it, tossing it around like a ball, accelerating and decelerating it, and so forth - as I could devise.

In the meantime, in the several months that elapsed between the date of the commission and my allotted three-month stint in the studio, a miracle took place: the Technical Directorate of the WDR had finally relented to pressures from Stockhausen and the permanent studio technician and had agreed to the purchase of a new analog synthesiser, an EMS Synthi 100 from the stables of the VCS3. And to top it all off, they also agreed to the purchase of a digital reverberation-delay unit, a Timeline. This was a miracle because up until then, the Technical Directorate (which was in charge of the maintenance of all electronic equipment in the radio) had always blocked the purchase of any commercially available music kit, maintaining that it was all of technically inferior quality, that they didn't want it in their house and, anyway, if something went wrong with it, they wouldn't be able to fix it.

In order to achieve my goal of making complex sounds that seemed to be 'objects', that were identifiable as being one 'thing' and not just mixtures of many things, I began to experiment by making mixtures whose components were phase-related, i.e. complex sounds consisting of many layers of different frequencies whose phase-relationship remained constant. During earlier examination of sounds, I had noticed that mixtures consisting of 'phase-synchronous' or 'phase-locked' frequencies sound as if they emanate from a single source, from, for example, an instrument or a vibrating object. It seems that although our ear cannot distinguish the absolute phase of a sound, it is very sensitive to any change in the phase relationship between its components. (Stockhausen's groundbreaking work in KONTAKTE is a prime illustration of this: much of the basic sound material of that piece was made by recording rhythms of pulses of varying intensities on tape, making loops of the tapes and then accelerating the loops until the rhythms turned into sounds.)

The EMS Synthi 100 had a provision for locking the phase of many of its oscillators over a fairly wide range of frequencies: I was able to make phase-synchronous mixtures of the whole-number multiples and fractions of a single frequency up to the 13 th harmonic and 13 th subharmonic. These mixtures were, however not particularly stable and did not like being transposed from the original pitch. (One of the problems with the Synthi 100 was that it would drift; often I had to resort to setting up and tuning a mixture, closing all the windows and turning off the air conditioning and letting the thing sit for a few hours until it had found the temperature at which it was stable for a while - a good excuse to take a walk.)

The mixtures were made with 4 phase-locked oscillators controlled by the onboard digital sequencer running at approximately 10 pitches/second. The output of the oscillators was simultaneously feedback with itself over the read-record heads of the tape deck and filtered by bandpasses also under control of the sequencer. The resulting mixtures were soft, colourful, spatialised 'natural' sounds.

(In the following descriptions I will not go into detail on how exactly the sounds were made, as the techniques are analog, very old fashioned and probably no one who has not worked with this kind of equipment would be able recreate them. For any one who is interested, however, the techniques are described in detail in 5 Deutsche Taenze Darmstaedter Beitraege zur Neuen Musik, B.Schotts Sons, Mainz, XVI:1976 ; digital technology makes room for far greater ease and scope in the manipulation and transformation of sound.)

Another important process I examined was the trusty ring modulator - the old standby of any analog studio (a device developed for telecommunications in order to increase the bandwidth telephone wires: several different signals can be transmitted on the same wire if they each - with a limited bandwidth - are transposed to a frequency region outside the others; a ringmod. has two inputs - signal and carrier - and one output; the output signal consists of the sums and differences of the two input signals). I focussed on the decays of extremely low or high resonating bandpass filters which had been set into motion with short impulses. The resulting sharp chirps and clicks are to be found in Dance I and Dance V. Furthermore, I found another way to create very 'concrete' sounds - uncannily like those of Scottish pipes - with the ring modulator. These may be found in Dance IV where they were used to record the following four different traditional pieces taken from "Highland Pipe Music" by Angus McKay: Cumha Mhic Shimidh 1746, Blar Bhaterloo 1815, Iomarbhadh Mhicleod 1603 and Cumha Chraobh Na'n Teud (ancient). These recordings were then slightly blurred and treated so as to move about spatially as if they were carried in by the wind from a great distance.

A further process - one which was to become the main tool of transformation - was multiple 4-channel variable phase shifting of white noise (the process is akin to how 'moiré antique' is made: two layers of cloth of similar weave, one of which is slightly rotated and shifted in respect to the other, are pressed with great force into each other, leaving behind the pattern of the other's weave in each, resulting on the characteristic shimmering, liquid pattern). Now commonly known as 'flanging' , when this is done repeatedly with the same material, employing white noise as the starting material, the resulting sound begins to become coloured, almost sculpted, as fairly distinct frequencies gradually emerge to come briefly into the foreground and disappear again.

The Synthi 100 had a great advantage over all previous analog synthesisers: it had an 8-channel digital sequencer for control voltages. This allowed me make sounds that I called 'vector' sounds: sounds which behave as if the were objects realistically moving about in 3-dimensional space, emitting sounds. These were all made in real time, by voltage controlling sets of oscillators, bandpass filters, reverberation units and amplifiers, with the following sets of voltages (relative amplitude of signals):

When these 5 voltages were used as the control voltages to the stereo output amplifiers (1. & 2.), the Q of stereo bandpass filters (also 1. & 2.), the relative frequency of the oscillators producing the sound in question (3.) and the amplitude and depth of the reverberation of the sound (4. & 5.), and the sounds are heard by a person placed in the middle of a 4-channel speaker system, then these sounds would move realistically (as objects emitting sound) in orbits around the listener, passing through the middle of the head at that point marked 'middle' in the diagram, and disappearing into the distance, only to return again, as below:

The final structure of the piece is as follows: five sections - the 5 Dances -, each focussing on a different approach to sound as object.

Dance 1: sounds made of mixtures both harmonic and subharmonic, most of them massive, with complex spatialities and inner rhythms, as if they were the sounds resulting from some strange processes of mechanical production; sound as a completely synthetic object.

Dance 2: a single sound with a natural harmonic spectrum of 32 harmonics, spatialised, rhythmicised and analysed into its components; a sound as an object that can be sculpted and shaped, the process also revealing its structure.

Dance 3: all sounds moving in orbits of different sizes, each with a lifetime of its own, all produced with the 'vector' process mentioned above, embeded in a complex moving, spatialised white-pink noise, in order to give the listeners the impresson that they themselves are also moving; sound as emitted from moving objects.

Dance 4: a complex 'soup' of sounds - superimposed traditional melodies - all with the same subharmonic spectrum; sound created from found, cultural objects.

Dance 5: sounds made of mixtures both harmonic and subharmonic, all derived from the sounds of the human voice through multiple flanging, superposition and transposition; recogniseable concrete sounds (e.g. speech) transformed into different recogniseable sounds (e.g. birdsong).

Listen to 5 German Dances on Sound Cloud ►