Alistair Riddell

By using special electro-mechanical attachments to the piano action, the strings of the piano may be struck using a computer. And because there is no human intermediary, the composer is able to modify the composition almost immediately, then re-submit it for performance. The instrument also performs beyond the limits of human physical capability - in terms of the number of notes struck, speed of note attacks, tempi and the duration of the work. These extended performance techniques largely depend on the instrument used and whether the mechanism can withstand the activity. Another important feature is that two or more instruments can have their performances synchronised.

Riddell's work has helped to extend both the musical possibilities of the piano and the range of composition for the instrument. In addition, he is exploring new ways to make immediate, real-time changes to his music, allowing greater scope for improvisation.

His musical education began at age 14 when Riddell began studying classical guitar; and for the following five years he also took private lessons in music theory and developed a general interest in stringed instruments.

Although the classical guitar possesses a wide range of sonorities and allows an extensive repertoire of techniques, Riddell also took an interest in techniques and performances of other guitar styles, such as rock, folk, jazz and blues. His appreciation of the guitar's inherent flexibility also led to an interest in other fretted instruments, particularly the lute.

Scattered throughout the repertoire of music for classical guitar are numerous transcriptions of Medieval, Renaissance and Baroque lute pieces. The guitar allows a contemporary interpretation of these works, which might otherwise be neglected, even though they sound different when interpreted on guitar rather than the original instrument.

Riddell acquired his first lute in 1973 and began to study and play Early Music by such composers as Ferrabosco, Howitt, Dowland, Ballard and Cutting. His interest in the lute culminated in a trip to England in 1975, where he attended the Lute Society's summer school. He remembers the experience as important for the contact it provided with European and American enthusiasts and performers.

In 1978 he attended La Trobe University. Originally, he had not intended to study music formally, but changed his mind prior to first year, after being inspired by a jazz arrangers' seminar hosted by the university's music department.

Riddell's interest in composing was furthered by both the structure of the course and the music department, which included on its staff composers such as Lawrence Whiffin, Graham Hair, Keith Humble, Warren Burt and Jeff Pressing. Riddell had previously seen himself as an instrumentalist and performer. But even in performance, he realised, he was principally attracted to compositional ideas. Composers had - at least in theory - a wide range of instruments at their disposal, without the need of a matching instrumental facility; and there remained a valuable distinction between conception and implementation.

The course at La Trobe took a balanced approach, contrasting the music of the past with that of the present. The formal, pedagogical teaching of composition was relieved by talks and workshops by visiting performers such as Bertram Turetsky (double bass) and composers such as Wolfgang Hufschmidt (Germany). Emphasis was placed on technique, discipline and originality. Early in the course, Riddell was exposed to computer music - both in a practical sense and as a hailed new direction for contemporary music. He programmed micro-computers to produce sound and was able to observe the more substantial work by his colleagues, such as Graeme Gerrard, Brian Parish, and David Hirst, who were working in the department under Jim Sosnin. They approached computer music using `Software Synthesis', a technique whereby the computer constructs digital data representative of sounds which may then be input to equipment (such as a synthesizer) which realises these sounds. Riddell gained a passing experience of this approach, but was not at the time generally interested in computer music.

After leaving university in 1981, however, he immediately launched upon a computer-based music project which was to result in the computer-controlled piano - work with which he has been associated ever since.

Riddell was originally inspired by his research into the music of Mexican composer Conlon Nancarrow.

Born in Texarkana in the USA in 1912, Nancarrow has been a resident of Mexico City since the early 1940's and has produced a large number of Studies for player piano. The volume and consistently high quality of his work has made him internationally distinguished in new music circles in recent years, although for most of his composing life he has lived in relative obscurity.

Nancarrow's staggering corpus of player piano studies - the result of more than 40 years work - had influenced Riddell profoundly. In Nancarrow's work he perceived ideas and techniques that could be further explored through the application of computer technology to the piano. On a simple and practical level, this would establish a more versatile composer-to-instrument relationship.

Because the strings of the piano are able to be struck in rapid succession, with great precision, and from all points along the keyboard, it allows the instrument to be explored and extended in ways that are not possible with a human player. Also, the absence of a human player/interpreter helps collapse together the polarities of conception and execution.

Riddell's first working system consisted of a small Kawai piano fitted with a commercially available Marantz `Pianocorder' system. It was installed by Joe Kade, a piano technician, and the first computer interface was achieved by John Luke.

The Marantz system is basically an electronic player piano mechanism and in its original form required some minor modification to be able to communicate with a computer. Data representing the notes, dynamics and fundamental operation were also inefficiently organised , but nevertheless provided a means for the instrument to be played from the computer. The player mechanism operated when a series of electro-magnetic solenoids - one for each key - responded to certain codes. These solenoids are quite sensitive and can be controlled with some degree of precision, thus allowing a more expansive and versatile dynamic range than is possible with the traditional player piano.

His early compositions reflect the limited technological resources at his disposal, as much as his unique perspective on the piano. Works composed in 1981 and 1982 are short and rigid in execution. Riddell's compositions had to be manually translated from graphic scores into very low-level information that could be keyed into the computer. The process of composition, therefore, was not facilitated by the computer, and on the whole was very time-consuming. Riddell recalls the experience with a mixture of awe and frustration. At the time, software that would have facilitated the procedure was not available, and there was no alternative to this tedious method of working. However, the results gave great encouragement.

Early works like One 1, Two 1a and Existential Constellations hinted at possibilities to come and, although short and succinct, were fascinating to watch in performance. Recordings of some of these pieces appeared on the New Music Articles cassette series (NMATAPES 1). In 1982 and 1983 they were also broadcast on ABC FM and AM radio, and on Melbourne public radio stations.

In 1983 Riddell examined ways in which an improved piano action could be designed to allow a greater diversity of sounds through even faster string attacks and by modifying the hammer and dampening operations. In May of that year he received a Jacobena Angliss music award for computer research, becoming one of the inaugural recipients of this biennial prize. This provided a significant boost to the project, and a new system was completed within the year. It consisted of new software, a new method of interfacing the computer to the piano, a new computer system and a new piano.

The software, written by Chris Vaughan, defined an environment in which the instrument could be played and the score examined. It translated a higher level symbolic representation of the score to the level at which the machine operated. It was indeed a quantum leap from Riddell's earlier approach to the piano. New interface hardware was built by Harry Tagaris and Riddell to provide a more satisfactory link between computer and piano, which allowed a greater degree of performance capability than before.

Two `interfaces' were constructed, one for the `traditional' and the other for the `modified' piano. The latter instrument was constructed at the Brashes/Allans warehouse, West Melbourne, under the supervision of Joe Kade. The important difference between it and other pianos was that it could only be played by the computer. It had wooden hammers, giving it a sound similar to a harpsichord or harp, and a different means by which the strings were dampened.

This instrument was intended to provide a vehicle suitable for the consideration and application of some of Nancarrow's techniques; specifically, composition with `aggregates', which involve dense clusters of complex note structures. Since the system was very new, experimentation into the behavior of the instrument and the electro-mechanical action is clearly evident in the early works.

The project was finally complete with the addition of a new computer system based on the Z80 processor with 128K bytes of memory which ran under the C/PM 3.0 operating system.

Riddell completed his first composition, Atlantic Fears, using the new system in November, 1983, at the same time he had been invited to the Interface Exhibition - A Survey of Art and Technology, held as part of the 1984 Adelaide Festival of the Arts.

Another work from this period, was Variations For Two Pianos, completed in 1984, which explores the timbral diversity on the new modified piano, and its relationship with the conventional instrument.

Variations is characterised by the opening pitched drumming effect from the modified piano, with variations produced by the addition of new material at each repetition of an original sequence. The instruments are presented separately and then playing together. The main motive activates the instruments in unison, then homophonically - by delaying the performance of one instrument only enough to cause a kind of enhanced reverberation, which is often clearly audible as a temporal separation of identical musical fragments.

All Riddell's works from summer of 1983/84 were confined in length. Data had to be stored in the computer's memory prior to performance - otherwise it might be interrupted at random when the computer accessed a disc for additional information. In other words, longer or more complex works would always pose a problem, due in part to the process whereby the score is converted into information sent to the pianos. (This translation procedure is also common to other forms of computer music.)

As this problem and other technical issues arose, Riddell realised he required a substantial background in computer science, and in 1985 completed a postgraduate diploma in the subject at La Trobe University.

This brought many new insights and, armed with the skill to tackle his earlier problems, an alternative method of composition, which largely overcame the memory impasse, emerged in 1986.

A piece called A Tale from Transitions resulted - which was assisted by the Music Board of the Australia Council. It consisted of a collection of carefully chosen musical fragments which could be linked seamlessly and performed in realtime. Although the fragments had been created and refined at an earlier time, their arrangement could be decided during the performance. The choice of each fragment was determined in context; that is, by the fragment that had preceded it. This process - where history determines the present - is known in the terminology of computer science as `state machine'.

The order of fragments could be biased towards certain favorable combinations, resulting in recurring clusters. Identical repetitions of fragments were unlikely, because of the changing context.

After working on the `state machine' system, Riddell considered how he could gain access to and influence a computer-generated performance of his piano system. Was it possible, for example, to let the computer perform while occasionally causing some change to the performance, in effect overriding the computer's decisions? Although Riddell has yet to develop such a system, his interest has become centered on real-time computer music.

Riddell then developed a system focusing on the behavior of the modified piano. Improvisatory in nature, it involved only the 24 bass strings of the instrument. By chance, some years before, he had discovered that when the solenoids were left on those strings, the effect was similar to that of a bowed instrument, generating sounds that varied from cello-like tones to a continuous harsh buzzing. Different pitches could also be obtained from the one string and the transition between these pitches was characterised by a delicate legato.

The system that controlled this behavior was completed in early 1987. Importantly, its operation was entirely controlled via the computer keyboard, and was therefore directly related to the operator's or performer's actions. While certain computer keys had been programmed to activate the solenoids, others controlled the dynamics or (in this case) the energy allotted to them.

It is not the first time such an approach to computer-generated music has been taken. However, it is unique because of what is being controlled. The strings, for the most part, behave somewhat unpredictably. Discovering how to produce and control desired sounds remains one of the challenges and attractions of a real-time system.

Such direct manipulation of sound from a computer-controlled instrument perhaps indicates the way some composers might choose to work and, indeed, real- time computer music is attracting considerable attention.

However, while computer-controlled acoustic instruments may not be a significant direction for the future of computer music, Riddell believes he has gained some valuable insights into the function of computers in composition. On the other hand, he is certain the piano will remain an important medium for experiment for some time yet. From his own experience, it has abundant resources.

Riddell's recent activities include the completion of a Masters Degree in computer music at La Trobe University, while working as a data communications programmer for Telecom Australia.

RECORDINGS

"ONE 1", "ONE 1A", "TWO 1", "TWO 1A" (1982), NMATAPES 1, NMA Publications, 1982.

"Core Image", Computer Music, NMA Publications, 1983.

"A Tale from Transitions", hear NOW, National Programming Service - Public Broadcasting Association of Australia, Sydney. 1986.

22 Contemporary Australian Composers index

NMA Main Page