The two projects Agora (2003) and Boundary Conditions (2002) were the first collaborations between Birger Sevaldson and myself. These two projects are closely related in terms of concept, installation space and sound space.
Despite Boundary Conditions being the first that was exhibited, the Agora project was the source for most of its elements. The main different between the two projects concerns their presentation contexts: Boundary Conditions is a public space installation of unfixed duration; Agora a one-hour electroacoustic experimental theatre work.
Agora is an event where spatial elements are the structural and theatrical driving force. There is one soprano performer, but no 'acting out' of a strict theatrical structure. Instead the spatial disposition of sound via loudspeakers, music, physical resonating objects as carriers of sound, and the motion of the soprano through the space, assemble a musical-theatrical structure. The work extends the aesthetics of, on one hand, the music theatre of Kagel (such as his Staatstheater collection), and on the other, the consideration of delineated space defined in Oscar Schlemmer's theory of performance. From these starting points Agora gradually grew into an experimental architectural domain.
The "agora" in classical Greece was a political, economic and social meeting place. The Agora project presented a convergence in time and place of theatre, spatial architecture installation and electroacoustic sound. One of the main premises was that spatial information would be the structural driving force behind all aspects of the work – the music, the installation design, and the physical presence of the performer and the interaction of all these elements in three-dimensional geometry. In practice this meant that an audience would sit in the centre of the performance space, and all elements – including the installation – would be active through the three dimensions of the space. The starting point was therefore to find an appropriate spatial model.
The spatial model
Inspired by the previous ON project Chamberworks, which I had come across in an unrelated circumstance, I approached Sevaldson with the idea of developing a similar type of construction that would serve the basic framework outlined in the Agora concept. After various diverse investigations of different spatial models ranging from Įyvind Hammer modelling interference patterns of three-dimensional propagating sound waves to Sevaldson developing an idea similar to that of Chamberworks, we explored ways to model human motion in a defined space – this being far closer to the idea of an Agora.
The decision making process we use to avoid colliding with other people is relatively complex. However, the underlying principles for collision avoidance algorithms are readily available (and are frequently used when calculating crowd behaviour in enclosed spaces). Įyvind Hammer created a simple programme where 16 virtual "agents" or virtual people were each given personalities and goals that determined how they moved through a virtual space. This virtual space was a template for the performance space where audience and loudspeaker locations created fixed obstacles. The programme also allowed the user to specify a random distribution of virtual agent start-points and the number of time-steps for which the model would run. The output of this programme was then visualised as a time developing process in two dimensions.
In addition to the time-developing visualisation, the model created a matrix file of time and x-y co-ordinates for each agent. The underlying rules lead to a consistency such that regardless of agents start points the resulting visualisations showed similar agent interactions and resulting patterns. It was therefore safe to choose one single visualisation as the template for sound spatialisation, sound transformation and for physical installation design. The virtual model therefore became a real manifestation in the performance space.
Output from the model and influence on sound composition.
The materials for Agora came from many of recordings ranging from non-linguistic vocalisations, spoken texts (in English, Norwegian, Old Norse and Old English), environmental materials and resonating metal, stone, wood and plastic. Compositionally these sounds were developed exploring six psychological conditions: fettered, fear, play, calm, longing and anger. Although the composition involved many facets inappropriate to detail here, there are two ways in which the model influenced the sound that are relevant to describe: spatialisation and transformation.
The time-x-y matrix from the model controlled spatial trajectories of the composed sound material. Normal amplitude panning techniques do not create a sound trajectory in free space audible for anyone other than a single listener positioned very accurately (within 10 cm) in the centre of the loudspeaker array. To achieve the spatial control and audience coverage needed in Agora I use second order ambisonics projected over 16 loudspeakers. Ambisonics is a three-dimensional sound field technology that approximates the sound field as it is in reality, rather than using unstable phantom imagery characteristic of amplitude panning techniques. In the electroacoustic part of Agora many clear sound trajectories follow the curves of the installation – the sound developing through real-time while the installation is fixed.
A sound transformation method known as granulation was used for certain materials. Granulation involves a process of segmenting a sound into many short fragments (grains) and assembling new sound from these fragments. The way the grains are cut and re-assembled is controlled by many parameters such as grain size, splice window, pitch, filter, density, read and write pointer range, time variation and the randomisation of a variable within a defined range. Changes in these parameters result in varying degrees of densification and complexity in the resulting sound. The data created by the model was used to control these granulation processes such that an increases in spatial complexity resulted in a parallel increase in sound complexity.
Electromechanics and resonance
Electromechanics were used to set the installation into acoustic resonance with the aim of filling the continuum between spatial electroacoustic sound projected over loudspeakers, moving acoustic sound made by the soprano and the spatial physicality of the installation itself. The choice of some installation materials therefore needed to consider resonance properties. Solid aluminium rods, when freely suspended, were tested amongst many metals and found to produce loud, clearly pitched sounds of lengthy sustain. Lexon foils under uniform tension were found to capture sound vibrations in a similar way to a drum skin.
Two types of computer controlled electromechanics where then choreographed with the music:
(a) Small electronic actuators received voltage inputs from audio amplifiers. These devices inputted sound making vibrations into eight tuned tension lexon membranes. These membranes amplified the vibrations and created highly coloured audible sound.
(b) Computer controlled electromagnets providing a 50 Newton force were used to hit long aluminium bars (3 to 6 meters long) suspended above the audience.50 Newtons is a significant force that caused a large number of installation elements to quiver with energy. The lengths of the bars were chosen to result in non-harmonic ratios and thus avoid a direct connection to a musical instrument.
Although in Agora the installation and music composition were significantly integrated, part of the project involved finding ways to dislocate these elements in time and space. It was therefore particularly interesting to attempt a version of the work that would function in a public space in relation to the goal-orientated flow of people going about their everyday. This revision of Agora became known as Boundary Conditions, installed for the period of a month across a low trafficked elevated public walkway in the very centre of the Oslo Central Station.
Boundary Conditions involved a number of important changes from Agora. Here I will discuss those changes connected to sound.
The Oslo Central Station has a relatively low but constant background noise level (due to the size and shape of the building). Frequent and regular announcements are given over the station system and need to be heard by the public. There are also a number of restaurants and shops in the immediate area. These factors meant that the sound needed to be of lower volume than Agora and in a tighter sound-space. Instead of 16 large loudspeakers located throughout the space (as in Agora), eight small loudspeakers were used in a cubic array positioned centrally on the walkway. An ambisonics sound field was projected over this array such that the sound level propagating outside the array was significantly less than with normal playback, while when inside the array the volume appeared greater. The sound was so tightly controlled that four additional loudspeakers were hung underneath the bridge playing very low volume sound. This gave people a clue that there was in fact something happening above other than the visual aspect of the installation.
Boundary Conditions ran continuously throughout the day and the dramatic musical structure of Agora was inappropriate. Instead a new sound composition was made from the Agora materials. This work consisted of layers of small details and gestures, and focused on vocal sounds and shorter temporal developments while maintaining many of the spatial characteristics that mimicked the installation's forms. These layers of sound were played in random permutation controlled by a simple computer programme. An infrared sensor activated the sound each time a person entered the walkway, after which the sound would last for 5 minutes if no one else entered the area. Each time the sensor was triggered a short permutation in the electromechanics was also activated and served to inject both acoustic vibration and real motion life into the installation.