HVCCA Version

*asterisk was commissioned for a festival of public art in Peekskill NY. The piece took six months to complete, the first stage being about one hundred hours of observation and audio recording around downtown Peekskill, along with recording “staged” sound events. The second stage involved the editing and composing of the audio material (sound library), creation of the score, design of the software to realize the score, design of the sculpture, and installation/testing.
The idea behind *asterisk, and several other public sound projects I’ve taken on, has been about finding ways to observe and then work with the localized sound memory of a public’s relationship to a place, extracting and reconfiguring this aurality. In *asterisk, I created a life and memory cycle for these sounds, extracting them from place, processing them through a sculpture built from objects associated with the place, and then transmitting the output of this sculpture back into the environment.
*asterisk uses upturned broken pianos as resonant bodies, as vessels. Sound is projected through the pianos with loudspeakers coupled to their soundboards (the wood that amplifies the very quiet natural sound of the struck strings). On the other side of the soundboard are the strings. Above these I suspended microphones that would listen to the filtered/embodied sound of the piano, rather than the loudspeakers directly. In the first version, installed at the town’s historic Ford Pianos, there were enormous plate-glass storefront windows, with the sculpture visible from the street. I attached special vibration transducers that turned the entire storefront into a low quality loudspeaker. Pianos, with the dampers raised will sustain the last sound that has been made, usually the latest strings to be struck. In the case of my piece, the pianos were broken and I was “reanimating” these bodies with speakers, more or less using them as a means to embody sound at life size scale, something speakers can not do on their own. As the composition is played from a computer, my software was also sensing how loud the resonance in the pianos was — and when this resonance had built up to a sufficient volume, the speakers would stop, leaving only the resonance of the pianos audible — this sound alone, a ghost-like spectra created by the synergy between the broken instrument and the accumulation of resonances created by the loudspeakers, was transmitted through the storefront windows into the street. Given this mode of transmission, it was hard to tell where exactly the sound was coming from.

About the Score

I won’t go into extensive detail about the score (software), which in this case is a set of 179 sound files or “sound events” organized into “states” (collections of events that are allowed to occur simultaneously). Instead, I’ll describe it generally and then zoom in on the way each sound was interpolated as one example of depth. For a reader who is not a composer this might seem overly technical or esoteric.
The algorithmic composition was structured as a conditional probability structure which results in a series of modular “bagatelles” — or short interwoven episodes — punctuated by silences after sufficient acoustic accumulation was sensed by the microphones within each piano. They were highly layered, and each sound was used as a semi-transparent surface through which other sounds can peek through, modifying the meaning of another through mutualized perforation.
The 179 sound files ranged from a few seconds to five minutes. A sound could be a simple timbre, a piano note, or a two minute collage of interweaving events. I shaped each sound file within the library into a “Sound Event” that encompasses the entire manner in which a sound exists within the sculpture, its signature. This allows it to maintain a more unique identity in many contexts (since most sounds were part of a cluster of different “states”). Below are the variables I composed for each sound, the sum total of which creates the identity. Each variable below was not a static value but followed individual morphologies that I composed as graphs.
Attacks per Second

Number of movements from silence to sound within one second. This corresponds to starting and stopping a method of playback and allows for very perforated iterations, the playback of an entire file, and everything in between. An indication of density within the duration of the entire Event — the number of links in the chain.

Attack Duration

Duration, in seconds, that the the audio file player will play for each attack. The length of each link in the chain.
Volume Envelope

Volume envelope for the entire Event. This allows dynamic entrances and exits (emergence, disappearance, where a sound causes the most resonance, etc)
Event Cycle Duration How long the entire event is, the scale for interpolating all aspects of the Sound Event.
Spatial Assignment

A constellation of output channels. A Sound Event may be sent to all output channels (tutti) or just one (solo). Every sound has a specific manner in which they are embodied by the pianos.