Noise Sculptor
The Noise Sculptor is a noise-based synthesizer developed within the Max/MSP programming environment. During the course of my applied composition studies with Dr. Cohen Elias at the University of Iowa, I began to develop an idea about building a complex spectrum from highly filtered strands of noise. More than this, I wanted to see if I could build a spectrum based on the profiles of sampled acoustic instruments— strings, church organs, etc.— and then interpolate that spectrum’s states back and forth between those various models. The result of this experimentation was the Noise Sculptor.
The Noise Sculptor is, basically, a very elaborate and intensive fleet of filter banks based on the [fb~] object. There are forty-eight individual filter banks, and each is situated around a conventional musical note within the range of concert C2 - C6. The Q value for each "note" can be calibrated such that tones of varying clarity can be produced.
​
Every filter bank within the fleet is programmed to put out not just one but twenty individual “tones" mathematically arranged according to the overtone series, so that the output resembles a complex tone of twenty partials. The amplitude of each partial can be calibrated, allowing the user to produce sounds of varying timbres. These two features are the core of the instrument: intensively filtered noise that can be sculpted to produce complex sounds with up to twenty partials. Hence the name, Noise Sculptor.​​
The instrument has several other features as well, including settings for the tuning of the partials and the tuning of the chromatic scale. The tuning of the partials can be altered by raising or lowering the multiplication factor of the fundamental frequency for each individual partial. And in addition, the notes of the chromatic scale can be raised and lowered by the cent without limit.
​
I have been largely very pleased with the results of my work on the Noise Sculptor. As one might expect, the ability of the device to recreate the spectral profiles of acoustic instruments is limited– there’s more to it than the amplitude and tuning of partials, or the details of envelope. Even so, some of the states do recognizably approach the models they were based on. The “organ” state is the most aurally similar to its model, followed by the “electric guitar” (timbrally more than the envelope with its particular a!ack transients). The “strings” se!ing bears similarities to samples found in other synthesizers, but is a good deal further from its model than the “organ.”
​
Beyond this, the settings can be arranged freely as well, not based on any existing model but calibrated according to the user’s ear. The user can interpolate not just linearly but multi-dimensionally, using the [nodes] object paired with the [pattrstorage] object’s preset configuration. This means that the user can produce sounds at various positions between multiple states as well.
While developing the primary instrument, ideas for two alternative versions have also come into being. The first is above on the left, called the Noise Blaster. Instead of a fleet of filters, three filters are used to partition a sound into three more or less aurally equal bands. These bands can then be enveloped, individually or in varying combinations. This allows the user to play back an audio file of their choice and “play” it across its spectrum. If the sample used is a noise drone, then the output is blasts of noise at differing spectral positions, i.e., a Noise Blaster. And the second is a percussion instrument, producing four sounds modeled on common electronic percussion instruments (a bass drum, a snare drum, and two clap sounds– again, all made from noise).
​
Rather than "new sounds" for novelty's sake, with all of these instruments my interest has been in the use of transformation as a mode of artistic expression: sounds that can transform into something else, and music that uses transformation as a key element in its music-poetic language. These ideas are what lie upon the next horizons for me with the Noise Sculptor, and toward which I intend to work. In addition to this, I hope to next explore other technical aspects of the instrument's design:
​
- possible alternative or more variable modes of filtration (what kind of control can I assert over the waveforms produced by filtration, beyond center of frequency or value of Q?)
​
- use of the poly~ objects as potential CPU saving components (as opposed to 48 individual instances of the primary patch)
​
- a wider variety of mathematical arrangements for the "harmonics" of the generated tones, including: arrangements and mathematical deviations more closely modeled to specific values common to existing instruments; and more mathematically novel arrangements of the harmonics.