Growing Musical Forms With L-systems: Interpretations, Sound Synthesis and Control Schemata (2007)
(available upon request)
Abstract:
This paper introduces Lindenmayer systems (or L-systems in short) as an algorithmic method for designing emergent musical structures in all the time-levels of a composition, from macroforms to waveforms. The algorithm, a formalism related to abstract automata, fractals and formal grammars, consists of different sets of rules used to recursively transform and interpret symbol strings. The model described here can generate multiple networks of musical agents, each capable of performing any musical task supported by the implementing software. These agents move in a hierarchic parameter space while acting upon the compositional environment and interacting with each other. Most weight is given to the implementation of the model and its possible interpretations, with a special focus on the methods for generating waveforms. Extensions of the model are discussed, such as incorporating genetic algorithms, designing hierarchical L-systems and L-system networks and opening the system to real-time interaction.
-----------------------------------------------------
Musical L-systems (Master's Thesis, Institute of Sonology, 2006)
download here! - (sound examples available upon request)
Abstract:
This dissertation presents an extensive framework for designing Lindenmayer systems that generate musical structures. The algorithm is a recursive, biologically inspired string rewriting system in the form of a formal grammar, which can be used in a variety of different fields. The various sub-categories of L-systems are presented, following a review of the musical applications by other researchers. A new model for musical L- systems is shown, based on the author’ s implementation in Max/MSP/Jitter. The focus is on the musical interpretation and mapping of the generated data for electronic music, as well as in expanding the basic model to include various types of control schemata. Some compositional strategies are suggested for generating musical structures from the macro- level to the sample level, supported by a number of audio examples.