LSL: A specification language for program auralization

Abstract

The need for specification of sound patterns to be played during program execution arises in contexts where program auralization is useful. We present a language named LSL (Listen Speciikation Language) designed for specifying program auralization. Specifications written in LSL and included in the program to be auralized are preprocessed by an LSL preprocessor. The preprocessed program when compiled and executed generates MIDI or voice data sent through a MIDI interface to a synthesizer module, or via audio channels, to an audio processor, which is transformed into audible sound. LSL has the generality to specify auralization of a variety of occurrences during program execution. It derives its broad applicability born its few generic elements, that when adapted to any procedural programming language, enable the use of LSL specifications for auralizing sequential, parallel, or object oriented programs. We view LSL as a useful tool for building general purpose multimedia applications, auditory displays, and for research in program auralization.