Abstract
In this paper we discuss the use of a recently developed audio compression approach: Audio Compression Exploiting Repetition (ACER) as a compositional tool for glitch composition and remixing. ACER functions by repeating similar sections of audio where they occur in a file and discarding the repetitive data. Thresholds for similarity can be defined using this approach, allowing for various degrees of (dis)similarity between materials identified as 'repetitive'. Through our initial subjective evaluation of ACER, we unexpectedly discovered that the compression method produced musically interesting results on some materials with higher levels of compression. Whilst listeners demonstrate this level of loss of fidelity to be unacceptable for the purposes of compression, it shows potential as a performance or production tool. When applied to pop songs the predicable form of the music was disrupted, introducing moments of novelty, while retaining the songs quantized rhythmic structure. In this paper we propose the use of ACER as a suitable method for producing sonic materials for 'glitch' composition. We present the use of ACER for this purpose with regards to a variety of materials that may be suitable for glitch or electroacoustic composition and using ACER in several different ways to process and reproduce musical audio.