Publications — Computer Music, CSIRAC and Algorithmic Composition

The Music of CSIRAC: Australia’s First Computer Music. Common Ground Publishing, Melbourne. Includes a  CD of precisely reconstructed CSIRAC music.
Available from the author or Amazon 

A PDF of the book is available here.  

Full project documentation, reconstructed recordings and image gallery on the CSIRAC page.

(Co-curator & contributor) Musica ex Machina — Machines Thinking Musically. Exhibition guide and catalogue, EPFL Pavilions, Lausanne, Switzerland, 2024–2025. With Sarah Kenderdine.
Exhibition page

“Strategies for Teaching Audio Production Online.” In Music, Technology, Innovation: Industry and Educational Perspectives. Routledge, co-authored with Jason Torrens.

“A Chronology of Computer Music and Related Events.” In The Oxford Handbook of Computer Music, ed. Roger T. Dean. Oxford University Press, pp. 557–584.

This chapter is the basis for the expanded Chronology of Electronic and Computer Music on this site.

“Early Hardware and Ideas in Computer Music.” In The Oxford Handbook of Computer Music, ed. Roger T. Dean. Oxford University Press.

“Rethinking Assessment Design and Academic Integrity in the Age of AI: From Artefacts to Processes.” Innovative Teaching and Learning Vol. 8 No. 1 (2026), pp. 116-136 DOI: http://doi.org/10.4208/itl.20260107  https://www.global-sci.com/itl/article/view/24252

"A Framework for Evaluating Artistic Significance in the Age of Generative AI" Leonardo, MIT Press. Published 06 July 2026. DOI: https://doi.org/10.1162/LEON.a.2757

This essay presents a framework for evaluating artistic significance, developed through decades of algorithmic art practice, and applies it to contemporary generative AI art.

The framework identifies five characteristics of profound artworks: significant effort in creation, masterful technique, conceptual coherence, intellectual challenge, and deep human engagement rooted in lived experience. When applied to current prompt-driven AI generation tools, this framework reveals limitations in communicative intent, embodied technique, and the value of creative struggle. While acknowledging pioneering computational artists and emerging sophisticated AI tools, the analysis suggests that much contemporary generative AI art prioritizes rapid output over the developmental processes that infuse traditional art with depth. This framework offers artists and audiences a tool for articulating value and navigating aesthetic judgment as generative technologies evolve.

“Early Computer Music Experiments in Australia and England.” Organised Sound 22(2), pp. 297 - 307. Cambridge University Press. DOI: https://doi.org/10.1017/S1355771817000206  

A comparative examination of the earliest computer music activity in Australia (CSIRAC, Sydney and Melbourne) and England (Ferranti Mark 1, Manchester), drawing on primary sources. Contextualises CSIRAC within the global history of computer music and examines the parallel, independent development of musical computation in the early 1950s. Related events are documented in the Chronology of Electronic and Computer Music.

“How Australia created the world’s first computer to play music.” The Conversation.

"Pre-composition and Algorithmic Composition: Reflections on Disappearing Lines in the Sand" Context: Journal of Music Research, Nos. 29 & 30 (2005), p. 47. University of Melbourne.

A reflection on the role of pre-compositional thinking in algorithmic composition generally, examining practice across multiple composers. Considers how the distinction between designing a generative system and composing a work becomes increasingly blurred as algorithmic methods are employed.

“Computer Sound Synthesis in 1951: The Music of CSIRAC.” Computer Music Journal 28(1), pp. 10–25. MIT Press.  DOI: 10.1162/comj.2006.30.3.83

 Download PDF

The primary research paper documenting the CSIRAC music reconstruction project. Describes the computer’s sound-production method — raw bit-stream pulses converted to audio through a loudspeaker — and the process of reconstructing the music from original punched-paper program tapes, circuit diagrams and the recollections of surviving engineers. Establishes CSIRAC as the first computer in the world to play music. See also the CSIRAC page and reconstructed recordings.

“Composers' views on mapping in algorithmic composition.” Organised Sound 7(2). Cambridge University Press . DOI: https://doi.org/10.1017/S1355771802002066

“Early Computer Music Experiments in Australia, England and the USA.” Proceedings of the International Conference on Music and Sonic Art (MuSA), Karlsruhe, Germany.

An expanded comparative examination of early computer music activity in Australia (CSIRAC), England (Ferranti Mark 1, Pilot ACE) and the United States, drawing on primary sources. Analyses why the early developments in Australia and England — which preceded Bell Labs by approximately six years — did not lead to further advances, and how the involvement of composers was central to the success of the Bell Labs programme under Max Mathews. See also the CSIRAC page and related Chronology entries.

“The Music of CSIRAC, Some Untold Stories.” ISEA2013: 19th International Symposium on Electronic Art.

“The Application of Mapping in Composition and Design” Proceedings of Australasian Computer Music Conference.

“A Brief Survey of Mapping in Algorithmic Composition” Proceedings of ICMC, Göteborg. International Computer Music Association.

“Presence and Sound; Identifying Sonic Means to 'Be there'” Conscious Reframed, Beijing.

Several entries on computer music, CSIRAC, and Australian electronic music history. SAGE Music in the Social and Behavioral Sciences: An Encyclopedia. DOI: https://doi.org/10.4135/9781452283012

The papers below are currently in publication or must be loaded only to a personal website. Copies available upon request.

"Sonic Verisimilitude in Virtual Heritage: Lessons from Two Decades of PLACE-Hampi" peer-reviewed and accepted for the MIT Press' Computer Music Journal special issue on Sonic Practices in Virtual Worlds in March 2026.

Abstract: Over nearly two decades of exhibition, PLACE-Hampi has provided a rare longitudinal study in spatial audio for virtual heritage. This paper examines how the project's foundational principle of sonic verisimilitude, defined as the precise correspondence between acoustic and visual space, continues to shape contemporary immersive practice. By tracing the evolution of PLACE-Hampi from its 2006 first-order ambisonic implementation to its latest experimental recontextualization within modern game-engine environments, the study demonstrates that audiovisual coherence, rather than technological complexity or ambisonic order alone, most informs perceptual realism and audience engagement. Drawing on field recordings, system documentation, and both quantitative and qualitative audience data, the paper positions sonic verisimilitude within a broader discourse on spatial audio's role in cultural representation and curatorial practice. Emerging technologies such as AI-assisted production, real-time acoustic modeling, and personalized binaural rendering are discussed in relation to long-term preservation and ethical practice. The findings suggest that sonic verisimilitude functions not only as a technical criterion but as a design ethic and an epistemology of presence, linking sound, place, and cultural meaning across a changing technological landscape. 

 

"Mad Music Disease: Prion Pathology as a Model for Creative Collapse in Sampling and AI" submitted to Leonardo Journal March 2026, unrevised text. 

Abstract: The prion pathology conceptual model provides a useful biological framework for understanding creative collapse in self-referential systems. Drawing on bovine spongiform encephalopathy and related prion diseases, it viscerally demonstrates that when creative systems consume primarily from within themselves, they exhibit dynamics analogous to protein misfolding and systemic degeneration. The model applies across sample-based music that draws predominantly from its own stylistic conventions, and artificial intelligence trained on synthetic data. In both domains, the absence of external stimulus leads to homogenization, loss of diversity, and creative exhaustion. Recent empirical evidence demonstrates that human oversight cannot prevent this collapse. Humans routinely accept AI outputs without critical scrutiny, a phenomenon termed cognitive surrender, making structural intervention rather than individual psychological correction necessary for systems where AI models increasingly train on outputs from other AI models.