Algorithmic composition methods must prove themselves within real-world musical contexts to more firmly solidify their adoption in musical practice. The present project is an automatic composing program trained on a corpus of songs from musical theater to create novel material, directly generating a scored lead sheet of vocal melody and chords. The program can also produce output based upon phonetic analysis of user-provided lyrics. The chance to undertake the research arose from a television documentary funded by Sky Arts that considered the question of whether current-generation, computationally creative methods could devise a new work of musical theater (the research described here provides but one strand within that project). Allied with the documentary, the resultant musical had a two-week West End run in London and was itself broadcast in full. Evaluation of the project included both design feedback from a musical theater composer team, and critical feedback from audiences and media coverage. The research challenges of the real-world context are discussed, with respect to the compromises necessary to get such a project to the stage.

This article presents Autocousmatic, an algorithmic system that creates electroacoustic art music using machine-listening processes within the design cycle. After surveying previous projects in automated mixing and algorithmic composition, the design and implementation of the current system is outlined. An iterative, automatic effects processing system is coupled to machine-listening components, including the assessment of the “worthiness” of intermediate files to continue to a final mixing stage. Generation of the formal structure of output pieces utilizes models derived from a small corpus of exemplar electroacoustic music, and a dynamic time-warping similarity-measure technique drawn from music information retrieval is employed to decide between candidate final mixes. Evaluation of Autocousmatic has involved three main components: the entry of its output works into composition competitions, the public release of the software with an associated questionnaire and sound examples on SoundCloud, and direct feedback from three highly experienced electroacoustic composers. The article concludes with a discussion of the current status of the system, with regards to ideas from the computational creativity literature, among other sources, and suggestions for future work that may advance the compositional ability of the system beyond its current level and towards human-like expertise.