A Case-Based Reasoning Approach to Plugin Parameter Selection in Vocal Audio Production

The field of intelligent systems for music production aims to produce co-creative tools to aid and support musicians’ decision-making while targeting a specific aesthetic in their musical artifact . While case-based reasoning (CBR) approaches have been used to assist music generation and recommendation, music production has not yet been explored. This paper proposes using CBR within a co-creative agent to assist musicians in their aesthetic goals through a vocal audio plugin. Results show that although participants were interested in using a co-creative agent throughout the production process, they acted against the vocal plugin parameter recommendations set by the agent. Participants showed frustration when the co-creative agent acted in a way that deviated from set expectations. From this research, we posit that explainability is an essential aspect of effective CBR models within co-creative agents.