Moody Music Generator: Characterising Control Parameters Using Crowdsourcing

• Published in: Evolutionary and Biologically Inspired Music, Sound, Art and Design Vol. 9027; pp. 200 - 211
• Main Authors: Scirea, Marco, Nelson, Mark J., Togelius, Julian
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2015; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Algorithms & data structures; Artificial intelligence; High Arousal; Music Clip; Music Generator; Musical Feature; Musical Style; User interface design & usability
• ISBN: 331916497X; 9783319164977
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-16498-4_18

We characterise the expressive effects of a music generator capable of varying its moods through two control parameters. The two control parameters were constructed on the basis of existing work on valence and arousal in music, and intended to provide control over those two mood factors. In this paper we conduct a listener study to determine how people actually perceive the various moods the generator can produce. Rather than directly attempting to validate that our two control parameters represent arousal and valence, instead we conduct an open-ended study to crowd-source labels characterising different parts of this two-dimensional control space. Our aim is to characterise perception of the generator’s expressive space, without constraining listeners’ responses to labels specifically aimed at validating the original arousal/valence motivation. Subjects were asked to listen to clips of generated music over the Internet, and to describe the moods with free-text labels. We find that the arousal parameter does roughly map to perceived arousal, but that the nominal “valence” parameter has strong interaction with the arousal parameter, and produces different effects in different parts of the control space. We believe that the characterisation methodology described here is general and could be used to map the expressive range of other parameterisable generators.