Algorithmic music generation using quantification of cognitive properties, and utilisation in the DAW environment

This thesis is concerned with cognitive concepts related to melodic preferences, and attempts to capture quantifiable elements of what makes melodies sound pleasing to listeners, through the use of statistical properties and analysis of listener preferences. By using both empirical findings and drawing from literature, these elements were subsequently used as a basis for the development of software that generates musical sequences. The thesis included two studies investigating the perception of melodies, where the concept of a Uniformity Principle (a preference of listeners for melodies with distributionally uniform pitches), and the relation between working memory and liking were examined. Further, a study was conducted with the aim of understanding the usefulness and quality of the software we developed, showing positive results in both of those dimensions.