LightShow: Abstract Representations of Music Lighting In Python

Abstract

This thesis explores music lighting and ways in which music lighting can be generated automatically. We attempted to use videos of prior concerts as training data for a machine learning model, but ultimately this proved unsuccessful. Instead, a useful abstraction for representing, designing, and implementing light shows based on audio was designed, implemented in Python, and used to generate lighting in a few contexts. The abstraction designed in this thesis ultimately focuses on allowing developers to easily expand on the package and reuse code, with the restriction that audio data must be known ahead of time. While the current abstraction does not support live audio streams, the future work section outlines how this could be implemented.