DJs use stage lighting that is controlled using audio interfaces, but they often have limited customizability. These interfaces have audio active modes that use basic decompositions such as volume thresholds or require extensive manual programming and lighting engineers to activate and control the lights during a performance. We look to develop a dynamic lighting system that extracts different features from the audio input to activate and control lights. Our scope is to create a system that can take in an audio input, decompose it into key components and identify features, and control one set of lighting in close to real time for multiple genres.