Drums and Bass Interlocking

This thesis presents the technical and theoretical details of a generative system that considers the stylistic rhythmic interaction (interlocking) between bass and drums in the context of Electronic Dance Music (EDM). The concept of rhythmic interlocking, commonly known as groove, has been vaguely studied and is still an open research field. Before designing the generative system it has been necessary to materialize a rhythmic analysis algorithm that allows to study the rhythmic interlocking between instruments. The proposed rhythmic analysis extends recent work on rhythmic similarity that computes syncopation in a rhythmic pattern based on techniques derived from rhythm perception and cognition literature. The outcome of the analysis makes possible to visualize rhythmic patterns in a low-dimensional space defined by rhythmic features relevant to groove conception (i.e. syncopation and density). In addition, we model both temporal and interlocking context of a given corpus using Markov Chains. Taking advantage of the low-dimensional “cognitive-based” representation of rhythmic patterns, the system is parametrized by global syncopation and density controls for pattern generation. The most relevant contributions of this work are the compilation of literature on the topics of groove and stylistic algorithmic composition, the comparison between different generative approaches, and the software implementation of the proposed methods for analysis, visualization and generation. Results show that our proposals are useful for stylistic modelling and give better comprehension of the problem of rhythmic interlocking. Moreover, the generative system has been implemented and partially evaluated with promising results that encourage further research on the topic