MIR Notes by Andrew

Motivation

Why Audio?

Sound has always mattered to me in a personal way. After serving in the military, I came back with partial hearing loss — and that experience made me appreciate how much of life travels through audio. This notebook is my attempt to understand it computationally.

Music Information Retrieval (MIR) is the field that sits at the intersection of signal processing, machine learning, and musicology. The goal is to teach machines to understand sound the way humans do — to recognize genre, detect tempo, separate instruments, and more. These chapters represent my journey to understand it step by step.

Chapter 1 Basic Musical Representations and MIR Introduction to Music Information Retrieval — how computers represent, analyze, and understand music across music theory, signal processing, and machine learning. Open notebook Chapter 2 Audio Representations Exploring how audio signals are represented digitally — waveforms, spectrograms, and the fundamental building blocks of audio analysis. Open notebook Chapter 3 Features and Applications Extracting meaningful features from audio and applying them to real-world MIR tasks such as genre classification, beat tracking, and more. Open notebook