220c Final Project - Guitar Chord Detection

larrywu@ccrma.stanford.edu
MST student, CCRMA, Department of Music, Stanford University

Outline

Playing the guitar is popular hobby. As long as you know how to press the basic chords and have a list of chords for a song, you can enjoy playing guitar and sing by yourself. The problem arises when the list of the chords for a desirable song is not available. Then a tough task is faced- selecting appropriate chords.

This project presents spectral analysis of the sound, used for the guitar chord recognition. Each chord, played by the guitar, is represented by the unique set of spectral components. Therefore can be developed spectral analysis algorithm for the identification of the played chords.

Steps

1. Spectral analysis algorithm
• Sampling.
• Filtering.
• Decimation (Downsampling)
• Windowing.
• FFT
• Detecting played notes
2. Chord selection algorithm

System

Demo Video

Download

Reference

1. Takuya Yoshioka, Tetsuro Kitahara, Kazunori Komatani, Tetsuya Ogata, and Hiroshi G. Okuno "Automatic Chord Transcription with Concurrent Recognition of Chord Symbols and Boundaries"
2. Alexander Sheh and Daniel P.W. Ellis, "Chord Segmentation and Recognition using EM-Trained Hidden Markov Models"
3. Christopher A. Harte and Mark B. Sandler "Automatic Chord Identification Using a Quantised Chromagram"
4. Paulius Dambrauskas "Real-Time Guitar chord recognition"
5. Takuya Fujishima "Realtime chord recognition of musical sound: A system using Common Lisp Music"

Links

1. http://www-classes.usc.edu/engr/ise/599muscog/2004/projects/chuan-zhu/
2. http://www-scf.usc.edu/~chinghuc/pitch_detection_algorithms.htm
3. http://www.cs.tut.fi/~klap/iiro/index.html