Event Description
Large quantities of jazz recordings are now available and accessible in digital form, and more will come. Yet Music Information Retrieval has just begun to adapt its techniques to jazz and to the special challenges of documenting these improvised musical performances. A project known as J-DISC is exploring the use of MIR techniques in jazz. Signal processing and machine learning techniques can help provide valuable data on many common practices of improvised music, especially in the online space where industry metadata standards are low. A prime goal of J-DISC is to link conventional jazz discographic information with the recordings it describes in the digital space, which has never been practical in the past. Our work is also building the foundations for using MIR tools to help analyze and discuss the musical practices of jazz in any style. J-DISC is a collaboration between the Center for Jazz Studies, the Center for Digital Research and Scholarship, the Computer Music Center, the Gabe M. Wiener Music & Arts Library, and the Department of Electrical Engineering, all at Columbia University. It is funded by the Andrew W. Mellon Scholarly Communications Program. |
