Jump to Navigation

About Us

J-DISC is an extension and enhancement of the educational website Jazz Studies Online, a project of the Center for Jazz Studies at Columbia University; and it is funded by the Andrew W. Mellon Foundation and theGladys Krieble Delmas Foundation. John Szwed is Project Director and Principal Investigator of J-DISC; the Project Manager is Tad Shull. A blog on a newer phase of the J-DISC project that deals with Music Information Retrieval is Jazz Information Retrieval.

The Center for Jazz Studies aims to incite discussion and learning about jazz, its evolution, and its impact on other fields of human endeavor. Exploring the creative possibilities in the interaction of jazz and technology is central to the Center’s programming, research, and artistic interests. Jazz itself has always kept abreast of and been shaped by advances in technology—the automobile, the transistor, and, of course, the microphone—and there is no reason that jazz studies should not also do so. The Center began to create a dynamic scholarly presence on the Internet in the realm of research and education with Jazz Studies Online, with generous support from the Ford Foundation, in 2008, and is moving this agenda forward with the online jazz discography now known as J-DISC.

The idea for the collaborative, extensible online database of jazz recordings is the product of a decade-long discussion among The Center, its Board, and noted scholars at jazz studies departments in the United States and Europe. In early 2009, we decided to create an enhancement to Jazz Studies Online that would provide ready access to discographic information and address the inherent limitations in editing print-based discographic works.

J-DISC is designed to make the vast existing body of available reference material about jazz recordings easier to use, and as a model resource in which scholars may edit, interpret, and share the insights they gain from that material. It is also intended as a means to preserve discographic data from print and digital sources that are no longer being generated or carefully preserved. Our ultimate goal is to create an online reference work that will enable the emerging field of jazz studies to mine large amounts of data on jazz history and jazz improvisation generated in the production and dissemination of jazz recordings, and enable researchers, teachers and students to communicate effectively about how they may use the data.

The collaborative dimension of J-DISC was strongly informed by conversations with the Mellon Foundation’s Scholarly Communications Program and their interest in the possibilities of the Internet for allowing many scholars to make valuable editions and annotations to a single online repository of sources relevant to a given area of study. (See MSC Program Officer Donald J. Waters’ “Archives, Edition-Making, and the Future of Scholarly Communication” at <<http://msc.mellon.org/staff-papers/EditionMakingPaper>>). A grant project to build a prototype online discography funded by The Mellon Foundation was approved in March of 2010, and that work has culminated in the launch of J-DISC in June 2010. The Center received a two-year continuation grant for work on J-DISC in June 2012. In the 2012-2014 grant period, the J-DISC project team will continue to explore ways to identify and access jazz recordings in the digital and online music environment.

Since the grant project began, a discussion among our team and outside advisors on how to account for digital and digital-only releases of jazz recordings touched on how the media and content management and monitoring systems work on iTunes or YouTube, and the paucity of discographic work on digitally distributed jazz. After asking many experts about the subject, we realized that we had the benefit of Columbia’s Computer Music Center, and of LabROSA, whose practitioners work in the area of Music Information Retrieval, which could pose many solutions to the questions we were asking.

Music Information Retrieval (MIR) refers to a collection of techniques and applications in which computers are used to help organize and access digital music archives. An exciting new phase of the J-DISC project during the continuation grant period will be to research the applications of and feasibility of MIR to jazz and jazz discography in conjunction with the Computer Music Center and LabROSA. To our knowledge the specific concerns of improvised music have rarely been explored and tested using MIR. Yet we believe it is essential to address the challenges for traditional discography in documenting and accounting for digital, and especially digital-only, audio files. J-DISC anticipates a not-so-distant future in which physical sound carriers may be used less and less to disseminate jazz recordings.

Furthermore, most existing work in MIR has focused on applications of these techniques in contemporary popular music, or the classical orchestral and piano repertoires. Jazz, however, presents a different and interesting set of challenges, stemming from the importance of improvisation in jazz, the relatively narrow range of instrumentation (while spanning a wide range of styles), and the availability of important longitudinal collections, where we can study the relationships between multiple recordings of the same piece by the same performer recorded at different times.

Our research in MIR will focus on exploring only musical factors distinctive to jazz and the needs of jazz discography—that is, accounting for and identifying recording sessions where improvisations take place. We will ask, for example, how MIR techniques can help identify, compare, and distinguish audio files of jazz recordings that may be available on the Internet but lack the recording session data that are the prime access point and crucial identifying criteria used in jazz discography. Developing MIR techniques can help identify certain musical parameters that may be important to jazz expression, such as complex rhythms and rich, multilayered timbres, which are difficult to analytically separate from other auditory information such that a computer could recognize them. With the appropriate tools, performers, instruments, and even venues could be distinguished, helping identify and categorize files in vast numbers.

The application of MIR research to jazz will be of interest to audio engineering, musicology, and library sciences fields. Using content-based analysis of audio signals to help the discovery of audio files on the Web can also help musicians or music producers identify and document instances where their work has been issued or reissued, and may allow libraries to help users access widely distributed digital audio files and link them to their own holdings.

Another new phase of the J-DISC project is to work with Columbia University’s Center for Digital Research and Scholarship to incorporate Linked Open Data (LOD) into the application. Opening the J-DISC data to semantic query possibilities will enhance the scholarly usability and research potential of J-DISC and will enhance the intellectual contribution back to both the jazz and LOD communities. A semantically enabled J-DISC can link or operate smoothly with other online resources on jazz recordings, artists, and repertoire that have already incorporated Semantic Web content or that look to do so. This newly enabled resource will thereby suggest research questions or latent trends and patterns on jazz performance and history which can be explored using linked web resources and which J-DISC data can thus help illuminate.

Finally, the renewed grant project includes developing a formal sustainability plan that will address the long-range needs and costs of operating J-DISC, and what options may be available to allow it to become a permanent digital reference resource. Weare pleased to announce that we will be working with aconsulting organization named Ithaka S+R  to evaluate user needs, cost models, and business scenarios that will help sustain J-DISC through proposed grant project period (now scheduled to be complete by July 2014) and beyond.

J-DISC was built using Drupal, an open source framework. If you would like to install JDISC in your environment (that is, the files used to support the J-DISC website, not the data or expressive content itself), please contact jsotbs@gmail.com. J-DISC's application of code is covered by a GPL 3.0 license.

The design and expressive content is covered by a Creative Commons Attribution Non-commercial 3.0 US license.