Let's Encode! Crowd-Encoding and Validation of Music Scores

Abstract:

Music encoding for the masses!

Digital music scores have become commonplace in the activities of musicians, music scholars, and music enthusiasts, providing ubiquitous access to large music collections. However, in their most common form as PDFs (scanned images of notation documents), their content cannot be readily processed by machines; their meaning is immediately apparent to the music-literate viewer, but to a computer, they are mere binary blobs, undifferentiated from arbitrary image files.

The Music Encoding Initiative (MEI) is an open-source community effort to explicitly model musical meaning in notation documents to allow machines to 'understand' the score. Paired with Semantic Web technologies that allow external information to be interlinked with (elements of) the score, this opens up rich new possibilities for processing and interaction. Recent research has brought together musicians, composers, musicologists, music librarians, music informatics researchers, Web scientists, and software developers to explore the intersection of these technologies, building prototypes interconnecting live performances, audiovisual recordings, computational analyses, commentaries and annotations, concert programmes, bibliographical references and Web resources with music scores at note-level granularity.

However, these approaches are limited by a lack of available music encodings in the MEI format, which remains largely unsupported by commercial notation software. Writing MEI from scratch is a laborious process requiring technical expertise as well as music literacy. To make this process more pleasant while lowering the learning curve, we have created mei-friend (with significant contributions from MEI community members!): a free, open-source, browser-based music encoding editor. 

In Let's Encode! we will extend this editor, placing it within a Web-based infrastructure for crowd encoding and validation of music scores in which interested musicians, scholars, and music enthusiasts take part in, but also start and direct, music encoding projects – with minimal requirements in terms of technical expertise, and no administrative, financial, or technological requirements aside from a standard Web browser. 

In a series of trial encoding campaigns, we will investigate how to effectively empower users to make the most of the affordances of semantic music encoding technologies. We will accompany chamber-orchestra musicians and their associates as they encode performance scores of their repertoire, freeing them to perform these public-domain works without requiring significant licensing fee payments to obtain performance rights. We will work with communities of music lovers as they preserve important cultural heritage: the Electronic Corpus of Lute Music and their long-running project in collaboration with the UK Lute Society; the Klezmer Institute and its Klezmer Archive, assembled by an international community of klezmer musicians and fans; and the Rebalancing the Music Canon initiative, promoting the inclusion of the works of under-represented composers in our digital cultural repertoire. Finally, we will collaborate with the Repertoire International de Source Musicale, the world's largest catalogue of music sources, which will allow the enormous RISM Online user base to contribute new incipits (encodings of initial melodic phrases, highly useful in catalogue searching) to improve its collections.

Project Partners:

• Dagmar Abfalter, Department of Cultural Management and Gender Studies, mdw
• Michael Staudinger, University Library, mdw
• Werner Goebl, Department of Music Acoustics – Wiener Klangstil, mdw

Board of Advisors:

• Anna E. Kijas, Lilly Music Library, Tufts University - USA
• Christina Crowder, Klezmer Institute - USA
• Clara Byom, Klezmer Institute - USA
• David Lewis, Oxford e-Research Centre, University of Oxford and Goldsmiths, University of London - UK
• Laurent Pugin, RISM Digital Center - Switzerland

Funded by