Citations: The Renaissance Imitation Mass Project


CRIM: Quotable Music

The CRIM project relies on several key technologies that make it possible for us to create digital encodings of music scores, address them with note-level precision, and surround these citations with complex metadata about what they mean (and who made them). Through them, CRIM is creating a new kind of quotable text for musical scores.

The Music Encoding Initiative: From Sibelius to Verovio

Encodings of musical works begin with Sibelius proprietary music engraving software. Sibelius files are used to prepare digitally engraved PDF scores of the highest graphical quality for use by performers and analysts. The Sibelius transcriptions are also exported as open-source XML files conforming to the Music Encoding Initiative (MEI) schema using a Sibelius Plug-in Extension. Although Sibelius is proprietary software, the plug-in itself is open source, and freely available for others to use and adapt; it is under continuous improvement by the MEI community. The resulting XML files are validated against the MEI schema, which assures their suitability for further transformation as part of rendering, citation, and analysis projects. Various Python “massaging” routines in turn prepare the MEI files for use with CRIM.

Verovio is a flexible Javascript kit that will render MEI (or MusicXML) easily in any html browser, without special software. It creates beautifully engraved images, PDFs, and can even render audio versions of the scores it reads. And since Verovio represents scores as a collection of an SVG objects, individual notes are in fact directly connected with the underlying MEI they represent, making it possible to render scores with selected highlights.

To learn more about the pathway from Sibelius to MEI to Verovio (and the web) see This Demonstration.

Enhancing Music Addressability: Portable Musical Citations

Note-level citations are made possible by the Open MEI Addressability Service (OMAS) developed in conjunction with the Enhancing Music Notation Addressability API (EMA), funded by an NEH Start-Up Grant. With EMA we can point to

  • any combination of beats
  • in any combination of staves
  • in any combination of measures
  • in any common music notation score

Used together with MEI scores and Verovio, the EMA+OMAS system makes it possible for CRIM users to select notes directly in the browser, store them as EMA citations and then return them as valid MEI (complete with highlights for the chosen passages).

To learn more about the pathway from Sibelius to MEI to Verovio (and the web) see This Demonstration.

Open Annotation and Linked Open Data

We explore the CRIM repertory in ways ideally suited to the very concept that animates these pieces: through collaborative citation and commentary involving scholars and students alike. Linked Open Data technologies such as the Open Annotation Collaboration are the keys to this aspect of the project. EMA references will thus be surrounded by an array of information about the person who made the reference (or "observation" as we call them in the context of CRIM), the musical schema found here (based on our controlled vocabulary), the nature of the borrowing, plus any commentary or explanation the analyst might like to add. These assertions are encapsulated as digital publications (using the NanoPublication standard), each with a permanent and unique URL. They inaugurate a new kind of permanent and attributable status for analytic claims about music, using a data model that can be adapted for use by other projects.

Want to know more about CRIM and Open Annotation? See our Brief Guide to CRIM Open Annotation, or the explore the Technical Pages in the CRIM Workspace and Editorial Hub

Django Content Management System

CRIM data are managed through a Django application, which assembles and stores the metadata and relevant MEI coordinates of the musical passages (using the OMAS service noted below) in a searchable relational database. A Solr indexing system continuously updates the array of pieces and analytic observations. Such systems are robust and scalable: a similar system created for The Lost Voices Project efficiently and rapidly manages over 11,000 analytic observations (including over a dozen different facets of metadata, from who made each one and what it is meant to represent), along with user accounts, private notes, and threaded discussions that update continuously. All data can also be exported as JSON objects for back-up and re-use in other contexts.

See the Code pages for details of these and other technologies used in CRIM.