Background

In online archival spaces, music scores are typically available as PDF or JPEG files. Most newly created music scores are PDFs; if they are archived, it is usually in PDF or PDF/A format [1]. PDF formats—except PDF/UA—are frequently inaccessible to Disabled users because they are incompatible with many assistive technologies. As a result, most archives’ digital music scores are inaccessible to many Disabled users. That needs to change, so I devised a workflow that allows archivists (and any other interested parties) to create accessible music scores using MuseScore, a free, open-source music writing application.

The Workflow

An archivist, librarian, student worker, or other involved party manually enters the contents of each score into a free app called MuseScore. This is still tedious, but there are initiatives underway that might expedite or even automate this process. Once the music is in the app, it can be exported to any number of file formats, including MEI and musicXML, which are excellent for preservation. They can also be exported into formats that are more accessible, like Braille music notation, modified stave notation, or audio files.

MEI encoding every music score an archive holds is an incredibly high bar to reach for—but new tools can help speed up the process, and the accessibility benefits from it cannot be understated. 

What I Did

To test this workflow, I used MuseScore to create MEI files for 14 scores by Maria Theresia von Paradis (1759–1824), herself a Blind woman. I used those files to create Braille Music Notation, modified stave notation, and audio files. The modified stave notation files are made using an example set of settings that are not representative of all modified stave notation scores or user needs, but my aim with them was to meet at least some basic needs. After asking participants in a survey and set of usability tests where they find their music scores most often, I decided to upload the finished scores to IMSLP, GitHub, and Zenodo [link forthcoming] to accommodate as many preferences and file types as possible.

I also created guidelines for following the workflow and specifically for creating modified stave notation scores in MuseScore 4.X.X.

What I'm Doing

Now, I'm trying to figure out how to improve the workflow to make it more practical for archivists to implement. The primary challenge with it is the manual note-by-note entry of each original score into the MuseScore application. If that portion of the workflow can be automated, creating each score will take minutes instead of hours. To do that, we need advances in OMR--especially Audiveris, which is MuseScore's current OMR engine.

But archivists can use the workflow as it currently stands, so I'll trust the people working on OMR to keep working on it, and I'll work on getting more people motivated to create and make available accessible music scores, especially in archival settings.