Musical manuscripts and prints are an important part of the written heritage of the Middle Ages and the Renaissance.
The Ricercar musicological programme of the CESR has developed digital tools for the visualisation and exploitation of corpora and the reconstruction of sound spaces, and has built up a very rich documentary collection.
1/ A cartography of musical heritage and its metadata
Ricercar circumscribes the written musical heritage of the Middle Ages and the Renaissance, excluding the immense corpus of Christian monodic liturgical chant (plainchant or "Gregorian" chant) recorded in liturgical books (with or without musical notations).
The RISM (Répertoire International des Sources Musicales) and its online database provide basic bibliographic information on all printed music produced up to 1600.
For manuscripts, a digital tool maintained at Oxford (DIAMM) provides information that is just as exhaustive but of greater granularity than RISM, and monitors the availability of sources from libraries (4000 musical manuscripts up to the 16th century).
Ricercar monitors digitised sources online (8000 Zotero records) and makes available a few specific corpora of digitised early scores (Corpus des luthistes).
In the course of redesigning its databases, Ricercar has established data models for the description of musical sources and works that take into account their specificities. Its ongoing work on ontologies for the description of musical works will be completed within the framework of Biblissima+.
Regarding the medieval period, the IRHT intends to enrich the Wala database by increasing inventory operations, identifying new sources and indexing or editing the repertoires of Latin liturgical chants.
Within the Relicantus project, the aim will be to inventory, digitise and index dispersed musical fragments.
2/ Musical works: encoding and mining of musical data
Ricercar has musical editions encoded in MEI format (Music Encoding Initiative) and actively participates in the development of this open and scholarly encoding standard for musical notation based on TEI.
The availability of open MEI encodings (Gesualdo online, etc.) makes it possible to move from the image of the old score to various forms of displays:
- modern score (depending on the needs: online, in downloadable pdf, in XML-MEI or in a proprietary format - Sibelius, Finale...)
- or other graphic visualizations, static or animated,
- possibly synchronised with digital, synthetic or recorded audio files.
The project wants to go further in terms of annotation and mining of musical data, essentially for similarity recognition, with or without artificial intelligence. The goal of a music search engine (audio or written) is the research horizon of this work, which academic research shares with cultural industries.