Modeling and Editing Cross-Modal Synchronization on a Label Web Canvas

Louis Garczynski (1, 2) , Mathieu Giraud (1, 2) , Emmanuel Leguy (1, 2) , Philippe Rigaux (3)

1 Algomus
2 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
3 CEDRIC - VERTIGO - CEDRIC. Données complexes, apprentissage et représentations

We present how the open-source web framework Dezrann enables to hear, study, and annotate music by interacting with synchronized views such as image scores, rendered scores, videos, and representations of audio as waveforms or spectrograms. We encode as unit conversions the cross-modal synchronization between these music representations. Composing these conversions allows us to conveniently translate musical time, audio time, and graphical positions, as well to synchronize annotation labels across views. Usages of the Dezrann platform or of some of its components include corpus annotation for musicology or computer music research, music education, as well as collaborative score edition and correction in the CollabScore project.

Domaines: Son [cs.SD]; Musique, musicologie et arts de la scène; Interface homme-machine [cs.HC]

Louis Garczynski, Mathieu Giraud, Emmanuel Leguy, Philippe Rigaux. Modeling and Editing Cross-Modal Synchronization on a Label Web Canvas. Music Encoding Conference (MEC 2022), 2022, Halifax, Canada. ⟨hal-03583179⟩