Following big data analysis trends in the humanities and social sciences, we have begun development of a ground-breaking database that will increase artistic, research, theoretical, and pedagogical innovation by providing an environment for knowledge discovery about orchestration. .
A major achievement of previous McGill-led grants, Orchard is a queryable database derived from work in music information research.4,39,118 It currently integrates score annotations and music clips according to our perception-based taxonomy. Orchard will be expanded over the course of the grant and beyond and made open-access to the full Partnership and other interested users on our Partner Calcul Québec/Compute Canada's platform, which provides the necessary storage space and high-powered computing for database searches, data mining and machine learning.5,119 Large-scale data from the Analysis Axis will be entered into Orchard, including machine-readable scores and texts, annotations of scores and texts, modes of playing, and perceptual effects, recorded or rendered audio clips, acoustic analyses, and perceptual experimentation on audio excerpts. Metadata on the time-varying acoustic properties of sounds produced by musical instruments playing at different musical dynamics and pitches with various play techniques will also be integrated into Orchard to provide acoustic information for the data mining algorithms, seeking, for example, to determine the degree of spectral overlap between sounds that conditions how strongly they will blend together.
Thesaurus of Orchestration
(More info coming soon...)