ChoraleWind
An expressive wind-quartet dataset and reproducible baseline for score-to-audio rendering.
ChoraleWind is an expressive wind-quartet dataset for end-to-end rendering from symbolic music notation to synthesized audio. It is based on 312 four-part chorales from the Neues Thüringer Choralbuch and covers the full path from score encoding to performance-level rendition and multitrack audio.
The dataset provides aligned symbolic representations, performance annotations, isolated voice stems, and ensemble mixes. Its rule-based performance model adds expressive timing, dynamics, articulation, metric accents, structural accents, and phrase-end gestures before rendering the material with physical-modeling wind-instrument synthesis.
ChoraleWind is designed as a transparent baseline and controlled testbed for studying expressive performance generation, timbre modeling, and reproducible evaluation of wind-quartet rendering systems. It is compatible with the ChoraleBricks data structure and can therefore be used with the same Python toolbox.
Website: ChoraleWind Dataset
Data: Zenodo
Code: stefan-balke/choralebricks