This pull request adds the following functionality:
Export a partitura object to an audio file using Fluidsynth.
This functionality is available in pretty_midi, where it only works for MIDI files.
Such a method is useful for cross-modal tasks (e.g., transcription, audio-to-score alignment, etc.).
This implementation allows for the method to work on all partitura note-based objects (Score, Performance, Part, PerformedPart, note arrays), and thus for all formats supported by partitura.
Fluidsynth is treated as an optional import. Instead of adding to the repository a default soundfont (as is the case in pretty_midi, which includes a very small one), the default soundfont is set to be downloaded from the MuseScore soundfonts (which are distributed with the MIT license).
Fixes a few minor issues in slice_ppart_by_time adding previously missing MIDI messages and fixing the computation of onset and offset times in ticks.
This pull request adds the following functionality:
Export a partitura object to an audio file using Fluidsynth.
pretty_midi, where it only works for MIDI files.Score,Performance,Part,PerformedPart, note arrays), and thus for all formats supported by partitura.pretty_midi, which includes a very small one), the default soundfont is set to be downloaded from the MuseScore soundfonts (which are distributed with the MIT license).Fixes a few minor issues in
slice_ppart_by_timeadding previously missing MIDI messages and fixing the computation of onset and offset times in ticks.