Sungdrum is a (nascent) project to standardize and improve the representation of lyrical and linguistic data in computational musicology datasets, focusing on (and advocating for) data encoded in the humdrum syntax. Sungdrum is based in "classic" humdrum syntax/toolkit, but is (or will be) particularly integrated with humdrumR.
This repo contains several documents containing recommendations and standards for lyric/linguistic encodings and workflows, as well as some scripts to help with the annotation of lyrics in humdrum data. The main contribution at this time, is a set of new standards for encoding lyrical and linguistic data in humdrum syntax.
- Issues with lyric data (problems)
- Humdrum representations (solutions)
This project is still underdevelopment. I'll hopefully be refining my recommendations, and pushing some useful tools this Fall (2025).
Sungdrum is a project of the Georgia Tech Computational and Cognitive Musicology lab, led by PI Nat Condit-Schultz.