VocalSet

dataset·singing voice·active

A cappella singing recordings across singers, vowels, registers and vocal techniques, with a corrected annotated derivative.

Verified facts

Official site	https://zenodo.org/records/10200775 ↗
Geography	United States
doi	10.5281/zenodo.1193957
size	vowels: 12 · singers: 30 · registers: 4
scope	Solo singing voice analysis, synthesis and technique recognition.
genres	vocal exercises; singing techniques
license	CC BY 4.0
steward	Northwestern Interactive Audio Lab and deposit maintainers
creators	Jennifer Wilkins; Prem Seetharaman; Alison Wahl; Bryan Pardo
versions	VocalSet 1.0; Annotated VocalSet 1.0.0; corrected combined deposit
languages	vowels and sung material; exact language mix not normalized
modalities	audio; singer metadata; technique labels; pitch/note annotations in derivative
geographies	United States
access method	Open Zenodo deposit.
intended uses	singing voice synthesis; vocal technique classification; timbre analysis
related papers	VocalSet: A Singing Voice Dataset; Annotated-VocalSet
annotation method	Original labels plus corrected manual note and file annotations in Annotated VocalSet.
collection method	Thirty singers recorded isolated vocal materials across registers and techniques.
current availability	Combined corrected deposit available.
commercial use limits	Attribution required; no dataset-level noncommercial restriction stated.
us market scope basis	Publicly available or materially used in US-facing music/audio-AI research.
limitations biases disputes	Small singer population and constrained studio vocal tasks.
redistribution restrictions	CC BY 4.0 attribution applies.