The RAVDESS is a validated multimodal database of emotional speech and song. The set of 7356 files can be downloaded from the RAVDESS dataset. The set of recordings were evaluated by 319 individuals who were characteristic of untrained research participants from North America. High levels of emotional validity and intra-individual test-retest intrarater reliability were reported. All recordings are made freely available under a Creative Commons non-commercial license (CC BY-NA-SC 4.0). Examples of RAVDESS videos are available on YouTube.
The RAVDESS community will curate all datasets related to the RAVDESS project. This includes the original RAVDESS dataset, and in the future, analysis datasets generated from the RAVDESS, including but not limited to: facial motion tracking metrics, vocal acoustic analyses, additional perceptual validation data, psychometric tests, and canonical stimulus sets.
Datasets that are generated as a result of using the original RAVDESS stimulus dataset will be included in the community. The community will be curated by the RAVDESS authors, Dr. Steven R. Livingstone and Dr. Frank A. Russo.