Frederik Bous, Axel Roebel
UMR9912 STMS | IRCAM - CNRS - Sorbonne Université | Paris, France
This website contains supplementary material for our paper ``Analysis and transformations of voice level in singing voice'', currently under review for Interspeech 2022.
Below are samples from the test set of our singing dataset. Click on a sample to open a summary page for that sample containing transformations using the two models from the paper.
The available transformations have been selected based on the voice level of the original recordings, that is, the values are limited such that they remain in the range of voice level available in the dataset. In particular for recordings in fortissimo we only remove voice level and for recordings in pianissimo we only increase voice level.