Singing Synthesis from MIDI file

This script relies on the sinsy.jp website from the Nagoya Institute of Technology which implements a HMM-based Singing Voice Synthesis System.

You can find a sample merged with the instrumental audio here.

Requirements

• musescore: It's used to convert midi to musicxml
• python 3

Installation

pip install git+git://github.com/mathigatti/midi2voice.git

Usage

You can use it running the installed module using python -m. It has five parameters, the lyrics_file, midi_file, singer sex (optional), tempo (optional) and destination folder (optional).

Usage example

Check the midi and text samples here.

# Print help
python3 -m midi2voice -h

# Generate the voice given a midi file and a text file with the lyrics
python -m midi2voice -l shallow.txt -m shallow.mid -g female -t 96

Try it on Colab

If you don't have python installed or you just want to check it quickly you can try it online here.

Credits

This source code was developed by Mathias Gatti (@mathigatti) if you use it please remember to cite me. For scientific publications you can use this DOI.

Gatti, M. (2020). mathigatti/midi2voice v1.0.0 (v1.0.0) [Computer software]. Zenodo. https://doi.org/10.5281/ZENODO.3969003

Support my work

If you want to help me to keep going developing and maintaining open-source projects you can contribute buying me some ko-fi.

License

MIT