Real Time Singing Synthesizer project made from sinsy-NG. The idea was to generate vocal audio samples on real time easily for live coding performances.
Here is a video demonstration using the program to synthesize samples and load them into the FoxDot live coding environment.
Software contained in synthesisSoftware must be installed.
You can check the Colab notebook for a detailed guide about how to install it on Ubuntu.
If you want to check it quickly you can try it online here.
Build it
Run it
Once that's done you can try test-api.py
script as a small singing synthesis example. After running this an output wav file will be generated containing the specified vocals.
python test-api.py
All the parameters are optional, they contain a default value in case they are not specified. You can specify them in any order when running the program in the command line.
notes: The numerical value of notes in the scale, in C Major would be something like 0:C, 1:D, 2:E, 3:F and so on... The scale is C major by default, it's hardcoded in the main.py file and can be modified.
dur: The duration of each note in BPM (beats per minute)
lyrics: The text to be synthesized. Each word is mapped to one note from the "dur" and "notes" parameters
lang: The language code, "es" for spanish and "en" for english. There are several languages supported, You can check them here
tempo: The tempo in BPM
file: The name of the output file
If you want to help me to keep going developing and maintaining open-source projects you can contribute buying me some ko-fi.