[PYTHON] About "spleeter" that can separate vocals and musical instruments from music data

Development environment

OS:Windows10 Command: Use Git Bush

What is spletter

It seems that music data composed of vocals and multiple musical instruments can be output as a file classified into each sound using machine learning. [Details] (https://gigazine.net/news/20191107spleeter/#group=nogroup&photo=9)

It doesn't work as written: frowning2:

[Reference link] (https://gigazine.net/news/20191107spleeter/#group=nogroup&photo=9)

If you do this, you will get an error when you run "conda env create -f spleeter / conda / spleeter-cpu.yaml"

Cause

To the link that was git cloned in the first place Because there is no path called conda / spleeter-cpu.yaml

solution

There was a fix link on spletter's gitHub!

However, the content described in Quick start of README.md does not work. Put the modified version below

https://github.com/sigsep/spleeter-musdb

conda env create -f spleeter-musdb/conda/spleeter-cpu.yaml

conda activate spleeter-cpu

spleeter separate -i spleeter-musdb/audio_example.mp3 -p spleeter:2stems -o output

In the fourth "spleeter separate -i spleeter-musdb / audio_example.mp3 -p spleeter: 2stems -o output", -i sound source path -o output destination path -p spleeter: 2 (separation of vocal and accompaniment) means

For more information here

In addition, since the fourth separatener cannot be used with the command prompt or PowerShell on windows, writing python -m spleeter separate ~ should work, maybe.

The explanation is quite broken, so if you have any mistakes or questions, please use the comment section below: point_down: