There is a new research paper published from people at Deezer R&D, which I find quite promising, called Spleteer.

Here is an example courtesy of https://news.ycombinator.com/item?id=21432328:

First, the original: https://mwcampbell.us/tmp/spleeter-demo/jonathan-coulton-re-your-brains/original.mp3

Extracted vocals: https://mwcampbell.us/tmp/spleeter-demo/jonathan-coulton-re-your-brains/stems/vocals.mp3
Extracted accompaniment: https://mwcampbell.us/tmp/spleeter-demo/jonathan-coulton-re-your-brains/stems/accompaniment.mp3

You can find the software code itself at https://github.com/deezer/spleeter which everybody can run with a little effort.

 

Whit pyrhon can do it ? I stared learn.

 

You can use Anaconda Python distribution

 

@M S I'm not a big fan of python so I'll just use a cobra

 

Too advanced for me, despite being interested in AI. Thanks anyway Deezer

 

This is amazing tool. But not for newbies in python I guess.
So I must only:
1. Download Anaconda Python
2. Download Setup.py from Github
3. Open Python then run Setup.py script
...
4. ? Show up a green GUI with button says open file in the centre and I can open a mp3 or wav file and then show up options separate acompaniament or separate vocals?

Sorry I am dumb greenie with this stuff

 

Im noob in python...only if you run script works?

 

Is necesary anacanda ? I try whit visual Studio run setup.py whit dos ..but dont open

 

Is necesary anacanda ? I try whit visual Studio run setup.py whit dos ..but dont open

 

You can skip using Anaconda of course but it will get very hard to setup things.
After installing Anaconda, just open an Anaconda console (assuming you are on windows) and follow the instructions on the Quick Start sections of the github page. I tried and it worked perfect.

 

Someone can make clearer this? Step by step? For humanity

Quick start
Want to try it out ? Just clone the repository and install a Conda environment to start separating audio file as follows:

git clone https://github.com/Deezer/spleeter
conda env create -f spleeter/conda/spleeter-cpu.yaml
conda activate spleeter-cpu
spleeter separate -i spleeter/audio_example.mp3 -p spleeter:2stems -o output
You should get two separated audio files (vocals.wav and accompaniment.wav) in the output/audio_example folder.

For a more detailed documentation, please check the repository wiki

Reference
If you use Spleeter in your work, please cite:

@misc{spleeter2019,
title={Spleeter: A Fast And State-of-the Art Music Source Separation Tool With Pre-trained Models},
author={Romain Hennequin and Anis Khlif and Felix Voituret and Manuel Moussallam},
howpublished={Late-Breaking/Demo ISMIR 2019},
month={November},
year={2019}
}


What means "if you use Spleeter please citie"? I must copy this text and paste somewhere?
And what means this first text code under the github spleeter link:
"conda env create -f spleeter/conda/spleeter-cpu.yaml
conda activate spleeter-cpu
spleeter separate -i spleeter/audio_example.mp3 -p spleeter:2stems -o output"
I must copy this too and paste somewhere?

 

The part about "if you use Spleeter please cite" is about referring to the publication if you are using it in your research or in an app. Don't worry about that.

If you want to try it out, you have to install Anaconda and git. Then following the 4 step instruction in the Quick Start should work for you.

This is a research project. It is a little different from install, next, next which we are all used to, unfortunately.

 

I do it all and work. Download sleeter in my user ( folder spleeter). And download some udpates...but when im put
spleeter separate -i spleeter/audio_example.mp3 -p spleeter:2stems -o output"
Says : the ordinary 242 could not be located in the dynamic link library. Libiomp5md.dll

 

I can say that i do it.
Download git ( in site have link )
Download minianaconda
Then open anaconda pumt ( is like dos windows )
Then copy and paste comands that say site.