Ultimate Vocal Remover

Description
Ultimate Vocal Remover (UVR) is a free and open-source program that, given a master audio track, separates vocals from instruments, powered by AI (artificial intelligence). It runs fastest with a modern graphics card that supports PyTorch, though it produces identical results with any hardware, albeit more slowly. It may be used through cloud computing services such as Google Colab. Several models and parameters are included with UVR: different options may work better with different songs.

UVR is in active development. This information is accurate as of April 2021.

Google Colab
At least as of April 2021, the Google Colab cloud computing platform can be used to run UVR for free.


 * How to use Vocal Remover 5 in Colab! (video, 18 Apr 2021)
 * Online link to Colab form

"Foolproof" installation method via Windows Subsystem for Linux
This method may suit users with limited or no command line experience in either Windows or Linux, bypassing any idiosyncrasies of using Python on Windows.

sudo apt update sudo apt upgrade sudo apt install python3-pip sudo apt install ffmpeg sudo apt install python-is-python3 cd /mnt/c/uvr/ultimatevocalremovergui-5-beta-cml/ pip3 install --no-cache-dir -r requirements.txt pip3 install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio===0.8.1 -f https://download.pytorch.org/whl/torch_stable.html python inference.py -g 0 -m 2band_32000.json -P models/MGM-v5-2Band-32000-BETA1.pth -i MySong.mp3
 * 1) Enable Windows Subsystem for Linux (WSL).
 * 2) Install a Linux distribution from the Microsoft Store - this tutorial uses Ubuntu 20.04 LTS.
 * 3) Download the source for UVR v5 from GitHub - this tutorial assumes it is downloaded to C:\uvr\ultimatevocalremovergui-5-beta-cml\, which is accessed at /mnt/c/uvr/ultimatevocalremovergui-5-beta-cml/ in WSL.
 * 4) Download UVR model files to the models subfolder within the UVR installation folder.
 * 5) Open the Ubuntu terminal and run these commands to prepare for installing UVR's dependencies:
 * 1) Run these commands to install the required applications and packages: 
 * 1) Process your first track! Place an audio file (e.g. MySong.mp3) in the installation folder and run the following command. All the parameters are described in the GitHub manual. 
 * 1) The output separated audio tracks are saved to the separated subfolder within the installation folder. The vocal track is isolated, not discarded, allowing for portions to be mixed back in with an editor such as Audacity.
 * 2) If results are unsatisfactory, different parameters may perform better. Results from different models may also be edited together, if different ones are better suited to certain sections of a track. UVR's ensemble feature performs this automatically.

Tutorials

 * Ultimate Vocal Remover Installation
 * Ultimate Vocal Remover Basic Use

Gallery
 File:UVR Banner.png|UVR Banner File:UVR Model.png|UVR interface