Audio Editors

Demucs GUI

5  /  79 Reviews
27,734 Downloads
Nov 3, 2024 Last updated

Downloads

Older Versions

# Change Log

About the files:

Demucs-GUI_1.3b1_cpu.7z: For Windows users without accelerators (e.g. GPU).

Demucs-GUI_1.3b1_cuda_mkl.7z: For Windows users with NVIDIA GPU (whose compute capability is greater than 3.5) or Intel GPU (Intel Arc & Intel Xe).

Demucs-GUI_1.3b1_macOS_x86_64_Rosetta2.dmg: For macOS users (Both Intel Mac and Apple Silicon Mac). You may need to install Rosetta 2 on Apple Silicon Mac. MPS acceleration is available on Apple Silicon Mac.

Demucs-GUI_1.3b1_macOS_ARM64.dmg: For macOS users with Apple Silicon CPU. MPS acceleration is available.

Updates

1. Added new audio tag variables for output file name and ffmpeg command

2. Added clip mode "tanh"

Fixes

1. Package issues (which means that no code changes)

(1) Fixed an issue that models can't be loaded on macOS due to incompatible numpy version

(2) Fixed an issue that the output audio of GPU version on Windows is full of NaNs due to PyTorch compilation issue

(3) Fixed an issue that the GPU version on Windows can't be started due to missing DLLs

2. Now file names retrieved from URL could be URL decoded

Description

Demucs GUI is a GUI for python project demucs (Demucs Music Source Separation) with optimized memory usage in CUDA.

Please read release notes on GitHub Release before using!

The project aims to let users without any coding experience separate tracks without difficulty. Since the original project Demucs used a scientific library torch,the packed binaries with the environment are massive. If you have questions about usage or the project, please open an issue on GitHub to tell us.

Note: This project includes the code of Demucs under the MIT license.

System requirements

For Windows: At least Windows 8

For Mac: At least macOS 10.15

For Linux: We currently have no intention to pack binaries for Linux.

Hardware

Memory: At least 6GB memory,but an 8GB swap may be required.At least 3GB of private memory (not shared memory) is required.The longer the track you want to separate, the more memory will be required.

GPU: Only NVIDIA GPUs are available.

Found this software useful? Please consider a donation to the author.