Hugging Face's logo

Eddycrack864
/
AI-Cover-Maker

Audio-to-Audio
AI
rvc
ai-cover
voice-conversion
audio-separation
vocal-remover
Model card Files
xet
 Community

🎤 AI Cover Maker 🎤

madewithlove

Open In Colab Licence Discord

Make AI covers in one click: the song is separated with python-audio-separator, the vocals are converted with RVC (Applio) and everything is mixed back together. The rewrite of RVC AI Cover Maker UI, built on UVR5 UI. Originally created for the AI HUB community.

Features:

  • User friendly interface: a song (file or link) in, a cover out.
  • Separation with the best UVR models (Roformer, MDX23C, MDX-NET, VR Arch): vocals/instrumental, lead/backing vocals, de-reverb, de-echo and denoise. Each step only lists the models made for it, with a recommended model and three presets (Fast, Balanced, Best quality).
  • Voice conversion with RVC (Applio): rmvpe/fcpe/crepe/swift, index, protect, autotune, automatic pitch, formant shifting, custom embedders, noise cleanup.
  • Convert the backing vocals with a second voice, or keep them.
  • Change the key of the whole song keeping its tempo.
  • Vocal effects (high-pass, compressor, reverb) and a mix that never clips.
  • Every step is cached: trying another voice, effect or mix on the same song only recomputes what changed.
  • Each cover is saved in its own folder with every stem made along the way, for your own remixes.
  • Add voice models by link (Hugging Face, Google Drive, direct links) or upload them.
  • Songs from all sites supported by yt_dlp. Check the complete list here.
  • Batch covers.
  • Save your settings, themes and 19 languages.
  • Colab support.
  • Windows/Linux support.

Requirements

Hardware Requirements:

  • NVIDIA GTX 10 series or newer (CUDA 12.8).
  • About 15 GB of disk space.

CPU works, but it will be very slow. If you don't meet the hardware requirements use our Colab.

Prerequisites:

  • Git. You can download Git here.
  • FFmpeg. You can download FFmpeg here or you can use my automated installation script (for Windows).
  • For linux users, run this command on a terminal: (Debian and Ubuntu users): sudo apt install ffmpeg git build-essential (For Arch linux users): sudo pacman -S ffmpeg git base-devel (For Fedora users): sudo dnf install ffmpeg git gcc gcc-c++ (Some distributions already come with Git, FFmpeg and a C compiler preinstalled so this step may be optional. The compiler is needed to build one of the dependencies during installation.)

FFmpeg has to be added to the PATH. (only needed on Windows)

Getting Started

Clone the repository (git needed) or download the source code of the latest release here

git clone https://github.com/Eddycrack864/AI-Cover-Maker.git

Then continue with the steps described below

1. Installation

Run the installation script based on your operating system:

  • Windows: Double-click AI-Cover-Maker-installer.bat (DONT RUN AS ADMINISTRATOR 🚧).
  • Linux: Run AI-Cover-Maker-installer.sh with chmod +x AI-Cover-Maker-installer.sh and ./AI-Cover-Maker-installer.sh.

It creates a private Python 3.12 environment in the env/ folder, so your system Python is not touched.

2. Running AI Cover Maker

Start AI Cover Maker using:

  • Windows: Double-click run-AI-Cover-Maker.bat.
  • Linux: Run run-AI-Cover-Maker.sh with chmod +x run-AI-Cover-Maker.sh and ./run-AI-Cover-Maker.sh.

Extra options can be added to the run scripts (or to python app.py):

Option Description
--share Create a temporary public Gradio link
--listen-port 8080 Use another port (default: 7755)
--server-name 0.0.0.0 Make the UI reachable from other devices on your network
--open Open the UI in your browser (the run scripts already do this)

Example: run-AI-Cover-Maker.bat --share --listen-port 8080

3. Update AI Cover Maker (If you want/need it)

Update AI Cover Maker using (git needed):

  • Windows: Double-click AI-Cover-Maker-updater.bat.
  • Linux: Run AI-Cover-Maker-updater.sh with chmod +x AI-Cover-Maker-updater.sh and ./AI-Cover-Maker-updater.sh.

Your settings (user_data/), models (models/) and covers (outputs/) are kept; updating never touches them.

If you find an error when installing or running the program please consult this troubleshooting file first, if your error is not described there please create an issue

4. Debug (If you want/need it)

Check what AI Cover Maker runs on (GPU, CUDA, FFmpeg):

  • Windows: Double-click info/status-checker.bat.
  • Linux: Run info/status-checker.sh with chmod +x info/status-checker.sh and ./info/status-checker.sh.

How to use

  1. Add a voice model in the Voices tab (a link or the .pth and .index files), then select it in the Cover tab.
  2. Drop a song, or paste a link in Download a song by link.
  3. Set the Pitch if needed (see below) and press Make cover!.

Your cover is saved in outputs/<song> (<voice> Cover)/, together with every stem made along the way.

Separation steps

Choose a preset (Fast, Balanced or Best quality) or the model of each step. Each list only shows the models made for that step, the recommended one marked with a ⭐ (Show all models lists them all).

Step What it does
Vocals / instrumental model Separates the vocals from the instrumental. Required
Lead / backing vocals model Optional. Converts only the lead vocals and keeps the backing vocals apart
De-reverb model Optional. Removes the reverb: RVC sounds much cleaner on dry vocals
De-echo model Optional. Removes echo/delay from the lead vocals
Denoise model Optional. Removes noise from the lead vocals

Voice conversion settings (RVC)

Setting What it does
Pitch Semitones. +12 for a male voice with a female model, -12 for the opposite
Pitch extraction method How the melody is detected. rmvpe is the best for singing
Search feature ratio How much the .index file is used: more accent of the model, but more artifacts
Protect voiceless consonants Keeps breaths and consonants clean. 0.5 disables it
Volume envelope 1 keeps the loudness of the converted voice, 0 follows the original singer
More conversion settings Autotune, automatic pitch, embedder model, formant shifting and noise cleanup

Vocal effects and mix

Setting What it does
High-pass filter Removes rumble below the cutoff frequency
Compressor Evens out the vocal volume
Reverb Puts the dry converted vocals back in a room
Change the key of the song Semitones for the whole song (instrumental and vocals), keeping the tempo
Lead / backing vocals / instrumental volume The volume of each part in the cover, in dB

Every setting is explained in detail in the docs. You can save your settings in Settings > Save settings so they are loaded every time AI Cover Maker starts.

Precompiled Version

The precompiled version already includes the Python environment, so there is nothing to install: download, extract and run.

  1. Get the precompiled version for your PC (the latest one is also linked in each release):

    • Windows (.zip)
    • Linux (.tar.gz, needs glibc 2.35 or newer: Ubuntu 22.04, Debian 12, Fedora 36 or later)

  2. Extract it, it contains an AI-Cover-Maker folder:

    • Windows: right click > "Extract All..." (or "Extract here" with 7-Zip/WinRAR).
    • Linux: tar -xzf AI-Cover-Maker-vX.X.X-linux.tar.gz

  3. Run run-AI-Cover-Maker.bat (Windows) or ./run-AI-Cover-Maker.sh (Linux). You can now use all the features of the normal installation.

FFmpeg is still needed (see the prerequisites).

Still, to update AI Cover Maker you need to install Git.

Development

The app is a Python package, ai_cover_maker/, organised in layers:

Path What lives there
app.py Thin launcher (python app.py [options])
ai_cover_maker/core/ Separation, RVC (rvc/), effects, mixer, and the cover pipeline/ with its stage cache. No Gradio imports, testable on its own
ai_cover_maker/ui/ Gradio layer, one module per tab in ui/tabs/
third_party/applio/ Inference subset of Applio, vendored by tools/sync_applio.py (never edit it by hand)
assets/ Data only: model catalog (models.json), pipeline presets, default settings, translations, themes
info/ Docs, troubleshooting, the FFmpeg installer and the status checkers

Common tasks:

# Install the test tools (once) and run the tests
env/python.exe -m pip install -r requirements-dev.txt   # Linux: env/bin/python
env/python.exe -m pytest                                 # --slow: real models, --network: online checks

# Check that every translatable string exists in en_US.json and in every other language
env/python.exe tools/i18n_scan.py --check --langs

# Model catalog (needs internet)
env/python.exe tools/sync_models.py --write              # add the models audio-separator supports
env/python.exe tools/tag_model_stages.py --write         # tag each model with the separation steps it fits

# Re-vendor Applio from a checkout in .reference/Applio
env/python.exe tools/sync_applio.py
  • Translations: the English text passed to i18n("...") is the key in every language file, so changing it breaks existing translations. A new string must be added to all the files in assets/i18n/languages/; tests/test_i18n_scan.py fails if a language is missing one.

Contributions

If you want to participate and help me with this project feel free to create an issue if something goes wrong or make a pull request to improve this project.

Any type of contribution is welcome 💖

If you like this project you can star this repository. I will appreciate a lot 💖💖💖

Credits

Only use voices and songs you have the rights to.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Audio-to-Audio
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support