Fadr vs Open-Unmix

Fadr is a music tool for separating vocals and 15 other instruments from songs, making remixes and DJ sets, and creating instruments and drum kits with AI. Fadr Basic offers unlimited stems, remixes, and DJ sets, with high-quality MP3 downloads; song uploads on Basic are limited to 10 minutes. Fadr Plus provides 18 stem types, uploads up to one hour, WAV downloads, plugins, API access, and unlimited storage for $10/mo.

Siren makes playable instruments from words and provides ADSR controls and low-pass, high-pass, band-pass, and notch filters. DrumGPT creates 16-element drum kits from text prompts, with pitch, volume, and amplitude-envelope controls. The Stems Plugin works with major DAWs and other software that supports VST3, with Windows VST3 and Mac VST3/AU support. Siren is available online, as an iPhone and iPad app, and as an audio plugin. The API separates songs into vocals, drums, bass, melodies, and instrumental, and detects MIDI, chords, key, and tempo. API use requires Fadr Plus; its Create Stem Task endpoint costs US$0.05 per minute of audio.

Open-Unmix is a PyTorch reference implementation for music source separation, aimed at researchers, audio engineers, and artists. Its ready-to-use models split pop music into vocals, drums, bass, and other instruments; a separate model, umxse, is for speech enhancement and was trained on the 28-speaker Voicebank+DEMAND corpus. Pre-trained choices include umxl, umxhq, umx, and umxse, with spectrogram variants also available. Each source model uses a three-layer bidirectional LSTM, a design the project says cannot run online or in real time. The code supports reproducing results, developing architectures, and adding data, while training is documented separately from the package. Install through pip or use the Docker container; pre-trained models can be loaded in Python or through torch.hub, and a Colab notebook supports online experimentation without local setup. The command line handles WAV, FLAC, and OGG, not MP3; stempeg can extend formats, including STEMS export. The software is MIT-licensed; umxl weights are non-commercial under CC BY-NC-SA 4.0.