Are files uploaded for this tool?

This browser tool processes the selected media locally on this device. The file is not uploaded to ArtPlayer servers for this workflow.

Will the export have a watermark?

No. ArtPlayer browser tools do not add watermarks to exported files.

Why can processing take time?

The work runs in your browser, so speed depends on your device, browser, memory, source size, and the export settings needed for the result.

Which browsers work best?

Recent Chrome, Edge, and other modern Chromium browsers usually work best for heavier browser media processing.

What should I try if a large file fails?

Try a smaller source, close other heavy tabs, reduce resolution or duration first, and keep the tab open until the export finishes.

Which files can I use and download?

Two stereo WAV files: vocals and accompaniment, at 44.1 kHz and 16-bit. Choose audio your browser can decode. A video must contain an audio track and use a format supported by the browser. Try a WAV audio source if another format fails.

Does it work offline?

The first run needs a connection to download the model and runtime. Later runs can reuse cached assets for local processing, but the browser can clear its cache.

Will the separation be perfect?

Some instruments, backing vocals, or reverb may leak between tracks. Listen to both outputs before using them.

Why can a long track or a phone run out of memory?

The tool checks estimated memory before processing. The model and decoded audio can exceed the device budget. Try a shorter clip or a desktop browser; CPU processing is slower when WebGPU is unavailable.

Audio and video tool

Vocal Separation

Your media stays privateDemucsWAV

Separate vocals and accompaniment locally in your browser with Demucs. Preview both tracks and download WAV files without uploading your media.

Source media

Drop a file here, or click to choose

AI mode

How it works

Three steps to finish the browser workflow.

1
Choose a file
Choose an audio or video file.
2
Start separation
Download the model and separate the audio locally.
3
Preview and Download
Preview and download the vocal and accompaniment WAV tracks.

Everything you need

Core details about privacy, browser support, output, and practical limits before you start.

Voice and music split

Separate the main vocal track from background music for editing, karaoke, or remix prep.

Instrumental output

Create a backing track for covers, practice, subtitle work, or timeline editing.

Your media stays private

Audio is processed on your device and never uploaded. No account or cloud quota is needed.

Input, output and limits

Check supported inputs, output formats and usage requirements before you start.

Supported input	Audio or video containing a readable audio track; codec support depends on the browser and runtime.
Output result	Two stereo WAV files: vocals and accompaniment, at 44.1 kHz and 16-bit.
Processing and privacy	Processing runs on this device. Source media is not uploaded for this workflow.
Browser requirements	The first run downloads model files. Processing needs compatible browser APIs and enough device memory.
Usage limits	Processing capacity depends on duration, dimensions, codecs, settings and available device memory. Large inputs can fail; keep the original file.
Will the separation be perfect?	Some instruments, backing vocals, or reverb may leak between tracks. Listen to both outputs before using them.