- Vocal Isolator
- Acapella Extractor
Acapella Extractor — Dry 24-bit WAV Vocals, BPM and Key
One upload returns the vocal stem as WAV — 44.1 or 48 kHz, up to 24-bit, sample-aligned with the original — plus the song's BPM and key, and one click to strip the reverb for a dry acapella.
- Tempo and key detected
- 24-bit WAV
- Optional dry vocal
Have several recordings? Choose them all at once, then sign in to process them in one go.
Making a cover? The instrumental can be shifted to your key.
Hear it on a real track: original ⇄ acapella, with the BPM and key read from that song.
What Is an Acapella Extractor?
An acapella extractor is a source-separation tool that isolates the vocals of a song, lead and backing together, with no instruments left, for remixes, samples and mashups. Vocal Isolator's acapella extractor uses a BS-RoFormer model and returns a 44.1 or 48 kHz WAV, 16- or 24-bit, sample-aligned with the original; it also finds the song's key and BPM and removes the reverb in one click.
How Do I Know the Acapella Is Clean Before I Download?
The acapella extractor lets you solo the vocals before you download: it splits the first 30 seconds of your song for free, and you can switch between the original, the acapella, and the instrumental. Listen to the gaps between lines: a clean acapella has no hi-hats, snare, or bass bleeding through.
If the intro has no singing, choose "Try another part" and the acapella extractor previews a verse or the chorus instead. Download only once it sounds right.
- AcapellaEvery vocal layer in the song: lead, harmonies, doubles, and ad-libs.
- InstrumentalThe instrumental the acapella extractor separated comes as its own WAV track.
- FormatBoth tracks as 44.1 kHz, 16-bit WAV, or 24-bit WAV at 44.1 or 48 kHz, sample-aligned with the original.
Why Do Ripped Acapellas Sound Bad in a Mix?
A ripped acapella usually fails in one of three ways. Hi-hats and snare hits bleed into the gaps between lines. Older models leave a watery, phasey sound on sustained notes. And the file comes back as a compressed MP3 that falls apart once you add effects.
Small upload limits make it worse, because they force you to compress the song before you even start. A good acapella extractor should take the full-quality file and return the voice without dragging the band along.
| Method | Typical result | Use it when |
|---|---|---|
| Phase inversion with the official instrumental | Near-perfect, but only if both files come from the exact same master | You own the official instrumental of that release |
| AI separation (this acapella extractor) | Clean vocals with little drum bleed; faint traces on live or heavily distorted audio | You want an acapella from any finished song |
| Older AI models such as Demucs | More bleed and a watery sound on sustained notes (below 8.4 dB vocal SDR) | You already run them offline to batch many songs |
| Official acapella release | Perfect, straight from the studio session | The label published one |
Why Use This Acapella Extractor?
Clean vocals with less bleed
The acapella extractor separates the voice in every frequency band, so drums and bass stay out of the vocal track instead of leaking through the quiet parts. It scores 11.79 dB vocal SDR on the MVSEP Multisong leaderboard.
In time with the original
The acapella starts on the same sample and runs the same length as your song, so it drops onto your DAW grid at the original BPM.
Lossless WAV output, up to 24-bit
Download the acapella as 44.1 kHz, 16-bit WAV, or as 24-bit WAV at 44.1 or 48 kHz to match your session. There is no MP3 compression to smear the high end when you EQ, pitch, or time-stretch it.
Tempo and key, read for you
The preview estimates the BPM and key of the song, so you can set your project tempo and pick a beat in a matching key before the acapella is even downloaded.
A dry acapella on request
Most acapellas carry the reverb of the original mix. Once yours is ready, one click sends it to reverb removal and returns a dry vocal you can put in your own space.
The instrumental comes with it
The acapella extractor also returns the matching instrumental from the same upload, ready for a mashup, an edit, or a reference.
Big files, up to 2 GB
Upload WAV masters, long live recordings, or a video straight from your camera without compressing them first.
Why Does the Acapella Hold Up in a Mix?
A higher vocal score than Demucs
The acapella extractor runs BS-RoFormer, which scores 11.79 dB vocal SDR on the public MVSEP Multisong leaderboard. The older Demucs model scores below 8.4 dB on the same test, and that gap is the bleed and the watery sound you hear.
See the MVSEP leaderboardEvery vocal layer, together
The acapella extractor keeps lead, harmonies, doubles, and ad-libs together in one acapella, in place. You get the full vocal arrangement the artist recorded, ready to chop or rebuild.
What Can You Extract an Acapella From?
Start from the highest-quality file you have. A WAV, AIFF, or FLAC gives a cleaner acapella than a low-bitrate MP3.
| Input | Details |
|---|---|
| Audio | MP3, WAV, M4A, AAC, OGG, OPUS, FLAC, M4B, MKA, AIFF |
| Video | MP4, MOV, M4V, WEBM, MKV |
| File size | Up to 2 GB per file |
| You download | Acapella and instrumental as 44.1 kHz, 16-bit WAV, 24-bit WAV at 44.1 or 48 kHz, or MP3 |
How to Extract an Acapella from a Song
- 1UploadDrop a song or video into the acapella extractor.
- 2PreviewSolo the vocals for the first 30 seconds free.
- 3AdjustNote the tempo and key, set each track level, or trim.
- 4DownloadSave the acapella, the instrumental, or both as 16- or 24-bit WAV.
Who Uses an Acapella Extractor, and for What?
Remixes and mashups
A DJ runs a hit through the acapella extractor and lays the vocal over a new beat at the same BPM.
A remix vocal that stays on the grid because it is sample-aligned with the original.
Vocal samples and chops
A producer pulls a hook or an ad-lib and slices it into a sampler.
Clean one-shots with no drums underneath, ready to pitch and chop.
Learning a vocal part
A singer or vocal coach isolates the voice to hear phrasing, breaths, and harmonies.
Every run and harmony audible on its own, which also helps when writing down lyrics.
Acapella Extractor or Karaoke Maker: Which Vocal Do You Need?
Both pull vocals out of a song. The difference is which voices you get.
| You want | Use | You get |
|---|---|---|
| All the vocals together: lead, harmonies, and ad-libs | Acapella extractor (this page) | One acapella with every vocal layer |
| The lead vocal alone, without harmonies | Karaoke maker | Download its Lead vocal track |
| The music without any vocals | Instrumental maker | A full instrumental, bass and drums kept |
| A clean spoken voice from a noisy recording | Remove background noise | Speech with hum, fans, and room echo removed |
| Separate drums, bass, guitar, or piano | — | Not supported. Songs split into two tracks: vocals and everything else. |
Try It Free, Pay Only for Full Tracks
Free preview
Hear the first 30 seconds of any song. No sign-in needed.
Free download
Sign in and download 1 minute free.
Full tracks
Minute packs: $1.99 for 20 minutes or $4.99 for 1 hour or $14.99 for 10 hours, and pack minutes never expire. Pro: $9.99/month for 20 hours every month. Minutes are charged by song length, rounded up to the whole minute.
What Happens to Your Song After Upload?
Your upload and the separated tracks are stored only to process and deliver them, and are deleted automatically after 24 hours. Guest uploads are tied to your browser and signed-in uploads to your account, so no other visitor can open them. You need an account only to download.
Acapella Extractor FAQ
Partly. The acapella extractor previews the first 30 seconds of any song free, with no account. Sign in and your first 1 minute of downloads is free too. Longer songs use minutes from a minute pack or a monthly plan, charged by song length and rounded up to the whole minute, and pack minutes never expire.
Very little on most songs. The acapella extractor uses BS-RoFormer, with 11.79 dB vocal SDR against below 8.4 dB for Demucs. Live recordings, heavy distortion, and low-bitrate files can still leave faint traces, so solo the preview before you download. A WAV or FLAC source gives the cleanest acapella.
Yes, with the karaoke maker. The acapella extractor keeps every vocal layer together: lead, harmonies, doubles, and ad-libs. The karaoke maker separates the lead singer from the backing vocals instead, and its Lead vocal track downloads on its own, which gives you a lead-only acapella for a remix.
Upload the song to the acapella extractor. While you preview, it estimates the tempo and key of the part you are listening to and shows them above the controls, for example "≈ 124 BPM" and "A minor". Songs that change tempo or key, or that have no clear beat, can read differently, so check against your ears before you build a track on it.
Yes. Once the acapella is ready to download, choose "Remove reverb" under the download button. The acapella goes on to reverb and echo removal as a new file, with its own free preview, and comes back dry so you can add your own space in the mix. It works best on moderate reverb; long, washed-out tails may leave a trace.
Yes. Choose 24-bit, and 44.1 or 48 kHz to match your session, before you download. The separation itself runs at 44.1 kHz, 16-bit WAV, so 24-bit gives your edits more headroom rather than more detail in the source. MP3 is there too when you only need a reference.
Yes. The acapella extractor returns two tracks exactly as long as your upload, both starting on the same sample as the original. Drop the acapella into any DAW or DJ software at the song's original BPM and it lines up with the instrumental and the original track without nudging.
Using an acapella for practice, study, or DJ sets you are licensed to play is fine. Releasing or monetizing a remix built on a copyrighted vocal needs the rights holder's permission. Ask the label or publisher for a remix license before you release it, and keep the remix private until you have one.
Yes. The acapella extractor runs entirely online, so it works in the browser on iPhone, Android, Windows, Mac, and Chromebook. There is no app or plugin to install and no graphics card needed, because the separation runs on cloud GPUs. Upload a file from your phone, preview it, and download the result.
Extract Your Acapella Now
Upload a song or video and solo the acapella extractor result free.
Upload a SongLast updated: