Vocal Isolator — Extract Acapella Vocals from Any Song
This vocal isolator pulls the lead and backing vocals out of any track as a clean acapella file, and hands you the instrumental as a second download. Free to try.
Tracks you will get: Vocals · Instrumental
- 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.
Want the music without the singer instead? Make a karaoke track
What Is a Vocal Isolator?
A vocal isolator is an AI tool that separates the singing from a finished song, so you get the vocals on their own as an acapella and the music without them as a beat. Stem Splitter AI's vocal isolator uses AI source separation rather than phase cancellation, keeps the lead vocal, harmonies, and ad-libs together, shows the song's tempo and key, can remove the reverb for a dry take, and saves both files as WAV (16 or 24 bit, 44.1 or 48 kHz) or MP3.
How do you check the acapella before you download?
The vocal isolator splits the first 30 seconds of your song for free, and you can switch between the original, the acapella, and the beat from the same spot. Play the loudest chorus: the vocal should come through with no drums or bass under it, and the beat should have no voice left in it.
If the song opens with an instrumental intro, choose "Try another part" and the vocal isolator previews a section with singing. Heavy reverb, live recordings, and low-bitrate MP3s leave the most traces, so upload the best copy you have and download only once the preview sounds right.
How does a vocal isolator compare with the original song?
A vocal isolator splits one mixed song into two tracks of the same length: the acapella, which holds every sung or rapped line, and the beat, which holds everything else.
Because the vocal isolator makes both files from the same separation, they add back up to the original mix. Lay the acapella over a new beat, or the old beat under a new vocal, and the timing lines up from the first sample.
| What it holds | Original song | Acapella | Beat |
|---|---|---|---|
| Holds | Vocals and every instrument | Lead vocal, harmonies, ad-libs, and the reverb on them | Drums, bass, chords, synths, and effects |
| Good for | Reference while you work | Remixes, mashups, vocal study, sampling a hook | Freestyles, covers, a new topline |
| Download | — | WAV (16/24-bit, 44.1/48 kHz) or MP3 | WAV (16/24-bit, 44.1/48 kHz) or MP3 |
Why do so many acapellas sound thin or watery?
Most people who search for a vocal isolator have already tried one that let them down. The old trick of cancelling the centre of a stereo file is not a real vocal isolator: it takes out the bass and kick along with the voice and leaves the hi-hats behind. Older AI models do better, but leave a swirl of cymbals under every held note and smear the consonants, so the acapella sounds like it was recorded underwater.
The download often makes it worse. A free vocal isolator tends to hand back a low-bitrate MP3, which adds its own artifacts on top of the separation, and a daily cap stops you after a couple of songs. Then you still have to find the tempo and key somewhere else before the vocal will sit in your project.
Why use this vocal isolator for your acapella?
One upload, two files
Every split from the vocal isolator gives you the acapella and the beat. It works as a beat remover when you want the voice, and as a vocal remover when you want the music.
Lossless WAV for your session
Files save at 44.1 kHz by default, or pick 24-bit and 48 kHz so the acapella matches a DAW project. MP3 is there when file size matters more.
Tempo and key while you listen
The preview reads the BPM and key of the song, so you know whether the vocal fits your track before you download it.
Dry vocal on request
If the singer was recorded with a lot of room or plate reverb, send the finished acapella through reverb removal for a drier take that sits in a new mix.
A preview from the busiest part
The free 30 seconds come from where the band and the singer are both going, not from a quiet intro, so you hear the hardest part first.
Why does separation quality decide whether an acapella is usable?
11.3 dB vocal SDR on a public multi-genre benchmark
Signal-to-distortion ratio measures how much of a separated track is the real voice and how much is leftover music or artifacts. A higher number means fewer cymbal ghosts under sustained notes and cleaner breaths and consonants, which is what lets an acapella from a vocal isolator stand on its own in a remix.
Files that line up sample for sample
Both files are the same length as your upload and start on the same sample. Drag either into a DAW at bar one and it plays in time with the original, with no nudging or time-stretching.
What files can you upload?
Upload the song file itself; the vocal isolator does not take links from streaming sites.
| Input | Details |
|---|---|
| Audio | MP3, WAV, FLAC, AIFF, M4A, AAC, OGG, Opus |
| Video | MP4, MOV, M4V, WebM, MKV; the soundtrack is separated |
| File size | Up to 2 GB |
| Free preview | 30 seconds, 3 songs a day without an account and 20 once signed in |
| Full song | Charged once by length; 1 free minute on sign-up, then packs from $1.99 |
| Output | Acapella and beat as WAV (16/24-bit, 44.1/48 kHz) or MP3 |
How do you make an acapella with the vocal isolator?
- 1Drop in the songChoose an audio or video file.
- 2Hear the previewSwitch between the acapella, the beat, and the original for 30 seconds.
- 3Pick a formatKeep 44.1 kHz WAV, choose 24-bit or 48 kHz, or save MP3.
- 4DownloadSave the full acapella, or the beat on its own.
Who uses an acapella maker, and what do they get?
Remix producer
Runs a released single through the vocal isolator at 24-bit 48 kHz, reads the key off the preview, and builds a new drop under the hook.
A full-length acapella on the grid of the original tempo.
Mashup DJ
Takes the vocal from one track and the beat from another. Every split gives a beat file too, so either song can be the bed.
Two acapellas and two beats ready to pitch and blend.
Singer or vocal coach
Uses the acapella extractor to pull the lead line out of a dense mix and hear breaths, runs, and harmony stacks that the band normally hides.
An isolated vocal to loop, slow down, and learn from.
Which song tool fits your job?
Each tool keeps a different part of the song; the vocal isolator keeps the voice.
| What you need | Tool | What you get |
|---|---|---|
| Only the singing, plus the beat | Vocal Isolator | Acapella and beat |
| Music to sing over, harmonies kept | Karaoke Maker | Backing track with scrolling lyrics |
| Every instrument on its own track | Stem Splitter | Vocals, drums, bass, guitar, piano, other |
| The song without drums | Drum Remover | Drumless mix and a drum stem |
| The vocal melody as notes | Audio to MIDI | A .mid file of the melody |
| The tempo and key only | BPM & Key Finder | BPM, key, and Camelot code |
What happens to the songs you upload?
Songs sent to the vocal isolator, and the separated files, are stored only to process them and are deleted after 24 hours. They are not shared, published, or used to train models. Only upload songs you own or have permission to use.
Vocal isolator FAQ
Is the vocal isolator free?
Yes, to try. Every song gets a free 30-second preview of the acapella and the beat, without an account. Downloading the full song is charged by its length; a new account includes 1 free minute, and minute packs start at $1.99 for 20 minutes.
Can I use it as a beat remover and keep only the vocals?
Yes. Removing the beat and isolating the vocals are the same split. Download the acapella for the voice alone, or the beat for the music alone; both come from one upload.
How does this vocal isolator compare with Audacity or a DAW trick?
Audacity's Vocal Reduction and Isolation effect and the center-channel trick in a DAW keep only what is panned to the middle, so they work on some stereo mixes and leave kick and bass in the vocal. An AI vocal isolator learns what a voice sounds like, so it also separates panned and reverb-heavy vocals. Desktop apps such as UVR are free, but need an install and a fast GPU.
Does the acapella include backing vocals and harmonies?
Yes. Every voice in the song is kept: the lead, harmonies, doubles, and ad-libs. If you want a track with the lead removed but the harmonies kept, use the karaoke maker instead.
How do I get a dry acapella without reverb?
The isolated vocal keeps the reverb that was on it in the mix. Once the acapella is ready, choose Remove reverb to run it through reverb removal; that runs as a new file with its own free preview.
What format does the vocal isolator save?
A stereo WAV at 44.1 kHz and 16 bit by default. You can choose 24-bit, 48 kHz, or MP3 before you download.
Can I isolate vocals from a YouTube or Spotify song?
Paste-a-link is not supported. Upload the audio or video file instead; video files such as MP4 and MOV work, and the soundtrack is separated.
How accurate are the tempo and key it shows?
They are estimates read from the preview, labelled as such. They are right for most pop, hip-hop, and dance songs; confirm by ear for songs that change key or tempo. For a full-length reading of several files, use the BPM and key finder.
Ready to try the vocal isolator?
Drop a song into the vocal isolator and hear the acapella in seconds. The 30-second preview is free.
Isolate vocals freeLast updated: