← Voltar

Como funciona Removedor de voz Stem splitter Perguntas frequentes

Acapella
Extractor

Extract a clean acapella from any track — the vocal on its own, the music gone

MP3, WAV ou FLAC · até 20 MB e 5 minutos · 25 créditos por faixa

How to extract an acapella

  1. Carregue a faixa. Any MP3, WAV or FLAC up to 20 MB and five minutes. The file it extracts from is the file you give it, so start from the best copy you have — a stream rip has already lost detail no extractor can put back.
  2. The AI pulls the voice out. A separation model trained on real isolated stems lifts the sung vocal away from the instrumental. This is not the old phase-inversion trick, which only ever cancelled whatever sat dead centre and took the kick and snare with it.
  3. It opens on the acapella. This page starts with the music fader at zero, so the first thing you hear is the extracted vocal by itself. Lift the music back in if you want to check the extraction against the original.
  4. Save the acapella. WAV if it is going into a DAW to be chopped or timestretched. MP3 if it is going into a phone or a video. The export is exactly the mix on screen.

What producers extract acapellas for

  • Remixes — the acapella over your own instrumental, at your own tempo.
  • Sampling and chopping — a phrase, an ad-lib or a single held note, lifted with nothing sitting on top of it.
  • Mashups — one record’s vocal on another record’s beat.
  • Edits de DJ — an a cappella intro or outro built from the vocal alone.
  • Reference vocals — hearing exactly how a vocal was arranged, doubled and stacked before you track your own.
  • Vocal chains — feeding a clean extraction into a pitch or formant tool instead of a full mix.

Perguntas frequentes

Is it spelled acapella or a cappella?
Both point at the same thing. A cappella is the original Italian and the correct spelling; acapella is what the music industry has spelled it for decades and what almost everybody types. This extractor answers to either.
Does extracting lose quality?
The extraction itself does not resample or re-encode anything — the vocal comes back at the source rate and bit depth, and a WAV export is lossless. What you can hear are artefactos de separação: faint smearing where the model had to guess what was voice and what was a synth doubling it. That is a property of every separator on the market, free or paid, and it is worst on dense, heavily limited masters.
Will backing vocals and ad-libs come out too?
Yes, in almost every case — the model separates sung voice as a category rather than hunting for the lead line. Doubles, harmonies and ad-libs come with it. The exceptions are vocal chops used as an instrument and heavily vocoded or talkbox parts, which the model reads as music, because musically that is what they are.
Can I release a track built on an extracted acapella?
The extractor gives you the audio, not the rights. An acapella pulled from somebody else’s record still needs clearance before any commercial release, however far you take it. Your own material, private practice and official remix stems are the safe ground.
What file types can I extract from?
MP3, WAV, FLAC, M4A and OGG, up to 20 MB and five minutes. Larger files bounce to a 192 kbps MP3 first — it makes no meaningful difference to what the model can separate. The conversor does that in your browser.

Uma só separação, seis portas de entrada

These are all the same AI separation with the faders in a different place. Whichever one you start from, the split runs once and you get both tracks.