Two passes recover the vocal
A primary pass establishes the vocal split. A complementary pass recovers useful detail, then conservative vocal rescue combines both without blindly replacing the first result.
Planetary RecordsStem separation takes one mixed song and returns its parts. Remove the vocals, keep the instrumental, or receive 2, 5, or 10 named lossless WAVs.
From $3.00
Basic handles vocals and backing music. Standard adds bass, melodic, and drums. Full opens drums into kick, snare, toms, hi-hat, ride, and crash while retaining the useful backing mix.
Optional paired reduction cleans the complete vocals and instrumental before creating deeper component stems, so the cleaned instrumental—not the original backing—feeds the next separator.
Each source moves through stages tuned for vocals, bass, melodic content, and drums. The process is designed to recover useful detail, return an exact manifest, and fail clearly when processing cannot complete.
A primary pass establishes the vocal split. A complementary pass recovers useful detail, then conservative vocal rescue combines both without blindly replacing the first result.
Standard separates bass, melodic content, and the drum bus. Full then opens the drums into kick, snare, toms, hi-hat, ride, and crash.
Every tier has an exact named WAV manifest on its result page. Incomplete runs are rejected, and terminal processing failures return the full debit.
Sign in and upload the song you want to separate. Planetary validates the audio before sending a job.
Pick Basic for vocals and instrumental, Standard for five edit groups, or Full for individually named drum components. Add paired artifact reduction when the source needs cleanup first.
When processing finishes, download the complete tier manifest from your result page, which only your account can open.
Splitting a track is often a step, not the finish. If the source is a rough export, run artifact reduction first. If the goal is a release, finish the result with AI music mastering.
Yes. Basic creates a complete vocal and a complete instrumental from one mixed track, which covers vocal-removal, karaoke, and backing-track use. The source can be AI-generated or conventionally recorded. No separator guarantees perfect isolation.
Basic creates vocals and instrumental. Standard creates vocals, instrumental, bass, melodic, and drums. Full creates vocals, instrumental, bass, melodic, kick, snare, toms, hi-hat, ride, and crash. Instrumental is a useful backing mix and overlaps the component stems; these are not additive studio multitracks.
Planetary first creates the complete vocal and instrumental pair, reduces artifacts in each, and then—when you choose 5 or 10 stems—creates the component stems from the cleaned instrumental. That order lets downstream components benefit from instrumental cleanup.
Yes. The result page offers a separate cleanup action beneath eligible complete vocal and instrumental stems. Because component stems already exist at that point, later instrumental cleanup does not change bass, melodic, or drum outputs.
Planetary uses complementary vocal passes with conservative vocal rescue. Standard and Full then add deeper separation for bass, melodic content, and drums. Results still depend on the source mix, so perfect isolation is not guaranteed.
No separator can guarantee perfect isolation. Reverb, stacked harmonies, distorted sounds, and instruments that overlap the vocal frequency range can leave audible bleed or artifacts.
No. The standalone splitter delivers separate WAV files. Ultra AI mastering uses a two-stem split internally, processes voice and music separately, then recombines them into one mastered track.