Basic tool profile

Saturalabs: Isolate a Described Sound from a Recording

Describe one audible voice, instrument or sound effect in an audio or video file. Saturalabs attempts to separate that source and returns both the isolated sound and the remaining audio.

Visit Saturalabs

Sources checked October 25, 2026 · No hands-on product test

Choose the output that matches the task

The official workflow takes MP3, WAV, MP4 or MOV up to 50 MiB. One pass targets one described source. Keep Isolated sound to extract it, or Background audio to keep the remainder with that source reduced.

Video inputs produce WAV audio downloads, not a finished edited video. Align the chosen audio with your footage in a video editor. This is not an automatic four-stem export.

Downloads and free utilities

AI downloads require an account and sufficient credits; check the quote on the result screen and the current download pricing. The pricing page says subscription credits reset monthly and top-up credits require an active subscription to use. Separate browser-only trimming and preparation utilities are free; that does not make AI WAV downloads free.

Listen before choosing

Describe one audible source, such as a ringing bell. Compare the attack and decay against the original at similar volume, and listen to both outputs for missing speech or unwanted bleed. Separation can change tone or introduce artifacts. Prefer original stems when available; we have not tested separation or downloads.