Hi,

The important point is you can not reliably do this in the CapCut mobile app.

What you're asking for is called cinematic audio source separation (or extracting the FX/effects stem while removing the music stem).

This is much harder than isolating speech because music and sound effects often overlap in frequency and time.

For your A Quiet Place example:

• Desired result: ambient sounds + footsteps + wind + creature noises
• Remove: background score/music only

Unfortunately, CapCut Mobile's AI audio tools are primarily designed for:

• Voice isolation
• Vocal/instrumental separation
• Basic noise reduction

They do not currently offer a "keep sound effects, remove music" mode.

CapCut's newer "Separate Audio" tools on desktop can split audio into vocals, instruments, drums, etc., but they aren't designed to produce a clean "sound effects only" track from a movie soundtrack, and availability also varies by platform and version.

If you're limited to CapCut Mobile or CapCut Pad, there really isn't a built-in feature that does what you want.

Based on our other community discussions, the closest option is stem separation, but it doesn't isolate ambient effects cleanly from film audio.

What actually works?

If you don't mind using another tool:

• Ultimate Vocal Remover (UVR 5) (desktop, free) often gives the best results for this type of separation.
• Some AI services are beginning to separate movie audio into Dialogue / Music / Effects (FX) stems, but quality varies depending on how the original soundtrack was mixed.

If you must stay on mobile

At the moment, there isn't a mobile app (including CapCut) that can consistently remove only the film score while preserving all ambient sound effects. You may find AI apps that advertise "music removal," but most are really designed for:

• keeping speech while removing music, or
• separating songs into vocals and instrumentals,

rather than extracting the movie's effects track.

If your source is a movie, there's one exception: if you can obtain a version with separate audio stems (for example, some Blu-ray or production sources have distinct dialogue, music, and effects tracks), you can remove the music perfectly.

But with a normal stereo movie clip, AI has to estimate the separation, so perfect results are rarely possible.