Get the newsletter
Samsung · AI / ML · Filed Dec 29, 2025 · USPTO published Jun 25, 2026 · Jun 25 release

Samsung Patents an Audio Editor That Lets You Adjust Tracks Before the AI Finishes Splitting Them

Samsung is patenting an audio editor that shows you individual volume controls for voice, music, and background noise the moment it recognizes those sounds exist, without making you wait for the full audio separation to finish first.

By · Updated · 3 min read

Samsung Patent: AI Audio Editing With Instant Track Controls, figure from US 2026/0178265 A1
Figure from the official USPTO publication.
Publication number
US 2026/0178265 A1(US20260178265A1)
Applicant
SAMSUNG ELECTRONICS CO., LTD.
Filed · Published
Dec 29, 2025 · Jun 25, 2026
Inventors
Jaesik SOHN, Seoyeon Jang, Joonhyun Choi, Wonguen Cho, Boyoung Kim, Dongwoo Kim
Status when we published
Waiting for an examiner
Document
20 claims
US classification
700/94
Parent application
Continuation of PCT/KR2025/022914 (filed Dec 26, 2025)(PCTKR2025022914)

What Samsung's per-track audio editing actually does

Imagine you record a video at a noisy cafe. You want to turn down the background chatter and keep your own voice loud and clear. Normally, any app that can do that has to fully process and separate every audio layer before it hands you any controls. That wait can feel like forever.

Samsung's patent describes a system that skips the waiting room. The moment the device's AI identifies that your recording contains, say, a voice layer and a background noise layer, it puts the volume sliders for those two things right on screen. You can start making adjustments while the deeper audio separation work is still running in the background.

The result is a faster, more responsive editing experience. You don't have to stare at a progress bar before you can do anything. The controls appear early, and the audio catches up to your adjustments as the processing completes.

How Samsung separates classification from separation

The patent describes two distinct AI modules working in sequence, and crucially, the user interface doesn't wait for the slower one to finish.

  • Audio classification module: This is the fast first pass. It listens to the recording and identifies what general types of sound are present, for example a human voice, background music, or ambient noise. This step is relatively quick.
  • Audio separation module: This is the heavier, slower process. It actually pulls apart the audio into isolated tracks, one per sound type, so each can be manipulated independently.

The key invention is that the on-screen editing controls (represented as objects for each sound type) appear after classification finishes, not after separation finishes. So when a user adjusts the volume slider for "voice," the device begins applying that change to the audio being separated in real time, or queues it to apply once that track is ready.

This approach means the user interface feels immediate and responsive even though the computationally expensive work is still ongoing. The patent covers the device, the software method, and the stored instructions that make this behavior happen.

What this means for video editing on Galaxy devices

For anyone who edits video on a phone, this kind of responsiveness matters a lot. The biggest friction point in mobile audio editing has always been waiting for processing to complete before you can touch anything. By decoupling the UI from the slowest part of the pipeline, Samsung could make on-device audio mixing feel much closer to a desktop experience.

This fits directly into Samsung's Galaxy AI push, where the company has been adding AI-powered video and audio tools to its flagship phones and tablets. A feature like this would be a natural addition to Samsung's native video or voice recorder apps, giving everyday users pro-level audio control without requiring them to export files to a separate editor.

Source: US 2026/0178265 A1 at the USPTO (PDF). Full patent text and figures are in the official publication.

Get our take in your Top Stories

Liked this breakdown? Add Patentlyze as a preferred source on Google, and our plain-English take shows up more often in your Top Stories the next time Samsung patent news breaks.

Add Patentlyze on Google One click.
There are more where this came from

We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.

More from Samsung

All 1,629 Samsung patents →

More in AI / ML

All 2,637 AI / ML patents →