Identify sounds in a video before editing its audio

How the tool works

Identify sounds analyzes the selected source locally and lists detected sounds with names translated into the app's languages. Detailed identities can distinguish a bark from a growl when the classifier finds enough evidence. Detection is an aid to review, not a guarantee that every sound will be recognized or independently removable.

Build it step by step

Open Audio, choose the take or added sound, and use Identify sounds. Select a passage where the source you want to inspect can actually be heard.

• Listen to the original passage first and note the sounds that compete with the explanation.
• Run identification for the selected source rather than treating the whole project as one recording.
• Compare the detected names with what you hear, including quiet or overlapping sounds.
• Use the supported mixer choices only when you need an edit and their requirements are available.
• Keep the original if the detected list does not describe the passage well enough to guide a useful change.

An example to adapt

Illustrative exercise: a spoken clip contains a dog bark and background music. Analyze a passage containing the bark, then listen to verify whether the labels match that moment.

A detailed label such as bark or growl may clarify the sound you heard, but it does not mean there is a separate control for every label. Choose the next editing action from the available controls, not from the label’s specificity.

Check before finishing

Identification accepts selected ranges up to five minutes. It uses the included classifier; voice isolation and category mixing need an additional separation model. Detection timestamps are not shown in the list, and recognition confidence should not be read as a promise of successful sound removal.

The timeline editor requires Starter or above. Check the app for current feature access and purchase options before planning your edit.