Choose a Starting Mode for Your Recording

Identify what must remain

Decide whether the recording is mainly a spoken explanation, conversation, video soundtrack or music. List important non-speech sounds too. A mode suited to dialogue may change ambience or musical detail that matters to the scene. Keep the source so you can reverse the decision.

Understand the visible options

Noise Remover offers Auto, Podcast, Video, Call, Voiceover and Music. The names help you start an experiment; they do not establish separate proprietary neural models or objectively superior results for every matching file. Auto is not evidence that every source problem was correctly diagnosed.

Choose by the wanted sound

Compare a quiet phrase, a noisy moment and any important music or effect. Use similar listening levels and check complete consonants and endings. If the result changes useful sound, try a different reversible approach, local edit or retake. Repeatedly processing a processed output can compound artifacts rather than improve it.

Continue with the focused guide

This page remains at its existing address. For the detailed workflow, read the focused guide.

Before processing, keep the original, check your current plan and available minutes, and confirm permission to upload the material. Compare the actual words at similar playback levels, not just the size of the waveform. Download the accepted result before its displayed expiry and reopen the saved file.