Audio Polish Studio (Loudness, Hum, Ducking)
Normalise a voice or video track to YouTube/podcast loudness, notch mains hum, gate room noise, and automatically duck a music bed under the voice. Real DSP on the samples, processed in your browser.
Tool inputs stay in your browser · optional AI search sends your query
Loading tool…
Good to know: Most editors do this in a DAW with a meter and a sidechain compressor. Here it is three clicks: pick the target loudness (-14 YouTube, -16 podcast, -23 broadcast), optionalled notch 50/60 Hz mains hum, high-pass the rumble, then process and download a 16-bit WAV. The ducking mode mixes a voice track and a music bed and turns the music down whenever the voice is speaking, with adjustable depth and speed. Loudness is the K-weighted, gated measurement streaming platforms use, and the page shows input and output numbers side by side. Video files are not read — extract the audio first with the Video to Audio tool. Added 2026-11-19 22:34.
How to use it (step by step)
- 1Choose the audio file — for a video, extract its audio first with the Video to Audio tool, then come back.
- 2Pick the loudness target: -14 LUFS for YouTube and Spotify, -16 for podcasts, -23 for broadcast. Add the rumble filter and the mains-hum notch if the recording needs them.
- 3Press Clean & normalise. You get the input and output loudness side by side, the gain applied, and a 16-bit WAV to download.
- 4For a video with music under the voice, switch to Duck music under voice, load the voice and the music, choose how many dB the music drops, and press Mix & duck.
Words to know
- Compression
- Making a file smaller, usually by lowering quality a little, so it uploads and sends faster.
- WAV / MP3
- Audio file types. WAV is full quality but big; MP3 and M4A are smaller.
Spotted a wrong result or an outdated rule? Report a mistake · How we check our tools · Get new tools by email
Related tools
See all tools in the Video Creator Toolkit: Cut, Narrate & Finish →