Loading

The problem? A low, persistent electrical hum from an old refrigerator and the buzz of a thousand bees outside bled into every word.

The final documentary screened at a small festival. An audience member told Lena, “I felt like I was sitting right next to Arthur in that shed.”

Most noise reduction plugins (like iZotope RX, Brusfri, or NS1) need a noise profile. Lena zoomed into a two-second gap between Arthur’s sentences—just the hum and bees. She set NS1 to “learn” from that selection. The plugin analyzed the specific frequency fingerprint of the noise.

She applied it to the clip, hit play—nothing. The hum was still there. Then she realized her mistake: she hadn’t trained the plugin.

Lena was editing a documentary about a beekeeper named Arthur. The footage was gorgeous—close-ups of honey dripping off a comb, slow motion of bees taking flight. But the centerpiece was Arthur’s interview, recorded in his wooden shed.

The hum vanished. The bees became a distant whisper, not a roar. Arthur’s voice was clear, natural, still sitting in the room’s acoustic space. No watery artifacts.

Loading
How long does the OET Test take?