Reduce rumble and hiss online Use high-pass and low-pass filters

No, and it is worth knowing before you start. These are two filters that cut a fixed range at each end of the spectrum. Sounds that share a frequency range with the voice, or that change from moment to moment, cannot be separated this way.

Steady low-frequency content below the cutoff: mains hum, air handling, a desk being knocked, the low thump of wind or handling on a microphone. Speech has little useful content under about 80 Hz, so cutting there is usually free.

Steady high-frequency content above the cutoff: tape hiss, the fine sizzle of a cheap preamp, some air-conditioning noise. Cutting too far down here also takes the clarity out of consonants, so move it in small steps.

The filters do not know which parts are the voice. A high-pass set too high hollows out the body of the sound; a low-pass set too low makes it dull. Use the bypass to compare and pull back until only the unwanted part has gone.

Yes. Bypass switches between the filtered signal and the original using the same filter mapping the export uses, so the comparison is the real thing rather than an approximation.

Filtering will help with the steady part — the hum of the room, the rumble of traffic through a wall. It will not help with voices or sudden sounds. If the recording matters, a second take in a quieter place beats any amount of processing.

Common audio and video containers are accepted: the editor tries native decoding first and falls back to a local FFmpeg build to pull the first compatible audio track. Exports are MP3 from 64 to 320 kbps, 16-bit PCM WAV, M4A/AAC at 192 kbps, and FLAC.

The audio is decoded and filtered locally in this browser tab and is not sent to the application server to perform the edit. The page itself still loads site resources and consented analytics.