Audio Noise Reducer — Remove Background Hiss & Hum
Reduce steady background noise (hiss, hum, fan) in audio files: a filter-and-gate chain in your browser, or a stronger FFT denoiser on our server (free Google sign-in).
About Audio Noise Reducer
An audio noise reducer suppresses steady background noise — fan hum, computer whirr, traffic rumble, hiss — in a recording, producing a cleaner voice track for podcasts, video voice-overs and shareable recordings. The ZTools Audio Noise Reducer has two modes. On this device, it runs a high-pass filter and a smoothed noise gate with the Web Audio API, driven by one strength control (with the individual filters in advanced settings), lets you compare the original and the result side by side, and exports a 16-bit WAV; nothing is uploaded. On our server, ffmpeg's FFT denoiser removes noise frequency by frequency at light, medium or strong settings, with a low-rumble filter that is on by default, and saves WAV, MP3, M4A or FLAC. The server mode takes files up to 100 MB and 30 minutes, needs a free Google sign-in, and deletes your file right after processing.
Use cases
- Podcast / interview cleanup. Recorded in a room with HVAC hum or laptop fan noise. Removing the constant background leaves the voice easier to listen to over a long episode.
- Field-recording cleanup. Outdoor recording with wind / traffic. Noise reduction can recover some intelligibility lost to steady ambient noise (within limits — heavy noise destroys the underlying audio).
- Voice-memo polish. A voice memo recorded near a noisy appliance. Light noise reduction lowers the drone without making the voice sound hollow.
- Pre-transcription cleanup. Speech-to-text accuracy drops with background noise. Cleaning a noisy recording first can help the transcript, especially for steady hum and hiss.
How it works
- Add an audio file. MP3, WAV, OGG, Opus, FLAC, AAC or M4A. The server mode also takes a video with sound.
- Choose where to process it. "On this device" cleans it in your browser with no upload. "On our server" uses a stronger FFT denoiser and needs a free Google sign-in.
- Pick the strength. Start with medium. Strong removes more noise but can make voices sound thin or watery.
- Preview. Listen to the result next to the original. Back off the strength if the voice starts to sound processed.
- Export. On this device: a 16-bit WAV. On our server: WAV, MP3, M4A or FLAC. The original file is never changed.
Examples
Input: 5-min podcast with HVAC hum
Output: Hum much lower; voice intact at medium strength.
Input: Interview with laptop fan whirr
Output: Fan noise reduced; the voice stands out more clearly.
Input: Outdoor recording with wind gusts
Output: Steady wind reduced; sudden gusts are harder to remove. Use a wind muff in the field for best results.
Frequently asked questions
What noise can it remove?
Steady noise: hum, hiss, fans, air conditioning, distant traffic. Changing noise (people talking, music, clattering) is much harder and stays partly audible.
What is "musical noise"?
An artifact of aggressive noise reduction — residual noise becomes pitched bursts that sound watery / metallic. Caused by over-suppressing frequencies. Light settings avoid it; strong settings risk it.
Can I remove voices in the background?
No — a voice has a complex, time-varying spectrum, similar to the foreground voice. Separating two human voices requires AI source separation, which this tool does not do.
Which mode removes more noise?
The server mode: its FFT denoiser works frequency by frequency, while the on-device mode filters low rumble and gates quiet passages.
Is the audio uploaded?
Only in the "On our server" mode, which sends it over HTTPS to our own processing server; the server deletes the file and the result right after processing. The "On this device" mode never uploads anything.
Why does my voice sound thin after reduction?
A strong setting also removed voice frequencies that overlapped with the noise. Lower the strength and try again.
Pro tips
- Always record clean before relying on cleanup. Reduction recovers some quality but never restores what was masked.
- Start with light or medium reduction; go stronger only if needed. Strong settings introduce more artifacts than mild noise.
- Keep the rumble filter on for traffic or air-conditioning noise below the voice range.
- For broadcast-quality results, follow noise reduction with EQ + compression to restore tonal balance.
- Keep the original file — a cleaned version cannot be turned back.
Reviewed by Ahsan Mahmood · Last updated 2026-09-11 · Part of ZTools.
For the full,
formatted version of this page, please enable JavaScript and reload
https://ztools.zaions.com/tools/audio-noise-reducer.