Audio Noise Remover

This noise remover cleans up voice recordings, podcasts, interviews and video audio in your browser. Drop one file or a folder, pick the RNNoise neural model (a 4.8 MB one-time download that's excellent on speech) or the built-in spectral noise gate, and download clean WAV files with before-and-after players to compare. Nothing is uploaded.

Which method to use

RecordingBest method
Voice with fan or trafficRNNoise
Voice with steady hissEither
Music or instrumentsNoise gate, low strength
Very noisy phone callRNNoise, mix 0.8

When to use it

  1. Clean podcast and interview recordings.
  2. Remove fan and air-conditioner noise.
  3. Fix lecture and meeting audio.
  4. Reduce hiss on old recordings.
  5. Batch-clean voice notes.

How to Use

  1. Pick RNNoise or the noise gate.
  2. Drop recordings.
  3. Clean the files.
  4. Compare and download.

Features

  1. RNNoise neural model
  2. Built-in spectral gate
  3. Choose a noise sample
  4. Wet/dry mix
  5. Batch with ZIP
  6. Clear downloaded models

FAQs

How does noise removal work?

The gate learns the noise from quiet moments and turns down the frequencies where it sits; RNNoise uses a small neural network trained to separate voice from noise.

Does it work on music?

The gate does, gently. RNNoise is tuned for speech and can dull music.

What formats can I use?

Anything your browser can decode, including MP3, WAV, M4A, OGG and video files; output is WAV.

Is anything uploaded?

No. Files and text are processed in your browser. Only the model itself is downloaded, once, when you ask for it.

Can I delete the downloaded model?

Yes. The Downloaded models panel lists what this tool stored and deletes it after a warning; you'd need to download it again to use it.

Technical Details

Audio is decoded by the browser at 48 kHz and mixed to mono. The built-in method is a spectral gate: a 2048-point STFT with 75% overlap, a noise profile averaged from the quietest 10% of frames (or a chosen time range), and a per-bin gain of (SNR − 1) ÷ SNR limited to a maximum reduction and smoothed over neighbouring bins and frames before overlap-add resynthesis. The neural method runs Xiph's RNNoise, compiled to WebAssembly, on 480-sample frames; it is downloaded from jsDelivr on request, checked against a SHA-384 fingerprint and kept in Cache Storage until cleared. A wet/dry mix and peak normalisation follow, and results are saved as WAV.

Audio Tools

Audio Tools

  • Audio to MIDI Converter
  • Audio Trimmer
  • Audio Merger
  • Audio to WAV Converter
  • Audio Speed & Pitch Changer
  • BPM Detector
  • Tone Generator
  • Voice Recorder