Audio Noise Reducer

Reduce steady hiss or hum using a selected noise-only profile and compare original and processed WAV playback.

On-device No sign-up Free to use

Export current result controls

  1. 1 Add files
  2. 2 Arrange/Options
  3. 3 Download

Drop your file here

or choose one from your device

PCM WAV, mono/stereo, up to 20 MiB and 60 seconds.

    Result

    Showing an example.

    Use the first 0.4 seconds of a two-second recording as a noise profile, then compare the later tone and noise.

    The first 0.4 seconds supplies the noise profile. The downloaded two-second WAV keeps the source rate and channels. Compare its later tone with the original to judge noise reduction and any unwanted changes.

    Loading the tool. Your result will appear here.

    https://usefreetools.net/audio-noise-reducer · 20 November 2026

    Was this helpful?

    How to use Audio Noise Reducer

    1. Choose a PCM WAV. Enter the start and end of a section containing only the noise you want to reduce, without speech or music.
    2. Select Preview noise reduction. Listen to both Original and Reduced, particularly quiet words, breaths and sustained notes.
    3. Lower Noise subtraction strength or choose a cleaner noise sample if wanted sounds become metallic or disappear. Preview again after changing settings.
    4. Run the tool and download noise-reduced.wav. Keep the original and check the full recording before using the edited copy.

    Example: Audio Noise Reducer

    Use the first 0.4 seconds of a two-second recording as a noise profile, then compare the later tone and noise.

    You add
    Noise-only sample start (seconds): "0" Noise-only sample end (seconds): "0.4" Noise subtraction strength: "1"
    You get
    The first 0.4 seconds supplies the noise profile. The downloaded two-second WAV keeps the source rate and channels. Compare its later tone with the original to judge noise reduction and any unwanted changes.

    Options

    Noise-only interval
    Choose a section whose hiss or hum also occurs behind the wanted audio. The profile needs at least 2,048 sample frames. Including a voice or instrument can cause the processor to reduce that sound throughout the recording.
    Noise subtraction strength
    Start at 1. Values from 0.25 to 3 change how much estimated noise power is subtracted. More strength can create metallic artifacts; judge the result by listening.

    Supported inputs and limits

    • Local frequency-dependent spectral subtraction using the existing FFT.
    • PCM WAV mono/stereo, up to 20 MiB, sixty seconds and 24 MiB decoded samples.
    • Noise profile needs at least 2,048 frames.
    • Strength 0.25–3.
    • A noise sample containing speech/music can remove wanted sound.
    More details

    Changing noise, wind and music may cause metallic artifacts.

    This is not a volume control or silence gate.

    Where your input is processed

    This tool does not upload your inputs. See the limits for anything saved on your device.

    Choose a profile that represents the noise

    A steady fan or hiss may have a similar frequency pattern during a pause and behind speech. A short pause can therefore provide a useful profile. A passing vehicle, changing wind or another voice may not match that pattern. Compare several quiet and active passages after processing. If the profile contains words or music, choose another interval before increasing strength.

    Questions about Audio Noise Reducer

    Which noise works best?

    Steady hiss and hum whose spectrum is represented by the selected noise-only interval.

    Why does speech sound metallic?

    The noise profile or strength may remove speech frequencies. Choose a cleaner profile or lower strength.

    Does it upload audio?

    No. The denoising and A/B WAV playback run in this browser.

    Page updated · Project managed by Tony Hines · Report a problem with Audio Noise Reducer