What is the difference between denoising and noise gating?

Denoising (spectral subtraction) removes continuous background noise throughout the entire recording. A noise gate silences audio that falls below a volume threshold — useful for removing noise between words but not during speech. Our tool uses spectral subtraction, which is better for continuous noise like hiss and hum.

Can I denoise audio without affecting voice quality?

Yes, with light to medium reduction settings. The key is to use the minimum reduction needed. Heavy reduction can make voices sound unnatural. Always preview before downloading.

Does denoising work on music recordings?

Yes, but use light settings (20–40%) on music to avoid affecting the natural sound of instruments. Spectral subtraction is most effective on voice recordings with stationary background noise.

Is there a limit on how many files I can process?

No limit. The tool is completely free with no account required and no usage restrictions.

What audio formats are supported?

MP3, WAV, FLAC, OGG, M4A, and AAC. Output is downloaded as a WAV file.

Denoise Audio Online Free — Remove Hiss, Hum & Background Noise

Clean up any audio recording by removing hiss, hum, fan noise, and room ambience. Uses spectral subtraction for natural-sounding results. No upload, no signup.

Spectral SubtractionHiss & Hum RemovalNo Upload100% Free

Drop your audio file here

MP3, WAV supported — processed locally, never uploaded

Browse file

MP3WAVFLACOGG

Reduction Strength

Spectral subtraction (Boll 1979): noise profile estimated from first 0.5s, subtracted bin-by-bin from 1024-bin FFT frames with 75% overlap. Original phase preserved.

About This Tool

This page targets the technical "denoise audio" query — covering what denoising is, spectral subtraction vs noise gate comparison, step-by-step workflow, and common mistakes to avoid. Unlike the podcast-specific noise pages, this page serves a broader audience including voice-over artists, musicians, and archival audio restorers. The spectral subtraction algorithm runs entirely in your browser with no upload required.

What Is Audio Denoising?

Audio denoising is the process of removing unwanted background noise from a recording while preserving the desired signal (voice, music, or instruments). It is an essential step in audio post-production for:

Podcast and voice-over production
Interview and field recording cleanup
Music recording noise floor reduction
Video production audio cleanup
Archival audio restoration

Spectral Subtraction vs Noise Gate: Which Is Better?

Two common approaches to denoising — here is how they compare:

MethodHow It WorksBest For

Spectral SubtractionEstimates and subtracts noise spectrum bin-by-binContinuous noise (hiss, hum, fan)

Noise GateSilences audio below a volume thresholdIntermittent noise between speech

Our tool uses spectral subtraction — the better choice for continuous background noise like hiss, hum, and fan noise that is present throughout the entire recording.

Step-by-Step: How to Denoise Audio Online

Prepare your recording: Ensure the first 0.5–1 second contains only background noise (no speech or music). This is used as the noise profile.
Upload your file: Drag and drop your MP3, WAV, FLAC, or OGG file onto the tool above.
Adjust the reduction strength: Start with medium (50%) and increase if needed.
Preview the result: Listen to the denoised audio before downloading.
Download: Save the cleaned WAV file for use in your project.

Common Denoising Mistakes to Avoid

Over-processing: Too much noise reduction creates a metallic, robotic sound. Use the minimum reduction needed to achieve a clean result.
No noise tail: If your recording starts immediately with speech, the noise profile estimation is less accurate. Always record 1 second of silence at the start.
Low-quality source: Denoising a heavily compressed MP3 (128 kbps or lower) may introduce additional artifacts. Use WAV or 320 kbps MP3 for best results.
Expecting perfection: Spectral subtraction is excellent for stationary noise but cannot perfectly remove all noise types. For complex noise environments, multiple passes may be needed.