How to Remove Background Noise from Audio: Free Guide

You’ve recorded the perfect take, only to hear an annoying air conditioner hum threading through your voice. That kind of background noise can sink a podcast, a client call, or a video project before it even starts. The good news: removing it doesn’t require expensive software or a sound engineering degree. Audacity gives you surgical control for free, while AI-powered online tools can clean up your audio in minutes without installing anything. Here’s how both approaches work, with the exact settings that actually get results.

Free Software Option: Audacity · AI Noise Removers: Cleanvoice, Lalal.ai · Official Audacity Steps: Select noise, Get Profile, Reduce · Online Free Trials: VEED, ElevenLabs

Four tools and settings you’ll encounter across every noise removal workflow:

Can I remove background noise from audio?

Yes. Audacity’s Noise Reduction effect analyzes a sample of background noise to create a sonic fingerprint, then removes that noise from your entire recording (Swell AI). AI-powered online tools like Cleanvoice can handle the same job in approximately 10 minutes without any software installation (Cleanvoice AI). The method you choose depends on how much control you need versus how quickly you need results.

Using free software like Audacity

Audacity gives you granular control by letting you manually select a “noise profile” — a sample of the unwanted sound — and telling the software exactly what to hunt down and remove from the rest of your track (SimpleClean). This surgical precision means it handles consistent noise like air conditioner hums, computer fans, and microphone hiss effectively without degrading your actual audio (Swell AI). You need a few seconds of “clean” background noise in your recording to build that profile.

AI-based online tools

Online tools like Cleanvoice and Lalal.ai use machine learning models trained to distinguish speech from background noise. They work directly in your browser and handle challenging conditions like wind, fans, air conditioning, and even drums (Cleanvoice AI). The main advantage is speed — upload, process, download — but you lose the fine-grained settings control that Audacity offers.

Is there an app that removes background noise from audio?

Several options exist across desktop and web platforms. Audacity is free and open-source; Cleanvoice and Lalal.ai are subscription-based AI services with free trial options.

Desktop apps: Audacity

Audacity’s Noise Reduction effect has three main sliders that control how aggressively the software operates (SimpleClean):

• Noise Reduction (dB): Default is 6 dB, but starting at 12 dB typically produces better results without artifacts (Boris FX)
• Sensitivity: Ranges from 0 to 24; a starting value of 6.00 keeps false removals minimal while still being effective (SimpleClean)
• Frequency Smoothing (bands): Setting this to 3 fills sonic holes left after noise removal, preventing that thin or tinny quality (SimpleClean)

Web apps: Cleanvoice and Lalal.ai

Cleanvoice removes background noise from audio and video in approximately 10 minutes, handling conditions like wind, fans, AC units, and screaming children (Cleanvoice AI). Lalal.ai focuses on voice isolation, cleaning vocals from music tracks. Both require accounts for full processing, though limited free trials exist.

How do I stop a microphone from picking up background noise?

Prevention and correction work together. Software solutions like Audacity’s Noise Gate remove intermittent sounds between words, while hardware noise-cancelling microphones reduce noise at the source.

Software noise gates

Audacity’s Noise Gate has four key parameters that control how it operates (Swell AI):

• Threshold: The volume level below which audio gets reduced
• Attack time: How quickly the gate opens when sound crosses the threshold
• Hold time: How long the gate stays open after sound drops below threshold
• Decay time: How gradually the gate closes

To use Audacity’s Noise Gate effectively, select a section containing just background noise, go to Effects > Noise Gate, click “Select Function: Analyse Noise Level,” and press OK to get an auto-recommended threshold (Audacity Support). For maximum effectiveness, apply Noise Reduction before Noise Gate — the first reduces constant background noise, then the gate handles remaining intermittent sounds (Swell AI).

Hardware noise-cancelling mics

Dedicated noise-cancelling microphones use directional pickup patterns and built-in electronics to minimize background noise before it reaches your recording software. These typically cost more than standard USB microphones and may still let through very loud ambient sounds.

Is there a free tool to remove background noise?

Audacity remains the gold standard for free noise removal, while VEED and similar services offer free trials of their AI-powered tools.

Audacity noise reduction

Audacity is completely free and open-source. The workflow is straightforward: import your audio, select a section with only background noise (no speech), go to Effects > Noise Reduction, click “Get Noise Profile,” then select your entire track and apply the effect with your chosen settings (Boris FX). A longer, cleaner noise sample gives the software more data to work with, producing better results (SimpleClean).

VEED free trial

VEED offers a free trial for its background noise removal feature, processing audio directly in your browser without software installation. The catch is that free users face usage limits, and full-resolution exports typically require a paid subscription.

Is there a way to remove background music from audio?

Yes, through AI stem separation and targeted noise reduction. Lalal.ai specializes in extracting clean vocals from music tracks, while Adobe Podcast and similar services handle vocal isolation from mixed recordings.

AI stem separation

Services like Lalal.ai use machine learning models trained on massive audio datasets to identify and isolate different instrument and voice components. Upload a mixed track, and the service returns separated stems — typically vocals, bass, drums, and other instruments — in seconds.

Noise reduction techniques

For mixed audio where music overlaps with speech, Audacity’s Notch Filter lets you target specific frequencies known to correspond to musical instruments — reducing the musical component without harming the vocal track (Audacity Support). This requires more technical knowledge than the standard noise profile approach.

Step-by-step: Removing background noise in Audacity

Follow these steps in order for best results:

1. Import your audio: Press Ctrl+O (Windows) or Command+O (Mac) to open your file
2. Select a noise-only section: Find a part of the waveform containing only background noise — a few seconds of silence before you speak works well (Boris FX)
3. Get the noise profile: Go to Effects > Noise Reduction and click “Get Noise Profile”
4. Select all audio: Press Ctrl+A (Windows) or Command+A (Mac) to select your entire track
5. Apply Noise Reduction: Return to Effects > Noise Reduction, set Noise Reduction to 12 dB, Sensitivity to 6.00, and Frequency Smoothing to 3, then press OK
6. Preview and adjust: Use the Residue toggle to hear what will be removed — but remember to switch back to “Reduce” before applying (Audacity Support)

What experts say

Audacity’s Noise Reduction effect excels at removing consistent background noise like computer fans, air conditioning hums, or microphone hiss.

— Swell AI, Educational content provider

By feeding it a pure sample of the sound you want to get rid of, you’re essentially giving it a sonic fingerprint to hunt down and remove from your entire recording.

— SimpleClean, Audio editing tutorial

The implication: Audacity rewards patience and methodical technique. The quality of your noise profile directly determines how cleanly your background noise disappears. Online AI tools trade that precision for convenience, which works well when you need results fast and can accept slightly less surgical outcomes.

Related reading: AI Image Generator Free Tools

Frequently asked questions