How to Reduce Background Hiss in Voice Recordings for Free With Audacity
Use Audacity's noise profile and noise reduction controls to clean steady hiss or fan noise while avoiding distorted speech.
Before you begin
An interview, school lecture or voiceover can be difficult to understand when an air conditioner, computer fan or microphone preamp adds steady hiss. Audacity is a free audio editor with a Noise Reduction effect designed for consistent background noise. The tool cannot perfectly reconstruct words buried under heavy traffic, music or another voice, and aggressive filtering can make speech sound metallic. The best results come from recording clean audio first and applying modest corrections afterward.
1Install Audacity and make an editable copy
Download Audacity through audacityteam.org or its official project links. Open your recording and immediately save a project copy in a clearly named working folder. Keep the raw audio file untouched for comparison. Listen to the complete recording once and note the type of problem: constant hiss, electrical hum, sudden clicks or fluctuating traffic noise. Different problems need different tools. A filter intended for steady hiss may damage speech when applied to a changing crowd background.
2Find a clean sample of the unwanted noise
Look for a short section where nobody is speaking but the background fan or hiss remains audible. Select that section in the waveform. Audacity's official manual explains that the Noise Reduction effect needs a representative region containing only the noise to estimate its frequency profile. Do not choose a section with quiet words, breathing you want to preserve or music, because the algorithm may later try to remove some of that sound from the rest of the recording.
3Capture the noise profile before processing
With only the noise-only segment selected, open Effect, Noise Removal and Repair, Noise Reduction, or the corresponding menu in your installed version. Click Get Noise Profile. Audacity explains that the profile helps estimate the noise floor across frequencies. Once captured, select the portion of audio you want to clean, commonly the entire voice track. A profile obtained from a very different microphone or recording session may not be suitable, so record a fresh sample for each significantly different noise environment.
4Apply gentle noise reduction and preview first
Reopen Noise Reduction and start with conservative settings. Audacity's manual discusses controls for reduction amount, sensitivity and frequency smoothing, plus a Preview button. Listen to a quiet sentence and a loud one before applying the effect across the full recording. If consonants sound watery, robotic or muffled, reduce the effect intensity. A cleaner background is not valuable if your listener can no longer distinguish spoken words. Undo and retry rather than stacking extreme treatments.
5Identify artifacts and choose a different effect when needed
Audacity notes that excessive reduction can create musical or tinkling artifacts. The effect also works poorly when the unwanted noise is louder than the voice or changes unpredictably. For a mains electrical hum, a Notch Filter may be more suitable; for short isolated clicks, click-removal or manual repair tools may help. Do not apply multiple automated effects without checking each result. When speech was recorded too quietly and clipped later, filtering cannot restore information the microphone never captured.
6Export and compare on a normal phone speaker
After cleaning, check the recording's overall loudness and avoid clipping when making any final level adjustments. Export a copy in a format suitable for your destination, such as WAV for an editing master or a compatible compressed format for sharing. Listen through headphones and a normal phone speaker; a fix that sounds good on expensive headphones may hide words on a small device. Keep both the original and cleaned file, and document the settings that worked for future recordings.
Frequently asked questions
Can Audacity remove all traffic noise from speech?
Not reliably. The standard Noise Reduction effect is primarily designed for steady noise and may distort speech if the interference varies.
Is it okay to increase reduction until the room is completely silent?
Aggressive settings can damage the voice. Use the lowest reduction that improves intelligibility.
Official sources and further reading
Interface labels, availability and policies may change. Confirm current details with the relevant service before taking action.