How to Remove Background Noise from a Voice Recording Without Uploading It

1. Import the recording you want to clean

Open Voice Cleaner and use Choose file to load your recording. On a computer, you can also drag and drop the file.

In addition to MP3, M4A and WAV, you can open AAC, OGG, WebM, MP4 and other formats. Actual format support varies by browser.

To make a new recording, choose Record and grant microphone permission.

Files are limited to 250 MB, and recordings to 60 minutes.

2. Try just 10 seconds from a noisy section

A waveform appears after you load the file.

Click or tap a noisy spot, or drag along the waveform to set the playback position.

Select Try 10 seconds from here to apply noise reduction to a 10-second segment starting at that position.

You do not need to process a long recording all at once. Start with a short section to see how much difference it makes.

3. Compare the original and cleaned audio

Once processing finishes, switch between Original and Cleaned. You can compare both at the same playback position to hear the difference.

Do not check only whether the background noise has decreased.

Even if fan or air-conditioning noise is quieter, speech may sound muffled or unnatural. Make sure the voice itself has not become harder to understand.

Voice Cleaner does not offer a noise-reduction strength control. If the result sounds unnatural, keeping the original recording may be better.

4. Process the entire file if the sample sounds good

If the 10-second sample sounds good, select Clean the whole file.

Progress is displayed while the file is processed, and you can cancel partway through.

The 10-second preview cannot be saved by itself. You must process the whole file before exporting a WAV.

Processing happens on your device. Long recordings or phones with limited memory may take more time or fail to finish.

5. Save as WAV

After the full-file process finishes, compare the result with the original again.

If it sounds good, check the filename and select Save WAV. Your original recording will not be overwritten.

Export is fixed at 16 kHz mono, 16-bit PCM WAV.

If you imported MP3 or M4A, you cannot export in that same format. A stereo recording is also converted to mono.

The same applies to video files such as MP4: only audio is saved. Voice Cleaner cannot write cleaned audio back into a video.

Noise it can reduce, and noise that is harder to remove

Voice Cleaner mainly targets background noise mixed with spoken words.

Examples include the steady sound of air conditioning or a computer fan, and the hiss picked up by a microphone.

Other people talking nearby, music, strong echoes and sudden loud noises may not be reduced sufficiently.

Reducing background sound may also change the quality of the voice. For important recordings, keep the original as well as the processed file.

Using Voice Cleaner on a phone

On a phone, switch between Audio, Clean and Save using the bottom navigation. Save becomes available only after the whole file has been processed.

Recording from the microphone requires browser permission.

Some browsers do not allow microphone access when you open a downloaded standalone HTML file directly. Use the HTTPS version if that happens.

If an audio file will not open, your browser may not support its format. You can try converting it to another format first.

Is the recording sent to a server?

Voice Cleaner handles loading recordings, reducing noise and generating WAV files inside your browser.

It does not upload imported recordings or cleaned audio to an external server for conversion.

Noise reduction uses an AI model called GTCRN. It is included in the HTML and runs on your device.

The web version needs a connection to load the page initially, but processing the audio itself does not require network access. The distributed standalone HTML version can work offline in a compatible browser.