Audio Noise Reduction

How to Use the Noise Reduction Tool

The tool does one thing: it reduces constant background noise in a recording and saves the cleaned-up audio as a file. The whole process takes just a few clicks.

1. Choose an audio file with the recording you want to clean up. You can select material from your computer or phone.
2. Set the reduction strength. Medium is selected by default; you also have mild and strong options.
3. Decide about the lowest frequencies. The field that removes hum below 80 Hz is checked by default. Leave it checked if you hear rumbling or hum in the background, and uncheck it when low tones matter in the recording.
4. Click the “Remove Noise” button. The tool analyzes the recording and processes it in full.
5. Listen to the result and save the file to your device.

The tip included with the tool is simple: you get the best results with constant background noise, such as microphone noise, a fan, air conditioning, or tape hiss. The noise profile is determined automatically from the quietest parts of the recording, so it helps if the material has at least a brief moment of silence—that is what the tool uses to learn what to treat as background.

How to Choose the Reduction Strength: Mild, Medium, or Strong

The reduction strength works like a dial: the higher it is, the more aggressively the tool affects the recording’s background. There are three levels, and each has its use.

Start with medium and listen to the result. If the noise still draws attention, use strong. If the background no longer gets in the way and you want to preserve as much of the recording’s natural character as possible, choose mild. It is worth processing the same file at two neighboring settings and comparing both results.

When to Enable the 80 Hz High-Pass Filter and What It Does

The “remove hum below 80 Hz” field controls the lowest register of the sound. When it is checked, components below 80 Hz are faded out of the recording, and only then does the actual noise reduction take effect. You hear the effect mainly as the disappearance of rumbling and hum.

Leave this field checked when:

• you hear constant rumbling in the background that makes speech harder to follow,
• the recording was made with air conditioning, a fan, or another device producing a low drone,
• the noise comes from tape or an older recording in which the low end is the most annoying part.

Uncheck it when the recording contains sounds you do not want to lose: bass guitar parts, kick drum, double bass, bass synth, a low male voice, or a wide soundscape. Everything below 80 Hz will be muted, and you will not get that part of the recording back after saving the file.

Noise Profile, Silence in the Recording, and Limits of Automatic Reduction

The tool does not require you to manually point to a section with noise. It determines the background profile on its own, based on the quietest parts of the recording. That means you do not need to know audio editing, but it helps to understand what this mechanism is based on.

That is exactly why a moment of silence in the recording is very valuable. If you have a few seconds without speech or other sounds at the beginning, middle, or end, the tool has something to use to build a background pattern. A recording without such a section can still be processed, but the reference point for the noise will be less reliable, and the result will be harder to predict.

The limitations are just as important as the capabilities. Reduction works well with constant background noise: microphone, fan, air conditioning, tape hiss. It will not remove variable sounds and one-off events—footsteps, a door slamming, conversations in the background, a dog barking, keyboard clicks, or wind gusts. These are not noise but specific events, and the tool does not cut them out point by point.

With the strong setting, the quietest parts of the recording can become even quieter and lose detail, so after processing, listen to the result from start to finish and pay attention to the beginnings and ends of sentences.

File Formats, Local Processing, and What the Tool Does Not Calculate

You select an audio file from your computer or phone as the input. In practice, you can choose material in a format your browser supports, most often MP3, WAV, OGG, or M4A. There is no separate data-upload step here: all processing happens locally on your device, and the result is saved as a WAV file. A long recording simply takes more time, because the processing speed depends on your device’s performance.

What does the tool not calculate? It does not provide any numeric values or statistics—it will not measure loudness in decibels, count the share of noise, or create a chart or report. It also does not separate tracks, remove reverb, cut out individual sounds, or edit the recording. It is not an audio editor, only a tool for one task: reducing noise in a recording.

Check the result yourself. Listen to the file before and after, and judge whether the voice sounds natural, whether necessary low tones have disappeared, and whether any individual sounds remain in the background.

Frequently Asked Questions

This description is based on a code excerpt; processing happens locally in your browser, so verify the settings and the 80 Hz filter by listening on a copy of your recording.