Noise Remover01 · STEP
Choose a recording
Select a WAV or MP3 from your device. The tool checks the file format, size and duration before loading the audio engine.
Noise Remover · local audio cleanup
Background Noise Remover
Clean up audio in your browser, listen before you save, and keep the original recording under your control.
WAV or MP3 input30 sec processed previewLocal processing
Local audio workspaceStep 01 / 03
Choose an audio file
WAV or MP3 · 25 MiB maximum · 5 minutes maximum · first 30 seconds become a WAV preview
No file selected
Choose a supported WAV or MP3 file to begin.
Gentle cleanup preserves more of the original sound, so some background noise may remain. Echo, reverb, speaker separation and clipping repair are not promised.
Clean Up Audio without losing the comparison.
A practical first pass for recordings where the background is steady enough to test.
Background noise can make a recording tiring to follow, but removing every quiet sound is not always the right goal. A voice has room tone, breaths and small changes in level. Noise Remover gives you a short, local preview so you can hear what changes before you decide what to keep.
The page works around a simple question: does the processed section make the speaker easier to follow while leaving the voice recognizable? You can switch between the original and processed audio, listen more than once, and download the WAV preview when the result is useful. The source file stays with you. Upload a representative clip or a copy trimmed in another editor; this tool processes only the first 30 seconds of the file you choose. Keep the unedited recording and compare speech, pauses and background sound at the same playback volume.
How to use the background noise remover.
The controls stay close to the file so the result is easy to inspect.
02 · STEP
Preview the change
Start a preview and wait for the local processing step to finish. The first 30 seconds become a real WAV result for a focused comparison.
03 · STEP
Compare, then save
Play Original and Processed one at a time. Keep the source file, decide whether the tradeoff works, and download the preview only when you are ready.
Start with noise that stays fairly steady.
Different recording problems need different tools. A quick preview can tell you which problem you are hearing.
Steady fan or AC
A continuous, fairly even bed of fan, air conditioner or computer noise is the clearest fit for this first pass.
Low room hum
A stable electrical or room tone may become less distracting when it stays separate from the voice.
Light background wash
Distant room activity or a quiet office bed can be worth testing when the speaker remains the main signal.
What it cannot repair
Clicks, clipped samples, heavy echo, reverb, overlapping speakers, music and sudden traffic are different problems.
The underlying RNNoise approach is described by its original research and demonstration materials at jmvalin.ca/demo/rnnoise. That link explains the technology; it is not a product endorsement or a quality guarantee.
A small tool for real recording decisions.
Use the preview as an editing checkpoint, not as a promise that every track can be repaired.
USE CASE
Podcast rough cuts
Check a short section before committing to a longer edit. Compare the processed preview with the original so a quieter background does not hide consonants or room detail.
USE CASE
Lessons and courses
A clean first pass can help you decide whether a lecture recording is usable. Test a section with the same microphone position and room tone as the full lesson.
USE CASE
Voice notes and interviews
Phone recordings often contain a steady bed from a room or device. Use the preview to judge whether the speaker stays natural before you save anything.
USE CASE
Screen recordings
If your voice sits over a stable computer fan, this tool gives you a quick local comparison. It does not promise to remove keyboard hits, notification sounds or music.
A/B listening
Keep what helps the speaker.
Listen to the same moment in Original and Processed. Start with words at the beginning and end of sentences, then check breaths, sibilants and quiet pauses. If the voice feels thinner, metallic or strangely close, the quieter background may not be worth the tradeoff.
Use headphones when possible, but also check the result on the device where people will hear it. A preview that sounds pleasant on headphones can still lose clarity on a small phone speaker. There is no automatic “clean enough” score here; your comparison is the useful test.
For a fair check, use the same playback volume and the same short passage. Listen once for the speaker, once for the background, and once for artifacts between words. If the result helps in one sentence but harms another, keep the original and try a different recording or edit instead of forcing the whole track through one setting.
Formats and limits stay visible.
Knowing the boundary helps you choose the right source before a long wait.
WAV16-bit PCM, mono or stereo, 44.1 or 48 kHz
MP3Decoded locally before the preview is made
25 MiB / 5 minMaximum audio input size and duration
48 kHz WAVProcessed output with up to 30 seconds
Video routeMP4 or MOV, H.264/AAC, 50 MiB, 720p, 30 fps, 2 min
Phone useTouch-ready layout; device memory and browser support still matter
Clearing the tool removes its working copies. Downloads remain yours and are not automatically deleted. On the production site, Google Analytics measures page visits and Microsoft Clarity records a masked heatmap/session view; selecting a file stops Clarity recording for this page. These services do not receive the selected media files.
Questions before the first take.
Answers describe the current local capabilities.
What files can I use?
Audio input is WAV or MP3. WAV must be 16-bit PCM, mono or stereo, at 44.1 or 48 kHz. MP3 is decoded before processing. Each audio file is limited to 25 MiB and 5 minutes.
What do I download?
The result is a processed WAV containing up to the first 30 seconds. There is no full-length export or paid unlock in the current release.
Does it remove echo?
No promise is made for echo or reverb. Echo changes the relationship between the voice and the room, while this tool is focused on steady background noise.
Are my files uploaded?
The processing work happens on your device. The selected audio or video is not uploaded by this tool. The hosting layer may receive ordinary request and performance data through Cloudflare hosting.
Can I use it on a phone?
The layout and file controls are designed for touch screens, but the available memory and browser codec support still matter. A real device test is the best check for a long or complex recording.
Is video supported?
Video has a separate workflow for supported MP4 or MOV files with H.264 video and AAC audio, up to 50 MiB, 720p, 30 fps and 2 minutes. It creates a 30-second MP4 preview.