Is noise reduction free, including the download?
Yes. You can process and download without a payment, account or credits, and there is no added DojoClip watermark. Each recording must be no longer than five minutes and no larger than 100 MiB.
Reduce distracting noise around speech in voice recordings, interviews and videos. Adjust the strength, compare the sound, then download cleaned audio or video. Free local processing, with no upload or account.
Processed on your device · No account · No added watermark
Choose audio or video
Best for spoken recordings with background noise. Music and heavy echo may not improve. Listen before downloading.
The noise-reduction model downloads when you start. Video export also downloads a processing library. Your recording stays on this device.
A fan, air conditioner or constant background hiss can distract from an otherwise useful recording. This free background noise remover applies RNNoise speech noise reduction in your browser. Use it for a spoken note, narration, an interview excerpt, a podcast clip or a short talking-head video.
Your file stays on your device. You do not need to install an audio editor, create an account or send a voice recording to a remote processing service. The browser still needs a connection to download the processing code and model.
You can choose the video directly; there is no need to manually extract and replace its audio. The noise remover processes the speech soundtrack locally and prepares the cleaned sound for download. Compatible video exports retain the original encoded picture instead of applying a visual filter or adding a watermark.
The output container is MP4 or WebM, depending on the source codec. Extra audio and subtitle tracks are not copied. If you need only the sound for a voice edit, use the WAV download. An MP3 input also produces WAV output; it is not automatically saved back to MP3.
For fan noise, hiss or room noise around speech: start here. Try a short, representative recording and compare at a comfortable listening volume. Turning up the finished audio also turns up any noise that remains.
For singing mixed with instruments: use the vocal remover and instrumental maker. It separates a music mix into vocals and accompaniment; speech denoising does a different job.
For a completely silent video: remove the audio track. That mutes everything, including speech. For pauses or unwanted sections, cut the audio rather than expecting noise reduction to shorten the recording.
Noise and speech often share frequencies. Strong suppression can remove parts of a voice along with the unwanted sound, especially on heavily compressed or distorted recordings. Keep the original file and check sentence endings, breaths and soft consonants before deciding which version to use. Severe wind, clipping, heavy echo or several people speaking at once may need a better source recording.
After cleanup, add subtitles to your video or use another of the free browser audio and video tools.
Yes. You can process and download without a payment, account or credits, and there is no added DojoClip watermark. Each recording must be no longer than five minutes and no larger than 100 MiB.
No. Your selected recording is processed on your device. The browser downloads RNNoise processing code and its model when needed; video export can also download processing libraries. These are downloads to your browser, not uploads of your media.
Choose a supported audio or video file with mono or stereo audio, up to five minutes and 100 MiB. Common examples include MP3, WAV, M4A, MP4 and WebM, but actual codec support depends on the browser. A video needs a decodable audio track.
Yes. Supported video export copies the original picture and replaces its primary soundtrack with the cleaned audio. You can also download the cleaned sound as WAV. Additional audio tracks and subtitle tracks are not copied. If video export fails, a completed audio download remains available.
RNNoise targets unwanted sound around speech, such as room hiss or fan noise. Results depend on how much the noise overlaps the voice. Strong wind, clipping, heavy room echo and overlapping speakers may remain; this is not a guaranteed repair for every recording.
Noise reduction can also affect speech details. Lower the reduction strength and process again, then compare Original and Cleaned on the same passage. Listen for consonants and quiet words as well as noise. A little remaining background sound can be preferable to a damaged voice.
This tool is for reducing noise around spoken audio, not splitting a song into vocals and accompaniment. Use DojoClip's vocal remover for that task. It also does not remove silences, mute the whole soundtrack or isolate one person from overlapping speakers.
No desktop app is required. Cleaned audio downloads as a 48 kHz, 16-bit WAV file. Supported video output is MP4 or WebM; the format is shown before download. Keep the browser tab open while processing and save your result before leaving. Speed and format support vary by device.