About SoundTidy

Why SoundTidy exists

Most people who have a video with bad sound have no good way to fix it. Audio editors are built for engineers, and many online tools ask you to pay before you have heard anything. SoundTidy was built on one idea: you should hear the result on your own recording before you pay.

How it works

You upload a video or audio file. SoundTidy takes out the audio, cleans it with DeepFilterNet, an open speech-enhancement model that runs on 48 kHz audio, and puts it back next to your original picture. The picture is never re-encoded, so quality does not change. Long files are split into overlapping pieces and cleaned in parallel, which is why an hour of video takes about four minutes.

What it does well, and what it does not

It is best at steady noise behind a voice: fans, hum, hiss, traffic, room noise and wind. It reduces other people talking but does not remove them, it cannot restore speech that was never recorded, and it is not a music or vocal separation tool. The free 30-second preview exists so you can check this against your own file.

Simple pricing

Minutes are sold in packs starting at $4.99 for 60 minutes. There is no subscription. Minutes last a year, and if a job fails they are returned automatically. See the pricing page for details.

Your files

Files are stored privately and deleted automatically after two days. Read the privacy policy for the full detail.

Try it

Go to the home page, drop in a video and listen to the difference.