How much does the audio enhancer cost?

Clean uses 2 credits a minute, and Studio voice 15 credits a minute, plus 2 credits a minute for its word check. Every price is shown on the button before you start, and a run that doesn’t finish gives the credits back. Your sign-up credits clean about 30 minutes. See pricing for credit packs and plans.

Can I try it on my own recording first?

Yes. Run the first minute of your recording, at the price of one minute, and hear the result. If you like it, enhance the rest: the minute you heard is reused and not charged again.

Can the audio enhancer remove echo?

Yes. Studio voice takes room echo away by rebuilding the voice without it, which works even for a stairwell or a classroom. Clean leaves the voice as recorded, so it reduces echo only a little.

Will my voice still sound like me?

With Clean, the voice is the one you recorded, with the noise taken away. Studio voice rebuilds it and keeps how it sounds, and its word check lists any place where a word came out differently, with the original one click away.

How do I remove background noise from a video?

Drop the video into the audio enhancer. Its sound is taken out on your device and cleaned, and you download the same video with the enhanced sound, the picture copied as it is. MP4, MOV, MKV and WEBM all work.

Does it keep background music?

Both modes take music under the voice away along with the noise. To keep a music bed, separate the music first with the vocal remover, clean the voice, and put the music back under it.

How long can a recording be?

Up to 2 hours with Clean, and up to 60 minutes with Studio voice. Long files are cleaned in parts at the same time and joined seamlessly.

Does it work in other languages?

Yes. The voice is cleaned whatever language it speaks, and the word check works in every language the transcription knows. One of our samples is an interview in Spanish.

What loudness should I choose?

Podcast (−16 LUFS) for shows and talks, video (−14 LUFS) for social and video sites, audiobook (−19 LUFS, peaks under −3 dB) for narration, and broadcast (−23 LUFS) for radio and TV. Keep the level leaves it as it was.

Does it clean a live call as it happens?

The audio enhancer works on recordings you already have: once the call or the session is recorded, upload the file and it is cleaned in minutes.

Are my recordings private?

Yes. Your recordings and their enhanced versions are visible only to you and stay in your library for 30 days.

AI Audio Enhancer: Remove Noise and Echo

keep the voice, or rebuild it: 2 ways
per recording, 60 min for studio voice: 2 h
to try before the whole file: First minute
every place the new voice differs: Word check
podcast or video loudness: −16 or −14 LUFS
same picture, new sound: Video kept

How do you remove background noise from audio?

To remove background noise from audio, upload the recording to the audio enhancer and choose Clean: the noise, hum, wind and traffic go, and the voice stays as it was. For echo or other people talking, choose Studio voice. Try the first minute, set how much of the room to keep and the loudness, and download.

How to use the audio enhancer

Three steps, and a minute to try before the whole recording is cleaned.

Upload a recording or a video
Drop in an MP3, WAV, M4A or FLAC file, or a video such as MP4 or MOV. From a video only the sound is sent, taken out on your device. Recordings run up to 2 hours.
Choose what to fix
Clean takes noise, hum, wind, traffic and music beds away and keeps the voice as recorded. Studio voice also takes away room echo and other voices. Run the first minute to hear it on your own recording.
Listen, set and download
Switch between the original and the enhanced sound at the same moment, bring back a little of the room if you like, pick the loudness and download an MP3, a WAV or the video.

What this AI audio enhancer does for you

More than an audio cleaner: a voice you can understand, words you can trust, and a file that is ready to publish.

A smooth sand dune with a sharp ridge at dawn

Two ways to clean: keep the voice, or rebuild it

Clean subtracts the noise and leaves every breath and syllable of the voice where it was. Studio voice rebuilds the voice without the room, which is what takes a heavy echo or a café’s chatter away. Samples below play both on the same recording, so you can hear the difference before you choose.

An optional word check for studio voice

A rebuilt voice can now and then say a word differently. With the word check on (+2 credits a minute), the recording is transcribed before and after, and every place where the words differ is listed with its time. Play it, and keep the original in that place with one click.

Keep as much of the room as you like

A voice with no room at all can sound flat. The background slider mixes a share of the original back under the cleaned voice, from none to all of it, and you hear it as you move it. Changing it costs nothing, however often you try.

Loudness set for where it is going

Pick podcast (−16 LUFS), video (−14 LUFS), audiobook (−19 LUFS with peaks under −3 dB) or broadcast (−23 LUFS). The file is measured and limited so it lands on the target without clipping, as streaming apps and video sites expect.

Remove background noise from a video, keep the picture

Drop in an MP4, MOV, MKV or WEBM file. Only its sound is cleaned, and the enhanced sound goes back into the same video with the picture copied as it is, ready to post.

Try the first minute before the whole file

Run the first minute of your own recording and hear the result. If you like it, enhance the rest: the minute you heard is reused, so it is never charged twice.

Audio enhancer samples: before and after

Each sample is one of our own recordings, made noisy on purpose, then cleaned by this audio enhancer exactly as it came out. Switch between the original and the result while it plays.

Audio enhancer guide: cleaner recordings in minutes

Remove background noise from audio: which mode to choose

Clean is the right start for most recordings: a fan, an air conditioner, a hum from the mains, wind on a phone, traffic outside or a music bed under the voice. It takes the noise away and keeps the voice exactly as it was recorded, at 2 credits a minute.

Studio voice is the voice enhancer for problems a cleaner alone leaves: a hard room that rings, a café full of other conversations, or a voice far from the microphone. It rebuilds the voice without the room, at 15 credits a minute, and its word check lists anything that came out differently.

Not sure? Run the first minute in each mode and compare them at the same moment.

Remove echo from audio recorded in a big room

Echo is the room answering the voice: bare walls, a stairwell, a classroom, a kitchen. It sits on top of every word, so the voice sounds far away even when nothing else is wrong.

Studio voice takes the echo away by rebuilding the voice without it. For a light echo, Clean with a little background kept may be enough. Next time, a smaller room with curtains, a rug or a bookshelf, and the microphone closer to the mouth, makes the biggest difference.

Voice enhancer settings for podcasts, interviews and lectures

Podcasts and interviews: pick Clean for a home studio with some hiss or hum, Studio voice for a guest who called in from a noisy place. Set the loudness to podcast (−16 LUFS) for every episode, so they all play at the same level.

Lectures and classes: big rooms echo, so Studio voice usually wins. Keep a little background if the room’s sound helps listeners feel they are there. Calls and meetings: Clean takes the line hiss, the hum and the room noise away.

1Choose the recording type, which sets the loudness.
2Run the first minute in the mode that fits the problem.
3Compare the original and the enhanced sound at the same moment.
4Enhance the rest, set the background and download.

Clean up the audio of a video

Most noisy recordings today are videos: a vlog in the wind, a talk filmed from the back of a room, an interview by a busy road. The audio enhancer takes the sound out of the video on your device, cleans it, and puts the enhanced sound back into the same file.

The picture is copied as it is, frame for frame, so nothing about the video changes but its sound. Pick video loudness (−14 LUFS) for social and video sites.

How loud should a podcast or a video be?

Streaming apps and video sites turn loud files down and leave quiet ones quiet, so a file mastered to their level sounds its best. Podcasts aim at −16 LUFS, videos at −14 LUFS, audiobooks at about −19 LUFS with peaks under −3 dB, and broadcast at −23 LUFS.

The audio enhancer measures the whole file and uses a true-peak limiter, so the loudness lands on the target and no peak clips, in MP3, WAV and video alike.

Check a cleaned recording before you publish

Listen to the busiest part of the recording, not only the start: where two people talk at once, or a loud noise sits on a word. Switch to the original at the same moment to hear what changed.

With Studio voice, read the word check first. Each listed place plays from just before it, and one click keeps the original there. Then download, and keep the original file next to the new one.