Audio editor

Trim, fade, normalize and convert sound, make ringtones and take it out of videos, many files at once, on your device. Free, with no account or uploads.

Drop sound or video files

Trim, fade and change the volume; convert them, or take the sound out of a video. Many at once, on your device. You can also paste with ⌘V or Ctrl+V.

Choose files

MP3
M4A
Opus
FLAC
WAV
AIFF
MP4
MOV
WebM

Stays on your device · up to 50 files at once

What you can do

Trim, fade and change the volumeTo the millisecond, on a waveform you can zoom; every edit can be undone, and your file is never changed.
Every common formatMP3, M4A (AAC), Opus, FLAC, WAV and iPhone ringtones, from those and from AIFF, Ogg and the sound of videos.
Loudness to a standardMeasured as ITU-R BS.1770 does, normalized to −14 LUFS or another target, with a true-peak limiter so nothing clips.
Sound out of videoKept exactly as it is, without being encoded again, whenever the format allows; or converted to the format you choose.
Checked before you downloadEvery file written is opened again and listened to: its length, rate and channels are held against what you asked.
Many files at onceTags and cover art kept, gapless MP3s, a size limit if you need one, and everything in one ZIP.
Nothing is uploadedYour files are read and written in this browser. No account, no limits.

A little help along the way

Follow the steps below. Every option is explained underneath.

Add your filesDrop them, choose them or paste them: MP3, M4A, Opus, Ogg, FLAC, WAV, AIFF, or videos. Up to 50 at once.
Edit the soundChoose a part on the waveform, then keep it, cut it out, silence it, fade it or change its volume. Each file keeps its own edits.
Choose the formatMP3, M4A, Opus, FLAC, WAV or an iPhone ringtone, a quality, and if you like a sample rate, mono or stereo, and a file size limit.
Listen, check and downloadEach file is written as soon as it is read or edited, and checked. Download one, or everything as one ZIP.

Every option explained

Choosing a part: Drag across the waveform, drag a selection's edge to move it, or type its start and end (as 7.25, 1:07.25 or 1:02:03). On the waveform, I and O start and end the selection at the playhead. Zoom in to choose to the millisecond: the buttons, Ctrl or ⌘ with the wheel, or a trackpad's pinch.
Keep only this, Cut out: Keep only this keeps the selection and drops the rest; Cut out takes the selection away and closes the gap. Wherever the sound was cut, it fades out and in over 4 milliseconds, too short to hear as a fade and enough to take away the click a cut can make.
Fades: Choose the part to fade over and press Fade in or Fade out. Even fades at a steady rate; Smooth keeps more of the sound until late; S-curve starts and ends softly.
Silence and volume: Silence makes the selection silent and keeps its length. Volume makes the selection, or all of it when nothing is chosen, louder or quieter by up to 24 dB. Louder can clip: listen before you download. Measuring and matching loudness comes in a later version.
Edits and undo: Your file is never changed. Each edit is a step in the list; the result is written from your file and the steps, so you can undo, redo or start again at any time. Each file in the list keeps its own edits; the format settings are shared.
Playing: Play and Play the selection play the sound as edited: cut parts are skipped and fades heard. This is a preview played from your file; the file written is exact to the sample and is checked. Loop repeats what is playing.
Encoded again: An edited sound is always encoded again: MP3 and AAC hold sound in blocks of about a fortieth of a second, so copying them could not cut where you chose. FLAC and WAV keep every sample.
Keys: listening: Space or K: Play or pause. P: Play the selection. ←: Move the playhead back 0.1 s. →: Move the playhead on 0.1 s. Shift + ←: Move the playhead back 1 s. Shift + →: Move the playhead on 1 s. J: Back 5 s. L: On 5 s. HOME: Go to the start. END: Go to the end.
Keys: choosing a part: I: Start the selection at the playhead. O: End the selection at the playhead. Ctrl or ⌘ + A: Select all of it. Esc: Clear the selection. M: Add a marker at the playhead.
Keys: editing: DELETE or BACKSPACE: Cut the selection out. Ctrl or ⌘ + Z: Undo. Ctrl or ⌘ + Shift + Z or Ctrl or ⌘ + Y: Redo.
Keys: zoom: + or =: Zoom in. -: Zoom out. 0: Show all of it.
Silence at the ends: Finds the silence at the start and the end of the sound as edited (under the level chosen, -60 dBFS, -50 dBFS, -40 dBFS) and keeps the sound between, leaving 0.1 s of each. It is one step, undone like any other.
Long pauses: Shortens every pause inside the sound longer than the length chosen (0.5 s, 1 s, 2 s) to 0.4 s, all in one step. Each cut is faded, so nothing clicks.
Ways to split: Every few minutes (the last part is what is left), into a number of equal parts (2 to 500), at markers you place, or at pauses longer than 0.5 s, 1 s, 2 s (each part ends at the middle of a pause, or the pauses can be left out). The parts are listed with their times before anything is written.
Markers: Put the playhead where a part should start and press Add a marker, or M on the waveform. Markers stay where they are in the file when you edit it; one in a part you cut out is not used. Remove one with its ×.
The parts: Each part is the file as you edited it, cut to its part to the sample, in the format and with the settings chosen, and checked like any result. They are named after the file with a number (talk-01, talk-02 …), and can be downloaded one by one or as one ZIP. The file itself is never changed.
The order: Files join in the order of the list; move each up or down with its arrows. Each joins as you edited it, so trim, fade or cut each one first if you like.
Between files: Straight on, with 0.5, 1, 2 s of silence, or crossfaded over 0.5, 1, 2, 4 s: the end of one fades out as the next fades in, at equal power so the loudness does not dip. A crossfade is never longer than half of either file.
Matching files: Files of different rates and channels are made the same (the first file's rate unless you choose one, and stereo when any is stereo). Choose Normalize loudness and each file is measured and brought to the target by itself, so they come out equally loud; one limiter keeps the peaks of the whole under the ceiling.
Chapters: An MP3 or M4A gets a chapter for each file, named after it: ID3 chapters in an MP3, which most players and podcast apps read; a Nero chapter list in an M4A, which VLC, FFmpeg and foobar2000 read. Apple's own apps read chapters of another kind, which is not written.
Speed, keeping the pitch: From 0.5 to 2 times as fast. The sound is rebuilt from short overlapping pieces (WSOLA: 40 ms, each placed where its waveform best matches the one before), so the notes stay where they are. Speech and most music sound natural up to about 1.5 times; much faster or slower, a little flutter can be heard.
Pitch, keeping the length: Up or down by up to 12 semitones (an octave): the sound is stretched by the pitch's ratio and then played faster or slower by it. Voices move with the music, so a large change sounds less natural.
Tape speed: With Keep the pitch off, faster is higher and slower is lower, as a tape or a record plays: the sound is resampled, nothing else is done to it, and nothing above the new top frequency folds back.
Channels: Swap left and right, put one side in both ears, or set the balance. Karaoke takes away what is in the middle of a stereo mix, where the voice often is, keeping the bass under 150 Hz: it is not a real vocal remover, works on some songs only and takes away anything else in the middle too. A mono result has no channels to change, and the page says so.
Backwards: The whole sound, from its end to its start. It is kept while it is written (as blobs the browser may hold on disk), so very long files need room. A joined file is reversed as a whole and gets no chapters.
Background noise: Taken away by RNNoise, a small neural network trained on speech (from Xiph.org and Mozilla), run on your device in your browser: hiss, fans, traffic and room noise between and under words. It is made for voices; on music it can take away notes and breaths you want, so listen, and lower the strength if it does. The strength mixes the cleaned sound with the sound as it was: 100% is RNNoise alone. The sound is taken to 48 kHz for it and back. The first time, its file (3.1 MB to download, 4.8 MB in all) comes from this site and is then kept for working offline. Your recording and its noise are never sent anywhere.
Mains hum: The steady buzz of the mains, at 50 Hz in most of the world and 60 Hz in the Americas and parts of Asia, and its harmonics up to 1 kHz: each is notched out 3 Hz wide, so the voice or the music beside it stays as it was.
Clicks: Short spikes, at most 2 ms long and far louder than the sound just around them, such as a crackle on a record or a bad cable, are drawn over with a smooth curve from the sound on either side. Drum hits and consonants last longer and are left alone. The page says how many clicks were drawn over.
Harsh "s" sounds: A de-esser: only the sound over 5 kHz is turned down, and only while it is loud, as in a sharp s, sh or t. Gentle turns it down by up to 6 dB, strong by up to 12 dB; the page says how far it went.
Tags: Title, artist, album, album artist, year, track and disc (each with how many there are in all), genre, comment and lyrics. A field left empty is taken out of the file; a field you do not touch is kept exactly as the file has it, with every other tag it holds (chapters, an encoder's notes, ReplayGain and the rest).
Many files at once: Choose All files and what you type goes into every file, such as one album and artist for a whole record; fields that differ between files say so and stay as each file has them until you type in them.
Cover: A JPEG or PNG picture, smaller than 16 MB, put in as the front cover. Taking it away takes every picture the file holds.
Without touching the sound: When the result is the same kind of file (Same as the file, with nothing else changed), only the part of the file that holds the tags is written: ID3 in an MP3, the item list in an M4A, the comment and picture blocks in a FLAC, the comment header in an Ogg or Opus file, and INFO with an id3 chunk in a WAV. The sound is byte for byte as it was, and the page checks it by listening to the result. Files of other kinds, or written in another format, get the tags as they are written.
Recording: Press Record and allow the microphone when the browser asks. The sound is taken as the browser hands it over, 32-bit floats at the device's own rate, so nothing is lost; it goes into the list of files as a WAV when you press Stop, to be trimmed, cleaned up, made louder and written in any format. Long takes are kept a second at a time and never held whole.
Pause and go on: Pause stops taking sound and Go on takes it again, in the same file. The sound fades out and in over 5 ms at each pause, so the join does not click.
Level: The meter shows how loud the microphone is. When the sound reaches full scale the page says so: move back, or turn the input down, since sound past full scale is cut off for good.
Microphone and channels: Choose the microphone once the browser has allowed one, and mono (for a voice) or stereo. Letting the browser clean up the voice turns on its echo cancelling, noise suppression and automatic level, which are made for calls; off, the sound is exactly what the microphone gives.
Keep a draft while recording: Off at first, every time. On, the take is kept on this device as it is recorded, so a closed tab or a crash does not lose it (up to the last second before it stopped); the next visit offers to add it to the files or delete it, and finishing a take deletes it. It is never sent anywhere.
Whisper: OpenAI's speech recognition model, run in your browser with Transformers.js and ONNX Runtime. The first time, the model and its engine are downloaded from this site (Base about 107 MB, Small about 280 MB) and then kept for working offline; the sound never leaves your device. Base takes about a third of the sound's length on a laptop and is good for clear speech; Small is slower and better with accents, noise and other languages.
Language: Whisper finds the language by itself, or you can say which it is, which helps with short or mixed speech. Translate into English writes what is said in English instead: a quick translation, not a careful one.
The transcript: Written in parts with their times, as Whisper hears them, 30 seconds at a time; each next window starts where the last whole sentence ended, so nothing is cut in two. Press a time to listen from there, find words in it, copy it, or save it as text, text with times, SRT or WebVTT subtitles. Whisper can mishear names, quiet words and speech over noise, and on silence or music it can write words that were not said: read it through before you rely on it.
Video: The sound as written with your settings (trimmed, cleaned, louder: whatever you chose), with a picture, as a video for places that take only video. MP4 with H.264 and AAC where the browser can write H.264, otherwise WebM with VP9 and Opus, and the page says which. Made on your device, opened again and checked before you can download it.
Shape and picture: Square (1080 × 1080), landscape (1920 × 1080) or portrait (1080 × 1920). The picture is the file's cover, one of your own (JPEG, PNG or WebP) or none, filling its place and cut evenly at the sides that do not fit.
Waveform and title: The sound's waveform with a playhead that moves as it plays (24 frames a second), and the title and artist from its tags. Without the waveform the picture stands still at one frame a second, so an hour takes moments.
Normalize loudness: Each file is measured as it will be written, after its edits, at its new rate and channels, as ITU-R BS.1770 measures loudness (K-weighted, gated), and made louder or quieter to reach the target: −14 LUFS for Spotify, streaming; −16 LUFS for Apple Podcasts; −23 LUFS for Broadcast (EBU R 128); −24 LUFS for US TV (ATSC A/85), or one of your own. Files normalized together come out equally loud. A file is made at most 30 dB louder.
The limiter and the ceiling: When a file is made louder, its loudest moments could go over full scale and clip. The limiter looks 5 ms ahead and lowers the gain smoothly just before them, letting go over 100 ms, so peaks stay under the ceiling, between samples too (true peak). −1 dBTP leaves room for MP3 and AAC encoders, which raise peaks a little. If the ceiling keeps a file from its target, the page says how loud it reached.
Normalize peaks: Makes each file louder or quieter so its highest true peak reaches the level chosen, with no limiter. Files keep their own loudness, so they will not sound equally loud.
Make louder: Makes each file louder by up to 30 dB, with the limiter. 6 dB doubles the height of the waveform; 10 dB sounds about twice as loud.
Loudness, range and peaks: The editor shows each file's loudness in LUFS, its loudness range in LU (how much it varies, as EBU Tech 3342 measures it) and its highest true peak in dBTP, for the whole file and for the part chosen, with its short-term loudness drawn under the waveform. The check measures the file written again.
Equaliser: Raise or lower the bass (a shelf under 120 Hz), the mid (around 1 kHz) and the treble (a shelf over 8 kHz) by up to 12 dB, and cut rumble under 80 or 120 Hz. Presets: Flat (The sound's own balance, unchanged); Voice clarity (Rumble under 80 Hz cut, a little less boom, a little more presence and air); More bass (6 dB more under 120 Hz); Cut rumble (Everything under 80 Hz cut: wind, handling noise, traffic); Telephone (Only 300 Hz to 3.4 kHz, as a phone line carries). These presets are KeepSimp's own choices.
Compressor: Turns the loudest moments down so loud and quiet parts sit closer together. Gentle: Evens out the loudest moments a little: 2:1 above −18 dBFS. Voice: Keeps speech at one level, near and far from the microphone: 3:1 above −24 dBFS, quick to act. Music: Holds a mix together, letting drums through: 2.5:1 above −16 dBFS, slower to act. A compressed sound is quieter, so choose Normalize loudness to bring it back up. When the equaliser or compressor is on and the loudness is left as it is, a limiter keeps peaks under −1 dBTP.
Same as the file: Keeps the file's own sound as it is, without decoding and encoding it again: AAC goes into an M4A, MP3 stays MP3, Opus and Vorbis go into Ogg, FLAC stays FLAC and plain samples become a WAV. Nothing is lost and it takes a moment. When the rate, the channels or a size limit you chose mean the sound must change, it is encoded again in its own kind; sound that cannot be kept (such as AC-3 from a film) becomes AAC, and the page says so.
MP3: Plays everywhere. Written with LAME at a constant bitrate, with the header that tells players the encoder's delay and padding, so the file is exactly as long as the sound. Below 32 kHz an MP3 holds at most 160 kbit/s.
M4A (AAC) and M4R: Smaller than MP3 for the same quality, and played by every phone and computer. Written with your browser's own AAC encoder, or one that comes with the page when it has none. The encoder's priming is measured and an edit list written, so the sound starts and ends where it should. M4R is the same for an iPhone ringtone, which can be at most 40 seconds long (an alert tone 30).
Opus: The best quality for its size, voice above all. It is always written at 48 kHz. Browsers, Android, VLC and recent Apple devices play it.
FLAC and WAV: Lossless: every sample is kept. FLAC is about half the size of WAV. Choose 16-bit (CD quality), 24-bit, or for WAV 32-bit floating point; FLAC holds at most 24 bits.
Quality: Small, Balanced or High set the bitrate for each format (MP3 96, 192 or 320 kbit/s; AAC 96, 160 or 256; Opus 48, 96 or 160), less for mono. Custom lets you pick the bitrate. These numbers are KeepSimp's own choices.
Sample rate and channels: Keep the file's own, or choose one. When a format cannot hold the rate asked, the nearest it can is used and the page says so. Lossy formats here are mono or stereo, so surround sound is mixed down to stereo for them.
Tags and cover art: Title, artist, album, year, genre and the cover picture go into the new file where its format holds them. Switch it off to write none.
File size limit: MP3, M4A and Opus are written at the highest bitrate that keeps each file under the limit, with room for the container; the page says what it chose. Lossless formats cannot be made to fit a size.
The check: Every file written is opened again and decoded from start to end. Its length, sample rate and channels are held against what was asked, an MP3's size against its bitrate, and a file that came out silent is caught. If something is not as asked, the page says what.
Files the browser cannot read: Some browsers' decoders lack a kind of sound (Safari's lacks FLAC in some places). The page then decodes it with the browser's own player, which plays more kinds, and goes on. AIFF is read by KeepSimp itself. A file nothing here can decode says so.
Privacy: Files are read and written in this browser and never uploaded. There is no account and no limit on how many you convert.

Read the step-by-step guide