Audio Editor

Free Online Audio Editor with Real Effects

This free online audio editor runs entirely inside your browser. Drop in an MP3, WAV, M4A, FLAC, OGG or AAC file and you can cut it, fade it in and out, change volume, normalize loudness, shift pitch, change speed, add reverb or echo, boost bass and treble, run a soft compressor, reverse it or trim silence — then export to WAV or MP3 without uploading anything to a server.

Everything happens with the Web Audio API. Your audio stays on your device, no account is needed, and there are no daily limits, file-size caps or watermarks. Open the page, drop a file, edit it, save it. That is the whole workflow.

What You Can Do With This Tool

• Cut and trim. Click and drag across the waveform to set a selection, then hit Play selection. Use Set start / Set end buttons to fine-tune the in and out points with sub-frame accuracy.
• Fade in and fade out. Linear fades with adjustable length up to ten seconds. Useful for podcast intros, music beds and avoiding abrupt cuts.
• Volume and normalize. A simple gain slider covers routine level changes. The normalize switch finds the loudest sample in your file and rescales everything so the peak hits a target dBFS, which is the right way to make a quiet recording audible without clipping.
• Bass and treble EQ. Two cascaded biquad shelves at 150 Hz and 4 kHz. Subtle values clean up voice recordings; aggressive values reshape music.
• Compressor. Threshold-driven dynamics control at a 4:1 ratio. Pulls down loud peaks and lifts quiet passages, which is what podcasters use to even out voice levels.
• Reverb and echo. Convolver-based reverb with four room sizes and a feedback delay line for echo. Stack them for ambient pads.
• Speed and pitch. Change tempo independently of pitch, or shift pitch while keeping the same duration. Useful for music practice, language learning and fixing slightly off-key recordings.
• Reverse and trim silence. Reverse playback for creative effects, plus an automatic silence trimmer that removes quiet gaps above 100 ms.
• Export to WAV or MP3. WAV renders in milliseconds. MP3 uses FFmpeg.wasm loaded on demand, with the standard LAME encoder and adjustable bitrate.

Why AI Chatbots Cannot Replace This

Audio editing is a real-time DSP problem. To apply a fade, the tool has to walk through millions of samples and multiply each by a ramp. To normalize, it has to scan the whole buffer for the peak, then divide every sample by a derived gain. A convolution reverb multiplies the input by an impulse response across multiple channels in real time. None of this can be done with text prediction.

AI assistants can describe what a fade is, suggest EQ settings or recommend a target loudness in LUFS. They can even write down a list of operations you should perform. But they cannot take your private audio file, render the changes and hand you back a finished WAV without seeing the bytes. The file is too large to paste into a chat window, and most users rightly refuse to upload private recordings to a third-party AI service. This editor solves that gap by doing the math locally.

Who This Tool Is For

• Podcasters trimming interviews, normalising hosts and guests, adding a short fade between segments.
• Musicians slowing down a riff for practice, shifting the key of a backing track, experimenting with reverb settings.
• Video editors who need a quick EQ pass, room-tone removal or de-essing alternative without launching a full DAW.
• Students and language learners slowing dictation tracks or speeding up lectures while keeping voices intelligible.
• Anyone with a recording who wants to clean it up, send it, and move on.

Tips for Best Results

• Normalize once, then compress. Normalizing first sets a known peak. Compressing after keeps the dynamics even.
• Use mild EQ. A bass cut of -3 to -6 dB and a treble boost of +2 to +4 dB will clean up most phone recordings without sounding processed.
• Reverb before echo, not after. Reverb on a dry signal then a subtle echo gives a natural room. Echo into reverb gives a metallic sound that rarely works.
• Speed and pitch interact. Speed change at 50% also drops the pitch by an octave. Use pitch shift if you only want to change one.
• MP3 is for sharing, WAV is for editing. Export WAV whenever you plan to do another pass. MP3 is lossy and accumulates artefacts each round trip.

How It Works Under the Hood

When you drop a file, the browser decodes it into a raw PCM buffer using AudioContext.decodeAudioData. From that point on, every effect is a deterministic transform on a Float32Array. Volume is a sample-by-sample multiply. Normalize scans for the peak once, divides by the resulting scale, and applies it. Reverb uses an ConvolverNode with a synthetic impulse response. Pitch shift uses a granular phase vocoder that overlaps short windows of the audio and resynthesises them at a new frequency while keeping the total length.

Export to WAV is a direct write of the PCM samples into a 16-bit RIFF container. Export to MP3 writes the same buffer into FFmpeg's virtual filesystem, then runs the LAME encoder with a quality preset, and finally reads the resulting MP3 bytes back out as a download. Nothing in the chain touches a server.

Privacy

Your audio file is read by your browser, processed by your browser, and saved to your disk by your browser. There is no upload step. There is no analytics on the audio content. The page only loads the Web Audio API (built in), the FFmpeg.wasm core (loaded only when you export MP3) and the static stylesheet. Closing the tab clears everything.

Frequently Asked Questions