Audio Merger & Volume Normalizer

About the audio merger

Joining audio files sounds like it should be simple, and the joining part is. What goes wrong is everything around it: one track is half the volume of the next, the files were recorded at different sample rates so the later ones play at the wrong speed, and there is an audible click at every join.

This tool handles those. Add your files, put them in order, and get one file back — with the loudness evened out, the sample rates properly reconciled, and optional fades so the joins are clean.

Everything is decoded and mixed in your browser. No audio is uploaded, which for interviews, recordings and unreleased music is the whole point.

How to use it

• Add audio files — pick several at once, or add more later
• Put them in order with the arrows, and remove anything you did not want
• Choose how to handle volume — match loudness, normalise peaks, or leave it alone
• Optionally set a gap between tracks and a fade at each join
• Press Merge into one file and download the result

Each track shows its length and its measured level in dB, plus the adjustment about to be applied. You can see what the tool is going to do before it does it.

The three volume options

This is the setting that decides whether the result is comfortable to listen to.

• Match loudness between tracks — the default, and usually what you want. Each track is measured by RMS, which is the average energy of the signal and tracks how loud something actually sounds. Every track is then brought to the level of the loudest one, so quiet tracks come up rather than everything being pulled down
• Normalise each to a peak level — brings each track's single loudest sample to the target you set. Predictable, but it does not fix perceived loudness: one stray click can sit at full scale while the track is quiet throughout, and peak normalising will then leave it quiet
• Leave the levels alone — a straight join with no gain applied. Right when the files were already mastered together

Whatever the setting, the finished file is checked once at the end and scaled down if anything went over full scale, so the result cannot clip.

Sample rates and channels

Audio files carry a sample rate — 44,100 Hz for CD-quality, 48,000 for video, sometimes 22,050 for voice recordings. You cannot simply string together files at different rates: the later ones would play at the wrong speed and pitch.

Everything here is rendered through a single offline audio context at the highest rate among your files, which resamples properly rather than reinterpreting the samples. Mixing mono and stereo works the same way: if any file is stereo the result is stereo, and mono files are placed in both channels.

Gaps and fades

• Gap inserts silence between tracks. Useful for audiobook chapters or separating takes; leave it at 0 for continuous music
• Fade applies a short fade in and out at each track boundary. Even a tenth of a second removes the click you get when one waveform ends mid-cycle and the next starts somewhere else

What people use it for

• Joining interview or podcast segments recorded in separate takes
• Combining audiobook chapters into one file for a player that has no playlist
• Stringing together voice memos into a single recording
• Making a continuous mix from several tracks
• Evening out the volume across recordings made on different devices
• Merging lecture recordings that were split by a recorder's file size limit

Good to know

• Works on Windows, macOS, Linux and ChromeOS in any modern browser, and unlike the folder tools it works on phones and tablets too, since it takes files rather than a folder
• Nothing is uploaded. Decoding, merging and encoding all happen in the page
• The output is a WAV — uncompressed, so it is lossless from the point of merging but large. Roughly 10 MB per minute in stereo. If you need an MP3, run the result through the Audio Converter
• Anything your browser can play can be merged: MP3, M4A/AAC, WAV, FLAC, OGG and Opus. A file it cannot decode is reported by name rather than silently skipped
• Very long merges use memory in proportion to the total length, since the whole result is assembled before it is written. An hour of stereo audio is around 600 MB in memory
• Levels are shown in dBFS, where 0 is the maximum. A typical track measures somewhere between −20 and −12 dB RMS

Related tools

• Audio Converter — turn the merged WAV into MP3 or M4A
• Audio Trim — cut the pieces before joining them
• Audio to Text — transcribe the finished recording
• Video to Audio — pull the sound out of a video first