The real jobs behind merging audio
If you're figuring out how to merge audio files into one, you almost certainly have a recording that exists in pieces. Maybe your phone split a lecture into two halves when a call came in, or an interview stopped and restarted three times because someone knocked on the door. The recording is fine. It just isn't one file yet.
The same problem shows up everywhere. Students end up with lecture halves. Journalists and researchers end up with interview parts. Musicians and writers end up with a folder of voice memo takes, where take three has the good verse and take seven has the good chorus. Podcasters record intros, interviews, and outros separately and need them stitched into one episode. Gym goers want a handful of tracks fused into one continuous workout mix so nothing stops mid-set. In every case the goal is identical: several files in, one continuous file out, with no glitch at the seams.
Why naive concatenation gives you chipmunks
The tempting shortcut is to glue the files together end to end, and plenty of quick tools and command line tricks do roughly that. It works only when every file has the exact same format, sample rate, and channel layout. The moment they differ, things go wrong in a very audible way.
Here is why. Digital audio is a stream of samples, and the sample rate tells the player how fast to read them. Music files are usually 44.1 kHz, audio that came from video is often 48 kHz, and voice memo apps sometimes record lower to save space. If a merged file declares one rate but contains segments recorded at another, the player reads those segments at the wrong speed. A 44.1 kHz segment read at 48 kHz plays fast and high pitched: the chipmunk effect. A 48 kHz segment read at 44.1 kHz plays slow and low, like the recording is melting.
Mismatched channel layouts cause trouble too: a mono voice memo glued against a stereo file can come out garbled or playing at the wrong speed. Proper merging avoids all of this by resampling first. Every input gets converted to one common sample rate and channel layout, and only then are the streams joined.
How to merge audio files into one, step by step
The free Audio Joiner handles the resampling for you. Open the page, add your files, and put them in the order you want. Before joining, it aligns every input to a common sample rate and channel layout, so a 44.1 kHz MP3, a 48 kHz M4A, and a mono voice memo merge cleanly into one file with no pitch or speed artifacts at the seams.
There is exactly one re-encode, into whatever output format you choose. That matters because every lossy encode costs a little quality, and a tool that re-encodes each file separately and then again for the merge throws away more than it needs to. Your files stay on your machine the whole time.
Order matters, so clean up before you join
A merge is only as good as the sequence you feed it. Double check the order before exporting, because file pickers often sort alphabetically, and "Recording 10" sorts before "Recording 2". Renaming with padded numbers (01, 02, 03) keeps things honest, and the order you set in the joiner is the final word. Listen to the first few seconds of each piece if you're not sure which half is which.
It also pays to trim each piece first. Lecture halves usually open with a minute of shuffling and mic checks. Voice memo takes have false starts. The Audio Cutter trims in the browser too. Its lossless mode cuts without re-encoding by snapping to the nearest frame boundary, which lands within a few hundredths of a second of your mark, close enough for trimming dead air. If an edit truly needs sample-accurate edges, its precise mode re-encodes instead. Either way, tidy each segment first, then join the clean pieces.
WAV, FLAC, or MP3: pick the output on purpose
If you plan to edit the merged file further, export WAV or FLAC. Both are lossless, so the single re-encode costs nothing audible, and future edits start from full quality. The tradeoff is size: WAV runs roughly 10 MB per minute at CD quality, so a two hour merged lecture lands well over a gigabyte. FLAC compresses substantially while staying lossless.
If the merged file is a finished product you'll send or upload somewhere, MP3 is the practical choice because everything plays it. At 320 kbps it runs roughly 2.4 MB per minute, about a quarter the size of WAV. Use 256 or 320 kbps and convert once; each lossy generation adds a small loss, so you don't want to re-encode the same audio repeatedly. For a deeper look at when lossless is worth the space, see WAV vs MP3.
If you later need the merged file in a different format, the Audio Converter handles MP3, WAV, M4A, FLAC, OGG, AIFF, and OPUS in the browser as well.
Why merging locally is the right default
Think about what these recordings actually are. Interviews contain other people's voices and unguarded answers. Voice memos hold unfinished songs and half formed ideas. Lectures can include classroom discussion that was never meant to be public. Uploading all of that to a random merge site, waiting in a queue, and trusting a server you know nothing about is a strange default for such personal material.
The Audio Joiner works differently: it runs in the browser, nothing uploaded, we never see the file. The merging happens on your own machine, using FFmpeg compiled to WebAssembly, the same engine many desktop audio tools are built on. When you close the tab, the job is done and gone. There is no server copy to worry about, because no server ever touched the audio.