Audio Joiner (Merge MP3 and WAV Files)

Gluing recordings together

Sometimes the pieces you have are not the piece you need. A podcast comes as an intro, an interview and an outro. A practice session was captured in four separate takes. A birthday playlist should run as one unbroken track on a USB stick. The joiner takes such parts and writes them as a single audio file, in the order you choose.

Up to 30 files fit in one job. Each one shows up in a list where you can move it up or down, take it out, or add further files at the end. Anything your browser can decode is accepted, from MP3 and WAV to M4A, Opus and FLAC, and formats can be mixed freely in the same list. The merged result is saved as MP3 or WAV.

This page works only with files that are already on your device. It does not collect audio from streaming services or from web addresses. Combine your own recordings, or tracks you are permitted to use.

Order, preview and the first file

Build the list in the order you want to hear it. The arrows beside each file shift it one place at a time, which is quick for a handful of tracks and still manageable for thirty. If you added a file by mistake, the remove button takes it off the list without touching the original.

The first file is special. Its sample rate and its stereo layout set the standard for the whole result, as the section on mixed sample rates explains below. So start the list with the file whose quality you care most about, or the one recorded at the higher sample rate.

Use the space bar to play the combined result. Check the joins before saving, particularly the one between two tracks that were recorded at different loudness. A quiet track following a loud one can feel like a hole in the audio even when the transition itself is clean.

Silence or crossfade between tracks

Two knobs control the transition, and they are alternatives. Silence between files runs from 0 to 5 seconds and inserts a gap of nothing at every join. A gap of one to two seconds suits recordings that are separate items, such as voice memos or lecture sections, because it tells the listener that something new is starting. Zero gives a direct butt joint, which is right for pieces that were cut from one continuous recording.

Crossfade runs from 0 to 8 seconds. Instead of a gap, the end of one file fades out while the start of the next fades in, and the two overlap. The overlap uses an equal-power curve, which keeps the combined loudness steady through the middle of the transition rather than dipping. As soon as the crossfade is above zero, the silence setting is ignored, so you never get both at once.

A crossfade shortens the final result, because the overlapped stretch is played once instead of twice. Two tracks of three minutes with a four-second crossfade add up to about five minutes fifty-six, not six. Plan around that when you are fitting audio to a fixed length.

Crossfades only sound natural when the material under them can take it. Overlapping two songs with a steady beat works if the tempos are close and the overlap covers whole bars. Overlapping two people speaking is a mess. For speech, choose a short silence and leave the crossfade at zero.

Crossfades that follow the beat

The table in the Numbers tab converts bars into seconds. In 4/4 time a bar lasts four beats, so at 120 BPM one bar is two seconds and two bars are four. A crossfade sounds tidy when its length equals a whole number of bars and both tracks run at the same tempo with their bars starting where the files start; the tool does not align beats for you.

A few examples show how to read it. For a house set at 128 BPM, four bars last 7.5 seconds, right below the 8-second ceiling of the Crossfade knob. For a hip-hop mix near 90 BPM, a single bar already gives 2.7 seconds, and two bars give 5.3. Drum and bass at 174 BPM is brisk: one bar is only 1.4 seconds. Since the knob stops at 8 seconds, longer blends of eight bars at slower tempos are out of range; use fewer bars.

If you do not know the tempos of your tracks, the BPM finder will estimate them. A crossfade only blends well when both tempos are close; if one track is at 100 and the other at 128, use the speed changer first so the two match, or choose the gap instead.

How different sample rates are handled

Audio files do not all use the same sample rate. Music is often 44.1 kHz, video soundtracks and many phone recordings are 48 kHz, and some voice memos are lower. A joined file can only have one rate, so the tool converts every file to the rate of the first file in the list.

The same rule applies to channels: all tracks are brought to stereo. A mono recording is placed equally in the left and right channels, so it plays from the middle and does not drift to one side. Nothing is lost in that step, though a file that was mono still sounds mono.

Conversion is done by resampling, which is nearly transparent when the rates are close, such as 44.1 and 48 kHz. If your list holds a file at 22.05 kHz and others at 48 kHz, put a 48 kHz file first. The 22.05 kHz file is stretched to 48 kHz but cannot gain the high frequencies it never had.

You do not have to do anything about this. It simply explains why the file you download may sound slightly different in character for a track that came from a lower-quality source.

Gapless albums, podcasts and practice takes

For a live recording split into numbered tracks, load them in order and turn both knobs to zero. The result is close to gapless, because each piece starts at the exact sample where the previous one ended; MP3 sources can carry a few milliseconds of encoder padding. If you hear a small click at a seam, the cut points of the original files were probably not at silent spots; a crossfade of a few tenths of a second smooths it.

For a podcast, put the intro first, then the episode, then the outro, with about half a second of silence at each join. Loudness matters more than the transition: if the voice was recorded at a different level than the music, run each file through the normalizer or volume booster before joining.

For rehearsal takes, a gap of two seconds between files makes the result easy to navigate, because you can hear where one take ends. And for a presentation soundtrack made of several short clips, a short crossfade of one to two seconds keeps it flowing without sounding like a mix.

Whichever pattern you pick, shorten each piece first. The audio trimmer is the natural step before the joiner, and trim away the dead air first, then assemble what is left.

Size, format and saving

The final size is shown before you save. Join times add up: ten tracks of four minutes make a forty-minute file, and the size estimate follows from the format you choose. At 192 kbps MP3, forty minutes takes around 57 MB. As WAV, it is several hundred megabytes, so WAV is a choice for further editing, not for sharing. Files taken from video work as pieces too, since the sound of most MP4, MOV and WebM files is readable; the joined result is audio.

Merged audio is encoded once, at the end. The individual files are decoded, combined as raw samples, and then compressed a single time into the MP3 you download. That is better than joining already-compressed MP3 files several times in a row. Pick the bitrate that suits the content: 128 kbps for speech, 192 to 320 kbps for music.

Files up to 250 MB each can be loaded. Many large files on a slow phone will take a while to decode, so if the page seems busy, give it a moment before you press anything again.

Where the files go

Everything stays on the device in front of you. The browser opens each file, the page's code mixes them in memory, and the merged file is handed to you for download. No server receives your tracks, which makes it a reasonable choice for private recordings such as interviews, family audio or unreleased music.

Beyond the download, none of your audio is kept: the joiner does not store your files, and the Continue with links hand the merged result to the next tool through your browser's own storage on this device. Your browser settings let you wipe it whenever you like. If you want to share the arrangement of your settings with someone, use Copy settings link; it stores the knob positions in the address and never the audio.