Why does each layer sound quieter in the mix?

Because the sum is scaled by one over the number of layers, so two inputs are each counted at half level. That is what guarantees the mix cannot clip where the sources did not. If a layer needs to dominate, its own level should be raised before mixing, or fewer layers used.

What happens if the files are different lengths?

The longest file sets the output length and the shorter ones simply run out, continuing as silence. No padding is written and nothing is trimmed, so a short bed under a long voice leaves the tail of the voice clear.

Can I offset one layer so it starts later?

Not here, every layer starts at zero. Cut the layer so that the material begins where it should, then add it; the trimmer or the fade tool can prepare that leading silence.

Does previewing change anything?

No. The preview plays the blend through the browser's own audio engine for checking balance, and it writes nothing. The rendered WAV is produced only when the mix and download control is pressed, and the two follow the same summing rule.

Can I keep the layers as separate files too?

The output is the mixed WAV only, since the purpose is to combine. If the individual parts matter as well, keep copies of the originals, which are untouched throughout, and use the slicer or cutter if they need dividing.

Audio Merger

You Might Also Need

Audio Joiner

Merge multiple audio files into one

Crossfade Creator

Create crossfades between audio clips

Audio Loop Maker

Create seamless audio loops

Audio Normalizer

Normalize audio levels for consistency

Audio Merger plays several files at the same time and renders the result as one mix, which makes it the tool for layering a voice over a bed or stacking takes. Order does not matter, because every file starts at once.

What is Audio Merger?

Add two or more files, preview the blend if you want, then mix and download. The panel reports how many layers are in play, how long the longest one is, and the level they are summed at.

The controls and syntax:

A drop zone that accepts several files and keeps adding to the stack.
A row per layer showing the basename and its duration, with a remove control.
A preview control that plays the blend without rendering it, so the balance can be checked first.
A mix-and-download control producing merged-mix.wav, plus a line naming the layer count, the longest duration and the summing level.

Mixing sums every input sample by sample and scales the result by one over the number of inputs, so four layers are each counted at a quarter of their level. That choice is deliberate and is the reason the tool cannot clip where the sources did not: if every input peaks at full scale, the sum peaks at the full-scale value rather than at four times it. The trade is that each individual layer comes out quieter than it went in, by roughly the number of layers in decibels, which is why the readout states the scale explicitly. Unequal lengths are handled by treating the longest file as the output length and letting the shorter ones run out into silence, so a short bed under a long voice does not shorten the result. Sample rates are again matched by the browser's decoder, so a set of mixed sources aligns without pitch correction. The preview path uses the browser's own mixer, which is the quickest way to hear the balance without writing a file. The rule is easy to state: every layer is summed sample by sample and the total is scaled by 1/n, where n is the number of layers, so the mix cannot exceed the level any single input reached. All layers begin simultaneously, which is what makes their order irrelevant, and a shorter layer simply runs out into silence while the longest continues to the end of the output.

What it does not do:

No per-layer fader or pan, so the balance is fixed at equal weighting.
No offset or delay per layer, so a layer cannot be nudged later than the start.
No ducking, so a voice does not automatically push a music bed down.
Reverb, gain and compression are separate tools, not options here.
Everything is held in memory, so many long layers are bounded by the browser.
The output is a plain WAV with no metadata and a fixed name.

Use it for the jobs where two things have to happen at once: a voiceover over an instrumental bed, a narration with a room tone underneath, or several takes stacked to thicken a part. When the same files should follow one another instead, the joiner is the right shape, and when the contents of a layer need work before it is combined, the gain, leveler and EQ tools act on it first.

How to use Audio Merger

Drop two or more files to build the layer stack.
Use the preview control to hear the blend before committing to it.
Remove any layer that does not belong and check the stated summing level.
Press mix and download, then listen through to confirm nothing is masked.

When to use Audio Merger vs related tools

Audio Merger is for playing files at the same time, so it suits layering a voice over music or stacking takes. Audio Joiner sequences files instead of layering them, and Crossfade Creator blends two files across a transition rather than placing them on top of each other. When a layer needs its level evened out first, Audio Leveler and Multi-Guest Leveler prepare it.

Privacy & Security

This tool runs entirely in your browser — no data ever leaves your device. There is no server round-trip, no upload, no logging, and no account required. Your input is processed locally using client-side JavaScript and is never stored, transmitted, or accessible to anyone else. When you close the tab, everything disappears.