How is a leveler different from a compressor?

A compressor reacts to peaks within milliseconds and reduces the loud parts, while a leveler rides the average level over seconds and moves both up and down. The result is that a leveler matches quiet and loud passages without touching transients, and a compressor tightens dynamics. They solve different problems and often work together.

What does the slew limit actually do?

It caps how fast the gain is allowed to change, in decibels per second. Without it the leveler would jump to the correct gain immediately, which sounds like a compressor pumping. With a limit of 24 decibels per second at hundred millisecond blocks, the gain can move about 2.4 decibels per step, which glides rather than jumps.

Why is there a maximum gain setting?

Because silence is the quietest thing there is. Without a cap, a leveler would try to lift near silent passages all the way to the target and turn the noise floor into the loudest part of the recording. The cap lets you decide how much lift is acceptable and verified behaviour keeps the boost within it.

Will it even out two different speakers?

It will reduce the difference between them over time, since it responds to the running average rather than to individual words, but it will not make them match word for word. For a conversation where the balance between speakers matters, the multi guest leveler is built for exactly that job.

Why does the beginning of my file sound too quiet?

Because the gain is limited in how quickly it can change, so it takes a moment to reach the target when the file starts quietly. Lowering the speed makes this worse and raising it makes the ride more audible. If the opening matters, the cleanest fix is to trim or fade the first second rather than to chase it with speed.

Audio Leveler

You Might Also Need

Loudness Normalizer

Normalize to LUFS loudness standards

RMS Normalizer

Normalize audio using RMS levels

Audio Compressor

Dynamic range compression for audio

Multi-Guest Leveler

Balance levels for multiple guests

Audio Leveler rides the level of a recording slowly so that quiet and loud passages end up matched, without squashing transients the way a compressor does. Gain moves at a limited speed and is capped at a maximum.

What is Audio Leveler?

Drop a file, set the level you want it to settle at and the limits on how fast and how far the gain may move, then run the pass. The footer reports the gain range that was actually applied across the file, so you can see how much work the leveler did.

What the panel gives you:

A target level from minus 30 to minus 12 dBFS in RMS terms, defaulting to minus 20.
A maximum gain from 6 to 30 decibels, defaulting to 18, which stops silence being amplified into a wall of noise.
A maximum speed from 6 to 60 decibels per second in steps of 3, defaulting to 24.
A footer reporting the gain range applied, from the smallest cut to the largest boost, together with the speed limit in use.
An output WAV named after the source with a leveled suffix, and a two channel limit.

The engine is a slow automatic gain control rather than a compressor. The signal is divided into hundred millisecond blocks, each block's RMS level is compared against the target, and the correction is applied as a gain that is allowed to change only at the rate you set, expressed in decibels per second. That slew limit is what makes the difference audible: a leveler glides while a compressor pumps, because the compressor reacts within milliseconds to individual peaks while this tool reacts over fractions of a second to the average. The gain is also capped, so a passage of silence or a very quiet room does not get boosted by the maximum amount and turn the noise floor into the loudest thing in the file. Verified behaviour is that the adjustments follow the slow trend of the material and that transients are left alone.

What it does not do:

It responds to RMS rather than to peaks, so a recording with sudden spikes is levelled on average while the spikes remain.
A low maximum speed can lag behind fast changes in level, leaving the first words after a quiet section too quiet.
A high maximum speed starts to sound like compression, which defeats the point of using a leveler instead.
The capped gain means a section that is far below the target is brought up only part of the way.
It works on the file as a whole, so two speakers with very different voices may still need manual attention afterwards.
There is no visual gain curve, only the reported range, so the shape of the ride is not shown.

Use it on the material where level drifts rather than on material where it is uneven bar by bar: lectures, interviews, voice memos and long recordings where someone leaned back from the microphone and then leaned in again. Start at the default target and speed, listen to the quietest and loudest passages, and lower the maximum speed if the ride is audible as it moves. Check the reported gain range too: a range that never exceeds a decibel or two means the file was already even and the pass changed almost nothing.

How to use Audio Leveler

Drop an audio file onto the panel; the tool processes up to two channels.
Set the target level in dBFS RMS, starting from the default of minus 20.
Adjust the maximum gain and the maximum speed, lowering the speed for a smoother ride.
Run the pass, read the applied gain range in the footer, then download the leveled WAV.

When to use Audio Leveler vs related tools

Audio Leveler is the tool for level that drifts slowly over a long recording, which is a different problem from a mix that needs its dynamics controlled. Loudness Normalizer applies one static gain to hit a measured loudness target, and Rms Normalizer does the same against an average level rather than a perceptual one. When the variation is bar to bar rather than minute to minute, Audio Compressor reacts quickly enough to catch it, and Multi-Guest Leveler is the version built for balancing several speakers at once.

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.