What does the ratio mean?

It describes how much the level above the threshold is reduced. At 3 to 1, every 3 decibels above the threshold becomes 1 decibel at the output, so the excess is cut by two thirds. At 6 to 1 the reduction is stronger, which is why the drum preset feels more controlled than the vocal one.

Why does attack time change the character so much?

Because it decides how much of a transient survives before the gain comes down. A 25 millisecond attack lets the initial hit of a drum through before compressing the body, which is what gives the drum preset its punch. A 5 millisecond attack catches the transient too, which is what broadcast speech wants.

It is a band around the threshold where compression is applied gradually rather than all at once. Without it, material hovering near the threshold switches between compressed and uncompressed, which is audible as a lurch. With a knee of 3 to 8 decibels the transition is smooth.

Does make-up gain change the dynamics?

No. It is a constant gain applied after compression, so it moves the whole file up without altering the reduced dynamic range. It exists because compression lowers the peak level, and without it the result would simply be quieter than the input rather than denser.

Why does the after figure look similar to the before figure?

Because the make-up gain is chosen to bring the average level back. What has changed is not the overall level but the distance between the loud and quiet parts, which is exactly what compression is for. If the two figures look identical, the preset is behaving as intended rather than doing nothing.

Dynamic Range Compressor

You Might Also Need

Audio Compressor

Dynamic range compression for audio

Multiband Compressor

Compress different frequency bands separately

Parallel Compressor

Blend compressed and dry signals

Audio Leveler

Automatically level audio volume

Dynamic Range Compressor narrows the gap between the loud and quiet parts of a recording, and leads with three presets rather than a wall of controls. The panel shows the actual curve numbers behind each preset instead of hiding them.

What is Dynamic Range Compressor?

Drop a file, choose a preset and read the parameter line beneath it, which states the ratio, threshold, knee, attack, release and make-up gain in full. Running the pass returns a new file plus a report comparing the average level before and after.

What each preset is set to:

Vocal leveler at 3 to 1 above minus 24 dBFS, with an 8 decibel knee, 10 millisecond attack, 150 millisecond release and plus 4 decibels of make-up gain.
Drum punch at 6 to 1 above minus 18 dBFS, with a 3 decibel knee, 25 millisecond attack, 100 millisecond release and plus 3 decibels make-up.
Broadcast at 4 to 1 above minus 20 dBFS, with a 6 decibel knee, 5 millisecond attack, 120 millisecond release and plus 5 decibels make-up.
A report line giving the average level before and after, the ratio and threshold in force, and a short description of the intended character.
An output WAV named after the source with a compressed suffix.

The engine uses a windowed RMS detector, a soft knee and gain smoothing rather than hard switching. An RMS detector responds to the energy in a window rather than to individual samples, which is what makes the result sound like levelling rather than like distortion. The soft knee means the transition into compression is gradual across a band around the threshold, so material that sits near the threshold does not switch between compressed and uncompressed states. Attack and release shape the gain over time, and the smoothing in the gain path is what keeps the changes from producing clicks. The make-up gain compensates for the level that compression removed, which is why the report's after figure is usually close to the before figure even though the dynamic range has narrowed. The drum preset deliberately uses a slower attack so the initial transient survives the compression and the body behind it gets denser.

What compression does not do:

It reduces dynamic range rather than raising loudness; the level only rises because of the make-up gain that follows.
It cannot make a recording sound better than it is, and on already compressed material it mostly removes what little dynamic range remains.
The presets are starting points rather than settings matched to your file, and there is no preview to hear the curve before committing.
A compressor reacts to level, not to quality, so a noisy quiet passage gets boosted along with everything else once make-up gain is applied.
Three presets is not a full compressor: there is no sidechain, no band splitting and no per channel linking control here.
No metering of gain reduction is shown while setting up, only the average level change in the report afterwards.

Use it to even out a voice before a podcast mix, to add density to a drum bus, or to make a spoken word recording sit at a consistent level for broadcast. Read the parameter line rather than trusting the preset name, because it tells you exactly what the curve is doing, and remember that compressing and then normalising is the usual order when the goal is a loud, controlled result.

How to use Dynamic Range Compressor

Drop an audio file onto the panel or click to choose one from disk.
Choose one of the three presets and read the parameter line to see the ratio, threshold and timings behind it.
Run the pass and compare the average level reported before and after compression.
Download the compressed WAV, and follow with a normaliser if the file needs to hit a specific loudness.

When to use Dynamic Range Compressor vs related tools

Dynamic Range Compressor is the preset-led route into compression, which suits anyone who wants a known result rather than a parameter hunt. Audio Compressor exposes the same idea as raw controls, and Multiband Compressor splits the signal so a boomy low end can be controlled without touching the top. When the aim is transparency rather than density, Audio Leveler rides the level slowly instead, and Parallel Compressor keeps the original punch underneath a compressed layer.

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.