Online Audio Compressor: Learn How Threshold, Ratio, Attack and Release Actually Work

Compression is one of the hardest audio processes to really understand — a textbook diagram can show you a static curve, but it can't show you how a compressor breathes with real music. This interactive compressor lets you load a sample track, move each control — threshold, ratio, attack, release and makeup gain — and watch input, output and the gain reduction curve respond together, in real time.

It's built to teach the process, not just illustrate it: a hands-on way to build an intuition for compression before applying it to a real mix, part of AudioFxLab's free collection of interactive audio tools.

Parameter Guide: How Each Compressor Control Works

Threshold

The threshold sets the level, in decibels, above which the compressor starts reducing gain. Every part of the signal below this line passes through untouched; everything above it gets compressed according to the ratio. Threshold and ratio always work together: a threshold set very close to 0 dB will only catch the loudest peaks, while a threshold set low (around -30 to -40 dB) will bring almost the entire signal under control, shaping the overall dynamics rather than just taming occasional spikes.

Where to start: watch the input waveform against the purple threshold line and place it just above the average level of the material you're feeding in, so only the louder passages cross into compression. If the signal barely touches the line, you'll see almost no gain reduction on the output; if the whole waveform sits above it, you're compressing continuously, which usually sounds squashed unless that's the effect you're after.

Ratio

Ratio defines how much of the signal above threshold gets reduced. A ratio of 2:1 means that for every 2 dB the input exceeds the threshold, only 1 dB comes through in the output — a gentle, mostly transparent effect. Higher ratios (8:1, 10:1) push more aggressively, flattening dynamics noticeably. At the far end of the scale, a very high ratio starts to behave like a limiter: virtually everything above threshold gets pinned close to that line.

Where to start: 2:1 to 4:1 for gentle dynamic control that still preserves the natural feel of the performance; 4:1 to 8:1 for a more noticeable, "glued" sound; above 10:1 only if you specifically want the aggressive, clamped character that comes with it.

Attack

Attack is the time the compressor takes to reach full gain reduction once the signal crosses the threshold. A fast attack (a few tenths of a millisecond) catches transients almost immediately, which controls peaks tightly but can also blunt the initial "punch" of a sound, since it reduces gain right as the transient happens. A slow attack (tens of milliseconds) lets the initial transient pass through uncompressed before the reduction kicks in, preserving punch and definition at the cost of tighter peak control.

Where to start: fast attack (1-10 ms) when the priority is control and consistency; slower attack (20-50 ms) when you want to preserve transient snap and let the compressor act mainly on the sustain of the sound. Use the Attack Area overlay to see exactly how much of each transient is passing through before the reduction engages.

Release

Release is the time the compressor takes to let go of the gain reduction once the signal drops back below threshold. A fast release (tens of milliseconds) returns to unity gain quickly, which can sound lively but risks audible "pumping" if it's timed close to the rhythm of the material. A slow release (several hundred milliseconds) smooths the transition out, avoiding pumping but potentially causing the compressor to still be reducing gain when the next transient arrives, reducing its impact.

Where to start: fast release (50-150 ms) for percussive or rhythmic material where responsiveness matters; slower release (300-800 ms) for sustained or legato material where a smooth, consistent level matters more than reacting to every fluctuation. Watch the Release Area overlay against the waveform to see whether the compressor has fully recovered before the next peak.

Makeup Gain

Compression, by definition, reduces the overall level of the signal — the harder you compress, the quieter the result gets on average. Makeup gain simply adds a fixed amount of gain back after compression, to bring the output back up to a comparable loudness to the input. It doesn't change how the compressor reacts to the signal, only the final volume of what comes out.

Where to start: after dialing in threshold, ratio, attack and release, increase makeup gain until the loudest parts of the output waveform reach roughly the same level as the original input peaks — this makes it easier to judge the effect of compression by ear without loudness itself being a factor.