Compressor

General Purpose Dynamic Audio Range Compressor

Windows XP or laterWindows 64bit
v7.0 (2024-09-07)
2.9MiB

Windows XP or laterWindows 32bit
v7.0 (2024-09-07)
2.8MiB

macOS 11.0 or later. ARM (M1, M2) or Intel 64bitmacOS
v7.0 (2024-11-15)
3.9MiB

OSX 10.6 or later. Intel 32bit, Intel 64bit, PPCOSX Universal
v7.0 (2024-11-15)
5.8MiB

Linux 64bit
v7.0 (2024-09-07)
5.5MiB

Linux 32bit
v7.0 (2024-09-07)
5.3MiB

Linux/ARM 32bit (Raspberry Pi)Linux ARM
v7.0 (2024-09-07)
5.0MiB

Linux ARM64
v7.0 (2024-09-07)
5.3MiB

Source Code
(all platforms)

Donate

(no licensing payment)

About

x42-compressor (aka darc.lv2) is a Dynamic Audio Range Compressor inspired by Adriaensen's zita-dc1. It is a very nice sounding downward compressor that focuses on workflow, exposing only a minimal set of controls. Compression responds to signal-power (RMS), loudness as perceived by humans, not medium related digital-peak.

The compressor features auto makeup-gain, fixed at -10 dBFS/RMS. This maintains perceived loudness over a wide range of thresholds and ratio settings.

A distinct «Hold» feature changes the release-time to be applied only as long as the signal exceeds the threshold. In other words: When «Hold» is enabled, the current gain (or attenuation) is frozen when the signal falls below the threshold.
This prevents the compressor from adding excessive gain (e.g. modulate the noise-floor) and it can counter-act excessive 'pumping' when using short release times.

The plugin is available as mono and stereo variant. In case of the latter, both inputs are summed and the signal power is averaged to determine the gain factor.

Screenshots

User Interface

The control knobs respond to mouse click and mouse-drag movements:

• Click + drag on knob: up/left to increase, down/right to decrease
• Shift + click: reset to default
• Right-click on knob: toggle current value with default, 2nd click restore

The Ctrl key allows for fine-grained control when dragging or using the mouse-wheel on a knob.

Hovering the mouse over label below a control displays a tooltip with information.

Parameters

• Input Gain: Gain applied before level detection or any other processing. Note that this is not visualized as x-axis offset in the curve display. The range is -10 to +30dB in steps of 1dB (0.2dB fine-gained).
• Threshold: Signal level (RMS) at which the compression effect is engaged. The range is -50 to -10dB in steps of 0.5dB (0.1dB fine-grained).
• Ratio: The amount of gain or attenuation to be applied (dB/dB above threshold). Unity is retained at -10dBFS/RMS (auto makeup-gain).
• Attack time: Time it takes for the signal to become fully compressed after exceeding the threshold.
• Release time: Minimum recovery time to uncompressed signal-level after falling below threshold.
• Hold (check-button): Retain current attenuation when the signal subceeds the threshold. This prevents modulation of the noise-floor and can counter-act 'pumping'.

Display

The horizontal meter at the bottom shows the gain that is applied to the signal. The display shows the min/max range in the last 50ms. Values to the left < 0 means the signal is attenuated (input is louder than -10 dBFS/RMS).

The graph at the top visualizes the input to output gain mapping in dB. When «Hold» is enabled the area where the gain is not releases is shaded blue. The display also shows the input signal level as shaded white area under the curve as well and the current output-level (50ms average) as dot.

Acknowledgments

darc.lv2 is inspired by the algorithm and UI of Fons Adriaensen's zita-dc1.