Transient shaper - independent control over attack and sustain. More...

#include <gin_transientshaper.h>

Public Member Functions
	TransientShaper ()=default

	~TransientShaper ()=default

void	setSampleRate (double sr)

void	setAttack (float a)
	Attack shaping, -4 to +4.

void	setSustain (float s)
	Sustain shaping, -8 to +8.

void	setDetectorBalance (float b)
	What the detector listens to: 0 is mid only, 0.5 both, 1 side only.

void	setProcessBalance (float b)
	Where the gain is applied: 0 is mid only, 0.5 both, 1 side only.

void	setBand (float lowCutHz, float highCutHz)
	Restricts shaping to a band.

void	setSmoothing (float s)
	Gain smoothing, 0 (snappy, 10 kHz) to 1 (smooth, 0.1 Hz).

void	setOutputGain (float db)

void	reset ()

void	process (juce::AudioSampleBuffer &buffer)

Detailed Description

Transient shaper - independent control over attack and sustain.

Unlike a compressor, there is no threshold. The gain comes from the ratio between three envelope followers running on the same detector signal: a fast one (100 Hz), a baseline (10 Hz) and a slow one (1 Hz). When the fast envelope runs ahead of the baseline the signal is in its attack; when the slow one sits above the baseline it is in its tail. Because both are ratios, the effect is level independent - it works the same on a quiet drum loop as on a hot one, with nothing to re-tune as the source changes.

Positive attack sharpens transients, negative softens them. Positive sustain lengthens the tail, negative shortens it - a quick way to dry out an over-reverberant room mic.

Key Features:

Attack and sustain shaped independently, with no threshold to set
Mid/side weighting for both detection and processing, so transients can be shaped on the sides only
Optional band limiting - the untouched part of the spectrum is passed through dry, so only the kick region gets shaped
Gain smoothing to trade snap against artefacts

Usage:

TransientShaper shaper;
shaper.setSampleRate (44100.0);
shaper.setAttack (2.0f);      // sharpen the attack
shaper.setSustain (-1.0f);    // and dry out the tail
 
shaper.process (buffer);      // stereo, in place

Note: stereo only - process() expects a two channel buffer.

See also: Dynamics, Maximizer

Constructor & Destructor Documentation

◆ TransientShaper()

TransientShaper::TransientShaper ( )

default

◆ ~TransientShaper()

TransientShaper::~TransientShaper ( )

default

Member Function Documentation

◆ setSampleRate()

void TransientShaper::setSampleRate ( double sr )

References reset().

◆ setAttack()

void TransientShaper::setAttack ( float a )

Attack shaping, -4 to +4.

Positive sharpens, negative softens, 0 is neutral.

◆ setSustain()

void TransientShaper::setSustain ( float s )

Sustain shaping, -8 to +8.

Positive lengthens the tail, negative shortens it.

◆ setDetectorBalance()

void TransientShaper::setDetectorBalance ( float b )

What the detector listens to: 0 is mid only, 0.5 both, 1 side only.

◆ setProcessBalance()

void TransientShaper::setProcessBalance ( float b )

Where the gain is applied: 0 is mid only, 0.5 both, 1 side only.

◆ setBand()

void TransientShaper::setBand	(	float	lowCutHz,
		float	highCutHz
	)

Restricts shaping to a band.

Everything outside it is passed through dry and unshaped, so a 40-200 Hz band shapes the kick and leaves the rest of the mix alone.

Parameters

lowCutHz	bottom of the shaped band, 0 for no low cut
highCutHz	top of the shaped band, 0 or >= nyquist for none

◆ setSmoothing()

void TransientShaper::setSmoothing ( float s )

Gain smoothing, 0 (snappy, 10 kHz) to 1 (smooth, 0.1 Hz).

◆ setOutputGain()

void TransientShaper::setOutputGain ( float db )

◆ reset()

void TransientShaper::reset ( )

References Biquad::reset().

Referenced by setSampleRate().

◆ process()

void TransientShaper::process ( juce::AudioSampleBuffer & buffer )

References Biquad::process().

The documentation for this class was generated from the following file:

gin_transientshaper.h

Public Member Functions