One LFO's running state: phase, delay and fade-in accumulators, stepped a control tick at a time. More...

#include <tabulasonora/lfo_engine.hpp>

Public Member Functions
	LfoRunner (const LfoEngine &engine, const LfoConfig &config)
const LfoConfig &	config () const noexcept
	The configuration this runner steps.
bool	is_applied () const noexcept
	Whether the last tick ran an update at all, so the LFO has an output to give.
void	tick (int rate_offset=0) noexcept
	Advances one control tick.
double	value (LfoDestination destination, int matrix_depth=0) const noexcept
	The modulation for one destination, after the last tick.
double	pitch_value (double wheel_depth) const noexcept
	The pitch modulation with a mod-wheel depth folded in, in milli-semitones.

Detailed Description

One LFO's running state: phase, delay and fade-in accumulators, stepped a control tick at a time.

All three destinations share one runner. The accumulators do not depend on which depth is being applied, so stepping once and reading three values is the same trajectory as running the LFO three times — and it is what the engine does, since a partial has one LFO object, not one per destination.

tick advances; the value is read afterwards, so the first tick already carries a phase increment.

Constructor & Destructor Documentation

◆ LfoRunner()

ts::LfoRunner::LfoRunner	(	const LfoEngine &	engine,
		const LfoConfig &	config )

inline

Member Function Documentation

◆ config()

const LfoConfig & ts::LfoRunner::config ( ) const

inlinenodiscardnoexcept

The configuration this runner steps.

◆ is_applied()

bool ts::LfoRunner::is_applied ( ) const

inlinenodiscardnoexcept

Whether the last tick ran an update at all, so the LFO has an output to give.

This is not "the delay has elapsed" — the delay holds the fade-in at zero but does not stop the LFO being applied, since a matrix depth is summed past the fade. What turns it off is an increment of nothing, which is the one case the module skips outright.

◆ tick()

void ts::LfoRunner::tick ( int rate_offset = 0 )

noexcept

Advances one control tick.

rate_offset is the control matrix's LFO rate destination, in the same per-tick increment units as the configured rate — the engine adds it to the rate straight out of the table (or, for LFO2, to the raw rate) and clamps the total to 0x28f6.

A total of zero or less does not merely stop the phase: the engine skips the whole update, so the depths are not applied either and the LFO's last output is held rather than decaying. Deep enough negative rate modulation is therefore an off switch, which is why this shares its early return with a configured rate of zero.

◆ value()

double ts::LfoRunner::value	(	LfoDestination	destination,
		int	matrix_depth = 0 ) const

nodiscardnoexcept

The modulation for one destination, after the last tick.

matrix_depth is the control matrix's depth destination for this LFO, in the destination's own units. It is summed after the fade-in — the fade scales the patch's depth, not the controller's — and the total is then clamped to the destination's limit. Those limits are exactly the matrix's own full-scale figures, so a matrix depth alone can reach the rail but never exceed it.

◆ pitch_value()

double ts::LfoRunner::pitch_value ( double wheel_depth ) const