Convolution engine. More...

Classes
struct	Convolver
	Combination of Input and Output. More...

struct	fft_node
	Two blocks of time-domain or FFT (half-complex) data. More...

struct	Filter
	Container holding a number of FFT blocks. More...

struct	Input
	Input stage of convolution. More...

class	Output
	Convolution engine (output part). More...

class	OutputBase
	Base class for Output and StaticOutput. More...

struct	StaticConvolver
	Combination of Input and StaticOutput. More...

class	StaticOutput
	Convolver output stage with static filter. More...

struct	Transform
	Helper class to prepare filters. More...

class	TransformBase
	Forward-FFT-related functions. More...

Functions
static size_t	min_partitions (size_t block_size, size_t filter_size)
	Calculate necessary number of partitions for a given filter length. More...

template<typename BinaryFunction >
void	transform_nested (const Filter &in1, const Filter &in2, Filter &out, BinaryFunction f)
	Apply `std::transform` to a container of fft_nodes. More...

Detailed Description

Convolution engine.

A convolution engine normally consists of an Input, a Filter and an Output / StaticOutput. There are also combinations: Input + Output = Convolver; Input + StaticOutput = StaticConvolver

Uses (uniformly) partitioned convolution.

TODO: describe thread (un)safety

Function Documentation

static size_t apf::conv::min_partitions	(	size_t	block_size,
		size_t	filter_size
	)

static

Calculate necessary number of partitions for a given filter length.

Definition at line 63 of file convolver.h.

template<typename BinaryFunction >

void apf::conv::transform_nested	(	const Filter &	in1,
		const Filter &	in2,
		Filter &	out,
		BinaryFunction	f
	)

Apply std::transform to a container of fft_nodes.

Definition at line 792 of file convolver.h.