pulp::signal::CombFilterT< SampleType > Class Template Reference

A fixed-capacity, integer-delay comb-filter family. More...

#include <comb_filter.hpp>

Public Member Functions
bool	prepare (std::size_t maximum_delay_samples)
bool	configure (const CombFilterConfig &candidate) noexcept
void	reset () noexcept
bool	prepared () const noexcept
bool	configured () const noexcept
std::size_t	maximum_delay_samples () const noexcept
const CombFilterConfig &	config () const noexcept
std::size_t	fault_count () const noexcept
void	clear_fault_count () noexcept
std::optional< std::size_t >	tail_samples () const noexcept
	Finite tail in samples, or nullopt for a recursive infinite tail.
CombFilterSampleResult< SampleType >	process (SampleType input) noexcept
CombFilterBlockResult	process (const SampleType *input, SampleType *output, std::size_t frames) noexcept

Static Public Member Functions
static constexpr std::size_t	processing_latency_samples () noexcept
	Comb structures retain a direct path and therefore add no fixed host compensation latency.

Static Public Attributes
static constexpr double	maximum_recursive_gain = 0.999
static constexpr std::size_t	minimum_delay_samples = 2u

Detailed Description

template<typename SampleType = float>
class pulp::signal::CombFilterT< SampleType >

A fixed-capacity, integer-delay comb-filter family.

The modes implement: feedforward: y[n] = x[n] + g x[n-D] feedback: y[n] = x[n] + g y[n-D] allpass: y[n] = w[n-D] - g w[n], w[n] = x[n] + g w[n-D]

prepare() is the only allocating operation and is transactional. configure() validates the complete candidate before committing it, then starts from a cold history. Processing, configure(), reset(), and fault recovery allocate nothing. Exact in-place block processing is supported. A nonfinite input or arithmetic overflow emits zero and discards recursive history in O(1).

Member Function Documentation

clear_fault_count()

template<typename SampleType = float>

void pulp::signal::CombFilterT< SampleType >::clear_fault_count ( )

inlinenoexcept

config()

template<typename SampleType = float>

const CombFilterConfig & pulp::signal::CombFilterT< SampleType >::config ( ) const

inlinenodiscardnoexcept

configure()

template<typename SampleType = float>

bool pulp::signal::CombFilterT< SampleType >::configure ( const CombFilterConfig & candidate )

inlinenodiscardnoexcept

configured()

template<typename SampleType = float>

bool pulp::signal::CombFilterT< SampleType >::configured ( ) const

inlinenodiscardnoexcept

fault_count()

template<typename SampleType = float>

std::size_t pulp::signal::CombFilterT< SampleType >::fault_count ( ) const

inlinenodiscardnoexcept

maximum_delay_samples()

template<typename SampleType = float>

std::size_t pulp::signal::CombFilterT< SampleType >::maximum_delay_samples ( ) const

inlinenodiscardnoexcept

prepare()

template<typename SampleType = float>

bool pulp::signal::CombFilterT< SampleType >::prepare ( std::size_t maximum_delay_samples )

inlinenodiscard

prepared()

template<typename SampleType = float>

bool pulp::signal::CombFilterT< SampleType >::prepared ( ) const

inlinenodiscardnoexcept

process() [1/2]

template<typename SampleType = float>

CombFilterBlockResult pulp::signal::CombFilterT< SampleType >::process ( const SampleType * input, SampleType * output, std::size_t frames )

inlinenodiscardnoexcept

process() [2/2]

template<typename SampleType = float>

CombFilterSampleResult< SampleType > pulp::signal::CombFilterT< SampleType >::process ( SampleType input )

inlinenodiscardnoexcept

processing_latency_samples()

template<typename SampleType = float>

constexpr std::size_t pulp::signal::CombFilterT< SampleType >::processing_latency_samples ( )

inlinestaticnodiscardconstexprnoexcept

Comb structures retain a direct path and therefore add no fixed host compensation latency.

D is an internal signal-path delay.

reset()

template<typename SampleType = float>

void pulp::signal::CombFilterT< SampleType >::reset ( )

inlinenoexcept

tail_samples()

template<typename SampleType = float>

std::optional< std::size_t > pulp::signal::CombFilterT< SampleType >::tail_samples ( ) const

inlinenodiscardnoexcept

Finite tail in samples, or nullopt for a recursive infinite tail.

Zero-gain feedback is dry; zero-gain allpass is a pure D-sample delay.

Member Data Documentation

maximum_recursive_gain

template<typename SampleType = float>

double pulp::signal::CombFilterT< SampleType >::maximum_recursive_gain = 0.999

staticconstexpr

minimum_delay_samples

template<typename SampleType = float>

std::size_t pulp::signal::CombFilterT< SampleType >::minimum_delay_samples = 2u

staticconstexpr

---

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

• /home/runner/work/pulp/pulp/core/signal/include/pulp/signal/comb_filter.hpp