class

FConvolutionEngine

Uniform partitioned convolution (overlap-save), the engine behind convolution reverb. The impulse response is split into blockSize partitions, each FFT'd once at Init; per input block one FFT + P complex multiply-accumulates + one IFFT replace the O(n·m) direct convolution. Mono — the effect node runs one engine per channel. Streaming contract: fixed latency of exactly blockSize samples regardless of how the input is chunked (the output FIFO is primed with one block of zeros, making underrun impossible for any feed pattern).

Module
Audio
Header
#include "Audio/ConvolutionEngine.h"
Namespace
Mowo

Public functions

ReturnsNameDescription
voidFFTRadix2In-place complex radix-2 FFT (n must be a power of two).
boolInit
voidReset
boolIsInitialized
uint32_tGetBlockSize
uint32_tGetPartitionCount
voidProcess

Function details

FFTRadix2

C++
static void FFTRadix2(float* re, float* im, uint32_t n, bool inverse)

In-place complex radix-2 FFT (n must be a power of two). Twiddles accumulate in double so long transforms stay audio-clean. The inverse transform includes the 1/n scale.

Init

C++
bool Init(const float* ir, uint32_t irFrames, uint32_t blockSize = 256)

Reset

C++
void Reset()

IsInitialized

C++
bool IsInitialized() const

GetBlockSize

C++
uint32_t GetBlockSize() const

GetPartitionCount

C++
uint32_t GetPartitionCount() const

Process

C++
void Process(const float* in, float* out, uint32_t frames)
Generated from Audio/ConvolutionEngine.h via audio.json · Oct 26, 2026Version 1.0.0