|
MayaFlux 1.0.0
Digital-First Multimedia Processing Framework
|
Settings of a WindowedKernelAnalyzer. More...
#include <WindowedKernelAnalyzer.hpp>
Public Member Functions | |
| WindowedKernelShape | shape () const noexcept |
| Public Member Functions inherited from MayaFlux::Yantra::OperationConfig | |
| template<typename Op> | |
| void | apply_to (Op &operation) const |
| Apply retained concrete settings to an operation through its parent contract. | |
| template<typename T> | |
| const T * | get () const |
| The struct of this type, or nullptr when absent. | |
| OperationConfig ()=default | |
| template<typename Config> requires (!std::same_as<std::remove_cvref_t<Config>, OperationConfig> && std::derived_from<std::remove_cvref_t<Config>, OperationConfig>) | |
| OperationConfig (Config &&config) | |
| Retain a concrete config whole when placed in a parent-config collection. | |
| OperationConfig (const OperationConfig &)=default | |
| OperationConfig (OperationConfig &&) noexcept=default | |
| OperationConfig & | operator= (const OperationConfig &)=default |
| OperationConfig & | operator= (OperationConfig &&) noexcept=default |
| template<typename T> | |
| void | set (T value) |
| Stores a struct under its type identity. | |
Public Attributes | |
| uint32_t | inputs { 1 } |
| Sequences the kernel reads. | |
| std::string | kernel |
| GLSL statements run for each window. | |
| uint32_t | lanes { 1 } |
| Invocations that share one window. | |
| uint32_t | outputs { 1 } |
| Results the kernel writes for each window. | |
| std::vector< float > | parameters |
| Floats the kernel reads as parameters[k]. | |
| std::string | shader_path |
| A .comp file following the windowed kernel layout, used instead of kernel. | |
| Public Attributes inherited from MayaFlux::Yantra::AnalyzerConfig | |
| bool | classification_enabled { false } |
| Whether to classify energy or value levels. | |
| AnalysisGranularity | granularity { AnalysisGranularity::RAW_VALUES } |
| Level at which the analyzer processes input. | |
| uint32_t | hop_size { 128 } |
| Distance between successive windows in samples. | |
| uint32_t | window_size { 256 } |
| Analysis window length in samples. | |
Settings of a WindowedKernelAnalyzer.
Exactly one of kernel and shader_path is set. kernel is the body of the per-window GLSL function, in the layout described above windowed_kernel_workgroup. shader_path names a whole .comp file written against the same layout.
lanes is how many invocations share a window, 1 or a power of two up to 1024. inputs is how many sequences the kernel reads. With one input every channel is analysed on its own. With more, the first inputs channels of the data are the inputs of one kernel run, with the same window and hop, and the analysis has one channel; windows run while every input has them.
Definition at line 338 of file WindowedKernelAnalyzer.hpp.