DspMath.h File Reference

Core mathematical utilities for digital signal processing. More...

#include <algorithm>
#include <cmath>
#include <concepts>
#include <numbers>

Include dependency graph for DspMath.h:

Go to the source code of this file.

Namespaces
namespace	dspark
	Main namespace for the DSPark framework.

Concepts
concept	dspark::FloatType
	Constrains a type to IEEE floating-point (float or double).

Functions
template<FloatType T>
T	dspark::decibelsToGain (T dB, T minusInfinityDb=T(-100)) noexcept
	Converts a value in decibels to linear gain.
template<FloatType T>
T	dspark::gainToDecibels (T gain, T minusInfinityDb=T(-100)) noexcept
	Converts a linear gain value to decibels.
template<FloatType T>
T	dspark::mapRange (T value, T inMin, T inMax, T outMin, T outMax) noexcept
	Maps a value from one range to another (linear interpolation).
template<FloatType T>
T	dspark::fastTanh (T x) noexcept
	Fast tanh approximation using Pade rational function.
template<FloatType T>
T	dspark::fastPow10 (T x) noexcept
	Fast approximation of 10^x using exp2.
template<FloatType T>
T	dspark::fastExp (T x) noexcept
	Fast approximation of e^x via std::exp2 (~2x faster than std::exp on MSVC).
template<FloatType T>
T	dspark::fastTan (T x) noexcept
	Fast approximation of tan(x) using a Pade [5,4] rational approximant.
template<FloatType T>
T	dspark::fastSin (T x) noexcept
	Fast sine approximation (degree-9 odd minimax polynomial).
template<FloatType T>
T	dspark::fastCos (T x) noexcept
	Fast cosine approximation. See fastSin() for accuracy notes (float error is ~7e-6 here: half an ulp more from the pi/2 offset).
template<FloatType T>
T	dspark::fastLog (T x) noexcept
	Fast natural logarithm approximation.
template<FloatType T>
T	dspark::wrapPhase (T phase) noexcept
	Normalises a phase value to the range [0, 2*pi).

Variables
template<FloatType T>
constexpr T	dspark::pi = std::numbers::pi_v<T>
	Pi (3.14159...) for the given floating-point type.
template<FloatType T>
constexpr T	dspark::twoPi = T(2) * std::numbers::pi_v<T>
	2 * Pi (6.28318...).
template<FloatType T>
constexpr T	dspark::invTwoPi = T(1) / twoPi<T>
	1 / (2 * Pi) (0.15915...). Useful for fast phase divisions.
template<FloatType T>
constexpr T	dspark::halfPi = std::numbers::pi_v<T> / T(2)
	Pi / 2 (1.57079...). Quarter period; sin/cos phase offset.
template<FloatType T>
constexpr T	dspark::sqrt2 = std::numbers::sqrt2_v<T>
	Square root of 2 (1.41421...).
template<FloatType T>
constexpr T	dspark::invSqrt2 = T(1) / std::numbers::sqrt2_v<T>
	1 / square root of 2 (0.70710...). Butterworth Q factor.

Detailed Description

Core mathematical utilities for digital signal processing.

Provides constants, unit conversions, and fast approximations commonly needed in audio DSP code. All functions are noexcept and suitable for real-time use.

Dependencies: C++20 standard library only.

Definition in file DspMath.h.