diff --git a/rtengine/curves.h.save-failed b/rtengine/curves.h.save-failed new file mode 100644 index 000000000..47ab0fc6f --- /dev/null +++ b/rtengine/curves.h.save-failed @@ -0,0 +1,1142 @@ +/* + * This file is part of RawTherapee. + * + * Copyright (c) 2004-2010 Gabor Horvath + * + * RawTherapee is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * RawTherapee is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with RawTherapee. If not, see . + */ +#ifndef __CURVES_H__ +#define __CURVES_H__ + +#include +#include +#include +#include "rt_math.h" +#include "../rtgui/mycurve.h" +#include "../rtgui/myflatcurve.h" +#include "../rtgui/mydiagonalcurve.h" +#include "color.h" +#include "procparams.h" +#include "pipettebuffer.h" + +#include "LUT.h" + +#define CURVES_MIN_POLY_POINTS 1000 + +#include "rt_math.h" + +#define CLIPI(a) ((a)>0?((a)<65534?(a):65534):0) + +using namespace std; + +namespace rtengine +{ +class ToneCurve; +class ColorAppearance; + +class CurveFactory +{ + + friend class Curve; + +protected: + + // functions calculating the parameters of the contrast curve based on the desired slope at the center + static double solve_upper (double m, double c, double deriv); + static double solve_lower (double m, double c, double deriv); + static double dupper (const double b, const double m, const double c); + static double dlower (const double b, const double m, const double c); + + // basic convex function between (0,0) and (1,1). m1 and m2 controls the slope at the start and end point + static inline double basel (double x, double m1, double m2) + { + if (x == 0.0) { + return 0.0; + } + + double k = sqrt ((m1 - 1.0) * (m1 - m2) * 0.5) / (1.0 - m2); + double l = (m1 - m2) / (1.0 - m2) + k; + double lx = xlog(x); + return m2 * x + (1.0 - m2) * (2.0 - xexp(k * lx)) * xexp(l * lx); + } + // basic concave function between (0,0) and (1,1). m1 and m2 controls the slope at the start and end point + static inline double baseu (double x, double m1, double m2) + { + return 1.0 - basel(1.0 - x, m1, m2); + } + // convex curve between (0,0) and (1,1) with slope m at (0,0). hr controls the highlight recovery + static inline double cupper (double x, double m, double hr) + { + if (hr > 1.0) { + return baseu (x, m, 2.0 * (hr - 1.0) / m); + } + + double x1 = (1.0 - hr) / m; + double x2 = x1 + hr; + + if (x >= x2) { + return 1.0; + } + + if (x < x1) { + return x * m; + } + + return 1.0 - hr + hr * baseu((x - x1) / hr, m, 0); + } + // concave curve between (0,0) and (1,1) with slope m at (1,1). sr controls the shadow recovery + static inline double clower (double x, double m, double sr) + { + return 1.0 - cupper(1.0 - x, m, sr); + } + // convex curve between (0,0) and (1,1) with slope m at (0,0). hr controls the highlight recovery + static inline double cupper2 (double x, double m, double hr) + { + double x1 = (1.0 - hr) / m; + double x2 = x1 + hr; + + if (x >= x2) { + return 1.0; + } + + if (x < x1) { + return x * m; + } + + return 1.0 - hr + hr * baseu((x - x1) / hr, m, 0.3 * hr); + } + static inline double clower2 (double x, double m, double sr) + { + //curve for b<0; starts with positive slope and then rolls over toward straight line to x=y=1 + double x1 = sr / 1.5 + 0.00001; + + if (x > x1 || sr < 0.001) { + return 1 - (1 - x) * m; + } else { + double y1 = 1 - (1 - x1) * m; + return y1 + m * (x - x1) - (1 - m) * SQR(SQR(1 - x / x1)); + } + } + // tone curve base. a: slope (from exp.comp.), b: black point normalized by 65535, + // D: max. x value (can be>1), hr,sr: highlight,shadow recovery + static inline double basecurve (double x, double a, double b, double D, double hr, double sr) + { + if (b < 0) { + double m = 0.5;//midpoint + double slope = 1.0 + b; //slope of straight line between (0,-b) and (1,1) + double y = -b + m * slope; //value at midpoint + + if (x > m) { + return y + (x - m) * slope; //value on straight line between (m,y) and (1,1) + } else { + return y * clower2(x / m, slope * m / y, 2.0 - sr); + } + } else { + double slope = a / (1.0 - b); + double m = a * D > 1.0 ? b / a + (0.25) / slope : b + (1 - b) / 4; + double y = a * D > 1.0 ? 0.25 : (m - b / a) * slope; + + if (x <= m) { + return b == 0 ? x * slope : clower (x / m, slope * m / y, sr) * y; + } else if (a * D > 1.0) { + return y + (1.0 - y) * cupper2((x - m) / (D - m), slope * (D - m) / (1.0 - y), hr); + } else { + return y + (x - m) * slope; + } + } + } + static inline double simplebasecurve (double x, double b, double sr) + { + // a = 1, D = 1, hr = 0 (unused for a = D = 1) + if (b == 0.0) { + return x; + } else if (b < 0) { + double m = 0.5;//midpoint + double slope = 1.0 + b; //slope of straight line between (0,-b) and (1,1) + double y = -b + m * slope; //value at midpoint + + if (x > m) { + return y + (x - m) * slope; //value on straight line between (m,y) and (1,1) + } else { + return y * clower2(x / m, slope * m / y, 2.0 - sr); + } + } else { + double slope = 1.0 / (1.0 - b); + double m = b + (1 - b) * 0.25; + double y = (m - b) * slope; + + if (x <= m) { + return clower (x / m, slope * m / y, sr) * y; + } else { + return y + (x - m) * slope; + } + } + } + + +public: + const static double sRGBGamma; // standard average gamma + const static double sRGBGammaCurve; // 2.4 in the curve + + + //%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + // accurately determine value from integer array with float as index + //linearly interpolate from ends of range if arg is out of bounds + static inline float interp(int *array, float f) + { + int index = CLIPI(floor(f)); + float part = (float)((f) - index) * (float)(array[index + 1] - array[index]); + return (float)array[index] + part; + } + //%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + // accurately determine value from float array with float as index + //linearly interpolate from ends of range if arg is out of bounds + static inline float flinterp(float *array, float f) + { + int index = CLIPI(floor(f)); + float part = ((f) - (float)index) * (array[index + 1] - array[index]); + return array[index] + part; + } + //%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + + static inline double centercontrast (double x, double b, double m); + + // standard srgb gamma and its inverse + static inline double gamma2 (double x) + { + return x <= 0.00304 ? x * 12.92 : 1.055 * exp(log(x) / sRGBGammaCurve) - 0.055; + } + static inline double igamma2 (double x) + { + return x <= 0.03928 ? x / 12.92 : exp(log((x + 0.055) / 1.055) * sRGBGammaCurve); + } + static inline float gamma2 (float x) + { + return x <= 0.00304 ? x * 12.92 : 1.055 * expf(logf(x) / sRGBGammaCurve) - 0.055; + } + static inline float igamma2 (float x) + { + return x <= 0.03928 ? x / 12.92 : expf(logf((x + 0.055) / 1.055) * sRGBGammaCurve); + } + // gamma function with adjustable parameters + static inline double gamma (double x, double gamma, double start, double slope, double mul, double add) + { + return (x <= start ? x*slope : exp(log(x) / gamma) * mul - add); + } + static inline double igamma (double x, double gamma, double start, double slope, double mul, double add) + { + return (x <= start * slope ? x / slope : exp(log((x + add) / mul) * gamma) ); + } + static inline float gamma (float x, float gamma, float start, float slope, float mul, float add) + { + return (x <= start ? x*slope : xexpf(xlogf(x) / gamma) * mul - add); + } + static inline float igamma (float x, float gamma, float start, float slope, float mul, float add) + { + return (x <= start * slope ? x / slope : xexpf(xlogf((x + add) / mul) * gamma) ); + } +#ifdef __SSE2__ + static inline vfloat igamma (vfloat x, vfloat gamma, vfloat start, vfloat slope, vfloat mul, vfloat add) + { +#if !defined(__clang__) + return (x <= start * slope ? x / slope : xexpf(xlogf((x + add) / mul) * gamma) ); +#else + return vself(vmaskf_le(x, start * slope), x / slope, xexpf(xlogf((x + add) / mul) * gamma)); +#endif + } +#endif + static inline float hlcurve (const float exp_scale, const float comp, const float hlrange, float level) + { + if (comp > 0.0) { + float val = level + (hlrange - 65536.0); + + if(val == 0.0f) { // to avoid division by zero + val = 0.000001f; + } + + float Y = val * exp_scale / hlrange; + Y *= comp; + + if(Y <= -1.0) { // to avoid log(<=0) + Y = -.999999f; + } + + float R = hlrange / (val * comp); + return log1p(Y) * R; + } else { + return exp_scale; + } + } + +public: + static void complexCurve (double ecomp, double black, double hlcompr, double hlcomprthresh, double shcompr, double br, double contr, + const std::vector& curvePoints, const std::vector& curvePoints2, + LUTu & histogram, LUTf & hlCurve, LUTf & shCurve, LUTf & outCurve, LUTu & outBeforeCCurveHistogram, ToneCurve & outToneCurve, ToneCurve & outToneCurve2, + + int skip = 1); + static void curveBW (const std::vector& curvePointsbw, const std::vector& curvePointsbw2, const LUTu & histogrambw, LUTu & outBeforeCCurveHistogrambw, + ToneCurve & customToneCurvebw1, ToneCurve & customToneCurvebw2, int skip); + + static void curveCL ( bool & clcutili, const std::vector& clcurvePoints, LUTf & clCurve, int skip); + + static void curveWavContL ( bool & wavcontlutili, const std::vector& wavclcurvePoints, LUTf & wavclCurve,/* LUTu & histogramwavcl, LUTu & outBeforeWavCLurveHistogram,*/int skip); + static void curveDehaContL ( bool & dehacontlutili, const std::vector& dehaclcurvePoints, LUTf & dehaclCurve, int skip, const LUTu & histogram, LUTu & outBeforeCurveHistogram); + static void mapcurve ( bool & mapcontlutili, const std::vector& mapcurvePoints, LUTf & mapcurve, int skip, const LUTu & histogram, LUTu & outBeforeCurveHistogram); + + static void curveToning ( const std::vector& curvePoints, LUTf & ToningCurve, int skip); + + static void complexsgnCurve ( bool & autili, bool & butili, bool & ccutili, bool & clcutili, const std::vector& acurvePoints, + const std::vector& bcurvePoints, const std::vector& cccurvePoints, const std::vector& lccurvePoints, LUTf & aoutCurve, LUTf & boutCurve, LUTf & satCurve, LUTf & lhskCurve, + int skip = 1); + static void complexLCurve (double br, double contr, const std::vector& curvePoints, const LUTu & histogram, LUTf & outCurve, LUTu & outBeforeCCurveHistogram, int skip, bool & utili); + + static void curveLightBrightColor ( + const std::vector& curvePoints, + const std::vector& curvePoints2, + const std::vector& curvePoints3, + const LUTu & histogram, LUTu & outBeforeCCurveHistogram, + const LUTu & histogramC, LUTu & outBeforeCCurveHistogramC, + ColorAppearance & outColCurve1, + ColorAppearance & outColCurve2, + ColorAppearance & outColCurve3, + int skip = 1); + static void RGBCurve (const std::vector& curvePoints, LUTf & outCurve, int skip); + +}; + +class Curve +{ + + class HashEntry + { + public: + unsigned short smallerValue; + unsigned short higherValue; + }; +protected: + int N; + int ppn; // targeted polyline point number + double* x; + double* y; + // begin of variables used in Parametric curves only + double mc; + double mfc; + double msc; + double mhc; + // end of variables used in Parametric curves only + std::vector poly_x; // X points of the faceted curve + std::vector poly_y; // Y points of the faceted curve + std::vector dyByDx; + std::vector hash; + unsigned short hashSize; // hash table's size, between [10, 100, 1000] + + double* ypp; + + // Fields for the elementary curve polygonisation + double x1, y1, x2, y2, x3, y3; + bool firstPointIncluded; + double increment; + int nbr_points; + + static inline double p00 (double x, double prot) + { + return CurveFactory::clower (x, 2.0, prot); + } + static inline double p11 (double x, double prot) + { + return CurveFactory::cupper (x, 2.0, prot); + } + static inline double p01 (double x, double prot) + { + return x <= 0.5 ? CurveFactory::clower (x * 2, 2.0, prot) * 0.5 : 0.5 + CurveFactory::cupper ((x - 0.5) * 2, 2.0, prot) * 0.5; + } + static inline double p10 (double x, double prot) + { + return x <= 0.5 ? CurveFactory::cupper (x * 2, 2.0, prot) * 0.5 : 0.5 + CurveFactory::clower ((x - 0.5) * 2, 2.0, prot) * 0.5; + } + static inline double pfull (double x, double prot, double sh, double hl) + { + return (1 - sh) * (1 - hl) * p00(x, prot) + sh * hl * p11(x, prot) + (1 - sh) * hl * p01(x, prot) + sh * (1 - hl) * p10(x, prot); + } + + void fillHash(); + void fillDyByDx(); + +public: + Curve (); + virtual ~Curve () {}; + void AddPolygons (); + int getSize () const; // return the number of control points + void getControlPoint(int cpNum, double &x, double &y) const; + virtual double getVal (double t) const = 0; + virtual void getVal (const std::vector& t, std::vector& res) const = 0; + + virtual bool isIdentity () const = 0; +}; + +class DiagonalCurve : public Curve +{ + +protected: + DiagonalCurveType kind; + + void spline_cubic_set (); + void NURBS_set (); + +public: + DiagonalCurve (const std::vector& points, int ppn = CURVES_MIN_POLY_POINTS); + virtual ~DiagonalCurve (); + + double getVal (double t) const; + void getVal (const std::vector& t, std::vector& res) const; + bool isIdentity () const + { + return kind == DCT_Empty; + }; +}; + +class FlatCurve : public Curve +{ + +private: + FlatCurveType kind; + double* leftTangent; + double* rightTangent; + double identityValue; + bool periodic; + + void CtrlPoints_set (); + +public: + + FlatCurve (const std::vector& points, bool isPeriodic = true, int ppn = CURVES_MIN_POLY_POINTS); + virtual ~FlatCurve (); + + double getVal (double t) const; + void getVal (const std::vector& t, std::vector& res) const; + bool setIdentityValue (double iVal); + bool isIdentity () const + { + return kind == FCT_Empty; + }; +}; + +class RetinextransmissionCurve +{ +private: + LUTf luttransmission; // 0xffff range + void Set(const Curve &pCurve); + +public: + virtual ~RetinextransmissionCurve() {}; + RetinextransmissionCurve(); + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return luttransmission[index]; + } + + operator bool (void) const + { + return luttransmission; + } +}; + +class RetinexgaintransmissionCurve +{ +private: + LUTf lutgaintransmission; // 0xffff range + void Set(const Curve &pCurve); + +public: + virtual ~RetinexgaintransmissionCurve() {}; + RetinexgaintransmissionCurve(); + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return lutgaintransmission[index]; + } + + operator bool (void) const + { + return lutgaintransmission; + } +}; + + + +class ToneCurve +{ +public: + LUTf lutToneCurve; // 0xffff range + + virtual ~ToneCurve() {}; + + void Reset(); + void Set(const Curve &pCurve, float gamma = 0); + operator bool (void) const + { + return lutToneCurve; + } +}; + +class OpacityCurve +{ +public: + LUTf lutOpacityCurve; // 0xffff range + + virtual ~OpacityCurve() {}; + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints, bool &opautili); + + // TODO: transfer this method to the Color class... + float blend (float x, float lower, float upper) const + { + return (upper - lower) * lutOpacityCurve[x * 500.f] + lower; + } + void blend3f (float x, float lower1, float upper1, float &result1, float lower2, float upper2, float &result2, float lower3, float upper3, float &result3) const + { + float opacity = lutOpacityCurve[x * 500.f]; + result1 = (upper1 - lower1) * opacity + lower1; + result2 = (upper2 - lower2) * opacity + lower2; + result3 = (upper3 - lower3) * opacity + lower3; + } + + operator bool (void) const + { + return lutOpacityCurve; + } +}; + +class WavCurve +{ +private: + LUTf lutWavCurve; // 0xffff range + void Set(const Curve &pCurve); + +public: + float sum; + + virtual ~WavCurve() {}; + WavCurve(); + void Reset(); + void Set(const std::vector &curvePoints); + float getSum() const + { + return sum; + } + + float operator[](float index) const + { + return lutWavCurve[index]; + } + operator bool (void) const + { + return lutWavCurve; + } +}; + +class WavOpacityCurveRG +{ +private: + LUTf lutOpacityCurveRG; // 0xffff range + void Set(const Curve &pCurve); +public: + virtual ~WavOpacityCurveRG() {}; + WavOpacityCurveRG(); + + void Reset(); + // void Set(const std::vector &curvePoints, bool &opautili); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return lutOpacityCurveRG[index]; + } + + operator bool (void) const + { + return lutOpacityCurveRG; + } +}; +class WavOpacityCurveBY +{ +private: + LUTf lutOpacityCurveBY; // 0xffff range + void Set(const Curve &pCurve); + +public: + virtual ~WavOpacityCurveBY() {}; + WavOpacityCurveBY(); + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return lutOpacityCurveBY[index]; + } + + operator bool (void) const + { + return lutOpacityCurveBY; + } +}; +class WavOpacityCurveW +{ +private: + LUTf lutOpacityCurveW; // 0xffff range + void Set(const Curve &pCurve); + +public: + virtual ~WavOpacityCurveW() {}; + WavOpacityCurveW(); + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return lutOpacityCurveW[index]; + } + + operator bool (void) const + { + return lutOpacityCurveW; + } +}; + +class WavOpacityCurveWL +{ +private: + LUTf lutOpacityCurveWL; // 0xffff range + void Set(const Curve &pCurve); + +public: + virtual ~WavOpacityCurveWL() {}; + WavOpacityCurveWL(); + + void Reset(); + void Set(const Curve *pCurve); + void Set(const std::vector &curvePoints); + float operator[](float index) const + { + return lutOpacityCurveWL[index]; + } + + operator bool (void) const + { + return lutOpacityCurveWL; + } +}; + +class NoiseCurve +{ +private: + LUTf lutNoiseCurve; // 0xffff range + float sum; + void Set(const Curve &pCurve); + +public: + virtual ~NoiseCurve() {}; + NoiseCurve(); + void Reset(); + void Set(const std::vector &curvePoints); + + float getSum() const + { + return sum; + } + float operator[](float index) const + { + return lutNoiseCurve[index]; + } + operator bool (void) const + { + return lutNoiseCurve; + } +}; + +class ColorGradientCurve +{ +public: + LUTf lut1; // [0.;1.] range (float values) + LUTf lut2; // [0.;1.] range (float values) + LUTf lut3; // [0.;1.] range (float values) + double low; + double high; + + virtual ~ColorGradientCurve() {}; + + void Reset(); + void SetXYZ(const Curve *pCurve, const double xyz_rgb[3][3], float satur, float lumin); + void SetXYZ(const std::vector &curvePoints, const double xyz_rgb[3][3], float satur, float lumin); + void SetRGB(const Curve *pCurve); + void SetRGB(const std::vector &curvePoints); + + /** + * @brief Get the value of Red, Green and Blue corresponding to the requested index + * @param index value in the [0 ; 1] range + * @param r corresponding red value [0 ; 65535] (return value) + * @param g corresponding green value [0 ; 65535] (return value) + * @param b corresponding blue value [0 ; 65535] (return value) + */ + void getVal(float index, float &r, float &g, float &b) const; + operator bool (void) const + { + return lut1 && lut2 && lut3; + } +}; + +class ColorAppearance +{ +public: + LUTf lutColCurve; // 0xffff range + + virtual ~ColorAppearance() {}; + + void Reset(); + void Set(const Curve &pCurve); + operator bool (void) const + { + return lutColCurve; + } +}; + +class Lightcurve : public ColorAppearance +{ +public: + void Apply(float& Li) const; +}; + +//lightness curve +inline void Lightcurve::Apply (float& Li) const +{ + + assert (lutColCurve); + + Li = lutColCurve[Li]; +} + +class Brightcurve : public ColorAppearance +{ +public: + void Apply(float& Br) const; +}; + +//brightness curve +inline void Brightcurve::Apply (float& Br) const +{ + + assert (lutColCurve); + + Br = lutColCurve[Br]; +} + +class Chromacurve : public ColorAppearance +{ +public: + void Apply(float& Cr) const; +}; + +//Chroma curve +inline void Chromacurve::Apply (float& Cr) const +{ + + assert (lutColCurve); + + Cr = lutColCurve[Cr]; +} +class Saturcurve : public ColorAppearance +{ +public: + void Apply(float& Sa) const; +}; + +//Saturation curve +inline void Saturcurve::Apply (float& Sa) const +{ + + assert (lutColCurve); + + Sa = lutColCurve[Sa]; +} + +class Colorfcurve : public ColorAppearance +{ +public: + void Apply(float& Cf) const; +}; + +//Colorfullness curve +inline void Colorfcurve::Apply (float& Cf) const +{ + + assert (lutColCurve); + + Cf = lutColCurve[Cf]; +} + + +class StandardToneCurve : public ToneCurve +{ +public: + void Apply(float& r, float& g, float& b) const; + + // Applies the tone curve to `r`, `g`, `b` arrays, starting at `r[start]` + // and ending at `r[end]` (and respectively for `b` and `g`). Uses SSE + // and requires that `r`, `g`, and `b` pointers have the same alignment. + void BatchApply( + const size_t start, const size_t end, + float *r, float *g, float *b) const; +}; + +class AdobeToneCurve : public ToneCurve +{ +private: + void RGBTone(float& r, float& g, float& b) const; // helper for tone curve + +public: + void Apply(float& r, float& g, float& b) const; +}; + +class SatAndValueBlendingToneCurve : public ToneCurve +{ +public: + void Apply(float& r, float& g, float& b) const; +}; + +class WeightedStdToneCurve : public ToneCurve +{ +private: + float Triangle(float refX, float refY, float X2) const; +#if defined( __SSE2__ ) && defined( __x86_64__ ) + vfloat Triangle(vfloat refX, vfloat refY, vfloat X2) const; +#endif +public: + void Apply(float& r, float& g, float& b) const; + void BatchApply(const size_t start, const size_t end, float *r, float *g, float *b) const; +}; + +class LuminanceToneCurve : public ToneCurve +{ +public: + void Apply(float& r, float& g, float& b) const; +}; + +class PerceptualToneCurveState +{ +public: + float Working2Prophoto[3][3]; + float Prophoto2Working[3][3]; + float cmul_contrast; + bool isProphoto; +}; + +// Tone curve whose purpose is to keep the color appearance constant, that is the curve changes contrast +// but colors appears to have the same hue and saturation as before. As contrast and saturation is tightly +// coupled in human vision saturation is modulated based on the curve's contrast, and that way the appearance +// can be kept perceptually constant (within limits). +class PerceptualToneCurve : public ToneCurve +{ +private: + static float cf_range[2]; + static float cf[1000]; + // for ciecam02 + static float f, c, nc, yb, la, xw, yw, zw, gamut; + static float n, d, nbb, ncb, cz, aw, wh, pfl, fl, pow1; + + static void cubic_spline(const float x[], const float y[], const int len, const float out_x[], float out_y[], const int out_len); + static float find_minimum_interval_halving(float (*func)(float x, void *arg), void *arg, float a, float b, float tol, int nmax); + static float find_tc_slope_fun(float k, void *arg); + static float get_curve_val(float x, float range[2], float lut[], size_t lut_size); + float calculateToneCurveContrastValue() const; +public: + static void init(); + void initApplyState(PerceptualToneCurveState & state, Glib::ustring workingSpace) const; + void BatchApply(const size_t start, const size_t end, float *r, float *g, float *b, const PerceptualToneCurveState &state) const; +}; + +// Standard tone curve +inline void StandardToneCurve::Apply (float& r, float& g, float& b) const +{ + + assert (lutToneCurve); + + r = lutToneCurve[r]; + g = lutToneCurve[g]; + b = lutToneCurve[b]; +} + +inline void StandardToneCurve::BatchApply( + const size_t start, const size_t end, + float *r, float *g, float *b) const { + assert (lutToneCurve); + assert (lutToneCurve.getClip() & LUT_CLIP_BELOW); + assert (lutToneCurve.getClip() & LUT_CLIP_ABOVE); + + // All pointers must have the same alignment for SSE usage. In the loop body below, + // we will only check `r`, assuming that the same result would hold for `g` and `b`. + assert (reinterpret_cast(r) % 16 == reinterpret_cast(g) % 16); + assert (reinterpret_cast(g) % 16 == reinterpret_cast(b) % 16); + + size_t i = start; + while (true) { + if (i >= end) { + // If we get to the end before getting to an aligned address, just return. + // (Or, for non-SSE mode, if we get to the end.) + return; +#if defined( __SSE2__ ) && defined( __x86_64__ ) + } else if (reinterpret_cast(&r[i]) % 16 == 0) { + // Otherwise, we get to the first aligned address; go to the SSE part. + break; +#endif + } + r[i] = lutToneCurve[r[i]]; + g[i] = lutToneCurve[g[i]]; + b[i] = lutToneCurve[b[i]]; + i++; + } + +#if defined( __SSE2__ ) && defined( __x86_64__ ) + for (; i + 3 < end; i += 4) { + __m128 r_val = LVF(r[i]); + __m128 g_val = LVF(g[i]); + __m128 b_val = LVF(b[i]); + STVF(r[i], lutToneCurve[r_val]); + STVF(g[i], lutToneCurve[g_val]); + STVF(b[i], lutToneCurve[b_val]); + } + + // Remainder in non-SSE. + for (; i < end; ++i) { + r[i] = lutToneCurve[r[i]]; + g[i] = lutToneCurve[g[i]]; + b[i] = lutToneCurve[b[i]]; + } +#endif +} + +// Tone curve according to Adobe's reference implementation +// values in 0xffff space +// inlined to make sure there will be no cache flush when used +inline void AdobeToneCurve::Apply (float& r, float& g, float& b) const +{ + + assert (lutToneCurve); + + if (r >= g) { + if (g > b) { + RGBTone (r, g, b); // Case 1: r >= g > b + } else if (b > r) { + RGBTone (b, r, g); // Case 2: b > r >= g + } else if (b > g) { + RGBTone (r, b, g); // Case 3: r >= b > g + } else { // Case 4: r >= g == b + r = lutToneCurve[r]; + g = lutToneCurve[g]; + b = g; + } + } else { + if (r >= b) { + RGBTone (g, r, b); // Case 5: g > r >= b + } else if (b > g) { + RGBTone (b, g, r); // Case 6: b > g > r + } else { + RGBTone (g, b, r); // Case 7: g >= b > r + } + } +} + +inline void AdobeToneCurve::RGBTone (float& r, float& g, float& b) const +{ + float rold = r, gold = g, bold = b; + + r = lutToneCurve[rold]; + b = lutToneCurve[bold]; + g = b + ((r - b) * (gold - bold) / (rold - bold)); +} + +// Modifying the Luminance channel only +inline void LuminanceToneCurve::Apply(float &r, float &g, float &b) const +{ + assert (lutToneCurve); + + float currLuminance = r * 0.2126729f + g * 0.7151521f + b * 0.0721750f; + const float newLuminance = lutToneCurve[currLuminance]; + currLuminance = currLuminance == 0.f ? 0.00001f : currLuminance; + const float coef = newLuminance / currLuminance; + r = LIM(r * coef, 0.f, 65535.f); + g = LIM(g * coef, 0.f, 65535.f); + b = LIM(b * coef, 0.f, 65535.f); +} + +inline float WeightedStdToneCurve::Triangle(float a, float a1, float b) const +{ + if (a != b) { + float b1; + float a2 = a1 - a; + + if (b < a) { + b1 = b + a2 * b / a ; + } else { + b1 = b + a2 * (65535.f - b) / (65535.f - a); + } + + return b1; + } + + return a1; +} + +#if defined( __SSE2__ ) && defined( __x86_64__ ) +inline vfloat WeightedStdToneCurve::Triangle(vfloat a, vfloat a1, vfloat b) const +{ + vfloat a2 = a1 - a; + vmask cmask = vmaskf_lt(b, a); + vfloat b3 = vself(cmask, b, F2V(65535.f) - b); + vfloat a3 = vself(cmask, a, F2V(65535.f) - a); + return b + a2 * b3 / a3; +} +#endif + +// Tone curve modifying the value channel only, preserving hue and saturation +// values in 0xffff space +inline void WeightedStdToneCurve::Apply (float& r, float& g, float& b) const +{ + + assert (lutToneCurve); + + r = CLIP(r); + g = CLIP(g); + b = CLIP(b); + float r1 = lutToneCurve[r]; + float g1 = Triangle(r, r1, g); + float b1 = Triangle(r, r1, b); + + float g2 = lutToneCurve[g]; + float r2 = Triangle(g, g2, r); + float b2 = Triangle(g, g2, b); + + float b3 = lutToneCurve[b]; + float r3 = Triangle(b, b3, r); + float g3 = Triangle(b, b3, g); + + r = CLIP(r1 * 0.50f + r2 * 0.25f + r3 * 0.25f); + g = CLIP(g1 * 0.25f + g2 * 0.50f + g3 * 0.25f); + b = CLIP(b1 * 0.25f + b2 * 0.25f + b3 * 0.50f); +} + +inline void WeightedStdToneCurve::BatchApply(const size_t start, const size_t end, float *r, float *g, float *b) const { + assert (lutToneCurve); + assert (lutToneCurve.getClip() & LUT_CLIP_BELOW); + assert (lutToneCurve.getClip() & LUT_CLIP_ABOVE); + + // All pointers must have the same alignment for SSE usage. In the loop body below, + // we will only check `r`, assuming that the same result would hold for `g` and `b`. + assert (reinterpret_cast(r) % 16 == reinterpret_cast(g) % 16); + assert (reinterpret_cast(g) % 16 == reinterpret_cast(b) % 16); + + size_t i = start; + while (true) { + if (i >= end) { + // If we get to the end before getting to an aligned address, just return. + // (Or, for non-SSE mode, if we get to the end.) + return; +#if defined( __SSE2__ ) && defined( __x86_64__ ) + } else if (reinterpret_cast(&r[i]) % 16 == 0) { + // Otherwise, we get to the first aligned address; go to the SSE part. + break; +#endif + } + Apply(r[i], g[i], b[i]); + i++; + } + +#if defined( __SSE2__ ) && defined( __x86_64__ ) + const vfloat c65535v = F2V(65535.f); + const vfloat zd5v = F2V(0.5f); + const vfloat zd25v = F2V(0.25f); + + for (; i + 3 < end; i += 4) { + vfloat r_val = LIMV(LVF(r[i]), ZEROV, c65535v); + vfloat g_val = LIMV(LVF(g[i]), ZEROV, c65535v); + vfloat b_val = LIMV(LVF(b[i]), ZEROV, c65535v); + vfloat r1 = lutToneCurve[r_val]; + vfloat g1 = Triangle(r_val, r1, g_val); + vfloat b1 = Triangle(r_val, r1, b_val); + + vfloat g2 = lutToneCurve[g_val]; + vfloat r2 = Triangle(g_val, g2, r_val); + vfloat b2 = Triangle(g_val, g2, b_val); + + vfloat b3 = lutToneCurve[b_val]; + vfloat r3 = Triangle(b_val, b3, r_val); + vfloat g3 = Triangle(b_val, b3, g_val); + + STVF(r[i], LIMV(r1 * zd5v + r2 * zd25v + r3 * zd25v, ZEROV, c65535v)); + STVF(g[i], LIMV(g1 * zd25v + g2 * zd5v + g3 * zd25v, ZEROV, c65535v)); + STVF(b[i], LIMV(b1 * zd25v + b2 * zd25v + b3 * zd5v, ZEROV, c65535v)); + } + + // Remainder in non-SSE. + for (; i < end; ++i) { + Apply(r[i], g[i], b[i]); + } +#endif +} + +// Tone curve modifying the value channel only, preserving hue and saturation +// values in 0xffff space +inline void SatAndValueBlendingToneCurve::Apply (float& r, float& g, float& b) const +{ + + assert (lutToneCurve); + + r = CLIP(r); + g = CLIP(g); + b = CLIP(b); + + const float lum = (r + g + b) / 3.f; + const float newLum = lutToneCurve[lum]; + + float h, s, v; + Color::rgb2hsvtc(r, g, b, h, s, v); + + float dV; + if (newLum >= lum) { + // Linearly targeting Value = 1 and Saturation = 0 + const float coef = (newLum - lum) / (65535.f - lum); + dV = (1.f - v) * coef; + s *= 1.f - coef; + } else { + // Linearly targeting Value = 0 + const float coef = (newLum - lum) / lum ; + dV = v * coef; + } + Color::hsv2rgbdcp(h, s, v + dV, r, g, b); +} + +} + +#undef CLIPI + +#endif diff --git a/rtengine/rt_math.h b/rtengine/rt_math.h index 4bdfe6f06..787e2e63f 100644 --- a/rtengine/rt_math.h +++ b/rtengine/rt_math.h @@ -25,6 +25,8 @@ constexpr double RT_NAN = std::numeric_limits::quiet_NaN(); constexpr float RT_PI_F = RT_PI; constexpr float RT_PI_F_2 = RT_PI_2; constexpr float RT_PI_F_180 = RT_PI_180; +constexpr float RT_1_PI_F = RT_1_PI; +constexpr float RT_2_PI_F = RT_2_PI; constexpr float RT_INFINITY_F = std::numeric_limits::infinity(); constexpr float RT_NAN_F = std::numeric_limits::quiet_NaN(); diff --git a/rtengine/sleef.c b/rtengine/sleef.c index 17dfccc0f..f03c9f1b3 100644 --- a/rtengine/sleef.c +++ b/rtengine/sleef.c @@ -923,9 +923,8 @@ __inline float mulsignf(float x, float y) { return intBitsToFloat(floatToRawIntBits(x) ^ (floatToRawIntBits(y) & (1 << 31))); } -__inline float signf(float d) { return mulsignf(1, d); } +__inline float signf(float d) { return copysign(1, d); } __inline float mlaf(float x, float y, float z) { return x * y + z; } -__inline float xrintf(float x) { return x < 0 ? (int)(x - 0.5f) : (int)(x + 0.5f); } __inline int xisnanf(float x) { return x != x; } __inline int xisinff(float x) { return x == rtengine::RT_INFINITY_F || x == -rtengine::RT_INFINITY_F; } @@ -984,7 +983,7 @@ __inline float xsinf(float d) { int q; float u, s; - q = (int)xrintf(d * (float)rtengine::RT_1_PI); + q = rint(d * rtengine::RT_1_PI_F); d = mlaf(q, -PI4_Af*4, d); d = mlaf(q, -PI4_Bf*4, d); @@ -1009,7 +1008,7 @@ __inline float xcosf(float d) { int q; float u, s; - q = 1 + 2*(int)xrintf(d * (float)rtengine::RT_1_PI - 0.5f); + q = 1 + 2*rint(d * rtengine::RT_1_PI_F - 0.5f); d = mlaf(q, -PI4_Af*2, d); d = mlaf(q, -PI4_Bf*2, d); @@ -1035,7 +1034,7 @@ __inline float2 xsincosf(float d) { float u, s, t; float2 r; - q = (int)rint(d * ((float)(2 * rtengine::RT_1_PI))); + q = rint(d * rtengine::RT_2_PI_F); s = d; @@ -1076,7 +1075,7 @@ __inline float xtanf(float d) { int q; float u, s, x; - q = (int)xrintf(d * (float)(2 * rtengine::RT_1_PI)); + q = rint(d * (float)(2 * rtengine::RT_1_PI)); x = d; @@ -1202,7 +1201,7 @@ __inline float xlogf(float d) { __inline float xexpf(float d) { if(d<=-104.0f) return 0.0f; - int q = (int)xrintf(d * R_LN2f); + int q = rint(d * R_LN2f); float s, u; s = mlaf(q, -L2Uf, d);