/* ======================================== * Hermepass - Hermepass.h * Copyright (c) 2016 airwindows, All rights reserved * ======================================== */ #ifndef __Hermepass_H #include "Hermepass.h" #endif void Hermepass::processReplacing(float **inputs, float **outputs, VstInt32 sampleFrames) { float* in1 = inputs[0]; float* in2 = inputs[1]; float* out1 = outputs[0]; float* out2 = outputs[1]; double overallscale = 1.0; overallscale /= 44100.0; overallscale *= getSampleRate(); long double fpOld = 0.618033988749894848204586; //golden ratio! long double fpNew = 1.0 - fpOld; double rangescale = 0.1 / overallscale; double cutoff = pow(A,3); double slope = pow(B,3) * 6.0; double newA = cutoff * rangescale; double newB = newA; //other part of interleaved IIR is the same double newC = cutoff * rangescale; //first extra pole is the same cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newD = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newE = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newF = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newG = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newH = cutoff * rangescale; //converge toward the unvarying fixed cutoff value double oldA = 1.0 - newA; double oldB = 1.0 - newB; double oldC = 1.0 - newC; double oldD = 1.0 - newD; double oldE = 1.0 - newE; double oldF = 1.0 - newF; double oldG = 1.0 - newG; double oldH = 1.0 - newH; double polesC; double polesD; double polesE; double polesF; double polesG; double polesH; polesC = slope; if (slope > 1.0) polesC = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesD = slope; if (slope > 1.0) polesD = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesE = slope; if (slope > 1.0) polesE = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesF = slope; if (slope > 1.0) polesF = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesG = slope; if (slope > 1.0) polesG = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesH = slope; if (slope > 1.0) polesH = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; //each one will either be 0.0, the fractional slope value, or 1 long double inputSampleL; long double inputSampleR; double tempSampleL; double tempSampleR; double correction; while (--sampleFrames >= 0) { inputSampleL = *in1; inputSampleR = *in2; if (inputSampleL<1.2e-38 && -inputSampleL<1.2e-38) { static int noisesource = 0; //this declares a variable before anything else is compiled. It won't keep assigning //it to 0 for every sample, it's as if the declaration doesn't exist in this context, //but it lets me add this denormalization fix in a single place rather than updating //it in three different locations. The variable isn't thread-safe but this is only //a random seed and we can share it with whatever. noisesource = noisesource % 1700021; noisesource++; int residue = noisesource * noisesource; residue = residue % 170003; residue *= residue; residue = residue % 17011; residue *= residue; residue = residue % 1709; residue *= residue; residue = residue % 173; residue *= residue; residue = residue % 17; double applyresidue = residue; applyresidue *= 0.00000001; applyresidue *= 0.00000001; inputSampleL = applyresidue; } if (inputSampleR<1.2e-38 && -inputSampleR<1.2e-38) { static int noisesource = 0; noisesource = noisesource % 1700021; noisesource++; int residue = noisesource * noisesource; residue = residue % 170003; residue *= residue; residue = residue % 17011; residue *= residue; residue = residue % 1709; residue *= residue; residue = residue % 173; residue *= residue; residue = residue % 17; double applyresidue = residue; applyresidue *= 0.00000001; applyresidue *= 0.00000001; inputSampleR = applyresidue; //this denormalization routine produces a white noise at -300 dB which the noise //shaping will interact with to produce a bipolar output, but the noise is actually //all positive. That should stop any variables from going denormal, and the routine //only kicks in if digital black is input. As a final touch, if you save to 24-bit //the silence will return to being digital black again. } tempSampleL = inputSampleL; tempSampleR = inputSampleR; //begin L channel if (fpFlip) { iirAL = (iirAL * oldA) + (tempSampleL * newA); tempSampleL -= iirAL; correction = iirAL; } else { iirBL = (iirBL * oldB) + (tempSampleL * newB); tempSampleL -= iirBL; correction = iirBL; } iirCL = (iirCL * oldC) + (tempSampleL * newC); tempSampleL -= iirCL; iirDL = (iirDL * oldD) + (tempSampleL * newD); tempSampleL -= iirDL; iirEL = (iirEL * oldE) + (tempSampleL * newE); tempSampleL -= iirEL; iirFL = (iirFL * oldF) + (tempSampleL * newF); tempSampleL -= iirFL; iirGL = (iirGL * oldG) + (tempSampleL * newG); tempSampleL -= iirGL; iirHL = (iirHL * oldH) + (tempSampleL * newH); tempSampleL -= iirHL; //set up all the iir filters in case they are used if (polesC == 1.0) correction += iirCL; if (polesC > 0.0 && polesC < 1.0) correction += (iirCL * polesC); if (polesD == 1.0) correction += iirDL; if (polesD > 0.0 && polesD < 1.0) correction += (iirDL * polesD); if (polesE == 1.0) correction += iirEL; if (polesE > 0.0 && polesE < 1.0) correction += (iirEL * polesE); if (polesF == 1.0) correction += iirFL; if (polesF > 0.0 && polesF < 1.0) correction += (iirFL * polesF); if (polesG == 1.0) correction += iirGL; if (polesG > 0.0 && polesG < 1.0) correction += (iirGL * polesG); if (polesH == 1.0) correction += iirHL; if (polesH > 0.0 && polesH < 1.0) correction += (iirHL * polesH); //each of these are added directly if they're fully engaged, //multiplied by 0-1 if they are the interpolated one, or skipped if they are beyond the interpolated one. //the purpose is to do all the math at the floating point exponent nearest to the tiny value in use. //also, it's formatted that way to easily substitute the next variable: this could be written as a loop //with everything an array value. However, this makes just as much sense for this few poles. inputSampleL -= correction; //end L channel //begin R channel if (fpFlip) { iirAR = (iirAR * oldA) + (tempSampleR * newA); tempSampleR -= iirAR; correction = iirAR; } else { iirBR = (iirBR * oldB) + (tempSampleR * newB); tempSampleR -= iirBR; correction = iirBR; } iirCR = (iirCR * oldC) + (tempSampleR * newC); tempSampleR -= iirCR; iirDR = (iirDR * oldD) + (tempSampleR * newD); tempSampleR -= iirDR; iirER = (iirER * oldE) + (tempSampleR * newE); tempSampleR -= iirER; iirFR = (iirFR * oldF) + (tempSampleR * newF); tempSampleR -= iirFR; iirGR = (iirGR * oldG) + (tempSampleR * newG); tempSampleR -= iirGR; iirHR = (iirHR * oldH) + (tempSampleR * newH); tempSampleR -= iirHR; //set up all the iir filters in case they are used if (polesC == 1.0) correction += iirCR; if (polesC > 0.0 && polesC < 1.0) correction += (iirCR * polesC); if (polesD == 1.0) correction += iirDR; if (polesD > 0.0 && polesD < 1.0) correction += (iirDR * polesD); if (polesE == 1.0) correction += iirER; if (polesE > 0.0 && polesE < 1.0) correction += (iirER * polesE); if (polesF == 1.0) correction += iirFR; if (polesF > 0.0 && polesF < 1.0) correction += (iirFR * polesF); if (polesG == 1.0) correction += iirGR; if (polesG > 0.0 && polesG < 1.0) correction += (iirGR * polesG); if (polesH == 1.0) correction += iirHR; if (polesH > 0.0 && polesH < 1.0) correction += (iirHR * polesH); //each of these are added directly if they're fully engaged, //multiplied by 0-1 if they are the interpolated one, or skipped if they are beyond the interpolated one. //the purpose is to do all the math at the floating point exponent nearest to the tiny value in use. //also, it's formatted that way to easily substitute the next variable: this could be written as a loop //with everything an array value. However, this makes just as much sense for this few poles. inputSampleR -= correction; //end R channel fpFlip = !fpFlip; //stereo 32 bit dither, made small and tidy. int expon; frexpf((float)inputSampleL, &expon); long double dither = (rand()/(RAND_MAX*7.737125245533627e+25))*pow(2,expon+62); inputSampleL += (dither-fpNShapeL); fpNShapeL = dither; frexpf((float)inputSampleR, &expon); dither = (rand()/(RAND_MAX*7.737125245533627e+25))*pow(2,expon+62); inputSampleR += (dither-fpNShapeR); fpNShapeR = dither; //end 32 bit dither *out1 = inputSampleL; *out2 = inputSampleR; *in1++; *in2++; *out1++; *out2++; } } void Hermepass::processDoubleReplacing(double **inputs, double **outputs, VstInt32 sampleFrames) { double* in1 = inputs[0]; double* in2 = inputs[1]; double* out1 = outputs[0]; double* out2 = outputs[1]; double overallscale = 1.0; overallscale /= 44100.0; overallscale *= getSampleRate(); long double fpOld = 0.618033988749894848204586; //golden ratio! long double fpNew = 1.0 - fpOld; double rangescale = 0.1 / overallscale; double cutoff = pow(A,3); double slope = pow(B,3) * 6.0; double newA = cutoff * rangescale; double newB = newA; //other part of interleaved IIR is the same double newC = cutoff * rangescale; //first extra pole is the same cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newD = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newE = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newF = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newG = cutoff * rangescale; cutoff = (cutoff * fpOld) + (0.00001 * fpNew); double newH = cutoff * rangescale; //converge toward the unvarying fixed cutoff value double oldA = 1.0 - newA; double oldB = 1.0 - newB; double oldC = 1.0 - newC; double oldD = 1.0 - newD; double oldE = 1.0 - newE; double oldF = 1.0 - newF; double oldG = 1.0 - newG; double oldH = 1.0 - newH; double polesC; double polesD; double polesE; double polesF; double polesG; double polesH; polesC = slope; if (slope > 1.0) polesC = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesD = slope; if (slope > 1.0) polesD = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesE = slope; if (slope > 1.0) polesE = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesF = slope; if (slope > 1.0) polesF = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesG = slope; if (slope > 1.0) polesG = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; polesH = slope; if (slope > 1.0) polesH = 1.0; slope -= 1.0; if (slope < 0.0) slope = 0.0; //each one will either be 0.0, the fractional slope value, or 1 long double inputSampleL; long double inputSampleR; double tempSampleL; double tempSampleR; double correction; while (--sampleFrames >= 0) { inputSampleL = *in1; inputSampleR = *in2; if (inputSampleL<1.2e-38 && -inputSampleL<1.2e-38) { static int noisesource = 0; //this declares a variable before anything else is compiled. It won't keep assigning //it to 0 for every sample, it's as if the declaration doesn't exist in this context, //but it lets me add this denormalization fix in a single place rather than updating //it in three different locations. The variable isn't thread-safe but this is only //a random seed and we can share it with whatever. noisesource = noisesource % 1700021; noisesource++; int residue = noisesource * noisesource; residue = residue % 170003; residue *= residue; residue = residue % 17011; residue *= residue; residue = residue % 1709; residue *= residue; residue = residue % 173; residue *= residue; residue = residue % 17; double applyresidue = residue; applyresidue *= 0.00000001; applyresidue *= 0.00000001; inputSampleL = applyresidue; } if (inputSampleR<1.2e-38 && -inputSampleR<1.2e-38) { static int noisesource = 0; noisesource = noisesource % 1700021; noisesource++; int residue = noisesource * noisesource; residue = residue % 170003; residue *= residue; residue = residue % 17011; residue *= residue; residue = residue % 1709; residue *= residue; residue = residue % 173; residue *= residue; residue = residue % 17; double applyresidue = residue; applyresidue *= 0.00000001; applyresidue *= 0.00000001; inputSampleR = applyresidue; //this denormalization routine produces a white noise at -300 dB which the noise //shaping will interact with to produce a bipolar output, but the noise is actually //all positive. That should stop any variables from going denormal, and the routine //only kicks in if digital black is input. As a final touch, if you save to 24-bit //the silence will return to being digital black again. } tempSampleL = inputSampleL; tempSampleR = inputSampleR; //begin L channel if (fpFlip) { iirAL = (iirAL * oldA) + (tempSampleL * newA); tempSampleL -= iirAL; correction = iirAL; } else { iirBL = (iirBL * oldB) + (tempSampleL * newB); tempSampleL -= iirBL; correction = iirBL; } iirCL = (iirCL * oldC) + (tempSampleL * newC); tempSampleL -= iirCL; iirDL = (iirDL * oldD) + (tempSampleL * newD); tempSampleL -= iirDL; iirEL = (iirEL * oldE) + (tempSampleL * newE); tempSampleL -= iirEL; iirFL = (iirFL * oldF) + (tempSampleL * newF); tempSampleL -= iirFL; iirGL = (iirGL * oldG) + (tempSampleL * newG); tempSampleL -= iirGL; iirHL = (iirHL * oldH) + (tempSampleL * newH); tempSampleL -= iirHL; //set up all the iir filters in case they are used if (polesC == 1.0) correction += iirCL; if (polesC > 0.0 && polesC < 1.0) correction += (iirCL * polesC); if (polesD == 1.0) correction += iirDL; if (polesD > 0.0 && polesD < 1.0) correction += (iirDL * polesD); if (polesE == 1.0) correction += iirEL; if (polesE > 0.0 && polesE < 1.0) correction += (iirEL * polesE); if (polesF == 1.0) correction += iirFL; if (polesF > 0.0 && polesF < 1.0) correction += (iirFL * polesF); if (polesG == 1.0) correction += iirGL; if (polesG > 0.0 && polesG < 1.0) correction += (iirGL * polesG); if (polesH == 1.0) correction += iirHL; if (polesH > 0.0 && polesH < 1.0) correction += (iirHL * polesH); //each of these are added directly if they're fully engaged, //multiplied by 0-1 if they are the interpolated one, or skipped if they are beyond the interpolated one. //the purpose is to do all the math at the floating point exponent nearest to the tiny value in use. //also, it's formatted that way to easily substitute the next variable: this could be written as a loop //with everything an array value. However, this makes just as much sense for this few poles. inputSampleL -= correction; //end L channel //begin R channel if (fpFlip) { iirAR = (iirAR * oldA) + (tempSampleR * newA); tempSampleR -= iirAR; correction = iirAR; } else { iirBR = (iirBR * oldB) + (tempSampleR * newB); tempSampleR -= iirBR; correction = iirBR; } iirCR = (iirCR * oldC) + (tempSampleR * newC); tempSampleR -= iirCR; iirDR = (iirDR * oldD) + (tempSampleR * newD); tempSampleR -= iirDR; iirER = (iirER * oldE) + (tempSampleR * newE); tempSampleR -= iirER; iirFR = (iirFR * oldF) + (tempSampleR * newF); tempSampleR -= iirFR; iirGR = (iirGR * oldG) + (tempSampleR * newG); tempSampleR -= iirGR; iirHR = (iirHR * oldH) + (tempSampleR * newH); tempSampleR -= iirHR; //set up all the iir filters in case they are used if (polesC == 1.0) correction += iirCR; if (polesC > 0.0 && polesC < 1.0) correction += (iirCR * polesC); if (polesD == 1.0) correction += iirDR; if (polesD > 0.0 && polesD < 1.0) correction += (iirDR * polesD); if (polesE == 1.0) correction += iirER; if (polesE > 0.0 && polesE < 1.0) correction += (iirER * polesE); if (polesF == 1.0) correction += iirFR; if (polesF > 0.0 && polesF < 1.0) correction += (iirFR * polesF); if (polesG == 1.0) correction += iirGR; if (polesG > 0.0 && polesG < 1.0) correction += (iirGR * polesG); if (polesH == 1.0) correction += iirHR; if (polesH > 0.0 && polesH < 1.0) correction += (iirHR * polesH); //each of these are added directly if they're fully engaged, //multiplied by 0-1 if they are the interpolated one, or skipped if they are beyond the interpolated one. //the purpose is to do all the math at the floating point exponent nearest to the tiny value in use. //also, it's formatted that way to easily substitute the next variable: this could be written as a loop //with everything an array value. However, this makes just as much sense for this few poles. inputSampleR -= correction; //end R channel fpFlip = !fpFlip; //stereo 64 bit dither, made small and tidy. int expon; frexp((double)inputSampleL, &expon); long double dither = (rand()/(RAND_MAX*7.737125245533627e+25))*pow(2,expon+62); dither /= 536870912.0; //needs this to scale to 64 bit zone inputSampleL += (dither-fpNShapeL); fpNShapeL = dither; frexp((double)inputSampleR, &expon); dither = (rand()/(RAND_MAX*7.737125245533627e+25))*pow(2,expon+62); dither /= 536870912.0; //needs this to scale to 64 bit zone inputSampleR += (dither-fpNShapeR); fpNShapeR = dither; //end 64 bit dither *out1 = inputSampleL; *out2 = inputSampleR; *in1++; *in2++; *out1++; *out2++; } }