engines/gig/Voice.cpp

/***************************************************************************
 *                                                                         *
 *   LinuxSampler - modular, streaming capable sampler                     *
 *                                                                         *
 *   Copyright (C) 2003, 2004 by Benno Senoner and Christian Schoenebeck   *
 *   Copyright (C) 2005 Christian Schoenebeck                              *
 *                                                                         *
 *   This program 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 2 of the License, or     *
 *   (at your option) any later version.                                   *
 *                                                                         *
 *   This program 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 this program; if not, write to the Free Software           *
 *   Foundation, Inc., 59 Temple Place, Suite 330, Boston,                 *
 *   MA  02111-1307  USA                                                   *
 ***************************************************************************/

#include "../../common/Features.h"
#include "Synthesizer.h"
#include "Profiler.h"

#include "Voice.h"

namespace LinuxSampler { namespace gig {

    const float Voice::FILTER_CUTOFF_COEFF(CalculateFilterCutoffCoeff());

    float Voice::CalculateFilterCutoffCoeff() {
        return log(CONFIG_FILTER_CUTOFF_MAX / CONFIG_FILTER_CUTOFF_MIN);
    }

    Voice::Voice() {
        pEngine     = NULL;
        pDiskThread = NULL;
        PlaybackState = playback_state_end;
        pLFO1 = new LFOUnsigned(1.0f);  // amplitude EG (0..1 range)
        pLFO2 = new LFOUnsigned(1.0f);  // filter EG (0..1 range)
        pLFO3 = new LFOSigned(1200.0f); // pitch EG (-1200..+1200 range)
        KeyGroup = 0;
        SynthesisMode = 0; // set all mode bits to 0 first
        // select synthesis implementation (currently either pure C++ or MMX+SSE(1))
        #if CONFIG_ASM && ARCH_X86
        SYNTHESIS_MODE_SET_IMPLEMENTATION(SynthesisMode, Features::supportsMMX() && Features::supportsSSE());
        #else
        SYNTHESIS_MODE_SET_IMPLEMENTATION(SynthesisMode, false);
        #endif
        SYNTHESIS_MODE_SET_PROFILING(SynthesisMode, Profiler::isEnabled());

        finalSynthesisParameters.filterLeft.Reset();
        finalSynthesisParameters.filterRight.Reset();
    }

    Voice::~Voice() {
        if (pLFO1) delete pLFO1;
        if (pLFO2) delete pLFO2;
        if (pLFO3) delete pLFO3;
    }

    void Voice::SetEngine(Engine* pEngine) {
        this->pEngine     = pEngine;
        this->pDiskThread = pEngine->pDiskThread;
        dmsg(6,("Voice::SetEngine()\n"));
    }

    /**
     *  Initializes and triggers the voice, a disk stream will be launched if
     *  needed.
     *
     *  @param pEngineChannel - engine channel on which this voice was ordered
     *  @param itNoteOnEvent  - event that caused triggering of this voice
     *  @param PitchBend      - MIDI detune factor (-8192 ... +8191)
     *  @param pDimRgn        - points to the dimension region which provides sample wave(s) and articulation data
     *  @param VoiceType      - type of this voice
     *  @param iKeyGroup      - a value > 0 defines a key group in which this voice is member of
     *  @returns 0 on success, a value < 0 if the voice wasn't triggered
     *           (either due to an error or e.g. because no region is
     *           defined for the given key)
     */
    int Voice::Trigger(EngineChannel* pEngineChannel, Pool<Event>::Iterator& itNoteOnEvent, int PitchBend, ::gig::DimensionRegion* pDimRgn, type_t VoiceType, int iKeyGroup) {
        this->pEngineChannel = pEngineChannel;
        this->pDimRgn        = pDimRgn;

        #if CONFIG_DEVMODE
        if (itNoteOnEvent->FragmentPos() > pEngine->MaxSamplesPerCycle) { // just a sanity check for debugging
            dmsg(1,("Voice::Trigger(): ERROR, TriggerDelay > Totalsamples\n"));
        }
        #endif // CONFIG_DEVMODE

        Type            = VoiceType;
        MIDIKey         = itNoteOnEvent->Param.Note.Key;
        PlaybackState   = playback_state_init; // mark voice as triggered, but no audio rendered yet
        Delay           = itNoteOnEvent->FragmentPos();
        itTriggerEvent  = itNoteOnEvent;
        itKillEvent     = Pool<Event>::Iterator();
        KeyGroup        = iKeyGroup;
        pSample         = pDimRgn->pSample; // sample won't change until the voice is finished

        // calculate volume
        const double velocityAttenuation = pDimRgn->GetVelocityAttenuation(itNoteOnEvent->Param.Note.Velocity);

        Volume = velocityAttenuation / 32768.0f; // we downscale by 32768 to convert from int16 value range to DSP value range (which is -1.0..1.0)

        Volume *= pDimRgn->SampleAttenuation;

        // the volume of release triggered samples depends on note length
        if (Type == type_release_trigger) {
            float noteLength = float(pEngine->FrameTime + Delay -
                                     pEngineChannel->pMIDIKeyInfo[MIDIKey].NoteOnTime) / pEngine->SampleRate;
            float attenuation = 1 - 0.01053 * (256 >> pDimRgn->ReleaseTriggerDecay) * noteLength;
            if (attenuation <= 0) return -1;
            Volume *= attenuation;
        }

        // select channel mode (mono or stereo)
        SYNTHESIS_MODE_SET_CHANNELS(SynthesisMode, pSample->Channels == 2);

        // get starting crossfade volume level
        switch (pDimRgn->AttenuationController.type) {
            case ::gig::attenuation_ctrl_t::type_channelaftertouch:
                CrossfadeVolume = 1.0f; //TODO: aftertouch not supported yet
                break;
            case ::gig::attenuation_ctrl_t::type_velocity:
                CrossfadeVolume = CrossfadeAttenuation(itNoteOnEvent->Param.Note.Velocity);
                break;
            case ::gig::attenuation_ctrl_t::type_controlchange: //FIXME: currently not sample accurate
                CrossfadeVolume = CrossfadeAttenuation(pEngineChannel->ControllerTable[pDimRgn->AttenuationController.controller_number]);
                break;
            case ::gig::attenuation_ctrl_t::type_none: // no crossfade defined
            default:
                CrossfadeVolume = 1.0f;
        }

        PanLeft  = 1.0f - float(RTMath::Max(pDimRgn->Pan, 0)) /  63.0f;
        PanRight = 1.0f - float(RTMath::Min(pDimRgn->Pan, 0)) / -64.0f;

        finalSynthesisParameters.dPos = pDimRgn->SampleStartOffset; // offset where we should start playback of sample (0 - 2000 sample points)
        Pos = pDimRgn->SampleStartOffset;

        // Check if the sample needs disk streaming or is too short for that
        long cachedsamples = pSample->GetCache().Size / pSample->FrameSize;
        DiskVoice          = cachedsamples < pSample->SamplesTotal;

        if (DiskVoice) { // voice to be streamed from disk
            MaxRAMPos = cachedsamples - (pEngine->MaxSamplesPerCycle << CONFIG_MAX_PITCH) / pSample->Channels; //TODO: this calculation is too pessimistic and may better be moved to Render() method, so it calculates MaxRAMPos dependent to the current demand of sample points to be rendered (e.g. in case of JACK)

            // check if there's a loop defined which completely fits into the cached (RAM) part of the sample
            if (pSample->Loops && pSample->LoopEnd <= MaxRAMPos) {
                RAMLoop            = true;
                loop.uiTotalCycles = pSample->LoopPlayCount;
                loop.uiCyclesLeft  = pSample->LoopPlayCount;
                loop.uiStart       = pSample->LoopStart;
                loop.uiEnd         = pSample->LoopEnd;
                loop.uiSize        = pSample->LoopSize;
            }
            else RAMLoop = false;

            if (pDiskThread->OrderNewStream(&DiskStreamRef, pSample, MaxRAMPos, !RAMLoop) < 0) {
                dmsg(1,("Disk stream order failed!\n"));
                KillImmediately();
                return -1;
            }
            dmsg(4,("Disk voice launched (cached samples: %d, total Samples: %d, MaxRAMPos: %d, RAMLooping: %s)\n", cachedsamples, pSample->SamplesTotal, MaxRAMPos, (RAMLoop) ? "yes" : "no"));
        }
        else { // RAM only voice
            MaxRAMPos = cachedsamples;
            if (pSample->Loops) {
                RAMLoop           = true;
                loop.uiCyclesLeft = pSample->LoopPlayCount;
            }
            else RAMLoop = false;
            dmsg(4,("RAM only voice launched (Looping: %s)\n", (RAMLoop) ? "yes" : "no"));
        }


        // calculate initial pitch value
        {
            double pitchbasecents = pDimRgn->FineTune + (int) pEngine->ScaleTuning[MIDIKey % 12];
            if (pDimRgn->PitchTrack) pitchbasecents += (MIDIKey - (int) pDimRgn->UnityNote) * 100;
            this->PitchBase = RTMath::CentsToFreqRatio(pitchbasecents) * (double(pSample->SamplesPerSecond) / double(pEngine->SampleRate));
            this->PitchBend = RTMath::CentsToFreqRatio(((double) PitchBend / 8192.0) * 200.0); // pitchbend wheel +-2 semitones = 200 cents
        }

        // the length of the decay and release curves are dependent on the velocity
        const double velrelease = 1 / pDimRgn->GetVelocityRelease(itNoteOnEvent->Param.Note.Velocity);

        // setup EG 1 (VCA EG)
        {
            // get current value of EG1 controller
            double eg1controllervalue;
            switch (pDimRgn->EG1Controller.type) {
                case ::gig::eg1_ctrl_t::type_none: // no controller defined
                    eg1controllervalue = 0;
                    break;
                case ::gig::eg1_ctrl_t::type_channelaftertouch:
                    eg1controllervalue = 0; // TODO: aftertouch not yet supported
                    break;
                case ::gig::eg1_ctrl_t::type_velocity:
                    eg1controllervalue = itNoteOnEvent->Param.Note.Velocity;
                    break;
                case ::gig::eg1_ctrl_t::type_controlchange: // MIDI control change controller
                    eg1controllervalue = pEngineChannel->ControllerTable[pDimRgn->EG1Controller.controller_number];
                    break;
            }
            if (pDimRgn->EG1ControllerInvert) eg1controllervalue = 127 - eg1controllervalue;

            // calculate influence of EG1 controller on EG1's parameters
            // (eg1attack is different from the others)
            double eg1attack  = (pDimRgn->EG1ControllerAttackInfluence)  ?
                1 + 0.031 * (double) (pDimRgn->EG1ControllerAttackInfluence == 1 ?
                                      1 : 1 << pDimRgn->EG1ControllerAttackInfluence) * eg1controllervalue : 1.0;
            double eg1decay   = (pDimRgn->EG1ControllerDecayInfluence)   ? 1 + 0.00775 * (double) (1 << pDimRgn->EG1ControllerDecayInfluence)   * eg1controllervalue : 1.0;
            double eg1release = (pDimRgn->EG1ControllerReleaseInfluence) ? 1 + 0.00775 * (double) (1 << pDimRgn->EG1ControllerReleaseInfluence) * eg1controllervalue : 1.0;

            EG1.trigger(pDimRgn->EG1PreAttack,
                        pDimRgn->EG1Attack * eg1attack,
                        pDimRgn->EG1Hold,
                        pDimRgn->EG1Decay1 * eg1decay * velrelease,
                        pDimRgn->EG1Decay2 * eg1decay * velrelease,
                        pDimRgn->EG1InfiniteSustain,
                        pDimRgn->EG1Sustain,
                        pDimRgn->EG1Release * eg1release * velrelease,
                        velocityAttenuation,
                        pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        // setup EG 2 (VCF Cutoff EG)
        {
            // get current value of EG2 controller
            double eg2controllervalue;
            switch (pDimRgn->EG2Controller.type) {
                case ::gig::eg2_ctrl_t::type_none: // no controller defined
                    eg2controllervalue = 0;
                    break;
                case ::gig::eg2_ctrl_t::type_channelaftertouch:
                    eg2controllervalue = 0; // TODO: aftertouch not yet supported
                    break;
                case ::gig::eg2_ctrl_t::type_velocity:
                    eg2controllervalue = itNoteOnEvent->Param.Note.Velocity;
                    break;
                case ::gig::eg2_ctrl_t::type_controlchange: // MIDI control change controller
                    eg2controllervalue = pEngineChannel->ControllerTable[pDimRgn->EG2Controller.controller_number];
                    break;
            }
            if (pDimRgn->EG2ControllerInvert) eg2controllervalue = 127 - eg2controllervalue;

            // calculate influence of EG2 controller on EG2's parameters
            double eg2attack  = (pDimRgn->EG2ControllerAttackInfluence)  ? 1 + 0.00775 * (double) (1 << pDimRgn->EG2ControllerAttackInfluence)  * eg2controllervalue : 1.0;
            double eg2decay   = (pDimRgn->EG2ControllerDecayInfluence)   ? 1 + 0.00775 * (double) (1 << pDimRgn->EG2ControllerDecayInfluence)   * eg2controllervalue : 1.0;
            double eg2release = (pDimRgn->EG2ControllerReleaseInfluence) ? 1 + 0.00775 * (double) (1 << pDimRgn->EG2ControllerReleaseInfluence) * eg2controllervalue : 1.0;

            EG2.trigger(pDimRgn->EG2PreAttack,
                        pDimRgn->EG2Attack * eg2attack,
                        false,
                        pDimRgn->EG2Decay1 * eg2decay * velrelease,
                        pDimRgn->EG2Decay2 * eg2decay * velrelease,
                        pDimRgn->EG2InfiniteSustain,
                        pDimRgn->EG2Sustain,
                        pDimRgn->EG2Release * eg2release * velrelease,
                        velocityAttenuation,
                        pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        // setup EG 3 (VCO EG)
        {
          double eg3depth = RTMath::CentsToFreqRatio(pDimRgn->EG3Depth);
          EG3.trigger(eg3depth, pDimRgn->EG3Attack, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        // setup LFO 1 (VCA LFO)
        {
            uint16_t lfo1_internal_depth;
            switch (pDimRgn->LFO1Controller) {
                case ::gig::lfo1_ctrl_internal:
                    lfo1_internal_depth  = pDimRgn->LFO1InternalDepth;
                    pLFO1->ExtController = 0; // no external controller
                    bLFO1Enabled         = (lfo1_internal_depth > 0);
                    break;
                case ::gig::lfo1_ctrl_modwheel:
                    lfo1_internal_depth  = 0;
                    pLFO1->ExtController = 1; // MIDI controller 1
                    bLFO1Enabled         = (pDimRgn->LFO1ControlDepth > 0);
                    break;
                case ::gig::lfo1_ctrl_breath:
                    lfo1_internal_depth  = 0;
                    pLFO1->ExtController = 2; // MIDI controller 2
                    bLFO1Enabled         = (pDimRgn->LFO1ControlDepth > 0);
                    break;
                case ::gig::lfo1_ctrl_internal_modwheel:
                    lfo1_internal_depth  = pDimRgn->LFO1InternalDepth;
                    pLFO1->ExtController = 1; // MIDI controller 1
                    bLFO1Enabled         = (lfo1_internal_depth > 0 || pDimRgn->LFO1ControlDepth > 0);
                    break;
                case ::gig::lfo1_ctrl_internal_breath:
                    lfo1_internal_depth  = pDimRgn->LFO1InternalDepth;
                    pLFO1->ExtController = 2; // MIDI controller 2
                    bLFO1Enabled         = (lfo1_internal_depth > 0 || pDimRgn->LFO1ControlDepth > 0);
                    break;
                default:
                    lfo1_internal_depth  = 0;
                    pLFO1->ExtController = 0; // no external controller
                    bLFO1Enabled         = false;
            }
            if (bLFO1Enabled) pLFO1->trigger(pDimRgn->LFO1Frequency,
                                             start_level_max,
                                             lfo1_internal_depth,
                                             pDimRgn->LFO1ControlDepth,
                                             pDimRgn->LFO1FlipPhase,
                                             pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        // setup LFO 2 (VCF Cutoff LFO)
        {
            uint16_t lfo2_internal_depth;
            switch (pDimRgn->LFO2Controller) {
                case ::gig::lfo2_ctrl_internal:
                    lfo2_internal_depth  = pDimRgn->LFO2InternalDepth;
                    pLFO2->ExtController = 0; // no external controller
                    bLFO2Enabled         = (lfo2_internal_depth > 0);
                    break;
                case ::gig::lfo2_ctrl_modwheel:
                    lfo2_internal_depth  = 0;
                    pLFO2->ExtController = 1; // MIDI controller 1
                    bLFO2Enabled         = (pDimRgn->LFO2ControlDepth > 0);
                    break;
                case ::gig::lfo2_ctrl_foot:
                    lfo2_internal_depth  = 0;
                    pLFO2->ExtController = 4; // MIDI controller 4
                    bLFO2Enabled         = (pDimRgn->LFO2ControlDepth > 0);
                    break;
                case ::gig::lfo2_ctrl_internal_modwheel:
                    lfo2_internal_depth  = pDimRgn->LFO2InternalDepth;
                    pLFO2->ExtController = 1; // MIDI controller 1
                    bLFO2Enabled         = (lfo2_internal_depth > 0 || pDimRgn->LFO2ControlDepth > 0);
                    break;
                case ::gig::lfo2_ctrl_internal_foot:
                    lfo2_internal_depth  = pDimRgn->LFO2InternalDepth;
                    pLFO2->ExtController = 4; // MIDI controller 4
                    bLFO2Enabled         = (lfo2_internal_depth > 0 || pDimRgn->LFO2ControlDepth > 0);
                    break;
                default:
                    lfo2_internal_depth  = 0;
                    pLFO2->ExtController = 0; // no external controller
                    bLFO2Enabled         = false;
            }
            if (bLFO2Enabled) pLFO2->trigger(pDimRgn->LFO2Frequency,
                                             start_level_max,
                                             lfo2_internal_depth,
                                             pDimRgn->LFO2ControlDepth,
                                             pDimRgn->LFO2FlipPhase,
                                             pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        // setup LFO 3 (VCO LFO)
        {
            uint16_t lfo3_internal_depth;
            switch (pDimRgn->LFO3Controller) {
                case ::gig::lfo3_ctrl_internal:
                    lfo3_internal_depth  = pDimRgn->LFO3InternalDepth;
                    pLFO3->ExtController = 0; // no external controller
                    bLFO3Enabled         = (lfo3_internal_depth > 0);
                    break;
                case ::gig::lfo3_ctrl_modwheel:
                    lfo3_internal_depth  = 0;
                    pLFO3->ExtController = 1; // MIDI controller 1
                    bLFO3Enabled         = (pDimRgn->LFO3ControlDepth > 0);
                    break;
                case ::gig::lfo3_ctrl_aftertouch:
                    lfo3_internal_depth  = 0;
                    pLFO3->ExtController = 0; // TODO: aftertouch not implemented yet
                    bLFO3Enabled         = false; // see TODO comment in line above
                    break;
                case ::gig::lfo3_ctrl_internal_modwheel:
                    lfo3_internal_depth  = pDimRgn->LFO3InternalDepth;
                    pLFO3->ExtController = 1; // MIDI controller 1
                    bLFO3Enabled         = (lfo3_internal_depth > 0 || pDimRgn->LFO3ControlDepth > 0);
                    break;
                case ::gig::lfo3_ctrl_internal_aftertouch:
                    lfo3_internal_depth  = pDimRgn->LFO3InternalDepth;
                    pLFO1->ExtController = 0; // TODO: aftertouch not implemented yet
                    bLFO3Enabled         = (lfo3_internal_depth > 0 /*|| pDimRgn->LFO3ControlDepth > 0*/); // see TODO comment in line above
                    break;
                default:
                    lfo3_internal_depth  = 0;
                    pLFO3->ExtController = 0; // no external controller
                    bLFO3Enabled         = false;
            }
            if (bLFO3Enabled) pLFO3->trigger(pDimRgn->LFO3Frequency,
                                             start_level_mid,
                                             lfo3_internal_depth,
                                             pDimRgn->LFO3ControlDepth,
                                             false,
                                             pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
        }


        #if CONFIG_FORCE_FILTER
        const bool bUseFilter = true;
        #else // use filter only if instrument file told so
        const bool bUseFilter = pDimRgn->VCFEnabled;
        #endif // CONFIG_FORCE_FILTER
        SYNTHESIS_MODE_SET_FILTER(SynthesisMode, bUseFilter);
        if (bUseFilter) {
            #ifdef CONFIG_OVERRIDE_CUTOFF_CTRL
            VCFCutoffCtrl.controller = CONFIG_OVERRIDE_CUTOFF_CTRL;
            #else // use the one defined in the instrument file
            switch (pDimRgn->VCFCutoffController) {
                case ::gig::vcf_cutoff_ctrl_modwheel:
                    VCFCutoffCtrl.controller = 1;
                    break;
                case ::gig::vcf_cutoff_ctrl_effect1:
                    VCFCutoffCtrl.controller = 12;
                    break;
                case ::gig::vcf_cutoff_ctrl_effect2:
                    VCFCutoffCtrl.controller = 13;
                    break;
                case ::gig::vcf_cutoff_ctrl_breath:
                    VCFCutoffCtrl.controller = 2;
                    break;
                case ::gig::vcf_cutoff_ctrl_foot:
                    VCFCutoffCtrl.controller = 4;
                    break;
                case ::gig::vcf_cutoff_ctrl_sustainpedal:
                    VCFCutoffCtrl.controller = 64;
                    break;
                case ::gig::vcf_cutoff_ctrl_softpedal:
                    VCFCutoffCtrl.controller = 67;
                    break;
                case ::gig::vcf_cutoff_ctrl_genpurpose7:
                    VCFCutoffCtrl.controller = 82;
                    break;
                case ::gig::vcf_cutoff_ctrl_genpurpose8:
                    VCFCutoffCtrl.controller = 83;
                    break;
                case ::gig::vcf_cutoff_ctrl_aftertouch: //TODO: not implemented yet
                case ::gig::vcf_cutoff_ctrl_none:
                default:
                    VCFCutoffCtrl.controller = 0;
                    break;
            }
            #endif // CONFIG_OVERRIDE_CUTOFF_CTRL

            #ifdef CONFIG_OVERRIDE_RESONANCE_CTRL
            VCFResonanceCtrl.controller = CONFIG_OVERRIDE_RESONANCE_CTRL;
            #else // use the one defined in the instrument file
            switch (pDimRgn->VCFResonanceController) {
                case ::gig::vcf_res_ctrl_genpurpose3:
                    VCFResonanceCtrl.controller = 18;
                    break;
                case ::gig::vcf_res_ctrl_genpurpose4:
                    VCFResonanceCtrl.controller = 19;
                    break;
                case ::gig::vcf_res_ctrl_genpurpose5:
                    VCFResonanceCtrl.controller = 80;
                    break;
                case ::gig::vcf_res_ctrl_genpurpose6:
                    VCFResonanceCtrl.controller = 81;
                    break;
                case ::gig::vcf_res_ctrl_none:
                default:
                    VCFResonanceCtrl.controller = 0;
            }
            #endif // CONFIG_OVERRIDE_RESONANCE_CTRL

            #ifndef CONFIG_OVERRIDE_FILTER_TYPE
            finalSynthesisParameters.filterLeft.SetType(pDimRgn->VCFType);
            finalSynthesisParameters.filterRight.SetType(pDimRgn->VCFType);
            #else // override filter type
            FilterLeft.SetType(CONFIG_OVERRIDE_FILTER_TYPE);
            FilterRight.SetType(CONFIG_OVERRIDE_FILTER_TYPE);
            #endif // CONFIG_OVERRIDE_FILTER_TYPE

            VCFCutoffCtrl.value    = pEngineChannel->ControllerTable[VCFCutoffCtrl.controller];
            VCFResonanceCtrl.value = pEngineChannel->ControllerTable[VCFResonanceCtrl.controller];

            // calculate cutoff frequency
            float cutoff = pDimRgn->GetVelocityCutoff(itNoteOnEvent->Param.Note.Velocity);
            if (pDimRgn->VCFKeyboardTracking) {
                cutoff *= exp((itNoteOnEvent->Param.Note.Key - pDimRgn->VCFKeyboardTrackingBreakpoint) * 0.057762265f); // (ln(2) / 12)
            }
            CutoffBase = cutoff;

            int cvalue;
            if (VCFCutoffCtrl.controller) {
                cvalue = pEngineChannel->ControllerTable[VCFCutoffCtrl.controller];
                if (pDimRgn->VCFCutoffControllerInvert) cvalue = 127 - cvalue;
                // VCFVelocityScale in this case means Minimum cutoff
                if (cvalue < pDimRgn->VCFVelocityScale) cvalue = pDimRgn->VCFVelocityScale;
            }
            else {
                cvalue = pDimRgn->VCFCutoff;
            }
            cutoff *= float(cvalue) * 0.00787402f; // (1 / 127)
            if (cutoff > 1.0) cutoff = 1.0;
            cutoff = (cutoff < 0.5 ? cutoff * 4826 - 1 : cutoff * 5715 - 449);
            if (cutoff < 1.0) cutoff = 1.0;

            // calculate resonance
            float resonance = (float) (VCFResonanceCtrl.controller ? VCFResonanceCtrl.value : pDimRgn->VCFResonance) * 0.00787f; // 0.0..1.0

            VCFCutoffCtrl.fvalue    = cutoff - 1.0;
            VCFResonanceCtrl.fvalue = resonance;
        }
        else {
            VCFCutoffCtrl.controller    = 0;
            VCFResonanceCtrl.controller = 0;
        }

        return 0; // success
    }

    /**
     *  Renders the audio data for this voice for the current audio fragment.
     *  The sample input data can either come from RAM (cached sample or sample
     *  part) or directly from disk. The output signal will be rendered by
     *  resampling / interpolation. If this voice is a disk streaming voice and
     *  the voice completely played back the cached RAM part of the sample, it
     *  will automatically switch to disk playback for the next RenderAudio()
     *  call.
     *
     *  @param Samples - number of samples to be rendered in this audio fragment cycle
     */
    void Voice::Render(uint Samples) {

        // select default values for synthesis mode bits
        SYNTHESIS_MODE_SET_LOOP(SynthesisMode, false);

        switch (this->PlaybackState) {

            case playback_state_init:
                this->PlaybackState = playback_state_ram; // we always start playback from RAM cache and switch then to disk if needed
                // no break - continue with playback_state_ram

            case playback_state_ram: {
                    if (RAMLoop) SYNTHESIS_MODE_SET_LOOP(SynthesisMode, true); // enable looping

                    // render current fragment
                    Synthesize(Samples, (sample_t*) pSample->GetCache().pStart, Delay);

                    if (DiskVoice) {
                        // check if we reached the allowed limit of the sample RAM cache
                        if (finalSynthesisParameters.dPos > MaxRAMPos) {
                            dmsg(5,("Voice: switching to disk playback (Pos=%f)\n", finalSynthesisParameters.dPos));
                            this->PlaybackState = playback_state_disk;
                        }
                    } else if (finalSynthesisParameters.dPos >= pSample->GetCache().Size / pSample->FrameSize) {
                        this->PlaybackState = playback_state_end;
                    }
                }
                break;

            case playback_state_disk: {
                    if (!DiskStreamRef.pStream) {
                        // check if the disk thread created our ordered disk stream in the meantime
                        DiskStreamRef.pStream = pDiskThread->AskForCreatedStream(DiskStreamRef.OrderID);
                        if (!DiskStreamRef.pStream) {
                            std::cout << stderr << "Disk stream not available in time!" << std::endl << std::flush;
                            KillImmediately();
                            return;
                        }
                        DiskStreamRef.pStream->IncrementReadPos(pSample->Channels * (int(finalSynthesisParameters.dPos) - MaxRAMPos));
                        finalSynthesisParameters.dPos -= int(finalSynthesisParameters.dPos);
                        RealSampleWordsLeftToRead = -1; // -1 means no silence has been added yet
                    }

                    const int sampleWordsLeftToRead = DiskStreamRef.pStream->GetReadSpace();

                    // add silence sample at the end if we reached the end of the stream (for the interpolator)
                    if (DiskStreamRef.State == Stream::state_end) {
                        const int maxSampleWordsPerCycle = (pEngine->MaxSamplesPerCycle << CONFIG_MAX_PITCH) * pSample->Channels + 6; // +6 for the interpolator algorithm
                        if (sampleWordsLeftToRead <= maxSampleWordsPerCycle) {
                            // remember how many sample words there are before any silence has been added
                            if (RealSampleWordsLeftToRead < 0) RealSampleWordsLeftToRead = sampleWordsLeftToRead;
                            DiskStreamRef.pStream->WriteSilence(maxSampleWordsPerCycle - sampleWordsLeftToRead);
                        }
                    }

                    sample_t* ptr = DiskStreamRef.pStream->GetReadPtr(); // get the current read_ptr within the ringbuffer where we read the samples from

                    // render current audio fragment
                    Synthesize(Samples, ptr, Delay);

                    const int iPos = (int) finalSynthesisParameters.dPos;
                    const int readSampleWords = iPos * pSample->Channels; // amount of sample words actually been read
                    DiskStreamRef.pStream->IncrementReadPos(readSampleWords);
                    finalSynthesisParameters.dPos -= iPos; // just keep fractional part of playback position

                    // change state of voice to 'end' if we really reached the end of the sample data
                    if (RealSampleWordsLeftToRead >= 0) {
                        RealSampleWordsLeftToRead -= readSampleWords;
                        if (RealSampleWordsLeftToRead <= 0) this->PlaybackState = playback_state_end;
                    }
                }
                break;

            case playback_state_end:
                std::cerr << "gig::Voice::Render(): entered with playback_state_end, this is a bug!\n" << std::flush;
                break;
        }

        // Reset delay
        Delay = 0;

        itTriggerEvent = Pool<Event>::Iterator();

        // If sample stream or release stage finished, kill the voice
        if (PlaybackState == playback_state_end || EG1.getSegmentType() == EGADSR::segment_end) KillImmediately();
    }

    /**
     *  Resets voice variables. Should only be called if rendering process is
     *  suspended / not running.
     */
    void Voice::Reset() {
        finalSynthesisParameters.filterLeft.Reset();
        finalSynthesisParameters.filterRight.Reset();
        DiskStreamRef.pStream = NULL;
        DiskStreamRef.hStream = 0;
        DiskStreamRef.State   = Stream::state_unused;
        DiskStreamRef.OrderID = 0;
        PlaybackState = playback_state_end;
        itTriggerEvent = Pool<Event>::Iterator();
        itKillEvent    = Pool<Event>::Iterator();
    }

    /**
     * Process given list of MIDI note on, note off and sustain pedal events
     * for the given time.
     *
     * @param itEvent - iterator pointing to the next event to be processed
     * @param End     - youngest time stamp where processing should be stopped 
     */
    void Voice::processTransitionEvents(RTList<Event>::Iterator& itEvent, uint End) {
        for (; itEvent && itEvent->FragmentPos() <= End; ++itEvent) {
            if (itEvent->Type == Event::type_release) {
                EG1.update(EGADSR::event_release, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
                EG2.update(EGADSR::event_release, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
            } else if (itEvent->Type == Event::type_cancel_release) {
                EG1.update(EGADSR::event_cancel_release, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
                EG2.update(EGADSR::event_cancel_release, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
            }
        }
    }

    /**
     * Process given list of MIDI control change and pitch bend events for
     * the given time.
     *
     * @param itEvent - iterator pointing to the next event to be processed
     * @param End     - youngest time stamp where processing should be stopped 
     */
    void Voice::processCCEvents(RTList<Event>::Iterator& itEvent, uint End) {
        for (; itEvent && itEvent->FragmentPos() <= End; ++itEvent) {
            if (itEvent->Type == Event::type_control_change &&
                itEvent->Param.CC.Controller) { // if (valid) MIDI control change event
                if (itEvent->Param.CC.Controller == VCFCutoffCtrl.controller) {
                    processCutoffEvent(itEvent);
                }
                if (itEvent->Param.CC.Controller == VCFResonanceCtrl.controller) {
                    processResonanceEvent(itEvent);
                }
                if (itEvent->Param.CC.Controller == pLFO1->ExtController) {
                    pLFO1->update(itEvent->Param.CC.Value);
                }
                if (itEvent->Param.CC.Controller == pLFO2->ExtController) {
                    pLFO2->update(itEvent->Param.CC.Value);
                }
                if (itEvent->Param.CC.Controller == pLFO3->ExtController) {
                    pLFO3->update(itEvent->Param.CC.Value);
                }
                if (pDimRgn->AttenuationController.type == ::gig::attenuation_ctrl_t::type_controlchange &&
                    itEvent->Param.CC.Controller == pDimRgn->AttenuationController.controller_number) {
                    processCrossFadeEvent(itEvent);
                }
            } else if (itEvent->Type == Event::type_pitchbend) { // if pitch bend event
                processPitchEvent(itEvent);
            }
        }
    }

    void Voice::processPitchEvent(RTList<Event>::Iterator& itEvent) {
        const float pitch = RTMath::CentsToFreqRatio(((double) itEvent->Param.Pitch.Pitch / 8192.0) * 200.0); // +-two semitones = +-200 cents
        finalSynthesisParameters.fFinalPitch *= pitch;
        PitchBend = pitch;
    }

    void Voice::processCrossFadeEvent(RTList<Event>::Iterator& itEvent) {
        CrossfadeVolume = CrossfadeAttenuation(itEvent->Param.CC.Value);
        #if CONFIG_PROCESS_MUTED_CHANNELS
        const float effectiveVolume = CrossfadeVolume * Volume * (pEngineChannel->GetMute() ? 0 : pEngineChannel->GlobalVolume);
        #else
        const float effectiveVolume = CrossfadeVolume * Volume * pEngineChannel->GlobalVolume;
        #endif
        fFinalVolume = effectiveVolume;
    }

    void Voice::processCutoffEvent(RTList<Event>::Iterator& itEvent) {
        int ccvalue = itEvent->Param.CC.Value;
        if (VCFCutoffCtrl.value == ccvalue) return;
        VCFCutoffCtrl.value == ccvalue;
        if (pDimRgn->VCFCutoffControllerInvert)  ccvalue = 127 - ccvalue;
        if (ccvalue < pDimRgn->VCFVelocityScale) ccvalue = pDimRgn->VCFVelocityScale;
        float cutoff = CutoffBase * float(ccvalue) * 0.00787402f; // (1 / 127)
        if (cutoff > 1.0) cutoff = 1.0;
        cutoff = (cutoff < 0.5 ? cutoff * 4826 - 1 : cutoff * 5715 - 449);
        if (cutoff < 1.0) cutoff = 1.0;

        VCFCutoffCtrl.fvalue = cutoff - 1.0; // needed for initialization of fFinalCutoff next time
        fFinalCutoff = cutoff;
    }

    void Voice::processResonanceEvent(RTList<Event>::Iterator& itEvent) {
        // convert absolute controller value to differential
        const int ctrldelta = itEvent->Param.CC.Value - VCFResonanceCtrl.value;
        VCFResonanceCtrl.value = itEvent->Param.CC.Value;
        const float resonancedelta = (float) ctrldelta * 0.00787f; // 0.0..1.0
        fFinalResonance += resonancedelta;
        // needed for initialization of parameter
        VCFResonanceCtrl.fvalue = itEvent->Param.CC.Value * 0.00787f;
    }

    /**
     *  Synthesizes the current audio fragment for this voice.
     *
     *  @param Samples - number of sample points to be rendered in this audio
     *                   fragment cycle
     *  @param pSrc    - pointer to input sample data
     *  @param Skip    - number of sample points to skip in output buffer
     */
    void Voice::Synthesize(uint Samples, sample_t* pSrc, uint Skip) {
        finalSynthesisParameters.pOutLeft  = &pEngineChannel->pOutputLeft[Skip];
        finalSynthesisParameters.pOutRight = &pEngineChannel->pOutputRight[Skip];
        finalSynthesisParameters.pSrc      = pSrc;

        RTList<Event>::Iterator itCCEvent = pEngineChannel->pEvents->first();
        RTList<Event>::Iterator itNoteEvent = pEngineChannel->pMIDIKeyInfo[MIDIKey].pEvents->first();

        if (Skip) { // skip events that happened before this voice was triggered
            while (itCCEvent && itCCEvent->FragmentPos() <= Skip) ++itCCEvent;
            while (itNoteEvent && itNoteEvent->FragmentPos() <= Skip) ++itNoteEvent;
        }

        uint killPos;
        if (itKillEvent) killPos = RTMath::Min(itKillEvent->FragmentPos(), pEngine->MaxFadeOutPos);

        uint i = Skip;
        while (i < Samples) {
            int iSubFragmentEnd = RTMath::Min(i + CONFIG_DEFAULT_SUBFRAGMENT_SIZE, Samples);

            // initialize all final synthesis parameters
            finalSynthesisParameters.fFinalPitch = PitchBase * PitchBend;
            #if CONFIG_PROCESS_MUTED_CHANNELS
            fFinalVolume = this->Volume * this->CrossfadeVolume * (pEngineChannel->GetMute() ? 0 : pEngineChannel->GlobalVolume);
            #else
            fFinalVolume = this->Volume * this->CrossfadeVolume * pEngineChannel->GlobalVolume;
            #endif
            fFinalCutoff    = VCFCutoffCtrl.fvalue;
            fFinalResonance = VCFResonanceCtrl.fvalue;

            // process MIDI control change and pitchbend events for this subfragment
            processCCEvents(itCCEvent, iSubFragmentEnd);

            // process transition events (note on, note off & sustain pedal)
            processTransitionEvents(itNoteEvent, iSubFragmentEnd);

            // if the voice was killed in this subfragment switch EG1 to fade out stage
            if (itKillEvent && killPos <= iSubFragmentEnd) {
                EG1.enterFadeOutStage();
                itKillEvent = Pool<Event>::Iterator();
            }

            // process envelope generators
            switch (EG1.getSegmentType()) {
                case EGADSR::segment_lin:
                    fFinalVolume *= EG1.processLin();
                    break;
                case EGADSR::segment_exp:
                    fFinalVolume *= EG1.processExp();
                    break;
                case EGADSR::segment_end:
                    fFinalVolume *= EG1.getLevel();
                    break; // noop
            }
            switch (EG2.getSegmentType()) {
                case EGADSR::segment_lin:
                    fFinalCutoff *= EG2.processLin();
                    break;
                case EGADSR::segment_exp:
                    fFinalCutoff *= EG2.processExp();
                    break;
                case EGADSR::segment_end:
                    fFinalCutoff *= EG2.getLevel();
                    break; // noop
            }
            if (EG3.active()) finalSynthesisParameters.fFinalPitch *= RTMath::CentsToFreqRatio(EG3.render());

            // process low frequency oscillators
            if (bLFO1Enabled) fFinalVolume *= pLFO1->render();
            if (bLFO2Enabled) fFinalCutoff *= pLFO2->render();
            if (bLFO3Enabled) finalSynthesisParameters.fFinalPitch *= RTMath::CentsToFreqRatio(pLFO3->render());

            // if filter enabled then update filter coefficients
            if (SYNTHESIS_MODE_GET_FILTER(SynthesisMode)) {
                finalSynthesisParameters.filterLeft.SetParameters(fFinalCutoff + 1.0, fFinalResonance, pEngine->SampleRate);
                finalSynthesisParameters.filterRight.SetParameters(fFinalCutoff + 1.0, fFinalResonance, pEngine->SampleRate);
            }

            // do we need resampling?
            const float __PLUS_ONE_CENT  = 1.000577789506554859250142541782224725466f;
            const float __MINUS_ONE_CENT = 0.9994225441413807496009516495583113737666f;
            const bool bResamplingRequired = !(finalSynthesisParameters.fFinalPitch <= __PLUS_ONE_CENT &&
                                               finalSynthesisParameters.fFinalPitch >= __MINUS_ONE_CENT);
            SYNTHESIS_MODE_SET_INTERPOLATE(SynthesisMode, bResamplingRequired);

            // prepare final synthesis parameters structure
            finalSynthesisParameters.fFinalVolumeLeft  = fFinalVolume * PanLeft;
            finalSynthesisParameters.fFinalVolumeRight = fFinalVolume * PanRight;
            finalSynthesisParameters.uiToGo            = iSubFragmentEnd - i;

            // render audio for one subfragment
            RunSynthesisFunction(SynthesisMode, &finalSynthesisParameters, &loop);

            const double newPos = Pos + (iSubFragmentEnd - i) * finalSynthesisParameters.fFinalPitch;

            // increment envelopes' positions
            if (EG1.active()) {

                // if sample has a loop and loop start has been reached in this subfragment, send a special event to EG1 to let it finish the attack hold stage
                if (pSample->Loops && Pos <= pSample->LoopStart && pSample->LoopStart < newPos) {
                    EG1.update(EGADSR::event_hold_end, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
                }

                EG1.increment(1);
                if (!EG1.toStageEndLeft()) EG1.update(EGADSR::event_stage_end, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
            }
            if (EG2.active()) {
                EG2.increment(1);
                if (!EG2.toStageEndLeft()) EG2.update(EGADSR::event_stage_end, pEngine->SampleRate / CONFIG_DEFAULT_SUBFRAGMENT_SIZE);
            }
            EG3.increment(1);
            if (!EG3.toEndLeft()) EG3.update(); // neutralize envelope coefficient if end reached

            Pos = newPos;
            i = iSubFragmentEnd;
        }
    }

    /**
     *  Immediately kill the voice. This method should not be used to kill
     *  a normal, active voice, because it doesn't take care of things like
     *  fading down the volume level to avoid clicks and regular processing
     *  until the kill event actually occured!
     *
     *  @see Kill()
     */
    void Voice::KillImmediately() {
        if (DiskVoice && DiskStreamRef.State != Stream::state_unused) {
            pDiskThread->OrderDeletionOfStream(&DiskStreamRef);
        }
        Reset();
    }

    /**
     *  Kill the voice in regular sense. Let the voice render audio until
     *  the kill event actually occured and then fade down the volume level
     *  very quickly and let the voice die finally. Unlike a normal release
     *  of a voice, a kill process cannot be cancalled and is therefore
     *  usually used for voice stealing and key group conflicts.
     *
     *  @param itKillEvent - event which caused the voice to be killed
     */
    void Voice::Kill(Pool<Event>::Iterator& itKillEvent) {
        #if CONFIG_DEVMODE
        if (!itKillEvent) dmsg(1,("gig::Voice::Kill(): ERROR, !itKillEvent !!!\n"));
        if (itKillEvent && !itKillEvent.isValid()) dmsg(1,("gig::Voice::Kill(): ERROR, itKillEvent invalid !!!\n"));
        #endif // CONFIG_DEVMODE

        if (itTriggerEvent && itKillEvent->FragmentPos() <= itTriggerEvent->FragmentPos()) return;
        this->itKillEvent = itKillEvent;
    }

}} // namespace LinuxSampler::gig