aboutsummaryrefslogtreecommitdiffstats
path: root/src/mdaDX10.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'src/mdaDX10.cpp')
-rw-r--r--src/mdaDX10.cpp131
1 files changed, 19 insertions, 112 deletions
diff --git a/src/mdaDX10.cpp b/src/mdaDX10.cpp
index bc4c8d1..8e8a5ca 100644
--- a/src/mdaDX10.cpp
+++ b/src/mdaDX10.cpp
@@ -18,6 +18,8 @@
#include "mdaDX10.h"
+#include "lv2/lv2plug.in/ns/ext/atom/util.h"
+
#include <stdio.h>
#include <stdlib.h> //rand()
#include <math.h>
@@ -90,7 +92,6 @@ mdaDX10::mdaDX10(audioMasterCallback audioMaster) : AudioEffectX(audioMaster, NP
voice[i].mod0 = voice[i].mod1 = voice[i].dmod = 0.0f;
voice[i].cdec = 0.99f; //all notes off
}
- notes[0] = EVENTS_DONE;
lfo0 = dlfo = modwhl = 0.0f;
lfo1 = pbend = 1.0f;
volume = 0.0035f;
@@ -240,7 +241,7 @@ bool mdaDX10::copyProgram(int32_t destination)
}
-int32_t mdaDX10::canDo(char* text)
+int32_t mdaDX10::canDo(const char* text)
{
if(strcmp(text, "receiveLvzEvents") == 0) return 1;
if(strcmp(text, "receiveLvzMidiEvent") == 0) return 1;
@@ -304,101 +305,22 @@ void mdaDX10::getParameterLabel(int32_t index, char *label)
}
}
-
-void mdaDX10::process(float **inputs, float **outputs, int32_t sampleFrames)
-{
- float* out1 = outputs[0];
- float* out2 = outputs[1];
- int32_t event=0, frame=0, frames, v;
- float o, x, e, mw=MW, w=rich, m=modmix;
- int32_t k=K;
-
- if(activevoices>0 || notes[event]<sampleFrames) //detect & bypass completely empty blocks
- {
- while(frame<sampleFrames)
- {
- frames = notes[event++];
- if(frames>sampleFrames) frames = sampleFrames;
- frames -= frame;
- frame += frames;
-
- while(--frames>=0) //would be faster with voice loop outside frame loop!
- { //but then each voice would need it's own LFO...
- VOICE *V = voice;
- o = 0.0f;
-
- if(--k<0)
- {
- lfo0 += dlfo * lfo1; //sine LFO
- lfo1 -= dlfo * lfo0;
- mw = lfo1 * (modwhl + vibrato);
- k=100;
- }
-
- for(v=0; v<NVOICES; v++) //for each voice
- {
- e = V->env;
- if(e > SILENCE) //**** this is the synth ****
- {
- V->env = e * V->cdec; //decay & release
- V->cenv += V->catt * (e - V->cenv); //attack
-
- x = V->dmod * V->mod0 - V->mod1; //could add more modulator blocks like
- V->mod1 = V->mod0; //this for a wider range of FM sounds
- V->mod0 = x;
- V->menv += V->mdec * (V->mlev - V->menv);
-
- x = V->car + V->dcar + x * V->menv + mw; //carrier phase
- while(x > 1.0f) x -= 2.0f; //wrap phase
- while(x < -1.0f) x += 2.0f;
- V->car = x;
- o += V->cenv * (m * V->mod1 + (x + x * x * x * (w * x * x - 1.0f - w)));
- } //amp env //mod thru-mix //5th-order sine approximation
- V++;
- }
- *out1++ += o;
- *out2++ += o;
- }
-
- if(frame<sampleFrames) //next note on/off
- {
- int32_t note = notes[event++];
- int32_t vel = notes[event++];
- noteOn(note, vel);
- }
- }
-
- activevoices = NVOICES;
- for(v=0; v<NVOICES; v++)
- {
- if(voice[v].env < SILENCE) //choke voices that have finished
- {
- voice[v].env = voice[v].cenv = 0.0f;
- activevoices--;
- }
- if(voice[v].menv < SILENCE) voice[v].menv = voice[v].mlev = 0.0f;
- }
- }
-
- K=k; MW=mw; //remember these so vibrato speed not buffer size dependant!
- notes[0] = EVENTS_DONE;
-}
-
-
void mdaDX10::processReplacing(float **inputs, float **outputs, int32_t sampleFrames)
{
float* out1 = outputs[0];
float* out2 = outputs[1];
- int32_t event=0, frame=0, frames, v;
+ int32_t frame=0, frames, v;
float o, x, e, mw=MW, w=rich, m=modmix;
int32_t k=K;
- if(activevoices>0 || notes[event]<sampleFrames) //detect & bypass completely empty blocks
+ LV2_Atom_Event* ev = lv2_atom_sequence_begin(&eventInput->body);
+ bool end = lv2_atom_sequence_is_end(&eventInput->body, eventInput->atom.size, ev);
+ if(activevoices>0 || !end) //detect & bypass completely empty blocks
{
while(frame<sampleFrames)
{
- frames = notes[event++];
- if(frames>sampleFrames) frames = sampleFrames;
+ end = lv2_atom_sequence_is_end(&eventInput->body, eventInput->atom.size, ev);
+ frames = end ? sampleFrames : ev->time.frames;
frames -= frame;
frame += frames;
@@ -444,11 +366,10 @@ void mdaDX10::processReplacing(float **inputs, float **outputs, int32_t sampleFr
*out2++ = o;
}
- if(frame<sampleFrames) //next note on/off
+ if(!end) //next note on/off
{
- int32_t note = notes[event++];
- int32_t vel = notes[event++];
- noteOn(note, vel);
+ processEvent(ev);
+ ev = lv2_atom_sequence_next(ev);
}
}
@@ -472,7 +393,6 @@ void mdaDX10::processReplacing(float **inputs, float **outputs, int32_t sampleFr
}
}
K=k; MW=mw; //remember these so vibrato speed not buffer size dependant!
- notes[0] = EVENTS_DONE;
}
@@ -528,28 +448,21 @@ void mdaDX10::noteOn(int32_t note, int32_t velocity)
}
-int32_t mdaDX10::processEvents(LvzEvents* ev)
+int32_t mdaDX10::processEvent(const LV2_Atom_Event* ev)
{
- int32_t npos=0;
+ if (ev->body.type != midiEventType)
+ return 0;
- for (int32_t i=0; i<ev->numEvents; i++)
- {
- if((ev->events[i])->type != kLvzMidiType) continue;
- LvzMidiEvent* event = (LvzMidiEvent*)ev->events[i];
- char* midiData = event->midiData;
+ const uint8_t* midiData = (const uint8_t*)LV2_ATOM_BODY(&ev->body);
switch(midiData[0] & 0xf0) //status byte (all channels)
{
case 0x80: //note off
- notes[npos++] = event->deltaFrames; //delta
- notes[npos++] = midiData[1] & 0x7F; //note
- notes[npos++] = 0; //vel
+ noteOn(midiData[1] & 0x7F, 0);
break;
case 0x90: //note on
- notes[npos++] = event->deltaFrames; //delta
- notes[npos++] = midiData[1] & 0x7F; //note
- notes[npos++] = midiData[2] & 0x7F; //vel
+ noteOn(midiData[1] & 0x7F, midiData[2] & 0x7F);
break;
case 0xB0: //controller
@@ -567,9 +480,7 @@ int32_t mdaDX10::processEvents(LvzEvents* ev)
sustain = midiData[2] & 0x40;
if(sustain==0)
{
- notes[npos++] = event->deltaFrames;
- notes[npos++] = SUSTAIN; //end all sustained notes
- notes[npos++] = 0;
+ noteOn(SUSTAIN, 0);
}
break;
@@ -596,10 +507,6 @@ int32_t mdaDX10::processEvents(LvzEvents* ev)
default: break;
}
- if(npos>EVENTBUFFER) npos -= 3; //discard events if buffer full!!
- event++;
- }
- notes[npos] = EVENTS_DONE;
return 1;
}
generated by cgit v1.2.1 (git 2.18.0) at 2024-05-26 19:04:16 +0000