2 files changed, 64 insertions(+), 24 deletions(-)

M lv2ttl/kantan.ttl
M src/main.c

@@ 37,18 37,18 @@
                 a lv2:InputPort ,
                         lv2:ControlPort ;
                 lv2:index 2 ;
-                lv2:symbol "gain" ;
-                lv2:name "Gain" ;
+                lv2:symbol "dist" ;
+                lv2:name "Distortion" ;
                 lv2:default 0.0 ;
                 lv2:minimum 0.0 ;
-                lv2:maximum 10.0 ;
+                lv2:maximum 100.0 ;
         ] , [
                 a lv2:InputPort ,
                         lv2:ControlPort ;
                 lv2:index 3 ;
-                lv2:symbol "offset" ;
-                lv2:name "Offset" ;
+                lv2:symbol "asym" ;
+                lv2:name "Asymmetry" ;
                 lv2:default 0.0 ;
                 lv2:minimum 0.0 ;
-                lv2:maximum 0.5 ;
+                lv2:maximum 100.0 ;
         ] .

@@ 10,11 10,12 @@
 typedef enum {
 	INPUT = 0,
 	OUTPUT = 1,
-	GAIN = 2,
-	OFFSET = 3,
+	DIST = 2,
+	ASYM = 3,
 } PortIndex;
 
 typedef struct {
+	const float r;
 	float prev_x;
 	float prev_y;
 } DCFilter;


@@ 22,22 23,58 @@ typedef struct {
 typedef struct {
 	const float *input;
 	float *output;
-	const float *gain;
-	const float *offset;
+	const float *dist;
+	const float *asym;
 
+	DCFilter dc_filter_rms;
 	DCFilter dc_filter;
 } Kantan;
 
 static float blockdc(const float x, DCFilter* filter)
 {
+	const float r = filter->r;
 	const float prev_x = filter->prev_x;
 	const float prev_y = filter->prev_y;
-	const float y = x - prev_x + 0.995f*prev_y;
+	const float y = x - prev_x + r*prev_y;
 	filter->prev_x = x;
 	filter->prev_y = y;
 	return y;
 }
 
+static float distort(const float x, const float d, const float a)
+{
+	float k;
+	if (x > 0.f) {
+		k = d*(1.f - a * 0.01);
+	} else {
+		k = d;
+	}
+
+	float y;
+	if (k < 0.01f) {
+		y = x;
+	} else {
+		y = 1 / k * atanf(k * x);
+	}
+
+	return y;
+}
+
+static float rms(const float d, const float a, DCFilter *filter)
+{
+	float accm = 0;
+	for (uint32_t i = 0; i < 49; i++) {
+		const float x = i / 48.f;
+		const float y = distort(0.1f * sin(2.f * M_PI * x), d, a);
+		const float y_nodc = blockdc(y, filter);
+		accm += y_nodc*y_nodc;
+	}
+	
+	const float result = sqrtf(accm / 49);
+	
+	return result;
+}
+
 static LV2_Handle instantiate(const LV2_Descriptor * descriptor,
 		double rate,
 		const char *bundle_path,


@@ 46,7 83,12 @@ static LV2_Handle instantiate(const LV2_Descriptor * descriptor,
 
 	Kantan *kantan = (Kantan *) calloc(1, sizeof(Kantan));
 
-	DCFilter dc_filter_init = {0,0};
+	const float wn = M_PI * 10.f / rate;
+	const float b0 = 1.f / (1.f + wn);
+	const float p = (1.f - wn) * b0;
+
+	DCFilter dc_filter_init = {p,0.f,0.f};
+	memcpy(&kantan->dc_filter_rms, &dc_filter_init, sizeof(DCFilter));
 	memcpy(&kantan->dc_filter, &dc_filter_init, sizeof(DCFilter));
 
 	return (LV2_Handle) kantan;


@@ 63,11 105,11 @@ static void connect_port(LV2_Handle instance, uint32_t port, void *data)
 		case OUTPUT:
 			kantan->output = (float *)data;
 			break;
-		case GAIN:
-			kantan->gain = (const float *)data;
+		case DIST:
+			kantan->dist = (const float *)data;
 			break;
-		case OFFSET:
-			kantan->offset = (const float *)data;
+		case ASYM:
+			kantan->asym = (const float *)data;
 			break;
 	}
 }


@@ 82,18 124,16 @@ static void run(LV2_Handle instance, uint32_t n_samples)
 
 	const float *const input = kantan->input;
 	float *const output = kantan->output;
-	const float gain = *(kantan->gain);
-	const float offset = *(kantan->offset);
+	const float dist = *(kantan->dist);
+	const float asym = *(kantan->asym);
+	DCFilter *dc_filter_rms = &kantan->dc_filter_rms;
 	DCFilter* dc_filter = &kantan->dc_filter;
 
-	const float g_pre = pow (10.0, (gain * 0.5f));
-	const float g_post = 0.1f / atan(0.1f * g_pre);
+	const float rms_ref = 0.1f / sqrtf(2);
+	const float g_post = rms_ref / rms(dist, asym, dc_filter_rms);
 
-	/**
-	  distortion using sigmoid function: 2 / pi * arctan (pi / 2 * x)
-	  */
 	for (uint32_t pos = 0; pos < n_samples; pos++) {
-		output[pos] = blockdc(atan((input[pos] * g_pre + offset)), dc_filter) * g_post;
+		output[pos] = blockdc(distort(input[pos], dist, asym), dc_filter) * g_post;
 	}
 }