From f5d1cccce68c20a135188327628c6b6cb146e5f0 Mon Sep 17 00:00:00 2001
From: zeyus <zeyus@zeyus.com>
Date: Mon, 24 Feb 2025 22:58:58 +0100
Subject: [PATCH] flip.ino is working ish (terribly) on the uno mega

---
 flip.ino        | 334 ++++++++++++++++++++++++++++++++++++++++++++++++
 sketch copy.ino | 113 ++++++++++++++++
 sketch.ino      | 151 ++++++++++------------
 3 files changed, 516 insertions(+), 82 deletions(-)
 create mode 100644 flip.ino
 create mode 100644 sketch copy.ino

diff --git a/flip.ino b/flip.ino
new file mode 100644
index 0000000..f3efc71
--- /dev/null
+++ b/flip.ino
@@ -0,0 +1,334 @@
+#include <Arduino.h>
+#include <lcdgfx.h>
+#include <FastTrig.h>
+
+// this is working on the mega
+
+DisplaySSD1306_128x64_I2C display(-1); 
+
+// Simulation constants
+#define SIM_WIDTH 128
+#define SIM_HEIGHT 64
+#define NUM_PARTICLES 80
+#define PARTICLE_RADIUS 2
+#define GRAVITY 0.9f
+#define DAMPING 0.92f
+#define PRESSURE_RADIUS 4.5f
+#define PRESSURE_FORCE 0.6f
+#define MAX_VEL 10.0f
+
+#ifdef __AVR__
+  typedef int16_t fixed_t;
+  #define FIX_SHIFT 6
+  #define TO_FIXED(x) ((fixed_t)((x) * (1<<FIX_SHIFT)))
+  #define TO_FLOAT(x) ((x) / (1<<FIX_SHIFT))
+  
+  struct Particle {
+    fixed_t x, y;
+    fixed_t vx, vy;
+  };
+  
+#else
+  typedef float f32 __attribute__((aligned(4)));
+  struct Particle {
+    f32 x, y;
+    f32 vx, vy;
+  };
+#endif
+
+inline float fast_sqrt(float x) {
+  union { float f; uint32_t i; } u;
+  u.f = x;
+  u.i = 0x5f375a86 - (u.i >> 1);
+  return u.f * (1.5f - 0.5f * x * u.f * u.f);
+}
+
+Particle particles[NUM_PARTICLES];
+uint8_t canvasData[SIM_WIDTH*(SIM_HEIGHT/8)];
+NanoCanvas1 canvas(SIM_WIDTH, SIM_HEIGHT, canvasData);
+
+
+#define GRID_SIZE 16
+#define CELL_SIZE (SIM_WIDTH/GRID_SIZE)
+
+struct GridCell {
+  uint8_t particles[10];
+  uint8_t count;
+};
+
+GridCell grid[GRID_SIZE][GRID_SIZE];
+
+void buildSpatialGrid() {
+  memset(grid, 0, sizeof(grid));
+
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    #ifdef __AVR__
+      float x = TO_FLOAT(particles[i].x);
+      float y = TO_FLOAT(particles[i].y);
+    #else
+      float x = particles[i].x;
+      float y = particles[i].y;
+    #endif
+    
+    int gx = constrain(x / CELL_SIZE, 0, GRID_SIZE-1);
+    int gy = constrain(y / CELL_SIZE, 0, GRID_SIZE-1);
+    
+    if(grid[gx][gy].count < 10) {
+      grid[gx][gy].particles[grid[gx][gy].count++] = i;
+    }
+  }
+}
+
+
+void setup() {
+  // String str; // Hhack
+  display.begin();
+  display.clear();
+  canvas.setMode(CANVAS_MODE_TRANSPARENT);
+  
+  // Serial.begin(115200);
+  // while (!Serial);
+
+  // Initialize particles in a droplet pattern
+  float cx = SIM_WIDTH/2;
+  float cy = SIM_HEIGHT/4;
+
+  #ifdef __AVR__
+    for(int i=0; i<NUM_PARTICLES; i++) {
+      float angle = random(360) * PI / 180.0f;
+      float radius = random(10);
+      
+      float x_offset = icos(angle) * radius; 
+      float y_offset = isin(angle) * radius;
+
+      // Constrain BEFORE fixed-point conversion
+      float xpos = constrain(cx + x_offset, 0.0f, SIM_WIDTH-1.0f);
+      float ypos = constrain(cy + y_offset, 0.0f, SIM_HEIGHT-1.0f);
+      
+      particles[i].x = TO_FIXED(xpos);
+      particles[i].y = TO_FIXED(ypos);
+
+      //Serial.print(str+"Particle "+i+": (angle)"+angle+", (radius)"+radius+" (offsets)"+x_offset+","+y_offset+" -> (x,y)"+particles[i].x+","+particles[i].y+"\n");
+      
+      // Ensure velocities are within range
+      particles[i].vx = TO_FIXED(constrain((random(-50,50)/25.0f), -2.0f, 2.0f));
+      particles[i].vy = TO_FIXED(constrain((random(-25,50)/25.0f), -1.0f, 2.0f));
+    }
+  #else
+    for(int i=0; i<NUM_PARTICLES; i++) {
+      float angle = random(360) * PI / 180.0;
+      float radius = random(10);
+      particles[i].x = cx + icos(angle) * radius;
+      particles[i].y = cy + isin(angle) * radius;
+      particles[i].vx = random(-50,50)/25.0;  // -2 to +2
+      particles[i].vy = random(-25,50)/25.0;   // -1 to +2
+    }
+  #endif
+}
+
+void applyPhysics() {
+#ifdef __AVR__
+  // Fixed-point implementation
+  const fixed_t gravity = TO_FIXED(GRAVITY);
+  const fixed_t damping = TO_FIXED(DAMPING);
+  const fixed_t pressure_radius = TO_FIXED(PRESSURE_RADIUS);
+  const fixed_t max_vel = TO_FIXED(MAX_VEL);
+
+  // Convert spatial grid to fixed-point
+  buildSpatialGrid();
+  const int32_t pressure_radius_sq = ((int32_t)pressure_radius * pressure_radius) >> FIX_SHIFT;
+
+  // Particle interactions
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    const int gx = TO_FLOAT(particles[i].x) / CELL_SIZE;
+    const int gy = TO_FLOAT(particles[i].y) / CELL_SIZE;
+
+    for(int dx=-1; dx<=1; dx++) {
+      for(int dy=-1; dy<=1; dy++) {
+        if(gx+dx < 0 || gx+dx >= GRID_SIZE) continue;
+        if(gy+dy < 0 || gy+dy >= GRID_SIZE) continue;
+        
+        GridCell &cell = grid[gx+dx][gy+dy];
+        for(int c=0; c<cell.count; c++) {
+          const int j = cell.particles[c];
+          if(j <= i) continue;
+
+          const fixed_t dx = particles[j].x - particles[i].x;
+          const fixed_t dy = particles[j].y - particles[i].y;
+          const fixed_t dist_sq = (dx*dx + dy*dy) >> FIX_SHIFT;
+
+          if(dist_sq < pressure_radius_sq && dist_sq > TO_FIXED(0.01f)) {
+            // Convert to float for precise calculation
+            float fx = TO_FLOAT(dx);
+            float fy = TO_FLOAT(dy);
+            float dist = sqrt(fx*fx + fy*fy) + 0.001f;
+            
+            // Normalize direction vector
+            float nx = fx / dist;
+            float ny = fy / dist;
+            
+            // Calculate force magnitude
+            float force = PRESSURE_FORCE * (1.0f - dist/TO_FLOAT(pressure_radius));
+            
+            // Apply force to both particles
+            float fx_impulse = force * nx;
+            float fy_impulse = force * ny * 0.7f;  // Reduce vertical effect
+            
+            particles[i].vx -= TO_FIXED(fx_impulse);
+            particles[i].vy -= TO_FIXED(fy_impulse);
+            particles[j].vx += TO_FIXED(fx_impulse); 
+            particles[j].vy += TO_FIXED(fy_impulse);
+          }
+        }
+      }
+    }
+  }
+
+  // Physics update with constraints
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    // Apply gravity
+    particles[i].vy += gravity;
+
+    // Update position
+    particles[i].x += particles[i].vx;
+    particles[i].y += particles[i].vy;
+
+    // Velocity constraints
+    particles[i].vx = constrain(particles[i].vx, -TO_FIXED(MAX_VEL*2), TO_FIXED(MAX_VEL*2));
+    particles[i].vy = constrain(particles[i].vy, -TO_FIXED(MAX_VEL*2), TO_FIXED(MAX_VEL*2));
+    // Simple floor collision
+    // if(TO_FLOAT(particles[i].y) > SIM_HEIGHT-1) {
+    //   particles[i].y = TO_FIXED(SIM_HEIGHT-1);
+    //   particles[i].vy = -particles[i].vy * damping;
+    // }
+    // Boundary checks
+    const float x = TO_FLOAT(particles[i].x);
+    const float y = TO_FLOAT(particles[i].y);
+    
+    if(x <= 0 || x >= SIM_WIDTH-1) {
+      particles[i].vx = (-damping * particles[i].vx) >> FIX_SHIFT;
+      particles[i].x = TO_FIXED(constrain(x, 1, SIM_WIDTH-2));
+    }
+    
+    if(y <= 0 || y >= SIM_HEIGHT-1) {
+      particles[i].vy = (-damping * particles[i].vy) >> FIX_SHIFT;
+      particles[i].y = TO_FIXED(constrain(y, 1, SIM_HEIGHT-2));
+    }
+  }
+#else
+  // Build spatial grid
+  buildSpatialGrid();
+
+  // Interactions using grid
+  const float PRESSURE_RADIUS_SQ = PRESSURE_RADIUS * PRESSURE_RADIUS;
+  
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    const int gx = particles[i].x / CELL_SIZE;
+    const int gy = particles[i].y / CELL_SIZE;
+
+    // Check 3x3 grid around particle
+    for(int dx=-1; dx<=1; dx++) {
+      for(int dy=-1; dy<=1; dy++) {
+        if(gx+dx < 0 || gx+dx >= GRID_SIZE) continue;
+        if(gy+dy < 0 || gy+dy >= GRID_SIZE) continue;
+        
+        GridCell &cell = grid[gx+dx][gy+dy];
+        for(int c=0; c<cell.count; c++) {
+          const int j = cell.particles[c];
+          if(j <= i) continue; // Avoid duplicate pairs
+
+          const float dx = particles[j].x - particles[i].x;
+          const float dy = particles[j].y - particles[i].y;
+          const float dist_sq = dx*dx + dy*dy;
+
+          if(dist_sq < PRESSURE_RADIUS_SQ && dist_sq > 0.01f) {
+            const float dist = fast_sqrt(dist_sq) + 0.001f;
+            const float force = PRESSURE_FORCE * (1.0f - dist/PRESSURE_RADIUS);
+            
+            // Only apply horizontal forces to preserve gravity
+            particles[i].vx -= force * dx/dist;
+            particles[j].vx += force * dx/dist;
+            
+            // Reduce vertical force impact
+            particles[i].vy -= force * dy/dist * 0.3f;
+            particles[j].vy += force * dy/dist * 0.3f;
+          }
+        }
+      }
+    }
+  }
+
+    // Gravity and movement
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    particles[i].vy += GRAVITY;
+    particles[i].x += particles[i].vx;
+    particles[i].y += particles[i].vy;
+    particles[i].vx = constrain(particles[i].vx, -MAX_VEL, MAX_VEL);
+    particles[i].vy = constrain(particles[i].vy, -MAX_VEL, MAX_VEL);
+    // X-axis
+    if(particles[i].x <= 0 || particles[i].x >= SIM_WIDTH-1) {
+      particles[i].vx *= -DAMPING;
+      particles[i].x = constrain(particles[i].x, 1, SIM_WIDTH-2);
+    }
+    
+    // Y-axis
+    if(particles[i].y <= 0 || particles[i].y >= SIM_HEIGHT-1) {
+      particles[i].vy *= -DAMPING;
+      particles[i].y = constrain(particles[i].y, 1, SIM_HEIGHT-2);
+    }
+  }
+#endif
+}
+
+// Direct canvas buffer access (replace drawParticles)
+void drawParticles() {
+  // Clear canvas by direct memory access
+  memset(canvasData, 0, sizeof(canvasData));
+  // canvas.clear();
+  // bool occupied[SIM_WIDTH][SIM_HEIGHT] = {false};
+  for(int i=0; i<NUM_PARTICLES; i++) {
+    #ifdef __AVR__
+      int x = static_cast<int>(TO_FLOAT(particles[i].x) + 0.5f);
+      int y = static_cast<int>(TO_FLOAT(particles[i].y) + 0.5f);
+    #else
+      int x = constrain(static_cast<int>(particles[i].x + 0.5f), 0, SIM_WIDTH-1);
+      int y = constrain(static_cast<int>(particles[i].y + 0.5f), 0, SIM_HEIGHT-1);
+    #endif
+
+    #if PARTICLE_RADIUS == 1
+      canvas.putPixel(x, y);
+    #else
+      canvas.drawCircle(x,y,PARTICLE_RADIUS);
+    #endif
+    // if(!occupied[x][y]) {
+    //   canvas.putPixel(x, y);
+    //   occupied[x][y] = true;
+      
+    //   // Add neighbor pixels if using 2x2
+    //   if(x < SIM_WIDTH-1 && !occupied[x+1][y]) {
+    //     canvas.putPixel(x+1, y);
+    //     occupied[x+1][y] = true;
+    //   }
+    //   if(y < SIM_HEIGHT-1 && !occupied[x][y+1]) {
+    //     canvas.putPixel(x, y+1);
+    //     occupied[x][y+1] = true;
+    //   }
+    // }
+  }
+  
+  display.drawCanvas(0,0,canvas);
+}
+
+
+void loop() {
+  static uint32_t last_frame = 0;
+  //lcd_delay(5000);
+  drawParticles();
+  //delay(1000);
+  if(millis() - last_frame >= 33) {
+    last_frame = millis();
+    applyPhysics();
+  } else {
+    lcd_delay(millis() - last_frame);
+  }
+}
diff --git a/sketch copy.ino b/sketch copy.ino
new file mode 100644
index 0000000..a0e635c
--- /dev/null
+++ b/sketch copy.ino	
@@ -0,0 +1,113 @@
+#include <lcdgfx.h>
+#include <nano_engine_v2.h>
+#include <nano_gfx_types.h>
+
+#define BAUD_RATE   115200
+#include <Arduino.h>
+#include <lcdgfx.h>
+// The parameters are  RST pin, BUS number, CS pin, DC pin, FREQ (0 means default), CLK pin, MOSI pin
+//DisplaySSD1325_128x64_SPI display(3,{-1, 4, 5, 0,-1,-1});   // Use this line for Atmega328p (3=RST, 4=CE, 5=D/C)
+//DisplaySSD1325_128x64_I2C display(-1);                    // or (-1,{busId, addr, scl, sda, frequency}). This line is suitable for most platforms by default
+//DisplaySSD1325_128x64_SPI display(22,{-1, 5, 21, 0,-1,-1}); // Use this line for ESP32 (VSPI)  (gpio22=RST, gpio5=CE for VSPI, gpio21=D/C)
+//DisplaySSD1325_128x64_SPI display(4,{-1, -1, 5, 0,-1,-1});  // Use this line for ESP8266 Arduino style rst=4, CS=-1, DC=5
+                                                          // And ESP8266 RTOS IDF. GPIO4 is D2, GPIO5 is D1 on NodeMCU boards
+
+
+DisplaySSD1306_128x64_I2C display(-1);
+
+#define PANEL_RES_X 128
+#define PANEL_RES_Y 64
+#define SIM_WIDTH (PANEL_RES_X)
+#define SIM_HEIGHT PANEL_RES_Y
+#define NUM_PARTICLES 255
+#define FIXED_SHIFT 8
+
+typedef int16_t fixed_t;
+#define TO_FIXED(x) ((fixed_t)((x) * (1 << FIXED_SHIFT)))
+#define TO_FLOAT(x) ((float)(x) / (1 << FIXED_SHIFT))
+
+struct Particle {
+  fixed_t x, y;
+  fixed_t vx, vy;
+};
+Particle particles[NUM_PARTICLES];
+
+const int canvasWidth = SIM_WIDTH; // Width must be power of 2, i.e. 16, 32, 64, 128...
+const int canvasHeight = SIM_HEIGHT; // Height must be divided on 8, i.e. 8, 16, 24, 32...
+const fixed_t gravity = TO_FIXED(0.4);
+const fixed_t damping = TO_FIXED(0.6);
+const fixed_t border = TO_FIXED(0.1);
+const uint32_t frame_time = 33; // ~30 FPS
+
+uint8_t canvasData[canvasWidth*(canvasHeight/8)];
+NanoCanvas1 canvas(canvasWidth, canvasHeight, canvasData);
+
+void updatePhysics() {
+  const fixed_t max_x = TO_FIXED(SIM_WIDTH-1);
+  const fixed_t max_y = TO_FIXED(SIM_HEIGHT-1);
+
+  for (int i = 0; i < NUM_PARTICLES; i++) {
+    particles[i].vy += gravity;
+    particles[i].x += particles[i].vx;
+    particles[i].y += particles[i].vy;
+
+    particles[i].x = constrain(particles[i].x, TO_FIXED(0), max_x);
+    particles[i].y = constrain(particles[i].y, TO_FIXED(0), max_y);
+
+    // Boundary collisions
+    if (TO_FLOAT(particles[i].x) <= TO_FLOAT(border) || TO_FLOAT(particles[i].x) >= (SIM_WIDTH - TO_FLOAT(border))) {
+      particles[i].vx = -particles[i].vx * damping;
+      particles[i].x = constrain(particles[i].x, border, TO_FIXED(SIM_WIDTH) - border);
+    }
+    if (TO_FLOAT(particles[i].y) < TO_FLOAT(border) || TO_FLOAT(particles[i].y) >= (SIM_HEIGHT - TO_FLOAT(border))) {
+      particles[i].vy = -particles[i].vy * damping;
+      particles[i].y = constrain(particles[i].y, border, TO_FIXED(SIM_HEIGHT) - border);
+    }
+
+  }
+}
+
+void drawParticles() {
+  canvas.clear();
+  for (int i = 0; i < NUM_PARTICLES; i++) {
+    int x = constrain(TO_FLOAT(particles[i].x), 0, SIM_WIDTH-1);
+    int y = constrain(TO_FLOAT(particles[i].y), 0, SIM_HEIGHT-1);
+    
+    canvas.putPixel(x, y);
+  }
+  display.drawCanvas(0,0,canvas);
+}
+
+// Setup, initialize 
+void setup() 
+{
+  //Serial.begin(115200);
+  display.begin();
+  display.clear();
+  canvas.setMode( CANVAS_MODE_TRANSPARENT );
+  // display.setColor( 65535 );
+  
+  // Initialize particles
+  for (int i = 0; i < NUM_PARTICLES; i++) {
+    particles[i].x = TO_FIXED(random(SIM_WIDTH-1));
+    particles[i].y = TO_FIXED(random(SIM_HEIGHT/2));
+    particles[i].vx = TO_FIXED((random(-100, 100)/100.0) * 0.7);
+    particles[i].vy = TO_FIXED((random(-50, 50)/100.0));
+  }
+}
+
+// Loop forever
+void loop() 
+{
+  // static uint32_t last_frame = 0;
+  
+  
+  //if (millis() - last_frame >= frame_time) {
+    //Serial.print(display.getColor());
+    //Serial.print("Hi");  
+    updatePhysics();
+    drawParticles();
+    //last_frame = millis(); 
+  //}
+  lcd_delay(frame_time);
+}
diff --git a/sketch.ino b/sketch.ino
index 79bcd8d..a0e635c 100644
--- a/sketch.ino
+++ b/sketch.ino
@@ -1,13 +1,25 @@
-#include <Arduino.h>
-#include <ESP32-HUB75-MatrixPanel-I2S-DMA.h>
+#include <lcdgfx.h>
+#include <nano_engine_v2.h>
+#include <nano_gfx_types.h>
 
-// Config
-#define PANEL_RES_X 64
-#define PANEL_RES_Y 32
-#define NUM_PANELS 2
-#define SIM_WIDTH (PANEL_RES_X * NUM_PANELS)
+#define BAUD_RATE   115200
+#include <Arduino.h>
+#include <lcdgfx.h>
+// The parameters are  RST pin, BUS number, CS pin, DC pin, FREQ (0 means default), CLK pin, MOSI pin
+//DisplaySSD1325_128x64_SPI display(3,{-1, 4, 5, 0,-1,-1});   // Use this line for Atmega328p (3=RST, 4=CE, 5=D/C)
+//DisplaySSD1325_128x64_I2C display(-1);                    // or (-1,{busId, addr, scl, sda, frequency}). This line is suitable for most platforms by default
+//DisplaySSD1325_128x64_SPI display(22,{-1, 5, 21, 0,-1,-1}); // Use this line for ESP32 (VSPI)  (gpio22=RST, gpio5=CE for VSPI, gpio21=D/C)
+//DisplaySSD1325_128x64_SPI display(4,{-1, -1, 5, 0,-1,-1});  // Use this line for ESP8266 Arduino style rst=4, CS=-1, DC=5
+                                                          // And ESP8266 RTOS IDF. GPIO4 is D2, GPIO5 is D1 on NodeMCU boards
+
+
+DisplaySSD1306_128x64_I2C display(-1);
+
+#define PANEL_RES_X 128
+#define PANEL_RES_Y 64
+#define SIM_WIDTH (PANEL_RES_X)
 #define SIM_HEIGHT PANEL_RES_Y
-#define NUM_PARTICLES 100
+#define NUM_PARTICLES 255
 #define FIXED_SHIFT 8
 
 typedef int16_t fixed_t;
@@ -18,109 +30,84 @@ struct Particle {
   fixed_t x, y;
   fixed_t vx, vy;
 };
-
 Particle particles[NUM_PARTICLES];
-MatrixPanel_I2S_DMA *dma_display;
 
-/*
-// Wokwi-compatible pin configuration
-HUB75_I2S_CFG::i2s_pins _pins = {
-  .r1 = 25, .g1 = 26, .b1 = 27, 
-  .r2 = 14, .g2 = 12, .b2 = 13, 
-  .a = 23, .b = 19, .c = 5, .d = 17, 
-  .e = -1, // Required for 64x64 panels
-  .lat = 4, .oe = 15, .clk = 16
-};
-*/
+const int canvasWidth = SIM_WIDTH; // Width must be power of 2, i.e. 16, 32, 64, 128...
+const int canvasHeight = SIM_HEIGHT; // Height must be divided on 8, i.e. 8, 16, 24, 32...
+const fixed_t gravity = TO_FIXED(0.4);
+const fixed_t damping = TO_FIXED(0.6);
+const fixed_t border = TO_FIXED(0.1);
+const uint32_t frame_time = 33; // ~30 FPS
 
-// ESP32-S3-WROOM-1 HUB75 Pin Mapping
-HUB75_I2S_CFG::i2s_pins _pins = {
-  .r1 = 1,  .g1 = 2,  .b1 = 3,
-  .r2 = 4,  .g2 = 5,  .b2 = 6,
-  .a = 7,   .b = 8,   .c = 9,
-  .d = 10,  .e = -1,  // 'e' only needed for 64x64 panels
-  .lat = 11, .oe = 12, .clk = 13
-};
-
-// Color palette
-const uint16_t COLORS[] = {
-  0x001F, 0x03FF, 0x07FF, 0x7FE0, 0x7F80, 0xFFE0, 0xFD20, 0xF800
-};
-const int NUM_COLORS = sizeof(COLORS)/sizeof(COLORS[0]);
-
-void setup() {
-
-  HUB75_I2S_CFG mxconfig(
-    PANEL_RES_X,
-    PANEL_RES_Y,
-    NUM_PANELS,
-    _pins,
-    HUB75_I2S_CFG::FM6126A,
-    false,
-    HUB75_I2S_CFG::HZ_10M,
-    true,
-    HUB75_I2S_CFG::SHIFTREG,
-    false,
-    0
-  );
-
-  mxconfig.double_buff = true;
-  dma_display = new MatrixPanel_I2S_DMA(mxconfig);
-  dma_display->begin();
-  dma_display->setBrightness(255);
-
-  // Initialize particles
-  for (int i = 0; i < NUM_PARTICLES; i++) {
-    particles[i].x = TO_FIXED(random(SIM_WIDTH));
-    particles[i].y = TO_FIXED(random(SIM_HEIGHT/2));
-    particles[i].vx = TO_FIXED((random(-100, 100)/100.0) * 0.7);
-    particles[i].vy = TO_FIXED((random(-50, 50)/100.0));
-  }
-}
+uint8_t canvasData[canvasWidth*(canvasHeight/8)];
+NanoCanvas1 canvas(canvasWidth, canvasHeight, canvasData);
 
 void updatePhysics() {
-  const fixed_t gravity = TO_FIXED(0.15);
-  const fixed_t damping = TO_FIXED(0.82);
-  const fixed_t border = TO_FIXED(2.0);
+  const fixed_t max_x = TO_FIXED(SIM_WIDTH-1);
+  const fixed_t max_y = TO_FIXED(SIM_HEIGHT-1);
 
   for (int i = 0; i < NUM_PARTICLES; i++) {
     particles[i].vy += gravity;
     particles[i].x += particles[i].vx;
     particles[i].y += particles[i].vy;
 
+    particles[i].x = constrain(particles[i].x, TO_FIXED(0), max_x);
+    particles[i].y = constrain(particles[i].y, TO_FIXED(0), max_y);
+
     // Boundary collisions
-    if (particles[i].x < border || particles[i].x >= TO_FIXED(SIM_WIDTH) - border) {
+    if (TO_FLOAT(particles[i].x) <= TO_FLOAT(border) || TO_FLOAT(particles[i].x) >= (SIM_WIDTH - TO_FLOAT(border))) {
       particles[i].vx = -particles[i].vx * damping;
       particles[i].x = constrain(particles[i].x, border, TO_FIXED(SIM_WIDTH) - border);
     }
-    if (particles[i].y < border || particles[i].y >= TO_FIXED(SIM_HEIGHT) - border) {
+    if (TO_FLOAT(particles[i].y) < TO_FLOAT(border) || TO_FLOAT(particles[i].y) >= (SIM_HEIGHT - TO_FLOAT(border))) {
       particles[i].vy = -particles[i].vy * damping;
       particles[i].y = constrain(particles[i].y, border, TO_FIXED(SIM_HEIGHT) - border);
     }
+
   }
 }
 
 void drawParticles() {
-  dma_display->fillScreen(0);
-  
+  canvas.clear();
   for (int i = 0; i < NUM_PARTICLES; i++) {
     int x = constrain(TO_FLOAT(particles[i].x), 0, SIM_WIDTH-1);
     int y = constrain(TO_FLOAT(particles[i].y), 0, SIM_HEIGHT-1);
     
-    // Simplified color selection
-    uint16_t color = COLORS[(abs(particles[i].vx) + abs(particles[i].vy)) % NUM_COLORS];
-    
-    dma_display->drawPixel(x, y, color);
+    canvas.putPixel(x, y);
+  }
+  display.drawCanvas(0,0,canvas);
+}
+
+// Setup, initialize 
+void setup() 
+{
+  //Serial.begin(115200);
+  display.begin();
+  display.clear();
+  canvas.setMode( CANVAS_MODE_TRANSPARENT );
+  // display.setColor( 65535 );
+  
+  // Initialize particles
+  for (int i = 0; i < NUM_PARTICLES; i++) {
+    particles[i].x = TO_FIXED(random(SIM_WIDTH-1));
+    particles[i].y = TO_FIXED(random(SIM_HEIGHT/2));
+    particles[i].vx = TO_FIXED((random(-100, 100)/100.0) * 0.7);
+    particles[i].vy = TO_FIXED((random(-50, 50)/100.0));
   }
 }
 
-void loop() {
-  static uint32_t last_frame = 0;
-  const uint32_t frame_time = 33; // ~30 FPS
-
-  if (millis() - last_frame >= frame_time) {
+// Loop forever
+void loop() 
+{
+  // static uint32_t last_frame = 0;
+  
+  
+  //if (millis() - last_frame >= frame_time) {
+    //Serial.print(display.getColor());
+    //Serial.print("Hi");  
     updatePhysics();
     drawParticles();
-    last_frame = millis();
-  }
+    //last_frame = millis(); 
+  //}
+  lcd_delay(frame_time);
 }