diff --git a/croppa/main.py b/croppa/main.py
index cf36c6f..4d985d1 100644
--- a/croppa/main.py
+++ b/croppa/main.py
@@ -6,9 +6,6 @@ import numpy as np
 from pathlib import Path
 from typing import Optional, Tuple, List
 import time
-import threading
-from concurrent.futures import ThreadPoolExecutor
-from queue import Queue
 
 
 class VideoEditor:
@@ -547,96 +544,45 @@ class VideoEditor:
             print("Error: Could not open video writer!")
             return False
         
+        # Simple sequential processing - the I/O is the bottleneck anyway
         total_output_frames = end_frame - start_frame + 1
-        frames_written = 0
         last_progress_update = 0
         
-        # Batch processing configuration
-        batch_size = min(50, max(10, total_output_frames // 20))  # Adaptive batch size
-        frame_queue = Queue(maxsize=batch_size * 2)
-        processed_queue = Queue(maxsize=batch_size * 2)
-        
-        def frame_reader():
-            """Background thread for reading frames"""
-            try:
-                for frame_idx in range(start_frame, end_frame + 1):
-                    self.cap.set(cv2.CAP_PROP_POS_FRAMES, frame_idx)
-                    ret, frame = self.cap.read()
-                    if ret:
-                        frame_queue.put((frame_idx, frame))
-                    else:
-                        break
-            finally:
-                frame_queue.put(None)  # Signal end of frames
-        
-        def frame_processor():
-            """Background thread for processing frames"""
-            try:
-                while True:
-                    item = frame_queue.get()
-                    if item is None:  # End signal
-                        break
-                    
-                    frame_idx, frame = item
-                    processed_frame = self._process_frame_for_render(frame, output_width, output_height)
-                    
-                    if processed_frame is not None:
-                        processed_queue.put((frame_idx, processed_frame))
-                    
-                    frame_queue.task_done()
-            finally:
-                processed_queue.put(None)  # Signal end of processing
-        
-        # Start background threads
-        reader_thread = threading.Thread(target=frame_reader, daemon=True)
-        processor_thread = threading.Thread(target=frame_processor, daemon=True)
-        
-        reader_thread.start()
-        processor_thread.start()
-        
-        # Main thread writes frames in order
-        expected_frame = start_frame
-        frame_buffer = {}  # Buffer for out-of-order frames
-        
-        while frames_written < total_output_frames:
-            item = processed_queue.get()
-            if item is None:  # End signal
+        for frame_idx in range(start_frame, end_frame + 1):
+            # Read frame
+            self.cap.set(cv2.CAP_PROP_POS_FRAMES, frame_idx)
+            ret, frame = self.cap.read()
+            
+            if not ret:
                 break
             
-            frame_idx, processed_frame = item
-            frame_buffer[frame_idx] = processed_frame
+            # Process and write frame directly (minimize memory copies)
+            processed_frame = self._process_frame_for_render(frame, output_width, output_height)
             
-            # Write frames in order
-            while expected_frame in frame_buffer:
-                out.write(frame_buffer.pop(expected_frame))
-                frames_written += 1
-                expected_frame += 1
+            if processed_frame is not None:
+                out.write(processed_frame)
                 
-                # Progress update (throttled to avoid too frequent updates)
+                frames_written = frame_idx - start_frame + 1
+                
+                # Throttled progress update
                 current_time = time.time()
-                if current_time - last_progress_update > 0.5:  # Update every 0.5 seconds
+                if current_time - last_progress_update > 0.5:
                     progress = frames_written / total_output_frames * 100
                     elapsed = current_time - start_time
-                    if frames_written > 0:
-                        eta = (elapsed / frames_written) * (total_output_frames - frames_written)
-                        fps_rate = frames_written / elapsed
-                        print(f"Progress: {progress:.1f}% | {frames_written}/{total_output_frames} | "
-                              f"FPS: {fps_rate:.1f} | ETA: {eta:.1f}s\r", end="")
+                    fps_rate = frames_written / elapsed
+                    eta = (elapsed / frames_written) * (total_output_frames - frames_written)
+                    print(f"Progress: {progress:.1f}% | {frames_written}/{total_output_frames} | "
+                          f"FPS: {fps_rate:.1f} | ETA: {eta:.1f}s\r", end="")
                     last_progress_update = current_time
-            
-            processed_queue.task_done()
-        
-        # Wait for threads to complete
-        reader_thread.join()
-        processor_thread.join()
         
         out.release()
         
         total_time = time.time() - start_time
-        avg_fps = frames_written / total_time if total_time > 0 else 0
+        total_frames_written = end_frame - start_frame + 1
+        avg_fps = total_frames_written / total_time if total_time > 0 else 0
         
         print(f"\nVideo rendered successfully to {output_path}")
-        print(f"Rendered {frames_written} frames in {total_time:.2f}s (avg {avg_fps:.1f} FPS)")
+        print(f"Rendered {total_frames_written} frames in {total_time:.2f}s (avg {avg_fps:.1f} FPS)")
         return True
     
     def _process_frame_for_render(self, frame, output_width: int, output_height: int):