Update server.py

server.py

@@ -86,8 +86,15 @@ def extract_audio_segment(video_path: str, start_seconds: float, end_seconds: fl
         "-ac", "1",
         output_wav
     ]
-    result = subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
-    return result.returncode == 0
+    result = subprocess.run(cmd, capture_output=True, text=True)
+    if result.returncode != 0:
+        print(f"  ✗ FFmpeg audio extraction failed: {result.stderr}")
+        return False
+    if not os.path.exists(output_wav):
+        print(f"  ✗ Output WAV file not created: {output_wav}")
+        return False
+    print(f"  ✓ Audio extracted successfully")
+    return True
 
 
 def transcribe_segment(audio_path: str) -> List[Tuple[float, float, str]]:
@@ -258,7 +265,10 @@ def process_video_segment(
         if audio_ok and whisper_model is not None:
             captions = transcribe_segment(temp_wav)
         else:
-            print("  Warning: Skipping transcription (audio failed or model not ready)")
+            if not audio_ok:
+                print("  ✗ Skipping transcription: audio extraction failed")
+            elif whisper_model is None:
+                print("  ✗ Skipping transcription: Whisper model not ready")
             captions = []
 
         frame_caption_map = build_frame_caption_map(captions, fps)
@@ -314,10 +324,8 @@ def process_video_segment(
             frame = cv2.resize(frame, (target_width, target_height), interpolation=cv2.INTER_LANCZOS4)
             frame = apply_color_grading_wedding_retro(frame)
 
-            current_caption = frame_caption_map.get(processed_frames, current_caption)
-            # Clear caption if this frame isn't in the map and the previous caption has ended
-            if processed_frames not in frame_caption_map:
-                current_caption = ""
+            # Set caption for this frame (empty if none).
+            current_caption = frame_caption_map.get(processed_frames, "")
 
             if current_caption:
                 frame = burn_captions_to_frame(frame, current_caption)
@@ -508,9 +516,10 @@ async def scan_and_process_videos():
         print("Video processing already running, skipping...")
         return
 
-    # Wait 3 minutes for Space to fully initialize
-    print("Waiting 3 minutes before starting video processing...")
-    await asyncio.sleep(180)
+    # Wait for Space to fully initialize (reduced for testing)
+    startup_delay = int(os.getenv("STARTUP_DELAY", 5))  # Default 5 seconds for testing
+    print(f"Waiting {startup_delay} seconds before starting video processing...")
+    await asyncio.sleep(startup_delay)
 
     processing_state["is_running"] = True
     print("\n" + "="*80)