Update app.py

app.py

@@ -249,9 +249,9 @@ def process_entry(entry, i, video_width, video_height, add_voiceover, target_lan
         logger.info("Methods in AudioFileClip:")
         for method in dir(audio_clip):
             logger.info(method)
-        audio_segment = audio_clip.with_start(entry["start"]).with_duration(entry["end"] - entry["start"])  # No subclip here
+        audio_segment = audio_clip.with_duration(entry["end"] - entry["start"])  # No subclip here
 
-    return txt_clip, audio_segment
+    return i, txt_clip, audio_segment
 
 def add_transcript_voiceover(video_path, translated_json, output_path, add_voiceover=False, target_language="en"):
     """
@@ -267,18 +267,21 @@ def add_transcript_voiceover(video_path, translated_json, output_path, add_voice
         futures = [executor.submit(process_entry, entry, i, video.w, video.h, add_voiceover, target_language) 
                    for i, entry in enumerate(translated_json)]
 
+        # Collect results with original index i
+        results = []
         for future in concurrent.futures.as_completed(futures):
             try:
-                txt_clip, audio_segment = future.result()
-                text_clips.append(txt_clip)
-                if add_voiceover and audio_segment:
-                    audio_segments.append(audio_segment)
+                i, txt_clip, audio_segment = future.result()
+                results.append((i, txt_clip, audio_segment))
             except Exception as e:
                 logger.error(f"Error processing entry: {e}")
-
-    # Sort text clips and audio segments based on their start times
-    text_clips.sort(key=lambda clip: clip.start)
+                
+    # Sort by original index i
+    results.sort(key=lambda x: x[0])
     
+    # Extract sorted clips
+    text_clips = [clip for i, clip, segment in results]
+
     final_video = CompositeVideoClip([video] + text_clips)
 
     logger.info("Methods in CompositeVideoClip:")
@@ -286,7 +289,7 @@ def add_transcript_voiceover(video_path, translated_json, output_path, add_voice
         logger.info(method)
     
     if add_voiceover and audio_segments:
-        audio_segments.sort(key=lambda segment: segment.start)
+        audio_segments = [segment for i, clip, segment in results if segment is not None]
         final_audio = CompositeAudioClip(audio_segments)  # Critical fix
         final_audio = final_audio.with_duration(video.duration)