concatenate only good video chunks

mix together

Author: frrobledo

main.py

@@ -3,6 +3,7 @@
 from tools.audio_splitter_ffmpeg import *
 from tools.video_downloader import *
 from tools.video_editing import *
+import sys
 
 if __name__ == '__main__':
     # Prepare folders
@@ -12,19 +13,23 @@
     os.makedirs('final_output', exist_ok=True)
     os.makedirs('logs', exist_ok=True)
 
+    # Redirect all print outputs to a log file
+    # log_output_file = open('logs/log.txt', 'w')
+    # sys.stdout = log_output_file
+
     # Set the multiprocessing start method to 'spawn'
     multiprocessing.set_start_method('spawn')
 
     # Download video
-    # url = input("Enter YouTube URL: ")
-    # video = video_download(url)
-    video = "downloads/xSh7PuWAxXU.mp4"
+    url = input("Enter YouTube URL: ")
+    video = video_download(url)
+    # video = "downloads/xSh7PuWAxXU.mp4"
 
     # # Extract audio
-    # audio = audio_extractor(video)
+    audio = audio_extractor(video)
 
     ## Load audio
-    audio = 'original_audios/xSh7PuWAxXU.wav'
+    # audio = 'original_audios/xSh7PuWAxXU.wav'
     original_audio_name = os.path.splitext(os.path.basename(audio))[0]
 
     ## Transcribe audio
@@ -36,32 +41,32 @@
     print("Audio split")
 
     ## Translate each segment
-    # new_segments = []
-    # i = 0
-    # total_segments = len(segments)
-    # for segment in segments:
-    #     if len(segment['text']) > 0:
-    #         translated_text = translate_deepl(segment['text'], 'es', detected_language)
-    #     else:
-    #         translated_text = ''
-    #     new_segments.append({'id': segment['id'], 
-    #                         'seek': segment['seek'],
-    #                         'start': segment['start'],
-    #                         'end': segment['end'],
-    #                         'text': translated_text
-    #                         })
-    #     percentage = (i+1)/total_segments*100
-    #     # Print the progress 
-    #     print(f'Translation progress: {percentage:.2f}%')
-    #     i += 1
-    # print(f"Audio translated")
+    new_segments = []
+    i = 0
+    total_segments = len(segments)
+    for segment in segments:
+        if len(segment['text']) > 0:
+            translated_text = translate_deepl(segment['text'], 'es', detected_language)
+        else:
+            translated_text = ''
+        new_segments.append({'id': segment['id'], 
+                            'seek': segment['seek'],
+                            'start': segment['start'],
+                            'end': segment['end'],
+                            'text': translated_text
+                            })
+        percentage = (i+1)/total_segments*100
+        # Print the progress 
+        print(f'Translation progress: {percentage:.2f}%')
+        i += 1
+    print(f"Audio translated")
 
     ## save new_segments as a pickle for later loading
-    # import pickle
+    import pickle
 
-    # with open('new_segments.pkl', 'wb') as f:
-    #     pickle.dump(new_segments, f)
-    # print("New segments saved")
+    with open('new_segments.pkl', 'wb') as f:
+        pickle.dump(new_segments, f)
+    print("New segments saved")
 
     ## Load new_segments from pickle
     import pickle
@@ -103,13 +108,17 @@
     )
 
     # Delete all content in output_audio folder
-    # for filename in os.listdir('output_audio'):
-    #     file_path = os.path.join('output_audio', filename)
-    #     if os.path.isfile(file_path):
-    #         os.remove(file_path)
+    for filename in os.listdir('output_audio'):
+        file_path = os.path.join('output_audio', filename)
+        if os.path.isfile(file_path):
+            os.remove(file_path)
 
 
     # Mix old video file and new audio file
     # video_path = f'downloads/{original_audio_name}.mp4'
     # audio_path = f'final_output/{original_audio_name}-{target_lang_code}.wav'
-    # replace_audio_in_video(video, audio_path, f'final_output/{original_audio_name}-{target_lang_code}.mp4')
+    # replace_audio_in_video(video, audio_path, f'final_output/{original_audio_name}-{target_lang_code}.mp4')
+
+    # Close log output file
+    # sys.stdout = sys.__stdout__
+    # log_output_file.close()

tools/video_editing.py

@@ -230,11 +230,13 @@ def adjust_video_speed_and_replace_audio(video_path, background_audio_path, star
         '-y',
         '-i', adjusted_video,
         '-i', synthesized_audio_path,
-        '-c:v', 'copy',
+        '-c:v', 'libx264',
         '-c:a', 'aac',
+        '-b:a', '192k',
         '-shortest',
         output_path
     ]
+
     print("Combining adjusted video and synthesized audio...")
     result = subprocess.run(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
     if result.returncode != 0:
@@ -327,22 +329,26 @@ def extract_video_segment(video_path, background_audio_path, start_time, duratio
 
 def create_concat_file(segment_video_paths, concat_file_path):
     """
-    Create a text file listing the video segments for FFmpeg concatenation.
+    Create a text file listing the video segments for FFmpeg concatenation,
+    including only files larger than a specified size.
 
     Parameters
     ----------
     segment_video_paths : list of str
         The list of paths to the video segments.
     concat_file_path : str
         The path to save the concat file.
-
-    Returns
-    -------
-    None
     """
+    min_file_size_bytes = 300  # Minimum file size in bytes
+
     with open(concat_file_path, 'w') as f:
         for segment_path in segment_video_paths:
-            f.write(f"file '{os.path.abspath(segment_path)}'\n")
+            # Check if the file exists and is larger than the minimum size
+            if os.path.isfile(segment_path) and os.path.getsize(segment_path) > min_file_size_bytes:
+                f.write(f"file '{os.path.abspath(segment_path)}'\n")
+            else:
+                print(f"Excluding segment {segment_path} due to insufficient file size.")
+
 
 def concatenate_segments(concat_file_path, output_video_path):
     """
@@ -367,7 +373,10 @@ def concatenate_segments(concat_file_path, output_video_path):
         '-f', 'concat',
         '-safe', '0',
         '-i', concat_file_path,
-        '-c', 'copy',
+        '-c:v', 'libx264',
+        '-c:a', 'aac',
+        '-strict', 'experimental',
+        '-b:a', '192k',
         output_video_path
     ]
     # subprocess.run(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE)