Shift audio to correct in camera audio misalignment

Author: antonxy

console.py

@@ -9,6 +9,7 @@ def main():
     parser.add_argument('ad', help='Audio directory')
     parser.add_argument('edl', help='EDL output directory')
     parser.add_argument('fps', help='Project frames per second')
+    parser.add_argument('--audio_shift', help='Shift audio X frames forward')
     args = parser.parse_args()
 
     audio_ret = analyse_directory(args.ad)
@@ -21,8 +22,9 @@ def main():
     program_logic.rename_files(video_ret, 'v')
 
     fps = float(args.fps)
+    audio_shift = int(args.audio_shift)
 
-    program_logic.generate_edls(video_ret, audio_ret, fps, args.edl)
+    program_logic.generate_edls(video_ret, audio_ret, fps, args.edl, audio_shift)
 
 
 def analyse_directory(directory):

gui.py

@@ -21,6 +21,7 @@ def __init__(self, parent):
         self.audio_progress_var = tk.IntVar()
         self.video_progress_var = tk.IntVar()
         self.fps_var = tk.StringVar()
+        self.audio_shift_var = tk.StringVar()
         self.start_button = None
 
         self.generate_ui()
@@ -41,6 +42,12 @@ def generate_ui(self):
         tk.Label(fps_frame, text='FPS:').pack(side=tk.LEFT)
         tk.Entry(fps_frame, textvariable=self.fps_var).pack(side=tk.LEFT)
 
+        audio_shift_frame = tk.Frame(self)
+        audio_shift_frame.pack()
+        tk.Label(audio_shift_frame, text='Shift Audio forward').pack(side=tk.LEFT)
+        tk.Entry(audio_shift_frame, textvariable=self.audio_shift_var).pack(side=tk.LEFT)
+        tk.Label(audio_shift_frame, text='frames').pack(side=tk.LEFT)
+
         cmd_frame = tk.Frame(self)
         cmd_frame.pack(fill=tk.X)
         self.start_button = tk.Button(cmd_frame, text='Start', command=self.execute)
@@ -53,7 +60,7 @@ def generate_ui(self):
 
     def center_window(self):
         w = 500
-        h = 120
+        h = 140
 
         sw = self.parent.winfo_screenwidth()
         sh = self.parent.winfo_screenheight()
@@ -91,21 +98,27 @@ def execute(self):
             messagebox.showerror(title='audiosync', message='FPS has to be decimal number')
             return
 
+        try:
+            audio_shift = int(self.audio_shift_var.get())
+        except ValueError:
+            messagebox.showerror(title='audiosync', message='Audio shift has to be integer')
+            return
+
         thread = Thread(target=self.thread_target,
-                        args=(self.audio_progress_var, self.video_progress_var, self.start_button, fps, directory))
+                        args=(self.audio_progress_var, self.video_progress_var, self.start_button, fps, directory, audio_shift))
         thread.start()
         self.start_button.config(state='disabled')
 
 
     @staticmethod
-    def thread_target(audio_progress_var, video_progress_var, start_button, fps, directory):
+    def thread_target(audio_progress_var, video_progress_var, start_button, fps, directory, audio_shift):
         video_ret = analyse_directory(os.path.join(directory, 'video'), video_progress_var)
         audio_ret = analyse_directory(os.path.join(directory, 'audio'), audio_progress_var)
 
         program_logic.rename_files(audio_ret, 'a')
         program_logic.rename_files(video_ret, 'v')
 
-        program_logic.generate_edls(video_ret, audio_ret, fps, os.path.join(directory, 'edl'))
+        program_logic.generate_edls(video_ret, audio_ret, fps, os.path.join(directory, 'edl'), audio_shift)
 
         audio_progress_var.set(0)
         video_progress_var.set(0)

program_logic.py

@@ -29,11 +29,11 @@ def rename_files(file_list, suffix):
                                           '%i-%i-%i_%s%s' % (sst[0], sst[1], sst[2], suffix, file_ext)))
 
 
-def generate_edls(videos, audios, fps, edl_dir):
+def generate_edls(videos, audios, fps, edl_dir, audio_shift=0):
     for v in videos:
         for a in audios:
             if v['sst'] == a['sst']:
-                generate_edl(v, a, fps, edl_dir)
+                generate_edl(v, a, fps, edl_dir, audio_shift)
 
 
 def generate_tc(seconds, fps):
@@ -49,7 +49,7 @@ def generate_tc(seconds, fps):
     return '{:02d}:{:02d}:{:02d}:{:02d}'.format(hrs, mins, secs, frames)
 
 
-def generate_edl(video_data, audio_data, fps, edl_dir):
+def generate_edl(video_data, audio_data, fps, edl_dir, audio_shift=0):
     sst = video_data['sst']
     sst_str = '-'.join(['%s' % e for e in sst])
     filename = sst_str + '.edl'
@@ -64,7 +64,7 @@ def generate_edl(video_data, audio_data, fps, edl_dir):
     len_sec_a = float(audio_data['length_samples'] / float(audio_data['sample_rate']))
     len_sec_v = float(video_data['length_samples'] / float(video_data['sample_rate']))
 
-    a_bef_v = sync_sec_a - sync_sec_v
+    a_bef_v = sync_sec_a - sync_sec_v - (audio_shift / fps)
     v_bef_a = -a_bef_v
 
     f.write('TITLE: %s   FORMAT: CMX3600\n' % sst_str)