silence

app.py

@@ -237,7 +237,27 @@ def run_pipeline(prompt_text, steps, image_paths, audio_file_path, max_duration
     # Prepare output
     output_dir = os.path.join(os.environ["PROCESSED_RESULTS"], session_id)
     os.makedirs(output_dir, exist_ok=True)
+    def add_silence_to_audio_ffmpeg(audio_path, tmp_audio_path, silence_duration_s=0.5):
+
+        command = [
+            'ffmpeg', 
+            '-i', audio_path,  
+            '-f', 'lavfi',  
+            '-t', str(silence_duration_s),  
+            '-i', 'anullsrc=r=16000:cl=stereo',  
+            '-filter_complex', '[1][0]concat=n=2:v=0:a=1[out]',  
+            '-map', '[out]', 
+            '-y', tmp_audio_path,
+            '-loglevel', 'quiet'
+        ]
+        
+        subprocess.run(command, check=True)
 
+    
+    tmp_audio_path = os.path.join(output_dir, "tmp_audio.wav")
+    
+    add_silence_to_audio_ffmpeg(audio_path, tmp_audio_path)
+    
     # Random filename
     filename = f"gen_{uuid.uuid4().hex[:10]}"
     width, height = 832, 480
@@ -247,7 +267,7 @@ def run_pipeline(prompt_text, steps, image_paths, audio_file_path, max_duration
     runner.inference_loop(
         prompt_text,
         img_paths,
-        audio_path,
+        tmp_audio_path,
         output_dir,
         filename,
         inference_mode,