lilited audio length

webgui.py

@@ -27,7 +27,7 @@ import argparse
 
 import gradio as gr
 from gradio_client import Client, handle_file
-
+from pydub import AudioSegment
 import huggingface_hub 
 
 huggingface_hub.snapshot_download(
@@ -320,8 +320,27 @@ with gr.Blocks() as demo:
             </div>
             """)
 
-    def generate_video(uploaded_img, uploaded_audio, width, height, length, seed, facemask_dilation_ratio, facecrop_dilation_ratio, context_frames, context_overlap, cfg, steps, sample_rate, fps, device, progress=gr.Progress(track_tqdm=True)):
+    def trim_audio(file_path, output_path, max_duration=10):
+        # Load the audio file
+        audio = AudioSegment.from_wav(file_path)
+    
+        # Convert max duration to milliseconds
+        max_duration_ms = max_duration * 1000
+    
+        # Trim the audio if it's longer than max_duration
+        if len(audio) > max_duration_ms:
+            audio = audio[:max_duration_ms]
+    
+        # Export the trimmed audio
+        audio.export(output_path, format="wav")
+        print(f"Audio trimmed and saved as {output_path}")
+        return output_path
 
+    def generate_video(uploaded_img, uploaded_audio, width, height, length, seed, facemask_dilation_ratio, facecrop_dilation_ratio, context_frames, context_overlap, cfg, steps, sample_rate, fps, device, progress=gr.Progress(track_tqdm=True)):
+        if is_shared_ui:
+            gr.Info("Trimming audio to max 10 seconds. Duplicate the space for unlimited audio length.")
+            uploaded_audio = trim_audio(uploaded_audio, "trimmed_audio.wav")
+            
         final_output_path = process_video(
             uploaded_img, uploaded_audio, width, height, length, seed, facemask_dilation_ratio, facecrop_dilation_ratio, context_frames, context_overlap, cfg, steps, sample_rate, fps, device
         )