update yt_download

app.py

@@ -43,7 +43,7 @@ def loudness_normalize(audio, sample_rate, target_loudness=-12.0):
 def process_youtube_url(url):
     try:
         audio, sr = download_youtube_audio(url)
-        return (sr, audio)
+        return (sr, audio), None
     except Exception as e:
         return None, f"Error processing YouTube URL: {str(e)}"
 
@@ -199,21 +199,23 @@ with gr.Blocks() as demo:
         gr.Markdown('<span style="color: lightgray; font-style: italic;">all output samples are normalized to -12dB LUFS</span>')
         
         with gr.Row():
-            output_audio_yt = gr.Audio(label="Output Audio", type='numpy')
-            param_output_yt = gr.Textbox(label="Predicted Parameters", lines=5)
+            with gr.Column():
+                output_audio = gr.Audio(label="Output Audio y'", type='numpy')
+                normalized_input = gr.Audio(label="Normalized Source Audio", type='numpy')
+            param_output = gr.Textbox(label="Predicted Parameters", lines=5)
         
         error_message_yt = gr.Textbox(label="Error Message", visible=False)
 
         def process_and_handle_errors(input_audio, input_youtube_url, reference_audio, reference_youtube_url):
             result = process_audio_with_youtube(input_audio, input_youtube_url, reference_audio, reference_youtube_url)
             if len(result) == 3 and isinstance(result[2], str):  # Error occurred
-                return None, None, gr.update(visible=True, value=result[2])
-            return result[0], result[1], gr.update(visible=False, value="")
+                return None, None, None, gr.update(visible=True, value=result[2])
+            return result[0], result[1], result[2], gr.update(visible=False, value="")
 
         process_button_yt.click(
             process_and_handle_errors, 
             inputs=[input_audio_yt, input_youtube_url, reference_audio_yt, reference_youtube_url], 
-            outputs=[output_audio_yt, param_output_yt, error_message_yt]
+            outputs=[output_audio_yt, param_output_yt, normalized_input, error_message_yt]
         )
 
     gr.Markdown("## Step 2: Inference Time Optimization (ITO)")

utils.py

@@ -3,10 +3,11 @@ import librosa
 import numpy as np
 
 def download_youtube_audio(url):
-    yt = YouTube(url, use_po_token=True)
+    # yt = YouTube(url, use_po_token=True)
+    yt = YouTube(url)
     stream = yt.streams.filter(only_audio=True).first()
     filename = stream.download()
     audio, sr = librosa.load(filename, sr=44100, mono=False)
     if audio.ndim == 1:
         audio = np.stack([audio, audio])
-    return audio.T
+    return audio.T, sr