Update app.py

app.py

@@ -34,6 +34,24 @@ st.set_page_config(
 )
 load_dotenv()
 
+# Add available English voices for Edge TTS
+EDGE_TTS_VOICES = [
+    "en-US-AriaNeural",  # Default voice
+    "en-US-GuyNeural", 
+    "en-US-JennyNeural",
+    "en-GB-SoniaNeural",
+    "en-GB-RyanNeural",
+    "en-AU-NatashaNeural",
+    "en-AU-WilliamNeural",
+    "en-CA-ClaraNeural",
+    "en-CA-LiamNeural"
+]
+
+# Add this to your session state initialization section:
+if 'tts_voice' not in st.session_state:
+    st.session_state['tts_voice'] = EDGE_TTS_VOICES[0]  # Default voice
+    
+
 # 🔑 2. API Setup & Clients
 openai_api_key = os.getenv('OPENAI_API_KEY', "")
 anthropic_key = os.getenv('ANTHROPIC_API_KEY_3', "")
@@ -361,28 +379,26 @@ def parse_arxiv_refs(ref_text: str):
     
     return results[:20]  # Ensure we return maximum 20 papers
 
+
+
 def create_paper_audio_files(papers):
     """
-    Create audio files for each paper's components and add file paths to paper dict.
+    Create audio files for each paper's content and add file paths to paper dict.
+    Only generates full audio file since it includes the title.
     """
     for paper in papers:
         try:
-            # Generate audio for title
-            title_text = clean_for_speech(paper['title'])
-            title_file = speak_with_edge_tts(title_text)
-            paper['title_audio'] = title_file
-
-            # Generate audio for full content
+            # Generate audio for full content only
             full_text = f"{paper['title']} by {paper['authors']}. {paper['summary']}"
             full_text = clean_for_speech(full_text)
-            full_file = speak_with_edge_tts(full_text)
+            full_file = speak_with_edge_tts(full_text, voice=st.session_state['tts_voice'])
             paper['full_audio'] = full_file
 
         except Exception as e:
             st.warning(f"Error generating audio for paper {paper['title']}: {str(e)}")
-            paper['title_audio'] = None
             paper['full_audio'] = None
 
+            
 def display_papers(papers):
     """
     Display papers with their audio controls using URLs as unique keys.
@@ -395,18 +411,11 @@ def display_papers(papers):
             st.markdown(f"*{paper['authors']}*")
             st.markdown(paper['summary'])
             
-            # Audio controls in columns
-            col1, col2 = st.columns(2)
-            
-            with col1:
-                if paper.get('title_audio'):
-                    st.write("🎙️ Title Audio")
-                    st.audio(paper['title_audio'])
-            
-            with col2:
-                if paper.get('full_audio'):
-                    st.write("📚 Full Paper Audio")
-                    st.audio(paper['full_audio'])
+            # Single audio control for full content
+            if paper.get('full_audio'):
+                st.write("📚 Paper Audio")
+                st.audio(paper['full_audio'])
+
 
 def perform_ai_lookup(q, vocal_summary=True, extended_refs=False, 
                      titles_summary=True, full_audio=False):
@@ -595,6 +604,21 @@ def display_file_manager_sidebar(groups, sorted_prefixes):
 # 🎯 11. Main Application
 def main():
     st.sidebar.markdown("### 🚲BikeAI🏆 Multi-Agent Research")
+    
+    # Add voice selector to sidebar
+    st.sidebar.markdown("### 🎤 Voice Settings")
+    selected_voice = st.sidebar.selectbox(
+        "Select TTS Voice:",
+        options=EDGE_TTS_VOICES,
+        index=EDGE_TTS_VOICES.index(st.session_state['tts_voice'])
+    )
+    
+    # Update session state if voice changes
+    if selected_voice != st.session_state['tts_voice']:
+        st.session_state['tts_voice'] = selected_voice
+        st.rerun()
+
+
     tab_main = st.radio("Action:",["🎤 Voice","📸 Media","🔍 ArXiv","📝 Editor"],horizontal=True)
 
     mycomponent = components.declare_component("mycomponent", path="mycomponent")