audi file format

app.py

@@ -110,29 +110,28 @@ if audio_option == 'Record Audio':
 
 # Option 2: Upload audio
 elif audio_option == 'Upload Audio':
-    audio_file = st.file_uploader("Upload audio file (WAV format)", type=['wav'])
-    
+    audio_file = st.file_uploader("Upload an audio file", type=["wav", "mp3", "ogg"])
+    # Assuming you're working with an uploaded file
+    # uploaded_file = st.file_uploader("Upload an audio file", type=["wav", "mp3", "ogg"])
+
     if audio_file:
-         
-        # When writing the audio data, specify the format explicitly
-        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp_file:
-            sf.write(tmp_file.name, audio_data, sr, format='WAV')
-            tmp_file_path = tmp_file.name    
         
-        audio_data, sr = librosa.load(tmp_file_path, sr=None)
+        # Read the uploaded audio file
+        audio_data, sr = librosa.load(audio_file, sr=None)  # sr=None preserves original sample rate
         
-        # Compute audio properties
-        audio_size = len(audio_data) * 2  # in bytes (16-bit PCM)
-        frame_rate = sr
-        duration = librosa.get_duration(y=audio_data, sr=sr)
-
-        # Display audio properties
-        st.write(f"Audio Size: {audio_size} bytes")
-        st.write(f"Frame Rate: {frame_rate} Hz")
-        st.write(f"Duration: {duration:.2f} seconds")
-
-        # Perform conversion using the selected model
-        st.subheader("Converting audio to text...")
+        # Display information about the audio file
+        st.write(f"Audio file size: {audio_file.size} bytes")
+        st.write(f"Sample rate: {sr}")
+        st.write(f"Duration: {len(audio_data) / sr:.2f} seconds")
+
+        # Convert the audio to WAV format and save as temporary file
+        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp_file:
+            sf.write(tmp_file.name, audio_data, sr, format='WAV')
+            tmp_file_path = tmp_file.name
+            
+        # Now you can proceed with the ASR model for conversion
+        st.write("Converting audio to text...")
+        # (Your ASR model conversion code goes here)
 
         start_time = time.time()