Update voice_processing.py

voice_processing.py

@@ -15,6 +15,8 @@ from lib.infer_pack.models import (
 )
 from vc_infer_pipeline import VC
 import uuid
+import tempfile  # Make sure to import tempfile
+import asyncio  # Import asyncio if not already imported
 
 config = Config()
 
@@ -121,27 +123,27 @@ def tts(
     use_uploaded_voice,
     uploaded_voice,
 ):
-    # Load models if not already loaded
-    load_hubert()
-    load_rmvpe()
-
-    # Default values for parameters used in EdgeTTS
-    f0_up_key = 0  # Default pitch adjustment
-    f0_method = "rmvpe"  # Default pitch extraction method
-    protect = 0.33  # Default protect value
-    filter_radius = 3
-    resample_sr = 0
-    rms_mix_rate = 0.25
-    edge_time = 0  # Initialize edge_time
-
-    edge_output_filename = get_unique_filename("mp3")
-    audio = None
-    sr = 16000  # Default sample rate
-
     try:
+        # Load models if not already loaded
+        load_hubert()
+        load_rmvpe()
+
+        # Default values for parameters used in EdgeTTS
+        f0_up_key = 0  # Default pitch adjustment
+        f0_method = "rmvpe"  # Default pitch extraction method
+        protect = 0.33  # Default protect value
+        filter_radius = 3
+        resample_sr = 0
+        rms_mix_rate = 0.25
+        edge_time = 0  # Initialize edge_time
+
+        edge_output_filename = get_unique_filename("mp3")
+        audio = None
+        sr = 16000  # Default sample rate
+
         if use_uploaded_voice:
             if uploaded_voice is None:
-                return "No voice file uploaded.", None, None
+                return {"error": "No voice file uploaded."}, None, None
 
             # Process the uploaded voice file
             with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
@@ -152,7 +154,6 @@ def tts(
             input_audio_path = uploaded_file_path
         else:
             # EdgeTTS processing
-            # Note: EdgeTTS code may need to be adjusted based on your implementation
             import edge_tts
             t0 = time.time()
             speed = 0  # Default speech speed
@@ -209,9 +210,7 @@ def tts(
         )
 
     except EOFError:
-        info = (
-            "Output not valid. This may occur when input text and speaker do not match."
-        )
+        info = "Output not valid. This may occur when input text and speaker do not match."
         print(info)
         return {"error": info}, None, None
     except Exception as e: