Update app.py

app.py

@@ -92,8 +92,16 @@ async def transcribe_audio(request: AudioRequest):
 
 @app.post("/synthesize")
 async def synthesize_speech(request: TTSRequest):
+    logger.info(f"Synthesize request received: text='{request.text}', language='{request.language}', speed={request.speed}")
     try:
-        logger.info(f"Synthesize request received: text='{request.text}', language='{request.language}', speed={request.speed}")
+        # Input validation
+        logger.info("Validating input parameters")
+        if not request.text:
+            raise ValueError("Text cannot be empty")
+        if request.language not in TTS_LANGUAGES:
+            raise ValueError(f"Unsupported language: {request.language}")
+        if not 0.5 <= request.speed <= 2.0:
+            raise ValueError(f"Speed must be between 0.5 and 2.0, got {request.speed}")
         
         logger.info("Calling synthesize function")
         result, filtered_text = synthesize(request.text, request.language, request.speed)
@@ -104,7 +112,7 @@ async def synthesize_speech(request: TTSRequest):
             raise ValueError("Synthesis failed to produce audio")
         
         sample_rate, audio = result
-        logger.info(f"Synthesis result: sample_rate={sample_rate}, audio_shape={audio.shape}, audio_dtype={audio.dtype}")
+        logger.info(f"Synthesis result: sample_rate={sample_rate}, audio_shape={audio.shape if isinstance(audio, np.ndarray) else 'not numpy array'}, audio_dtype={audio.dtype if isinstance(audio, np.ndarray) else type(audio)}")
         
         logger.info("Converting audio to numpy array")
         audio = np.array(audio, dtype=np.float32)
@@ -114,7 +122,8 @@ async def synthesize_speech(request: TTSRequest):
         max_value = np.max(np.abs(audio))
         if max_value == 0:
             logger.warning("Audio array is all zeros")
-        audio = audio / max(max_value, 1e-8)  # Avoid division by zero
+            raise ValueError("Generated audio is silent (all zeros)")
+        audio = audio / max_value
         logger.info(f"Normalized audio range: [{audio.min()}, {audio.max()}]")
         
         logger.info("Converting to int16")
@@ -136,16 +145,26 @@ async def synthesize_speech(request: TTSRequest):
         logger.info("FileResponse prepared successfully")
         
         return response
+
+    except ValueError as ve:
+        logger.error(f"ValueError in synthesize_speech: {str(ve)}", exc_info=True)
+        return JSONResponse(
+            status_code=400,
+            content={"message": "Invalid input", "details": str(ve)}
+        )
     except Exception as e:
-        logger.error(f"Error in synthesize_speech: {str(e)}", exc_info=True)
+        logger.error(f"Unexpected error in synthesize_speech: {str(e)}", exc_info=True)
         error_details = {
             "error": str(e),
+            "type": type(e).__name__,
             "traceback": traceback.format_exc()
         }
         return JSONResponse(
             status_code=500,
-            content={"message": "An error occurred during speech synthesis", "details": error_details}
+            content={"message": "An unexpected error occurred during speech synthesis", "details": error_details}
         )
+    finally:
+        logger.info("Synthesize request completed")
 
 @app.post("/identify")
 async def identify_language(request: AudioRequest):