Update piper_engine.py

piper_engine.py

@@ -1,5 +1,4 @@
 # piper_engine.py
-
 import io
 import numpy as np
 import soundfile as sf
@@ -8,29 +7,34 @@ from piper.voice import PiperVoice
 class PiperEngine:
 
     def __init__(self, model_path):
+        # Load the voice model
         self.voice = PiperVoice.load(model_path)
 
     def synthesize(
         self,
         text,
-        speed=1.0,
-        pitch=1.0,
+        speed=1.0,   # Default speed multiplier
+        pitch=1.0,   # Default pitch multiplier
         sample_rate=22050,
     ):
 
-        # Call synthesize using positional args for your Piper version
-        audio_gen = self.voice.synthesize(
-            text,
-            1 / speed,  # length_scale
-            pitch       # pitch_scale
-        )
-
-        # Convert generator to NumPy array
-        audio = np.fromiter(audio_gen, dtype=np.float32)
-
+        # -----------------------------
+        # Important: call synthesize EXACTLY as Piper expects
+        # -----------------------------
+        # Positional arguments only: text, length_scale, pitch_scale
+        # Do not use keywords like length_scale= or pitch_scale=
+        audio = self.voice.synthesize(text, speed, pitch)
+
+        # -----------------------------
+        # Convert to NumPy array and ensure correct shape
+        # -----------------------------
+        audio = np.array(audio, dtype=np.float32)
         if audio.ndim == 1:
             audio = audio[:, np.newaxis]
 
+        # -----------------------------
+        # Write to WAV buffer
+        # -----------------------------
         buffer = io.BytesIO()
         sf.write(buffer, audio, sample_rate, format="WAV")
         buffer.seek(0)