fix for Mac system voices

README.md

@@ -26,7 +26,7 @@ https://github.com/KoljaB/RealtimeTTS/assets/7604638/87dcd9a5-3a4e-4f57-be45-837
 
 ## Updates
 
-Latest Version: v0.3.32
+Latest Version: v0.3.34
 
 #### New Features:
 - new Engine: OpenAI TTS

RealtimeTTS/engines/system_engine.py

@@ -1,4 +1,6 @@
 from .base_engine import BaseEngine
+from pydub.utils import mediainfo
+from pydub import AudioSegment
 from typing import Union
 import tempfile
 import pyaudio
@@ -54,9 +56,8 @@ def get_stream_info(self):
                   - Sample Rate (int): The sample rate of the audio in Hz. 16000 represents 16kHz sample rate.
         """        
         return pyaudio.paInt16, 1, 22050
-
-    def synthesize(self, 
-                   text: str) -> bool:
+
+    def synthesize(self, text: str) -> bool:
         """
         Synthesizes text to audio stream.
 
@@ -67,12 +68,22 @@ def synthesize(self,
         self.engine.save_to_file(text, self.file_path)
         self.engine.runAndWait()
 
-        # Open the saved WAV file
+        # Get media info of the file
+        info = mediainfo(self.file_path)
+
+        # Check if the file format is AIFF and convert to WAV if necessary
+        if info['format_name'] == 'aiff':
+            audio = AudioSegment.from_file(self.file_path, format="aiff")
+            audio.export(self.file_path, format="wav")
+
+        # Now open the WAV file
         with wave.open(self.file_path, 'rb') as wf:
             audio_data = wf.readframes(wf.getnframes())
             self.queue.put(audio_data)
+            return True
 
-        return True
+        # Return False if the process failed
+        return False
 
     def get_voices(self):
         """

RealtimeTTS/text_to_stream.py

@@ -56,6 +56,7 @@ def __init__(self,
         self.abort_events = []
         self.tokenizer = tokenizer
         self.language = language
+        self.player = None
 
         self._create_iterators()
 
@@ -105,6 +106,7 @@ def load_engine(self,
             self.player = StreamPlayer(self.engine.queue, AudioConfiguration(format, channels, rate), on_playback_start=self._on_audio_stream_start)
         else:
             self.engine.on_playback_start = self._on_audio_stream_start
+            self.player = None
 
         logging.info(f"loaded engine {self.engine.engine_name}")
 
@@ -197,7 +199,9 @@ def play(self,
         self.stream_running = True
         abort_event = threading.Event()
         self.abort_events.append(abort_event)
-        self.player.mute(muted)
+
+        if self.player:
+            self.player.mute(muted)
 
         self.output_wavfile = output_wavfile
         self.chunk_callback = on_audio_chunk

setup.py

@@ -9,7 +9,7 @@
 
 setuptools.setup(
     name="RealTimeTTS", 
-    version="0.3.32",
+    version="0.3.34",
     author="Kolja Beigel",
     author_email="[email protected]",
     description="*Stream text into audio with an easy-to-use, highly configurable library delivering voice output with minimal latency.",

tests/coqui_test.py

@@ -7,17 +7,17 @@ def dummy_generator():
 
 
     # for normal use with minimal logging:
-    engine = CoquiEngine()
+    # engine = CoquiEngine()
 
     # test with extended logging:
-    # import logging
-    # logging.basicConfig(level=logging.DEBUG)    
-    # engine = CoquiEngine(level=logging.DEBUG)
+    import logging
+    logging.basicConfig(level=logging.INFO)    
+    engine = CoquiEngine(level=logging.INFO)
 
 
     stream = TextToAudioStream(engine)
 
     print ("Starting to play stream")
-    stream.feed(dummy_generator()).play()
+    stream.feed(dummy_generator()).play(log_synthesized_text=True)
 
     engine.shutdown()