From 5935062d5af2d8b00a38149d009d1b57dc68aa95 Mon Sep 17 00:00:00 2001
From: GHUFRAN-HYDER <65118907+GHUFRAN-HYDER@users.noreply.github.com>
Date: Thu, 14 Mar 2024 23:18:14 +0500
Subject: [PATCH] Update tts_stream.py

---
 tortoise/tts_stream.py | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)

diff --git a/tortoise/tts_stream.py b/tortoise/tts_stream.py
index 32a129d..9b8df6a 100644
--- a/tortoise/tts_stream.py
+++ b/tortoise/tts_stream.py
@@ -11,13 +11,19 @@ from utils.text import split_and_recombine_text
 import sounddevice as sd
 import queue
 import threading
+import pydub  # Using pydub for audio playback
+
+
 def play_audio(audio_queue):
-    while True:
-        chunk = audio_queue.get()
-        if chunk is None:
-            break
-        sd.play(chunk.cpu().numpy(), samplerate=24000)
-        sd.wait()
+  while True:
+    chunk = audio_queue.get()
+    if chunk is None:
+      break
+    # Convert PyTorch tensor to NumPy array and then to a WAV audio segment
+    audio_data = chunk.cpu().numpy().tobytes()
+    sound = pydub.AudioSegment(audio_data, frame_rate=24000, channels=1, sample_width=2)
+    # Play the audio segment
+    pydub.playback.play(sound)
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()