| import edge_tts | |
| import io | |
| import numpy as np | |
| from pydub import AudioSegment # Install with: pip install pydub | |
| class TextToSpeech: | |
| def __init__(self, voice="en-US-AriaNeural"): | |
| self.voice = voice | |
| async def synthesize(self, text): | |
| communicate = edge_tts.Communicate(text, self.voice) | |
| # Collect raw audio bytes | |
| audio_bytes = bytearray() | |
| async for chunk in communicate.stream(): | |
| if chunk["type"] == "audio": | |
| audio_bytes.extend(chunk["data"]) | |
| # Convert to numpy array using pydub | |
| audio = AudioSegment.from_file(io.BytesIO(audio_bytes), format="mp3") | |
| samples = np.array(audio.get_array_of_samples()) | |
| sample_rate = audio.frame_rate | |
| return (sample_rate, samples) |