ms_tts.py

# import cv2
# import numpy as np
# import os

# # 添加更多常见的视频编码器
# codecs = {
#     'X264': 'H.264 / AVC',
#     'MP4V': 'MPEG-4',
#     'AVC1': 'H.264 / AVC',
#     'DIVX': 'MPEG-4',
#     'XVID': 'MPEG-4',
#     'MJPG': 'Motion JPEG',
#     'VP80': 'VP8',
#     'VP90': 'VP9',
#     'HEVC': 'H.265 / HEVC',
#     'PIM1': 'MPEG-1',
#     'MPG1': 'MPEG-1',
#     'MPG2': 'MPEG-2'
# }

# # 输出MP4视频文件路径
# output_file = 'test_video.mp4'

# # 定义一个函数来测试是否支持某种编码器并输出 mp4 文件
# def test_codec(codec_fourcc, codec_name):
#     # 使用 cv2.VideoWriter 创建视频文件对象
#     fourcc = cv2.VideoWriter_fourcc(*codec_fourcc)
#     out = cv2.VideoWriter(output_file, fourcc, 25.0, (640, 480))

#     # 检查是否成功初始化
#     if out.isOpened():
#         print(f"[SUPPORTED] Codec {codec_name} ({codec_fourcc}) is supported.")
        
#         # 写入一些帧，生成简单的测试视频
#         for i in range(100):
#             # 创建一个简单的蓝色帧作为测试
#             frame = (255 * (i % 2) * np.ones((480, 640, 3), dtype=np.uint8))
#             out.write(frame)
        
#         out.release()  # 关闭视频写入器
        
#         # 检查文件是否生成
#         if os.path.exists(output_file):
#             print(f"Video file {output_file} successfully created with codec {codec_fourcc}.")
#         else:
#             print(f"Failed to create video file {output_file}.")
        
#         # 删除测试视频文件
#         os.remove(output_file)
#     # else:
#         # print(f"[NOT SUPPORTED] Codec {codec_name} ({codec_fourcc}) is not supported.")

# # 测试所有编码器
# for fourcc, name in codecs.items():
#     test_codec(fourcc, name)



import azure.cognitiveservices.speech as speechsdk

def ms_tts_gen(text, audio_path):
    # 创建 Speech 配置
    # speech_config = speechsdk.SpeechConfig(subscription="b45d4ca1f5eb4c96950666eb97b9be60", region="eastasia")
    speech_config = speechsdk.SpeechConfig(subscription="ef47bc3073b74e4e9837e886be2fb6f7", region="eastasia")
    speech_config.speech_synthesis_voice_name = 'zh-HK-HiuGaaiNeural'  # 粤语女声

    # 配置音频输出为 MP3 格式
    audio_config = speechsdk.audio.AudioOutputConfig(filename=audio_path)

    # 指定输出格式为 MP3
    speech_config.set_speech_synthesis_output_format(speechsdk.SpeechSynthesisOutputFormat.Audio16Khz32KBitRateMonoMp3)

    # 合成语音并输出到 MP3 文件
    synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=audio_config)
    synthesizer.speak_text_async(text).get()