AVAudioEngine录音崩溃, reason: ‘format.sampleRate == hwFormat.sampleRate

运行后出现import websockets import asyncio import json import base64 import uuid import requests import logging import struct import os import sys from typing import Dict, Optional, Union # 确保当前目录在Python路径中 sys.path.append(os.path.dirname(os.path.abspath(__file__))) try: from config import VOLCANO_TTS_CONFIG except ImportError: # 默认配置（仅用于演示） VOLCANO_TTS_CONFIG = { "appid": "YOUR_APP_ID", "token": "YOUR_API_TOKEN", "secret_key": "YOUR_SECRET_KEY", "http_voice": "zh_female_wanwanxiaohe_moon_bigtts", "websocket_voice": "BV701_streaming", "websocket_url": "wss://openspeech.bytedance.com/api/v2/tts", "http_url": "https://openspeech.bytedance.com/api/v2/tts", "log_level": "INFO" } print("警告: 未找到config.py，使用默认配置") # 配置日志 log_level = getattr(logging, VOLCANO_TTS_CONFIG.get("log_level", "INFO").upper()) logging.basicConfig( level=log_level, format="%(asctime)s [%(levelname)s] %(name)s:%(lineno)d - %(message)s", handlers=[logging.StreamHandler()] ) logger = logging.getLogger("VolcanoTTS") class VolcanoTTS: def __init__(self, uid: str = "default_uid"): """ 从config.py加载配置初始化TTS客户端 """ self.config = VOLCANO_TTS_CONFIG self.appid = self.config["appid"] self.token = self.config["token"] self.secret_key = self.config["secret_key"] self.uid = uid self.websocket_url = self.config["websocket_url"] self.http_url = self.config["http_url"] self.http_voice = self.config["http_voice"] self.websocket_voice = self.config["websocket_voice"] logger.info(f"TTS客户端初始化完成 | 应用ID: {self.appid[:3]}*** | HTTP音色: {self.http_voice} | WebSocket音色: {self.websocket_voice}") def _build_base_request(self, text: str, voice_type: str, operation: str, **kwargs) -> Dict: """构建基础请求参数""" request = { "app": { "appid": self.appid, "token": self.token, "cluster": "volcano_tts" }, "user": { "uid": self.uid }, "audio": { "voice_type": voice_type, "encoding": kwargs.get("encoding", "mp3"), "speed_ratio": kwargs.get("speed_ratio", 1.0), "rate": kwargs.get("rate", 24000), "volume": kwargs.get("volume", 1.0), "pitch_ratio": kwargs.get("pitch_ratio", 1.0), }, "request": { "reqid": str(uuid.uuid4()), "text": text, "operation": operation } } # 可选参数处理 if "emotion" in kwargs: request["audio"]["emotion"] = kwargs["emotion"] if "extra_param" in kwargs: request["request"]["extra_param"] = kwargs["extra_param"] return request async def websocket_tts(self, text: str, output_file: str, **kwargs): """ WebSocket流式语音合成（使用配置中的流式音色） """ voice_type = kwargs.pop("voice_type", self.websocket_voice) logger.info(f"开始WebSocket TTS合成 | 文本长度: {len(text)} | 音色: {voice_type}") # 构建请求数据 request_data = self._build_base_request(text, voice_type, "submit", **kwargs) headers = { "Authorization": f"Bearer; {self.token}", "X-Secret-Key": self.secret_key } # 直接发送JSON文本（火山引擎API要求） request_message = json.dumps(request_data) audio_data = b"" try: async with websockets.connect( self.websocket_url, extra_headers=headers, ping_interval=20, ping_timeout=20, close_timeout=10, max_size=10*1024*1024 ) as websocket: logger.debug(f"发送请求: {request_message[:200]}...") await websocket.send(request_message) # 接收初始响应 init_response = await websocket.recv() logger.debug(f"初始响应: {init_response}") # 验证初始响应 if isinstance(init_response, str): try: init_data = json.loads(init_response) if init_data.get("code") != 3000: logger.error(f"初始响应错误: {init_data}") return False except json.JSONDecodeError: logger.warning(f"无法解析初始响应: {init_response}") # 接收音频数据 while True: try: response = await asyncio.wait_for(websocket.recv(), timeout=30.0) if isinstance(response, bytes): # 二进制音频数据 audio_data += response logger.debug(f"收到音频数据块: {len(response)}字节") elif isinstance(response, str): # 文本控制消息 control_msg = json.loads(response) logger.debug(f"控制消息: {control_msg}") # 检查结束标志 if control_msg.get("status") == 2: logger.info("收到合成结束消息") break # 检查错误 if control_msg.get("code") != 3000: logger.error(f"服务端错误: {control_msg.get('message')}") return False except asyncio.TimeoutError: logger.error("接收响应超时") break except websockets.exceptions.ConnectionClosed as e: logger.error(f"连接异常关闭: 代码={e.code}, 原因={e.reason}") if e.code == 1006: logger.error("1006错误可能原因: 1.音色不支持流式 2.认证失败 3.协议错误") return False except Exception as e: logger.exception("处理响应时发生意外错误") return False except Exception as e: logger.exception("WebSocket连接失败") return False # 保存音频文件 if audio_data: encoding = kwargs.get("encoding", "mp3") self._save_audio(audio_data, output_file, encoding) logger.info(f"音频已保存至: {output_file} | 大小: {len(audio_data)}字节") return True else: logger.error("未收到有效音频数据") return False def http_tts(self, text: str, output_file: str, **kwargs): """ HTTP非流式语音合成（使用配置中的HTTP音色） """ voice_type = kwargs.pop("voice_type", self.http_voice) logger.info(f"开始HTTP TTS合成 | 文本长度: {len(text)} | 音色: {voice_type}") # 构建请求 request_data = self._build_base_request(text, voice_type, "query", **kwargs) headers = { "Authorization": f"Bearer; {self.token}", "X-Secret-Key": self.secret_key, "Content-Type": "application/json" } try: # 发送请求 response = requests.post( self.http_url, json=request_data, headers=headers, timeout=30.0 ) response.raise_for_status() # 解析响应 result = response.json() if result.get("code") != 3000: error_msg = result.get("message", "未知错误") logger.error(f"合成失败: {error_msg} (代码: {result.get('code')})") return False # 解码并保存音频 audio_data = base64.b64decode(result["data"]) encoding = kwargs.get("encoding", "mp3") self._save_audio(audio_data, output_file, encoding) logger.info(f"合成成功 | 时长: {result['addition']['duration']}ms | 保存至: {output_file}") return True except requests.exceptions.RequestException as e: logger.error(f"HTTP请求失败: {str(e)}") except (KeyError, ValueError) as e: logger.error(f"响应解析失败: {str(e)}") except Exception as e: logger.exception("处理HTTP响应时发生意外错误") return False def _save_audio(self, data: bytes, file_path: str, encoding: str): """ 保存音频数据到文件 """ # 确保文件扩展名匹配编码格式 ext_mapping = { "mp3": ".mp3", "wav": ".wav", "pcm": ".pcm" } # 获取扩展名 ext = ext_mapping.get(encoding.lower(), ".mp3") # 确保文件路径有正确的扩展名 if not file_path.endswith(ext): if "." in file_path: # 移除现有扩展名 base_path = os.path.splitext(file_path)[0] file_path = base_path + ext else: file_path += ext # 保存文件 with open(file_path, "wb") as f: f.write(data) logger.debug(f"音频保存成功 | 路径: {file_path} | 格式: {encoding} | 大小: {len(data)}字节") return file_path # 使用示例 if __name__ == "__main__": # 初始化TTS客户端 tts = VolcanoTTS(uid="app_user_001") # 示例1: HTTP合成 http_text = "欢迎使用火山引擎语音合成服务，这是HTTP合成示例" http_output = "http_tts_sample" tts.http_tts( text=http_text, output_file=http_output, encoding="mp3", speed_ratio=1.2 ) # 示例2: WebSocket流式合成 async def run_websocket_example(): ws_text = "这是WebSocket流式合成示例，已解决1006连接问题" ws_output = "websocket_tts_sample" success = await tts.websocket_tts( text=ws_text, output_file=ws_output, encoding="wav", emotion="happy" # 情感参数示例 ) if success: logger.info(f"WebSocket合成成功: {ws_output}") else: logger.error("WebSocket合成失败") asyncio.run(run_websocket_example())

import websockets import asyncio import json import base64 import uuid import requests import io import pyaudio from typing import Dict, Optional class VolcanoTTS: def __init__(self, uid: str = "default_uid"): """ 初始化火山引擎语音合成客户端 :param uid: 用户标识（非空字符串即可） """ from api.config import API_CONFIGS config = API_CONFIGS['volcano_tts'] self.appid = config['appid'] self.token = config['token'] self.secret_key = config['Key'] # 新增Secret Key self.uid = uid self.websocket_url = config['websocket_url'] self.http_url = config['http_url'] def _build_base_request(self, text: str, voice_type: str, operation: str, **kwargs) -> Dict: """构建基础请求参数""" request = { "app": { "appid": self.appid, "token": self.token, "cluster": "volcano_tts" }, "user": { "uid": self.uid }, "audio": { "voice_type": voice_type, "encoding": kwargs.get("encoding", "mp3"), "speed_ratio": kwargs.get("speed_ratio", 1.0), "rate": kwargs.get("rate", 24000), "loudness_ratio": kwargs.get("loudness_ratio", 1.0) }, "request": { "reqid": str(uuid.uuid4()), # 生成唯一请求ID "text": text, "operation": operation } } # 处理情感参数 if "emotion" in kwargs: request["audio"]["emotion"] = kwargs["emotion"] request["audio"]["enable_emotion"] = True # 处理额外参数 if "extra_param" in kwargs: request["request"]["extra_param"] = kwargs["extra_param"] return request async def websocket_tts(self, text: str, voice_type: str, output_file: str,** kwargs): """ WebSocket流式合成（边合成边返回音频） :param text: 合成文本 :param voice_type: 音色类型 :param output_file: 输出音频文件路径 """ import logging import struct logger = logging.getLogger(__name__) logger.info(f"开始WebSocket TTS转换，文本长度: {len(text)}，音色类型: {voice_type}") request_data = self._build_base_request(text, voice_type, "submit", **kwargs) headers = { "Authorization": f"Bearer; {self.token}", "X-Secret-Key": self.secret_key } # 构造WebSocket请求报文头 (4字节) # 协议版本(4bit)=0b0001, 报头大小(4bit)=0b0001(4字节) header_byte1 = 0b0001 << 4 | 0b0001 # 消息类型(4bit)=0b0001(full client request), 标志(4bit)=0b0000 header_byte2 = 0b0001 << 4 | 0b0000 # 序列化方法(4bit)=0b0001(JSON), 压缩方法(4bit)=0b0000(无压缩) header_byte3 = 0b0001 << 4 | 0b0000 # 保留字段(8bit)=0x00 header_byte4 = 0x00 request_header = struct.pack('>BBBB', header_byte1, header_byte2, header_byte3, header_byte4) # 序列化请求数据并添加报头 request_payload = json.dumps(request_data).encode('utf-8') request_message = request_header + request_payload async with websockets.connect(self.websocket_url, extra_headers=headers) as websocket: # 发送请求 await websocket.send(request_message) # 接收音频数据 audio_data = b"" while True: try: response = await websocket.recv() if isinstance(response, bytes): # 解析响应报文头 (4字节) if len(response) < 4: logger.error("响应报文过短，无法解析报头") continue header_byte1, header_byte2, header_byte3, header_byte4 = struct.unpack('>BBBB', response[:4]) protocol_version = (header_byte1 >> 4) & 0x0F header_size = (header_byte1 & 0x0F) * 4 message_type = (header_byte2 >> 4) & 0x0F flags = header_byte2 & 0x0F logger.debug(f"协议版本: {protocol_version}, 报头大小: {header_size}, 消息类型: {message_type}, 标志: {flags}") # 处理音频响应 (0b1011=11) if message_type == 0b1011: payload = response[header_size:] audio_data += payload logger.debug(f"接收到音频数据块，长度: {len(payload)} bytes") # 检查是否为最后一条消息 (sequence < 0) if flags & 0b0010: logger.info("接收到最后一块音频数据") break # 处理错误消息 (0b1111=15) elif message_type == 0b1111: error_payload = response[header_size:].decode('utf-8') logger.error(f"服务器错误响应: {error_payload}") break elif isinstance(response, str): logger.warning(f"接收到文本消息: {response}") if '"code":' in response: error_info = json.loads(response) logger.error(f"API错误: {error_info.get('message', '未知错误')} (错误码: {error_info.get('code')})") break except websockets.exceptions.ConnectionClosed as e: logger.error(f"WebSocket连接异常关闭: {e}") logger.error(f"连接关闭时状态码: {e.code}，原因: {e.reason}") break except Exception as e: logger.error(f"处理响应时发生异常: {str(e)}") break # 播放音频 if audio_data: self._play_audio(audio_data) logger.info("音频播放完成") # 保存音频文件 if output_file: with open(output_file, "wb") as f: f.write(audio_data) logger.info(f"音频已保存至: {output_file}") else: logger.error("未接收到有效的音频数据") if 'last_message' in locals(): logger.error(f"最后接收到的消息类型: {'str' if isinstance(last_message, str) else 'bytes'}") if isinstance(last_message, str): logger.error(f"最后接收到的文本消息: {last_message}") def _play_audio(self, audio_data: bytes): """ 播放音频数据 :param audio_data: 音频二进制数据 """ p = pyaudio.PyAudio() stream = p.open(format=pyaudio.paInt16, channels=1, rate=24000, output=True) # 将音频数据转换为流 audio_stream = io.BytesIO(audio_data) stream.write(audio_stream.read()) stream.stop_stream() stream.close() p.terminate() def http_tts(self, text: str, voice_type: str, output_file: str = None,** kwargs): """ HTTP非流式合成（全部合成后返回） :param text: 合成文本 :param voice_type: 音色类型 :param output_file: 输出音频文件路径 """ request_data = self._build_base_request(text, voice_type, "query", **kwargs) headers = { "Authorization": f"Bearer; {self.token}", "Content-Type": "application/json" } try: response = requests.post(self.http_url, json=request_data, headers=headers) response.raise_for_status() # 检查HTTP错误 result = response.json() if result.get("code") == 3000: # 解码并播放音频 audio_data = base64.b64decode(result["data"]) self._play_audio(audio_data) # 保存音频文件 if output_file: with open(output_file, "wb") as f: f.write(audio_data) print(f"音频合成成功，时长：{result['addition']['duration']}ms") return True else: error_msg = result.get("message", "未知错误") error_code = result.get("code", "未知错误码") print(f"合成失败：{error_msg}（错误码：{error_code}）") return False except requests.exceptions.RequestException as e: print(f"HTTP请求失败: {str(e)}") return False except (KeyError, ValueError) as e: print(f"响应解析失败: {str(e)}") return False # 使用示例 if __name__ == "__main__": # 从配置中获取默认音色类型 from api.config import API_CONFIGS VOICE_TYPE = API_CONFIGS['volcano_tts']['default_voice_type'] tts = VolcanoTTS() # 示例1：HTTP合成 tts.http_tts( text="欢迎使用火山引擎大模型语音合成API", voice_type=VOICE_TYPE, output_file="http_tts_result.mp3", encoding="mp3", speed_ratio=1.2 ) # 示例2：WebSocket流式合成（需异步运行） async def run_websocket_demo(): await tts.websocket_tts( text="这是一个WebSocket流式合成的示例", voice_type=VOICE_TYPE, output_file="websocket_tts_result.mp3", encoding="wav" ) asyncio.run(run_websocket_demo())