fix: 唤醒机制

main/xiaozhi-server/core/handle/sendAudioHandle.py

@@ -30,13 +30,14 @@ async def sendAudioMessage(conn, sentenceType, audios, text):
 
 
 # 播放音频
-async def sendAudio(conn, audios, pre_buffer=False):
+async def sendAudio(conn, audios, pre_buffer=False, frame_duration=60):
     """
     发送单个opus包，支持流控
     Args:
         conn: 连接对象
         opus_packet: 单个opus数据包
         pre_buffer: 快速发送音频
+        frame_duration: 帧时长（毫秒），匹配 Opus 编码
     """
     if audios is None or len(audios) == 0:
         return
@@ -60,7 +61,6 @@ async def sendAudio(conn, audios, pre_buffer=False):
                 "start_time": time.perf_counter(),
             }
 
-        frame_duration=60
         flow_control = conn.audio_flow_control
         current_time = time.perf_counter()
         # 计算预期发送时间
@@ -77,6 +77,35 @@ async def sendAudio(conn, audios, pre_buffer=False):
         # 更新流控状态
         flow_control["packet_count"] += 1
         flow_control["last_send_time"] = time.perf_counter()
+    else:
+        # 流控参数优化
+        start_time = time.perf_counter()
+        play_position = 0
+
+        # 执行预缓冲
+        pre_buffer_frames = min(3, len(audios))
+        for i in range(pre_buffer_frames):
+            await conn.websocket.send(audios[i])
+        remaining_audios = audios[pre_buffer_frames:]
+
+        # 播放剩余音频帧
+        for opus_packet in remaining_audios:
+            if conn.client_abort:
+                break
+
+            # 重置没有声音的状态
+            conn.last_activity_time = time.time() * 1000
+
+            # 计算预期发送时间
+            expected_time = start_time + (play_position / 1000)
+            current_time = time.perf_counter()
+            delay = expected_time - current_time
+            if delay > 0:
+                await asyncio.sleep(delay)
+
+            await conn.websocket.send(opus_packet)
+
+            play_position += frame_duration
 
 
 async def send_tts_message(conn, state, text=None):