Android通话状态下Kotlin实现音频路由至扬声器的问题排查

1. 当前音频路由至扬声器的实现是否正确？ #

不正确，存在几个关键问题：

• 权限缺失：必须在Manifest中添加MODIFY_AUDIO_SETTINGS权限，否则无法修改音频路由和模式
• 操作顺序错误：应该先配置MediaPlayer的AudioAttributes，再调整AudioManager的模式——AudioAttributes会影响系统的音频路由策略，提前设置才能让系统正确识别播放场景
• 未保存原始状态：直接覆盖系统的音频模式和扬声器状态，会破坏用户之前的设置，且现有恢复逻辑无法还原真实的初始状态
• AudioAttributes选型不当：USAGE_VOICE_COMMUNICATION是针对通话语音的，你的场景是播放音效，应该使用USAGE_ASSISTANCE_SONIFICATION或USAGE_MEDIA，配合CONTENT_TYPE_SONIFICATION，否则系统可能强制将音频路由到听筒

2. 音频播放与异常处理的改进方案及最佳实践 #

核心改进点： #

• 权限前置配置：在AndroidManifest.xml中添加权限：

  <uses-permission android:name="android.permission.MODIFY_AUDIO_SETTINGS" />
  

• 细粒度异常捕获：避免使用大而全的catch (e: Exception)，针对性捕获IOException（数据源错误）、IllegalStateException（MediaPlayer状态非法）等具体异常，便于精准定位问题
• MediaPlayer生命周期安全：
  • 使用prepareAsync()替代prepare()，避免阻塞主线程
  • 注册OnPreparedListener，在准备完成后再启动播放
  • 统一封装播放停止逻辑，确保在所有终止场景（播放完成、错误、焦点丢失、组件销毁）都能释放资源
• 音频焦点管理：通话状态下系统持有音频焦点，必须主动请求焦点，避免播放被系统中断：

  val focusRequest = AudioFocusRequest.Builder(AudioManager.AUDIOFOCUS_GAIN_TRANSIENT)
      .setAudioAttributes(audioAttributes)
      .setOnAudioFocusChangeListener { focusChange ->
          when (focusChange) {
              AudioManager.AUDIOFOCUS_LOSS -> stopPlayback(audioManager)
              AudioManager.AUDIOFOCUS_LOSS_TRANSIENT -> mediaPlayer?.pause()
              AudioManager.AUDIOFOCUS_GAIN -> mediaPlayer?.start()
          }
      }
      .build()
  

• 资源自动管理：使用Kotlin的use函数自动关闭RawResourceFd，避免资源泄漏：

  resources.openRawResourceFd(R.raw.gacha).use { afd ->
      mediaPlayer.setDataSource(afd.fileDescriptor, afd.startOffset, afd.length)
  }
  

• 完善状态监听：添加OnErrorListener处理播放过程中的错误，防止MediaPlayer进入无效状态

3. 音频播放结束后重置扬声器状态的有效方法 #

核心思路是先保存原始音频状态，再在所有播放终止场景中统一恢复，避免破坏用户原有设置：

1. 定义成员变量保存初始状态：

   private var originalAudioMode: Int = AudioManager.MODE_NORMAL
   private var originalSpeakerphoneState: Boolean = false
   

2. 在调整音频路由前保存原始状态：

   originalAudioMode = audioManager.mode
   originalSpeakerphoneState = audioManager.isSpeakerphoneOn
   

3. 封装统一的状态恢复方法：

   private fun restoreAudioState(audioManager: AudioManager) {
       try {
           audioManager.isSpeakerphoneOn = originalSpeakerphoneState
           audioManager.mode = originalAudioMode
       } catch (e: SecurityException) {
           Log.e("Tag", "无权限修改音频设置", e)
       } catch (e: Exception) {
           Log.e("Tag", "恢复音频状态失败", e)
       }
   }
   

4. 在所有播放终止场景调用恢复逻辑：包括OnCompletionListener、OnErrorListener、音频焦点丢失回调、组件销毁方法（如Service.onDestroy）

改进后的完整代码示例 #

private var mediaPlayer: MediaPlayer? = null
private var originalAudioMode: Int = AudioManager.MODE_NORMAL
private var originalSpeakerphoneState: Boolean = false
private var audioFocusRequest: AudioFocusRequest? = null

private fun playSound() {
    val audioManager = getSystemService(Context.AUDIO_SERVICE) as AudioManager
    // 保存原始音频状态
    originalAudioMode = audioManager.mode
    originalSpeakerphoneState = audioManager.isSpeakerphoneOn

    val audioAttributes = AudioAttributes.Builder()
        .setUsage(AudioAttributes.USAGE_ASSISTANCE_SONIFICATION)
        .setContentType(AudioAttributes.CONTENT_TYPE_SONIFICATION)
        .build()

    // 请求音频焦点
    audioFocusRequest = AudioFocusRequest.Builder(AudioManager.AUDIOFOCUS_GAIN_TRANSIENT)
        .setAudioAttributes(audioAttributes)
        .setOnAudioFocusChangeListener { focusChange ->
            when (focusChange) {
                AudioManager.AUDIOFOCUS_LOSS -> stopPlayback(audioManager)
                AudioManager.AUDIOFOCUS_LOSS_TRANSIENT -> mediaPlayer?.pause()
                AudioManager.AUDIOFOCUS_GAIN -> mediaPlayer?.start()
            }
        }
        .build()

    val focusResult = audioManager.requestAudioFocus(audioFocusRequest!!)
    if (focusResult != AudioManager.AUDIOFOCUS_REQUEST_GRANTED) {
        Log.w("Tag", "音频焦点请求被拒绝")
        return
    }

    mediaPlayer = MediaPlayer().apply {
        setAudioAttributes(audioAttributes)
        // 自动管理RawResourceFd
        resources.openRawResourceFd(R.raw.gacha).use { afd ->
            setDataSource(afd.fileDescriptor, afd.startOffset, afd.length)
        }
        setOnPreparedListener {
            // 准备完成后调整音频路由
            try {
                audioManager.mode = AudioManager.MODE_IN_COMMUNICATION
                audioManager.isSpeakerphoneOn = true
                it.start()
                Log.d("Tag", "音效播放开始")
            } catch (e: Exception) {
                Log.e("Tag", "开启扬声器失败", e)
                stopPlayback(audioManager)
            }
        }
        setOnCompletionListener {
            Log.d("Tag", "音效播放完成")
            stopPlayback(audioManager)
        }
        setOnErrorListener { _, what, extra ->
            Log.e("Tag", "播放错误：what=$what, extra=$extra")
            stopPlayback(audioManager)
            true // 标记错误已处理
        }
        prepareAsync() // 异步准备，避免阻塞主线程
    }
}

private fun stopPlayback(audioManager: AudioManager) {
    mediaPlayer?.let {
        if (it.isPlaying) it.stop()
        it.release()
        mediaPlayer = null
    }
    audioManager.abandonAudioFocusRequest(audioFocusRequest!!)
    audioFocusRequest = null
    restoreAudioState(audioManager)
}

private fun restoreAudioState(audioManager: AudioManager) {
    try {
        audioManager.isSpeakerphoneOn = originalSpeakerphoneState
        audioManager.mode = originalAudioMode
        Log.d("Tag", "音频状态已恢复")
    } catch (e: Exception) {
        Log.e("Tag", "恢复音频状态失败", e)
    }
}

// 组件销毁时调用（如Service的onDestroy）
override fun onDestroy() {
    super.onDestroy()
    val audioManager = getSystemService(Context.AUDIO_SERVICE) as AudioManager
    stopPlayback(audioManager)
}

内容的提问来源于stack exchange，提问作者Moaz