基于JavaScript实现聊天APP语音消息录制与持续监听技术咨询

Hey there! I’ve tackled similar voice recording challenges in chat apps before, so let’s walk through your questions with practical, actionable steps.

Building voice recording functionality boils down to choosing the right tools, handling core workflows, and addressing platform-specific requirements. Here’s how to approach it:

• Pick a platform-appropriate audio library

  • For native Android: Use MediaRecorder (simpler, pre-built) or AudioRecord (more control over raw audio data).
  • For native iOS: Use AVAudioRecorder from the AVFoundation framework.
  • For cross-platform apps (Flutter/React Native): Go with libraries like audioplayers (Flutter) or react-native-audio (React Native) — these abstract away platform differences.

• Core Recording Workflow

  1. Request permissions first: You’ll need to ask users for microphone access upfront. On Android, declare the RECORD_AUDIO permission in your manifest; on iOS, add NSMicrophoneUsageDescription to Info.plist explaining why you need the mic.
  2. Initialize the recorder: Configure settings like audio source (mic), output format (AAC/MP3 are ideal for chat apps due to small file sizes), and save path (use a temporary local file, then upload to your chat server later).
  3. Handle user controls: Wire up start/pause/stop buttons to trigger recorder methods. Update your UI to show recording state (e.g., a pulsating icon or live timer).
  4. Post-recording processing: Compress the audio file if needed, then upload it to your backend or pass it directly to your speech-to-text service.

The issue you faced — recording stopping after silence — usually stems from built-in silence detection in your library or system-level resource limits. Here’s how to fix it:

• Disable or adjust auto-stop silence detection
  Most pre-built recording libraries have default silence thresholds that trigger auto-stop. Override this:

  • For Android’s MediaRecorder, there’s no built-in auto-stop, but if you used a third-party library, look for settings like setSilenceTimeout and set it to a very high value or disable it entirely.
  • For iOS’s AVAudioRecorder, enable metering with meteringEnabled = true, then manually check audio levels instead of relying on auto-stop.

• Manually control recording with raw audio data
  For full control, use low-level audio APIs (like Android’s AudioRecord or iOS’s AVAudioEngine) to capture raw audio streams. You can then:

  1. Calculate audio amplitude from the raw data to detect silence.
  2. Set your own silence timeout (e.g., stop only if silence lasts 10+ seconds, not the default 2-3 seconds).
  3. Only stop recording when the user taps the stop button or your custom timeout is hit.

  Example snippet (Android AudioRecord):

  // Initialize recorder
  int sampleRate = 44100;
  int bufferSize = AudioRecord.getMinBufferSize(sampleRate, AudioFormat.CHANNEL_IN_MONO, AudioFormat.ENCODING_PCM_16BIT);
  AudioRecord recorder = new AudioRecord(MediaRecorder.AudioSource.MIC, sampleRate, AudioFormat.CHANNEL_IN_MONO, AudioFormat.ENCODING_PCM_16BIT, bufferSize);
  
  boolean isRecording = true;
  long silenceStart = System.currentTimeMillis();
  final long MAX_SILENCE = 10000; // 10 seconds
  
  // Start recording in a background thread
  new Thread(() -> {
      recorder.startRecording();
      byte[] buffer = new byte[bufferSize];
      while (isRecording) {
          int bytesRead = recorder.read(buffer, 0, bufferSize);
          if (bytesRead > 0) {
              // Save buffer to file
              writeToFile(buffer, bytesRead);
              // Calculate amplitude to check for sound
              double amplitude = getAmplitude(buffer, bytesRead);
              if (amplitude > 0.01) {
                  silenceStart = System.currentTimeMillis(); // Reset silence timer
              } else {
                  // Stop only if silence exceeds max duration
                  if (System.currentTimeMillis() - silenceStart > MAX_SILENCE) {
                      isRecording = false;
                      recorder.stop();
                  }
              }
          }
      }
  }).start();
  

• Prevent system resource recycling
  On mobile, systems may pause background audio to save resources. To avoid this:

  • Android: Run the recording service as a foreground service with a persistent notification, so the system knows it’s a high-priority task.
  • iOS: Enable the Background Audio capability in your app settings, and set the audio session category to playAndRecord.

• Adjust Voice Activity Detection (VAD) logic
  If you were using VAD to trigger recording stops, rework it: use VAD only to start recording when the user speaks (optional), but let the user manually stop it. Or set VAD’s silence threshold much lower so it doesn’t misinterpret soft background noise as silence.

内容的提问来源于stack exchange，提问作者Post Malone