Recherche avancée

Sur d’autres sites (5078)

• How to extract audio from a video in Flutter ?

  14 janvier, par Mohammed Bekele

  I have an image picker in Flutter to get the video from the device, and then I created a function to extract the audio using ffmpeg_kit_flutter package.

  


   Future<void> _convertVideoToAudio() async {&#xA;    if (_pickedVideo != null) {&#xA;      bool? permissionGranted = await _requestStoragePermission();&#xA;      if (permissionGranted != true) {&#xA;        print("Storage permission denied.");&#xA;        return;&#xA;      }&#xA;&#xA;      String videoPath = _pickedVideo!.path;&#xA;      _outputPath = await getOutputFilePath(); // Get platform-specific path&#xA;&#xA;      try {&#xA;        // Ensure the output directory exists&#xA;        await Directory(path.dirname(_outputPath)).create(recursive: true);&#xA;&#xA;        await FFmpegKit.execute(&#xA;            "-i $videoPath -vn -c:a libmp3lame -q:a 2 $_outputPath"); // FFmpeg command&#xA;        print("Video converted to audio successfully!");&#xA;        _showSuccessDialog(); // Display success dialog&#xA;&#xA;        try {&#xA;          final String fileName = path.basename(_outputPath);&#xA;          final transcription =&#xA;              await _sendAudioForTranscription(_outputPath, fileName);&#xA;&#xA;          if (transcription != null) {&#xA;            setState(() {&#xA;              _transcription = transcription;&#xA;            });&#xA;          } else {&#xA;            setState(() {&#xA;              _transcription = "Transcription failed";&#xA;            });&#xA;          }&#xA;        } catch (e) {&#xA;          print(&#x27;Error in transcription request: $e&#x27;);&#xA;          setState(() {&#xA;            _transcription = "Network request failed";&#xA;          });&#xA;        }&#xA;      } catch (e) {&#xA;        print("Error converting video: $e");&#xA;        _showErrorDialog(); // Display error dialog&#xA;      } finally {&#xA;        setState(() {&#xA;          _pickedVideo = null; // Clear selected video&#xA;        });&#xA;      }&#xA;    } else {&#xA;      print("Please pick a video first.");&#xA;    }&#xA;  }&#xA;</void>

  


  and for getting the path I have this function

  


   Future<string> getOutputFilePath() async {&#xA;    final directory = await getApplicationDocumentsDirectory();&#xA;    final downloadsDirectory = Directory(&#x27;${directory.path}/downloads&#x27;);&#xA;    if (!(await downloadsDirectory.exists())) {&#xA;      await downloadsDirectory.create(recursive: true);&#xA;    }&#xA;    final String fileName = path&#xA;        .basename(_pickedVideo!.path)&#xA;        .replaceAll(&#x27;.mp4&#x27;, &#x27;.mp3&#x27;); // Replace extension&#xA;    final filePath = &#x27;${downloadsDirectory.path}/$fileName&#x27;;&#xA;    return filePath;&#xA;  }&#xA;</string>

  


  but this is not working somehow. Because after I get the audio I'm uploading it to a server with http, then it displays that there is no path where the audio supposed to be.

  


• Video Ending prematurely with websocket and ffmpeg

  4 septembre 2024, par Zaid

  I am working on a screen recording application and it allows users to record their screen and I use websocket to send byte data in real time to my fastapi python server. The bytes are send every 2 second and I use ffmpeg to keep saving the bytes in the output mp4 video file. Everything was working fine when I had the server running on my local machine, however, I just deployed the server to EC2 instance us-ease-1 and when I try to record the videos, the videos are really short, IE, if I record a 30 second video, it only has the video for 3 second. Sometimes it saves 90% of the video and sometimes less. I am not sure what is the problem here.
I have been trying to debug the code for the past few days, but no success

  


  Here is my code :-

  


  FRONTEND

  


  const recorder = new MediaRecorder(stream, {
            mimeType: 'video/webm;codecs=H264',
            videoBitsPerSecond: 8000000
        });

recorder.ondataavailable = (e: BlobEvent) => {
    socketRef.socket.send(e.data)
}


  


  And here is my python code :-

  


  @router.websocket("/stream")
async def websocket_endpoint(websocket: WebSocket, token: str = Query(...), videoId: str = Query(...), authorize: AuthJWT = Depends()):
    await manager.connect(websocket)
    dataNumber = 1

    recordingFile = os.path.join(temp_dir, f"recording_{videoId}.mp4")

    command = [
        'ffmpeg', 
        '-y',
        '-i', 
        '-', 
        '-codec:v', 
        'copy', 
        '-f', 'mp4',
        recordingFile,
    ]

    process = subprocess.Popen(command, stdin=subprocess.PIPE)
    try:
        while True:
            try:
                data = await websocket.receive_bytes()
                if not data:
                    break
                process.stdin.write(data)
                await websocket.send_json({"chunkNumber": dataNumber, "status": 200})
                dataNumber = dataNumber + 1
            except RuntimeError:
                break      
    except WebSocketDisconnect:
        print(f"Client disconnected: {websocket.client.host}")
    finally:
        manager.disconnect(websocket)

        # Close stdin to signal EOF
        process.stdin.close()

        # Wait for FFmpeg to finish processing
        process.wait()

        # Ensure that the process is terminated
        process.terminate()  


  


  I also get this error in the console :-


  


• Write buffer to ffmpeg stdin and send websocket message after that

  6 septembre 2024, par alpecca

  I am working on a streaming application in which user can stream videos in real time to my server usig websocket and media recorder. I have written my backend code using fastapi python and I setup a websocket endpoint that would receive buffer data from the frontend every two second. But the problem I am facing is here :-

  


  process.stdin.write(data)
await websocket.send_json


  


  Here the code ffmpeg process write could take some time to take a buffer and write a mp4 for it, but the websocket send json won't wait for it and thus just send the message back to the client, which is causing videos being too short and currupted once the user stops the recording.

  


  Here is the full code

  


  @router.websocket("/stream")
async def websocket_endpoint(websocket: WebSocket, token: str = Query(...), videoId: str = Query(...), authorize: AuthJWT = Depends()):
    await manager.connect(websocket)
    dataNumber = 1

    recordingFile = os.path.join(temp_dir, f"recording_{videoId}.mp4")

    command = [
        'ffmpeg', 
        '-y',
        '-i', 
        '-', 
        '-codec:v', 
        'copy', 
        '-f', 'mp4',
        recordingFile,
        # "-"
        # f'output{queueNumber}.mp4',
    ]

    process = subprocess.Popen(command, stdin=subprocess.PIPE)

    try:
        while True:
            try:
                data = await websocket.receive_bytes()
                if not data:
                    break
                process.stdin.write(data)
                process.stdin.flush()
                await websocket.send_json({"chunkNumber": dataNumber, "status": 200})
                dataNumber = dataNumber + 1
            except RuntimeError:
                break      
    except WebSocketDisconnect:
        print(f"Client disconnected: {websocket.client.host}")
    finally:
        manager.disconnect(websocket)
        process.stdin.close()
        process.wait()
        process.terminate()


  


  What I want to do is on each buffer send from the client, I want to make sure that ffmpeg writes that to the filesystem compeletly and than only to send the websocket message back to the client. And also, as the request are coming from the client every 2 second no matter what, If the write is taking too long for the previous message, I want to make sure that that the ffmpeg first write that part to the file and send the message and than do the newer one