Recherche avancée

Sur d’autres sites (10537)

• Qt Open Source Product 2 - Vlc Demo [closed]

  3 juin 2021, par cool code

  Ⅰ. Preface

  


  The previous work was made by the FFmpeg kernel, and FFmpeg is too powerful for many beginners to understand. There are also a lot of users only need a simple video stream can be played, do not need to be involved in the responsible decoding and transcoding, so VLC came in handy, it directly made FFMPEG deep encapsulation, to provide a friendly interface. There's an MPV that does the same thing, and MPV is even better than VLC in that it's just one library file, and it looks like it's packaged as a static library, unlike VLC, VLC comes with a bunch of dynamic library files and plug-in files.
Of course, the simplicity of VLC is that it only needs a few lines of code to start, so that beginners immediately see the effect is very important, very excited, you can more quickly carry out the next step of coding, experience the fun of coding.

  


  Ⅱ. Code framework

  


  #include "ffmpeg.h"

FFmpegThread::FFmpegThread(QObject *parent) : QThread(parent)
{
    setObjectName("FFmpegThread");
    stopped = false;
    isPlay = false;

    frameFinish = false;
    videoWidth = 0;
    videoHeight = 0;
    oldWidth = 0;
    oldHeight = 0;
    videoStreamIndex = -1;
    audioStreamIndex = -1;

    url = "rtsp://192.168.1.128:554/1";

    buffer = NULL;
    avPacket = NULL;
    avFrame = NULL;
    avFrame2 = NULL;
    avFrame3 = NULL;
    avFormatContext = NULL;
    videoCodec = NULL;
    audioCodec = NULL;
    swsContext = NULL;

    options = NULL;
    videoDecoder = NULL;
    audioDecoder = NULL;

    //Initial registration, only register once in a software
    FFmpegThread::initlib();
}

//Only need to initialize once in a software
void FFmpegThread::initlib()
{
    static QMutex mutex;
    QMutexLocker locker(&mutex);
    static bool isInit = false;
    if (!isInit) {
        //Register all available file formats and decoders in the library
        av_register_all();
        //Register all devices, mainly for local camera playback support
#ifdef ffmpegdevice
        avdevice_register_all();
#endif
        //Initialize the network stream format, which must be executed first when using the network stream
        avformat_network_init();

        isInit = true;
        qDebug() << TIMEMS << "init ffmpeg lib ok" << " version:" << FFMPEG_VERSION;
#if 0
        //Output all supported decoder names
        QStringList listCodeName;
        AVCodec *code = av_codec_next(NULL);
        while (code != NULL) {
            listCodeName << code->name;
            code = code->next;
        }

        qDebug() << TIMEMS << listCodeName;
#endif
    }
}

bool FFmpegThread::init()
{
    //Before opening the code stream, specify various parameters such as: detection time/timeout time/maximum delay, etc.
    //Set the cache size, 1080p can increase the value
    av_dict_set(&options, "buffer_size", "8192000", 0);
    //Open in tcp mode, if open in udp mode, replace tcp with udp
    av_dict_set(&options, "rtsp_transport", "tcp", 0);
    //Set the timeout disconnection time, the unit is microseconds, 3000000 means 3 seconds
    av_dict_set(&options, "stimeout", "3000000", 0);
    //Set the maximum delay, in microseconds, 1000000 means 1 second
    av_dict_set(&options, "max_delay", "1000000", 0);
    //Automatically start the number of threads
    av_dict_set(&options, "threads", "auto", 0);

    //Open video stream
    avFormatContext = avformat_alloc_context();

    int result = avformat_open_input(&avFormatContext, url.toStdString().data(), NULL, &options);
    if (result < 0) {
        qDebug() << TIMEMS << "open input error" << url;
        return false;
    }

    //Release setting parameters
    if (options != NULL) {
        av_dict_free(&options);
    }

    //Get flow information
    result = avformat_find_stream_info(avFormatContext, NULL);
    if (result < 0) {
        qDebug() << TIMEMS << "find stream info error";
        return false;
    }

    //----------At the beginning of the video stream part, make a mark to facilitate the folding of the code----------
    if (1) {
        videoStreamIndex = av_find_best_stream(avFormatContext, AVMEDIA_TYPE_VIDEO, -1, -1, &videoDecoder, 0);
        if (videoStreamIndex < 0) {
            qDebug() << TIMEMS << "find video stream index error";
            return false;
        }

        //Get video stream
        AVStream *videoStream = avFormatContext->streams[videoStreamIndex];

        //Get the video stream decoder, or specify the decoder
        videoCodec = videoStream->codec;
        videoDecoder = avcodec_find_decoder(videoCodec->codec_id);
        //videoDecoder = avcodec_find_decoder_by_name("h264_qsv");
        if (videoDecoder == NULL) {
            qDebug() << TIMEMS << "video decoder not found";
            return false;
        }

        //Set up accelerated decoding
        videoCodec->lowres = videoDecoder->max_lowres;
        videoCodec->flags2 |= AV_CODEC_FLAG2_FAST;

        //Open the video decoder
        result = avcodec_open2(videoCodec, videoDecoder, NULL);
        if (result < 0) {
            qDebug() << TIMEMS << "open video codec error";
            return false;
        }

        //Get the resolution size
        videoWidth = videoStream->codec->width;
        videoHeight = videoStream->codec->height;

        //If the width and height are not obtained, return
        if (videoWidth == 0 || videoHeight == 0) {
            qDebug() << TIMEMS << "find width height error";
            return false;
        }

        QString videoInfo = QString("Video stream info -> index: %1  decode: %2  format: %3  duration: %4 s  Resolution: %5*%6")
                            .arg(videoStreamIndex).arg(videoDecoder->name).arg(avFormatContext->iformat->name)
                            .arg((avFormatContext->duration) / 1000000).arg(videoWidth).arg(videoHeight);
        qDebug() << TIMEMS << videoInfo;
    }
    //----------The video stream part starts----------

    //----------Start the audio stream part, mark it to facilitate the code folding----------
    if (1) {
        //Loop to find audio stream index
        audioStreamIndex = -1;
        for (uint i = 0; i < avFormatContext->nb_streams; i++) {
            if (avFormatContext->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO) {
                audioStreamIndex = i;
                break;
            }
        }

        //Some have no audio stream, so there is no need to return here
        if (audioStreamIndex == -1) {
            qDebug() << TIMEMS << "find audio stream index error";
        } else {
            //Get audio stream
            AVStream *audioStream = avFormatContext->streams[audioStreamIndex];
            audioCodec = audioStream->codec;

            //Get the audio stream decoder, or specify the decoder
            audioDecoder = avcodec_find_decoder(audioCodec->codec_id);
            //audioDecoder = avcodec_find_decoder_by_name("aac");
            if (audioDecoder == NULL) {
                qDebug() << TIMEMS << "audio codec not found";
                return false;
            }

            //Open the audio decoder
            result = avcodec_open2(audioCodec, audioDecoder, NULL);
            if (result < 0) {
                qDebug() << TIMEMS << "open audio codec error";
                return false;
            }

            QString audioInfo = QString("Audio stream information -> index: %1  decode: %2  Bit rate: %3  channel num: %4  sampling: %5")
                                .arg(audioStreamIndex).arg(audioDecoder->name).arg(avFormatContext->bit_rate)
                                .arg(audioCodec->channels).arg(audioCodec->sample_rate);
            qDebug() << TIMEMS << audioInfo;
        }
    }
    //----------End of audio stream----------

    //Pre-allocated memory
    avPacket = av_packet_alloc();
    avFrame = av_frame_alloc();
    avFrame2 = av_frame_alloc();
    avFrame3 = av_frame_alloc();

    //Compare the width and height of the last file. When changing, you need to reallocate the memory
    if (oldWidth != videoWidth || oldHeight != videoHeight) {
        int byte = avpicture_get_size(AV_PIX_FMT_RGB32, videoWidth, videoHeight);
        buffer = (uint8_t *)av_malloc(byte * sizeof(uint8_t));
        oldWidth = videoWidth;
        oldHeight = videoHeight;
    }

    //Define pixel format
    AVPixelFormat srcFormat = AV_PIX_FMT_YUV420P;
    AVPixelFormat dstFormat = AV_PIX_FMT_RGB32;
    //Get the decoded format through the decoder
    srcFormat = videoCodec->pix_fmt;

    //The SWS_FAST_BILINEAR parameter used by the default fastest decoding may lose part of the picture data, and you can change it to other parameters by yourself
    int flags = SWS_FAST_BILINEAR;

    //Open up a cache to store one frame of data
    //The following two methods are ok, avpicture_fill has been gradually abandoned
    //avpicture_fill((AVPicture *)avFrame3, buffer, dstFormat, videoWidth, videoHeight);
    av_image_fill_arrays(avFrame3->data, avFrame3->linesize, buffer, dstFormat, videoWidth, videoHeight, 1);

    //Image conversion
    swsContext = sws_getContext(videoWidth, videoHeight, srcFormat, videoWidth, videoHeight, dstFormat, flags, NULL, NULL, NULL);

    //Output video information
    //av_dump_format(avFormatContext, 0, url.toStdString().data(), 0);

    //qDebug() << TIMEMS << "init ffmpeg finsh";
    return true;
}

void FFmpegThread::run()
{
    while (!stopped) {
        //Perform initialization based on the flag bit
        if (isPlay) {
            this->init();
            isPlay = false;
            continue;
        }

        if (av_read_frame(avFormatContext, avPacket) >= 0) {
            //Determine whether the current package is video or audio
            int index = avPacket->stream_index;
            if (index == videoStreamIndex) {
                //Decode video stream avcodec_decode_video2 method has been deprecated
#if 0
                avcodec_decode_video2(videoCodec, avFrame2, &frameFinish, avPacket);
#else
                frameFinish = avcodec_send_packet(videoCodec, avPacket);
                if (frameFinish < 0) {
                    continue;
                }

                frameFinish = avcodec_receive_frame(videoCodec, avFrame2);
                if (frameFinish < 0) {
                    continue;
                }
#endif

                if (frameFinish >= 0) {
                    //Turn the data into a picture
                    sws_scale(swsContext, (const uint8_t *const *)avFrame2->data, avFrame2->linesize, 0, videoHeight, avFrame3->data, avFrame3->linesize);

                    //The following two methods can be used
                    //QImage image(avFrame3->data[0], videoWidth, videoHeight, QImage::Format_RGB32);
                    QImage image((uchar *)buffer, videoWidth, videoHeight, QImage::Format_RGB32);
                    if (!image.isNull()) {
                        emit receiveImage(image);
                    }

                    msleep(1);
                }
            } else if (index == audioStreamIndex) {
                //Decode the audio stream, it will not be processed here, and will be handed over to sdl to play
            }
        }

        av_packet_unref(avPacket);
        av_freep(avPacket);
        msleep(1);
    }

    //Release resources after the thread ends
    free();
    stopped = false;
    isPlay = false;
    qDebug() << TIMEMS << "stop ffmpeg thread";
}

void FFmpegThread::setUrl(const QString &url)
{
    this->url = url;
}

void FFmpegThread::free()
{
    if (swsContext != NULL) {
        sws_freeContext(swsContext);
        swsContext = NULL;
    }

    if (avPacket != NULL) {
        av_packet_unref(avPacket);
        avPacket = NULL;
    }

    if (avFrame != NULL) {
        av_frame_free(&avFrame);
        avFrame = NULL;
    }

    if (avFrame2 != NULL) {
        av_frame_free(&avFrame2);
        avFrame2 = NULL;
    }

    if (avFrame3 != NULL) {
        av_frame_free(&avFrame3);
        avFrame3 = NULL;
    }

    if (videoCodec != NULL) {
        avcodec_close(videoCodec);
        videoCodec = NULL;
    }

    if (audioCodec != NULL) {
        avcodec_close(audioCodec);
        audioCodec = NULL;
    }

    if (avFormatContext != NULL) {
        avformat_close_input(&avFormatContext);
        avFormatContext = NULL;
    }

    av_dict_free(&options);
    //qDebug() << TIMEMS << "close ffmpeg ok";
}

void FFmpegThread::play()
{
    //Let the thread perform initialization through the flag bit
    isPlay = true;
}

void FFmpegThread::pause()
{

}

void FFmpegThread::next()
{

}

void FFmpegThread::stop()
{
    //Stop the thread through the flag
    stopped = true;
}

//Real-time video display form class
FFmpegWidget::FFmpegWidget(QWidget *parent) : QWidget(parent)
{
    thread = new FFmpegThread(this);
    connect(thread, SIGNAL(receiveImage(QImage)), this, SLOT(updateImage(QImage)));
    image = QImage();
}

FFmpegWidget::~FFmpegWidget()
{
    close();
}

void FFmpegWidget::paintEvent(QPaintEvent *)
{
    if (image.isNull()) {
        return;
    }

    //qDebug() << TIMEMS << "paintEvent" << objectName();
    QPainter painter(this);
    painter.drawImage(this->rect(), image);
}

void FFmpegWidget::updateImage(const QImage &image)
{
    //this->image = image.copy();
    this->image = image;
    this->update();
}

void FFmpegWidget::setUrl(const QString &url)
{
    thread->setUrl(url);
}

void FFmpegWidget::open()
{
    //qDebug() << TIMEMS << "open video" << objectName();
    clear();

    thread->play();
    thread->start();
}

void FFmpegWidget::pause()
{
    thread->pause();
}

void FFmpegWidget::next()
{
    thread->next();
}

void FFmpegWidget::close()
{
    //qDebug() << TIMEMS << "close video" << objectName();
    if (thread->isRunning()) {
        thread->stop();
        thread->quit();
        thread->wait(500);
    }

    QTimer::singleShot(1, this, SLOT(clear()));
}

void FFmpegWidget::clear()
{
    image = QImage();
    update();
}



  


  Ⅲ. Renderings

  


  https://youtu.be/CLsGw2CJW-c

  


  Ⅳ. Open source code download URL

  


  1.download URL for dropbox :

  


  https://www.dropbox.com/sh/n58ucs57pscp25e/AABWBQlg4U3Oz2WF9YOJDrj1a?dl=0

  


  2.download URL for box :

  


  https://app.box.com/s/x48a7ttpk667afqqdk7t1fqok4fmvmyv

  


• ffmpeg library performance for decoding h.264 for embedded device

  2 avril 2015, par pasifus

  I have some confuse when tried to compile and run decode h.264 on ARM and MIPS architecture.

  I have two embedded devices. I tired to run simple code that read h264 format from file and decode it to h264 in loop in maximum speed (without sleep between frames)

  I found that it too slow in there devices.
  I tested HD video (720p/25fps)

  1. MIPS32® 1004K (700MHz) it was 4 fps average.
  2. Raspberry Pi Model B (700MHz) it also was 4 fps average. (i know that raspberry have GPU to decoding/encoding)

  A also check it on my virtual machine on ubuntu i686 (1300MHz) and it was 200 fps average.

  The question : why it so different preference ? Somebody know how to increase decoding preference on MIPS32® 1004K architecture ?

  #include 
  
  #include 
  
  #include 
  
  #include 
  
  
  
  #include <sys></sys>time.h>
  
  #include 
  
  
  
  #include "libavcodec/avcodec.h"
  
  #include "libavutil/mathematics.h"
  
  
  
  #define INBUF_SIZE 80000
  
  
  
  
  
  static long get_time_diff(struct timeval time_now) {
  
     struct timeval time_now2;
  
     gettimeofday(&amp;time_now2,0);
  
     return time_now2.tv_sec*1.e6 - time_now.tv_sec*1.e6 + time_now2.tv_usec - time_now.tv_usec;
  
  }
  
  
  
  int main(int argc, char **argv)
  
  {
  
      AVCodec *codec;
  
      AVCodecContext *c= NULL;
  
      AVCodecParserContext *parser = NULL;
  
      int frame, got_picture, len2, len;
  
      const char *filename;
  
      FILE *f;
  
      AVFrame *picture;
  
      char *arghwtf = malloc(INBUF_SIZE);
  
      uint64_t in_len;
  
      int pts, dts;
  
      struct timeval t,t2;
  
      float inv_fps = 1e6/23.98;
  
      AVPacket avpkt;
  
  
  
      // register all the codecs
  
      avcodec_register_all();
  
  
  
      // log level
  
      av_log_set_level(AV_LOG_PANIC|AV_LOG_FATAL|AV_LOG_ERROR|AV_LOG_WARNING);
  
  
  
      filename = argv[1];
  
  
  
      av_init_packet(&amp;avpkt);
  
  
  
      printf("Decoding file %s...\n", filename);
  
  
  
      // find the H.264 video decoder
  
      codec = avcodec_find_decoder(CODEC_ID_H264);
  
      if (!codec)
  
      {
  
          fprintf(stderr, "codec not found\n");
  
          exit(1);
  
      }
  
  
  
      c = avcodec_alloc_context3(codec);
  
      picture = avcodec_alloc_frame();
  
  
  
      // skiploopfilter=all
  
      c->skip_loop_filter = 48;
  
  
  
      AVDictionary* dictionary = NULL;
  
      if (avcodec_open2(c, codec, &amp;dictionary) &lt; 0)
  
      {
  
          fprintf(stderr, "could not open codec\n");
  
          exit(1);
  
      }
  
  
  
      // the codec gives us the frame size, in samples
  
      parser = av_parser_init(c->codec_id);
  
      parser->flags |= PARSER_FLAG_ONCE;
  
  
  
      f = fopen(filename, "rb");
  
      if (!f)
  
      {
  
          fprintf(stderr, "could not open %s\n", filename);
  
          exit(1);
  
      }
  
  
  
      frame = 0;
  
      gettimeofday(&amp;t, 0);
  
      if(fread(arghwtf, 1, INBUF_SIZE, f) == 0)
  
      {
  
          exit(1);
  
      }
  
  
  
      in_len = 80000;
  
      gettimeofday(&amp;t2, 0);
  
      while (in_len > 0 &amp;&amp; !feof(f))
  
      {
  
          len = av_parser_parse2(parser, c, &amp;avpkt.data, &amp;avpkt.size, arghwtf, in_len, pts, dts, AV_NOPTS_VALUE);
  
  
  
          len2 = avcodec_decode_video2(c, picture, &amp;got_picture, &amp;avpkt);
  
          if (len2 &lt; 0) {
  
              fprintf(stderr, "Error while decoding frame %d\n", frame);
  
              exit(1);
  
          }
  
  
  
          if (got_picture)
  
          {
  
              fprintf(stderr, "\rDisplaying %c %dx%d :frame %3d (%02d:%03d)...", av_get_picture_type_char(picture->pict_type), c->width, c->height, frame, (int)(get_time_diff(t)/1000000), (int)((get_time_diff(t)/1000)%1000));
  
              fflush(stderr);
  
  
  
              frame++;
  
          }
  
  
  
          memcpy(arghwtf, arghwtf + len, 80000-len);
  
          fread(arghwtf + 80000 - len, 1, len, f);
  
      }
  
  
  
      fclose(f);
  
      avcodec_close(c);
  
      av_free(c);
  
      av_free(picture);
  
      printf("\n");
  
      printf("Avarage fps: %d\n", (int)(((double)frame)/(double)(get_time_diff(t)/1000)*1000));
  
  
  
      return 0;
  
  }

• How to embed cover art into MP4 without errors ?

  7 mai 2020, par symonxd

  I want to add artwork to my MP4 file. How do I do this successfully / without errors ?
I've tried these methods with numerous MP4 files, none worked.
If you would like to replicate the errors / bugs, here are the sample files I've used for this question.

  



  I've tried the following :

  



  ---

  



  

  1. ffmpeg -i sample.mp4 -i sample.png -map 0 -map 1 -c copy -disposition:v:0 attached_pic sample_w_artwork.mp4
as stated by Lukas

  


  



  with this error :

  



  [mp4 @ 0000019ee4852280] Could not find tag for codec h264 in stream #0, codec not currently supported in container
Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument


  



  ffmpeg 4.2.2 (Windows build by Zeranoe)

  



  

  2. atomicparsley sample.mp4 --artwork sample.png --overWrite

  


  



  with no success ; AtomicParsley

  



  I get this message when executing the first time :

  



  Started writing to temp file.
 Progress: =============================================>100%|
 Finished writing to temp file.


  



  Executing the next (nth) time gives this :

  



  Updating metadata...   completed.


  



  

  3. Adding artwork through the GUI Tag Editor using both available versions : i686-w64 & x86_64-w64

  


  



  with no success ; it creates a sample.mp4.bak file, even though the software says that the operation was successful

  



  


  can't insert any more links, because I got too little rep apparently

  


  



  

  4. Mp3tag (mp3tag.de/en)

  


  



  with no success ; I can successfully add the image as a cover, but it's not visible in the File Explorer. After saving it says Saved tag in 1 of 1 files. The picture can been seen in Mp3tag though as if it's applied.

  



  

  5. Tag&Rename (softpointer.com)

  


  



  with no success ; same story as with Mp3tag, I can apply and it's visible in the app, but not in File Explorer.

  



  

  6. mp4v2 (https:// code.google.com/archive/p/mp4v2/downloads)

  


  



  mp4tags.exe -picture sample.png sample.mp4

  



  with no success ; it executes with no (error)message

  



  I found it on this thread

  



  (https:// forum.videohelp.com/threads/388025-How-to-set-the-thumbnail-of-a-video-clip).

  



  Have used the updated version which is the Windows build

  



  (https:// forum.doom9.org/attachment.php ?attachmentid=14314&d=1407985798).

  



  

  7. iTunes

  


  



  iTunes 12.10.6.2 (Microsoft Store version)

  



  iTunes doesn't load the file into the library. I can run it fine though in QuickTime Player.

  



  I didn't notice the 'Home Videos' tab. Now I can see it in there. I tried to apply the image, and it applied it only inside iTunes... So it's still not working for me.

  



  I tried to convert the MP4 to M4A and then adding it to iTunes. I was finally successfully able to change the cover and it was visible in File Explorer. But that's not what I want.

  



  I know for a fact that it's possible, because I've seen pictures on the Internet and YT and different threads where an answer is upvoted.

  



  

  8. tag

  


  



  https:// github.com/b4winckler/tag

  



  The description is : Simple command line audio tag editor... After having to install millions of libraries and a package manager I was stuck with the compilation of the software, followed all the steps, didn't work.

  



  I've also tried using MP4art as suggested by a comment on another issue about this. He didn't include a link to it so I went looking and found another issue about this. One comment recommended MP4art as well, with a link that's dead.

  



  ---

  



  I believe I've used some other methods as well but can't think of them right now. Will update if I recall.

  



  Any help is greatly appreciated.