Recherche avancée

Sur d’autres sites (11332)

• Measuring success for your SEO content

  20 mars 2020, par Jake Thornton — Uncategorized
  With over a billion searches every day in search engines, it’s hard to underestimate the importance of having your business present on page one (ideally in positions 1 – 3) ranking for the keywords that impact your sales and conversions.

  "In 2019, Google received nearly 2.3 trillion searches and on page one alone, the first five organic results accounted for 67.60% of all the clicks."

  So how is your business performing when it comes to ranking in the crucial top three spots of search for your most important keywords ?

  Accurately measuring the success of your content

  Once you’ve done your keyword research, created compelling content, optimised it to be search-friendly, and hit ‘publish’, you then need to accurately measure the success of your efforts.

  4 tips for measuring the success of your SEO content

  1. Create a custom segment for "Visitors from Search Engines only"

  By creating this custom segment, you’ll be able to analyse the behavioural patterns of the visitors who found your website through a search engine. 

  This way you can use many of Matomo’s powerful features (Visitors, Behaviour, Acquisition, Ecommerce, Goals etc.) focused entirely on search engine visitors only.

  

  Once you’ve created this segment, you can begin to see key metrics like which entry pages are responsible for referring visitors to your website. For example : Visit Behaviour – Entry Pages, this is a great way to analyse your most effective SEO pages.You may be surprised at what pages currently bring in the most traffic.

  As well as discovering which content resonates with your search audience, you will also be able to create more content focused on your targeted audience. Do this by learning which locations your search visitors are from, which device they use, what time of the day they visited your website and much more.

  >> Learn more about creating custom segments

  2. Website visits, time on site, pages per session, and bounce rate.

  “The top four ranking factors are website visits, time on site, pages per session, and bounce rate.”

  These four metrics set the benchmark for your SEO success.

  First, you need to get as many of the ‘right’ users to see your content. If you feel you’ve exhausted channels such as social media, email and possibly paid posts ; think about who your ideal audience is. Where are they likely to hang out online ? Are there community groups or forum sites that are interested in what you’re writing about ? 

  Whatever the case, putting yourself out there and getting more traffic to your website will help show search engines that people are interested in your website. As a result, they’ll likely rank you higher for that.

  When we say getting more of the ‘right’ users, we mean users who are generally interested in the topic/subject you’re writing about and interested in the work you do. 

  

  This is important for the next three metrics – increasing users time on your website, increasing the amount of pages your users explore on your website, and reducing the overall bounce rate for users who leave your website in a matter of seconds.

  To evaluate these metrics, go to Behaviour – Pages in your Matomo and see how these metrics vary on previous posts or pages you’ve created. Which pages are already showing you the best results ? Why do they get the results ? Can you focus on creating more content like this ?

  Understanding what content is resonating with your users through these metrics is easy and is the starting point for measuring the success of your SEO content strategy.

  >> Learn more about the Behaviour feature

  3. Row Evolution

  The Row Evolution feature embedded within the Search Engine Keywords Performance plugin lets you see how your ranking positions have changed over time for your important keywords. It also lets you see how the incoming traffic, related to your keywords, has changed over time.

  

  This is valuable when measuring the changes you’ve made to your landing pages to see if it has a positive or negative effect on your ranking efforts. 

  This also lets you see how search engine algorithm changes affect your search rankings over time, and to see if the effects of these algorithm updates are temporary or long lasting.

  Row evolution allows you to report on keyword performance with ease. If you only check your insights once a week or once a fortnight, you’ll see how ranking positions for your important keywords have changed daily (or even weekly, monthly or yearly however you prefer.)

  >> Learn more about Row Evolution

  4. What results are you getting from the lesser known search engines ?

  "In 2019 (to date), Google accounted for just over 75% of all global desktop search traffic, followed by Bing at 9.97%, Baidu at 9.34%, and Yahoo at 2.77%."

  For most of us, we want to be ranking in the top three spots in Google Search because that’s where the majority of search users are. However, don’t shy away from opportunities you could be missing with lesser known search engines.

  If you sell a product aimed at 55-65 year olds who use a PC computer, chances are they are using Bing. If you have customers in China the majority will be using Baidu, or in our case at Matomo, many of our loyal users use a privacy-friendly search engine like DuckDuckGo or Qwant.

  Some of your ideal customers might be finding you through these alternative search engines, so be sure to measure the impact that these referrals may have on your conversions.

  

  Strategically including important keywords that impact your business

  While search is an important acquisition channel for most businesses, it’s also one of the most competitive.

  We recommend analysing your keyword and content performance regularly and alter content that isn’t performing as well as you’d like. You need to continually learn from the content that is successful, and focus on creating more content like this. 

  The final thing to remember with search keyword performance is to be patient. If you have had little success in the past with attracting customers through search, it can take time to build this reputation with search engines.

• 4 ways to create more effective funnels

  24 février 2020, par Jake Thornton — Uncategorized

  Accurately measuring the success of your customer’s journey on your website is vital to increasing conversions and having the best outcome for your business. When it comes to website analytics, the Funnels feature is the best place to start measuring each touch point in the customer journey. From here you’ll find out where you lose your visitors so you can make changes to your website and convert more in the future.

  The funnels feature lets you measure the steps (actions, events and pages) your users go through to reach the desired outcomes you want them to achieve. This gives you valuable insights into the desired journey for your customers. 

  When creating a funnel with the funnels feature, you anticipate the customer journey that you want to measure, for example : 

  Step 1 – Visitor lands on your homepage and sees the promotion you’re offering. 
  Step 2 – They click the call-to-action (CTA) button which leads them to information on the product
  Step 3 – They add the product to their cart
  Step 4 – They fill in their personal information and credit card details
  Step 5 – They click the “pay now” button

  From here you can see exactly how many visitors you lose between each step. Then you can implement new techniques to decrease these drop-offs and evaluate the success of your changes over time.

  But what about the non-conventional routes to conversion ?

  That’s right, visitors can end up in all different directions on your website. It’s important to use other features in Matomo to discover these popular pathways your visitors may be taking before the point of conversion.

  Here are 4 Matomo features for discovering important alternative funnels on your website :

  Transitions

  The transitions feature lets you visualise mini funnels on selected pages. You can see how visitors landed on a specific page, and then where they moved on to from this specific page.

  First you need to identify the page(s) that sells your product or service the most. 

  Whether it’s your homepage, a product page or an information page on your services. The transitions feature will then show you the before and after pathways visitors are already taking to get from page to page

  

  The transitions feature is located under Behaviour – Pages. Find the important page you would like to analyse and click on the Transitions icon.

  

  In the example above, you’ll see 18% of visitors who entered from internal pages came from the homepage, which you may have already suspected as the first step in your conversion funnel.

  However, the exact same % of visitors are also entering through a blog post article called /best-of-the-best/

  In this case, it highlights the importance of creating funnels with popular blog posts as the first step in the funnel. Your visitors may have found this post through social media, a search engine etc. Whatever the case, your blog posts could be your biggest influencer for conversions on your website.

  >> Learn more about Transitions

  Page Overlay

  The overlay feature lets you see exactly where visitors are clicking on your landing pages which moves them either in the right or wrong direction in the conversion funnel. 

  If you see a high percentage of clicks to a page that’s off the beaten track from your desired conversion funnel, use the Funnels feature to follow this pathway and analyse how they get back to the pathway you initially intended them to take.

  

  The best thing about the page overlay feature is the visualisation showing the results on the landing page itself. This gives you an idea of where they may be getting distracted by the wrong content.

  You can locate the page overlay feature beside the transitions feature, shown in the screenshot below.

  

  The page overlay feature also gives you a summary of the pageviews, clicks, bounce rates, exit rates and average time spent on page, so you can measure the overall success of each page in the display menu.

  >> Learn more about Page Overlay

  Users Flow

  If you’re looking to see many of the most popular pathways your visitors are taking all at once, then Users Flow is a powerful feature which shows this visualisation.

  Note : For Matomo On-Premise users, Users Flow is a premium feature. More information here.

  

  The thicker the blue line between interactions means the more popular the pathway is. 

  Here you can see how visitors are navigating their way through your website before converting, this presenting clear steps in the conversion funnel that require monitoring and improving on to ensure your efforts are going into the right areas on your website.

  >> Learn more about Users Flow

  Row Evolution

  Another important feature to use which is integrated within the funnels feature, is row evolution which shows you important changes in your user’s behaviour over time.

  

  Having row evolution integrated within the funnels feature gives you a big advantage as it lets you measure the specific metrics and landing pages within your conversion funnel.

  

  You’ll be able to see the increases and decreases in entries and exits to your landing page, as well as increases and decreases in the number of visitors who proceed to the next step in the funnel, and the conversion rate %.

  You’ll also be able to add annotations so you can note all the changes you make to your landing pages over time and quickly identify how these changes impacted your conversion funnels.

  >>Learn more about Row Evolution

  Continually create more and more funnels !

  Measuring the success of the desired pathway you want your customers to take is crucial to ensure you are presenting the best possible user experience for your visitors.

  However, creating funnels for the less desired pathways is equally important. This way you’ll discover popular journeys your visitors are taking within your website you weren’t previously aware of, and can monitor them to make sure they still work in the future. You’ll be able to fix pain points easier and find faster ways to get visitors back on the right track to converting.

• Trying to cancel execution and delete file using ffmpeg C API

  6 mars 2020, par Vuwox

  The code below is a class that handle the conversion of multiples images, through add_frame() method, into a GIF with encode(). It also use a filter to generate and apply the palette. The usage is like this :

  Code call example

  std::unique_ptr gif_obj = nullptr;
  
  try
  
  {
  
      gif_obj = std::make_unique({1000,1000}, 12, "C:/out.gif",
  
                "format=pix_fmts=rgb24,split [a][b];[a]palettegen[p];[b][p]paletteuse");
  
  
  
      // Example: a simple vector of images (usually process internally)
  
      for(auto img : image_vector)
  
           gif_obj->add_frame(img);
  
  
  
      // Once all frame were added, encode the final GIF with the filter applied.
  
      gif_obj->encode();
  
  }
  
  catch(const std::exception& e)
  
  {
  
      // An error occured! We must close FFMPEG properly and delete the created file.
  
      gif_obj->cancel();
  
  }

  I have the following issue. If the code for any reason throw an exception, I call ffmpeg->cancel() and it supposes to delete the GIF file on disk. But this is never working, I assume there is a lock on the file or something like that. So here are my question :

  What is the proper way to close/free ffmpeg object in order to remove the file afterward ?

  ---

  Full class code below

  Header

  // C++ Standard includes    
  
  #include <memory>
  
  #include <string>
  
  #include <vector>
  
  
  
  
  
  // 3rd Party incldues
  
  #ifdef __cplusplus
  
  extern "C" {
  
  #include "libavformat/avformat.h"
  
  #include "libavfilter/avfilter.h"
  
  #include "libavutil/opt.h"
  
  #include "libavfilter/buffersrc.h"
  
  #include "libavfilter/buffersink.h"
  
  #include "libswscale/swscale.h"
  
  #include "libavutil/imgutils.h"
  
  }
  
  #endif
  
  
  
  #define FFMPEG_MSG_LEN 2000
  
  
  
  namespace px
  
  {
  
      namespace GIF
  
      {
  
          class FFMPEG
  
          {
  
          public:
  
              FFMPEG(const px::Point2D<int>&amp; dim,
  
                     const int framerate,
  
                     const std::string&amp; filename,
  
                     const std::string&amp; filter_cmd);
  
  
  
              ~FFMPEG();
  
  
  
              void add_frame(pxImage * const img);
  
              void encode();
  
              void cancel();
  
  
  
          private:
  
  
  
              void init_filters();            // Init everything that needed to filter the input frame.
  
              void init_muxer();              // The muxer that creates the output file.
  
              void muxing_one_frame(AVFrame* frame);
  
              void release();
  
  
  
              int _ret = 0;                   // status code from FFMPEG.
  
              char _err_msg[FFMPEG_MSG_LEN];  // Error message buffer.
  
  
  
  
  
              int m_width = 0;                // The width that all futur images must have to be accepted.
  
              int m_height = 0;               // The height that all futur images must have to be accepted.
  
  
  
              int m_framerate = 0;            // GIF Framerate.
  
              std::string m_filename = "";    // The GIF filename (on cache?)
  
              std::string m_filter_desc = ""; // The FFMPEG filter to apply over the frames.
  
  
  
              bool as_frame = false;
  
  
  
              AVFrame* picture_rgb24 = nullptr;           // Temporary frame that will hold the pxImage in an RGB24 format (NOTE: TOP-LEFT origin)
  
  
  
              AVFormatContext* ofmt_ctx = nullptr;        // ouput format context associated to the 
  
              AVCodecContext* o_codec_ctx = nullptr;      // output codec for the GIF
  
  
  
              AVFilterGraph* filter_graph = nullptr;      // filter graph associate with the string we want to execute
  
              AVFilterContext* buffersrc_ctx = nullptr;   // The buffer that will store all the frames in one place for the palette generation.
  
              AVFilterContext* buffersink_ctx = nullptr;  // The buffer that will store the result afterward (once the palette are used).
  
  
  
              int64_t m_pts_increment = 0;
  
          };
  
      };
  
  };
  
  </int></vector></string></memory>

  ctor

  px::GIF::FFMPEG::FFMPEG(const px::Point2D<int>&amp; dim,
  
                          const int framerate,
  
                          const std::string&amp; filename,
  
                          const std::string&amp; filter_cmd) :
  
      m_width(dim.x()),
  
      m_height(dim.y()),
  
      m_framerate(framerate),
  
      m_filename(filename),
  
      m_filter_desc(filter_cmd)
  
  {
  
  #if !_DEBUG
  
      av_log_set_level(AV_LOG_QUIET); // Set the FFMPEG log to quiet to avoid too much logs.
  
  #endif
  
  
  
      // Allocate the temporary buffer that hold the ffmpeg image (pxImage to AVFrame conversion).
  
      picture_rgb24 = av_frame_alloc();
  
      picture_rgb24->pts = 0;
  
      picture_rgb24->data[0] = NULL;
  
      picture_rgb24->linesize[0] = -1;
  
      picture_rgb24->format = AV_PIX_FMT_RGB24;
  
      picture_rgb24->height = m_height;
  
      picture_rgb24->width = m_width; 
  
  
  
      if ((_ret = av_image_alloc(picture_rgb24->data, picture_rgb24->linesize, m_width, m_height, (AVPixelFormat)picture_rgb24->format, 24)) &lt; 0)
  
          throw px::GIF::Error("Failed to allocate the AVFrame for pxImage conversion with error: " +
  
                               std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)),
  
                               "GIF::FFMPEG CTOR");   
  
  
  
      //printf("allocated picture of size %d, linesize %d %d %d %d\n", _ret, picture_rgb24->linesize[0], picture_rgb24->linesize[1], picture_rgb24->linesize[2], picture_rgb24->linesize[3]);
  
  
  
      init_muxer();   // Prepare the GIF encoder (open it on disk).
  
      init_filters(); // Prepare the filter that will be applied over the frame.
  
  
  
      // Instead of hardcoder {1,100} which is the GIF tbn, we collect it from its stream.
  
      // This will avoid future problem if the codec change in ffmpeg.
  
      if (ofmt_ctx &amp;&amp; ofmt_ctx->nb_streams > 0)
  
          m_pts_increment = av_rescale_q(1, { 1, m_framerate }, ofmt_ctx->streams[0]->time_base);
  
      else
  
          m_pts_increment = av_rescale_q(1, { 1, m_framerate }, { 1, 100 });
  
  }
  
  </int>

  FFMPEG Initialization (Filter and muxer)

  void px::GIF::FFMPEG::init_filters()
  
  {
  
      const AVFilter* buffersrc = avfilter_get_by_name("buffer");
  
      const AVFilter* buffersink = avfilter_get_by_name("buffersink");
  
  
  
      AVRational time_base = { 1, m_framerate };
  
      AVRational aspect_pixel = { 1, 1 };
  
  
  
      AVFilterInOut* inputs = avfilter_inout_alloc();
  
      AVFilterInOut* outputs = avfilter_inout_alloc();
  
  
  
      filter_graph = avfilter_graph_alloc();
  
  
  
      try
  
      {
  
          if (!outputs || !inputs || !filter_graph)
  
              throw px::GIF::Error("Failed to 'init_filters' could not allocated the graph/filters.", "GIF::FFMPEG init_filters");
  
  
  
          char args[512];
  
          snprintf(args, sizeof(args),
  
                   "video_size=%dx%d:pix_fmt=%d:time_base=%d/%d:pixel_aspect=%d/%d",
  
                   m_width, m_height,
  
                   picture_rgb24->format,
  
                   time_base.num, time_base.den,
  
                   aspect_pixel.num, aspect_pixel.den);
  
  
  
          if (avfilter_graph_create_filter(&buffersrc_ctx, buffersrc, "in", args, nullptr, filter_graph) < 0)
  
              throw px::GIF::Error("Failed to create the 'source buffer' in init_filer method.", "GIF::FFMPEG init_filters");
  
  
  
  
  
          if (avfilter_graph_create_filter(&buffersink_ctx, buffersink, "out", nullptr, nullptr, filter_graph) < 0)
  
              throw px::GIF::Error("Failed to create the 'sink buffer' in init_filer method.", "GIF::FFMPEG init_filters");
  
  
  
          // GIF has possible output of PAL8.
  
          enum AVPixelFormat pix_fmts[] = { AV_PIX_FMT_PAL8, AV_PIX_FMT_NONE };
  
  
  
          if (av_opt_set_int_list(buffersink_ctx, "pix_fmts", pix_fmts, AV_PIX_FMT_NONE, AV_OPT_SEARCH_CHILDREN) < 0)
  
              throw px::GIF::Error("Failed to set the output pixel format.", "GIF::FFMPEG init_filters");
  
  
  
          outputs->name = av_strdup("in");
  
          outputs->filter_ctx = buffersrc_ctx;
  
          outputs->pad_idx = 0;
  
          outputs->next = nullptr;
  
  
  
          inputs->name = av_strdup("out");
  
          inputs->filter_ctx = buffersink_ctx;
  
          inputs->pad_idx = 0;
  
          inputs->next = nullptr;
  
  
  
          // GIF has possible output of PAL8. 
  
          if (avfilter_graph_parse_ptr(filter_graph, m_filter_desc.c_str(), &inputs, &outputs, nullptr) < 0)
  
              throw px::GIF::Error("Failed to parse the filter graph (bad string!).", "GIF::FFMPEG init_filters");
  
  
  
          if (avfilter_graph_config(filter_graph, nullptr) < 0)
  
              throw px::GIF::Error("Failed to configure the filter graph (bad string!).", "GIF::FFMPEG init_filters");
  
  
  
          avfilter_inout_free(&inputs);
  
          avfilter_inout_free(&outputs);
  
      }
  
      catch (const std::exception& e)
  
      {
  
          // Catch exception to delete element.
  
          avfilter_inout_free(&inputs);
  
          avfilter_inout_free(&outputs);
  
          throw e; // re-throuw
  
      }
  
  }
  
  
  
  
  
  void px::GIF::FFMPEG::init_muxer()
  
  {
  
      AVOutputFormat* o_fmt = av_guess_format("gif", m_filename.c_str(), "video/gif");
  
  
  
      if ((_ret = avformat_alloc_output_context2(&ofmt_ctx, o_fmt, "gif", m_filename.c_str())) < 0)
  
          throw px::GIF::Error(std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)) + " allocate output format.", "GIF::FFMPEG init_muxer");
  
  
  
      AVCodec* codec = avcodec_find_encoder(AV_CODEC_ID_GIF);
  
      if (!codec) throw px::GIF::Error("Could to find the 'GIF' codec.", "GIF::FFMPEG init_muxer");
  
  
  
  #if 0
  
      const AVPixelFormat* p = codec->pix_fmts;
  
      while (p != NULL && *p != AV_PIX_FMT_NONE) {
  
          printf("supported pix fmt: %s\n", av_get_pix_fmt_name(*p));
  
          ++p;
  
      }
  
  #endif
  
  
  
      AVStream* stream = avformat_new_stream(ofmt_ctx, codec);
  
  
  
      AVCodecParameters* codec_paramters = stream->codecpar;
  
      codec_paramters->codec_tag = 0;
  
      codec_paramters->codec_id = codec->id;
  
      codec_paramters->codec_type = AVMEDIA_TYPE_VIDEO;
  
      codec_paramters->width = m_width;
  
      codec_paramters->height = m_height;
  
      codec_paramters->format = AV_PIX_FMT_PAL8;
  
  
  
      o_codec_ctx = avcodec_alloc_context3(codec);
  
      avcodec_parameters_to_context(o_codec_ctx, codec_paramters);
  
  
  
      o_codec_ctx->time_base = { 1, m_framerate };
  
  
  
      if (ofmt_ctx->oformat->flags & AVFMT_GLOBALHEADER)
  
          o_codec_ctx->flags |= AV_CODEC_FLAG_GLOBAL_HEADER;
  
  
  
      if ((_ret = avcodec_open2(o_codec_ctx, codec, NULL)) < 0)
  
          throw px::GIF::Error(std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)) + " open output codec.", "GIF::FFMPEG init_muxer");
  
  
  
      if ((_ret = avio_open(&ofmt_ctx->pb, m_filename.c_str(), AVIO_FLAG_WRITE)) < 0)
  
          throw px::GIF::Error(std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)) + " avio open error.", "GIF::FFMPEG init_muxer");
  
  
  
      if ((_ret = avformat_write_header(ofmt_ctx, NULL)) < 0)
  
          throw px::GIF::Error(std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)) + " write GIF header", "GIF::FFMPEG init_muxer");
  
  
  
  #if _DEBUG
  
      // This print the stream/output format.
  
      av_dump_format(ofmt_ctx, -1, m_filename.c_str(), 1);
  
  #endif
  
  }

  Add frame (usually in a loop)

  void px::GIF::FFMPEG::add_frame(pxImage * const img)
  
  {
  
      if (img->getImageType() != PXT_BYTE || img->getNChannels() != 4)
  
          throw px::GIF::Error("Failed to 'add_frame' since image is not PXT_BYTE and 4-channels.", "GIF::FFMPEG add_frame");
  
  
  
      if (img->getWidth() != m_width || img->getHeight() != m_height)
  
          throw px::GIF::Error("Failed to 'add_frame' since the size is not same to other inputs.", "GIF::FFMPEG add_frame");
  
  
  
      const int pitch = picture_rgb24->linesize[0];
  
      auto px_ptr = getImageAccessor(img);
  
  
  
      for (int y = 0; y < m_height; y++)
  
      {
  
          const int px_row = img->getOrigin() == ORIGIN_BOT_LEFT ? m_height - y - 1 : y;
  
          for (int x = 0; x < m_width; x++)
  
          {
  
              const int idx = y * pitch + 3 * x;
  
              picture_rgb24->data[0][idx] = px_ptr[px_row][x].ch[PX_RE];
  
              picture_rgb24->data[0][idx + 1] = px_ptr[px_row][x].ch[PX_GR];
  
              picture_rgb24->data[0][idx + 2] = px_ptr[px_row][x].ch[PX_BL];
  
          }
  
      }
  
  
  
      // palettegen need a whole stream, just add frame to buffer.
  
      if ((_ret = av_buffersrc_add_frame_flags(buffersrc_ctx, picture_rgb24, AV_BUFFERSRC_FLAG_KEEP_REF)) < 0)
  
          throw px::GIF::Error("Failed to 'add_frame' to global buffer with error: " +
  
                               std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)),
  
                               "GIF::FFMPEG add_frame");
  
  
  
      // Increment the FPS of the picture for the next add-up to the buffer.      
  
      picture_rgb24->pts += m_pts_increment;
  
  
  
      as_frame = true;
  
  }    

  Encoder (final step)

  void px::GIF::FFMPEG::encode()
  
  {
  
      if (!as_frame)
  
          throw px::GIF::Error("Please 'add_frame' before running the Encoding().", "GIF::FFMPEG encode");
  
  
  
      // end of buffer
  
      if ((_ret = av_buffersrc_add_frame_flags(buffersrc_ctx, nullptr, AV_BUFFERSRC_FLAG_KEEP_REF)) < 0)
  
          throw px::GIF::Error("error add frame to buffer source: " + std::string(av_make_error_string(_err_msg, FFMPEG_MSG_LEN, _ret)), "GIF::FFMPEG encode");
  
  
  
      do {
  
          AVFrame* filter_frame = av_frame_alloc();
  
          _ret = av_buffersink_get_frame(buffersink_ctx, filter_frame);
  
          if (_ret == AVERROR(EAGAIN) || _ret == AVERROR_EOF) {
  
              av_frame_unref(filter_frame);
  
              break;
  
          }
  
  
  
          // write the filter frame to output file
  
          muxing_one_frame(filter_frame);
  
  
  
          av_frame_unref(filter_frame);
  
      } while (_ret >= 0);
  
  
  
      av_write_trailer(ofmt_ctx);
  
  }
  
  
  
  void px::GIF::FFMPEG::muxing_one_frame(AVFrame* frame)
  
  {
  
      int ret = avcodec_send_frame(o_codec_ctx, frame);
  
      AVPacket *pkt = av_packet_alloc();
  
      av_init_packet(pkt);
  
  
  
      while (ret >= 0) {
  
          ret = avcodec_receive_packet(o_codec_ctx, pkt);
  
          if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF) {
  
              break;
  
          }
  
  
  
          av_write_frame(ofmt_ctx, pkt);
  
      }
  
      av_packet_unref(pkt);
  
  }

  DTOR, Release and Cancel

  px::GIF::FFMPEG::~FFMPEG()
  
  {
  
      release();
  
  }
  
  
  
  
  
  void px::GIF::FFMPEG::release()
  
  {
  
      // Muxer stuffs
  
      if (ofmt_ctx != nullptr) avformat_free_context(ofmt_ctx);
  
      if (o_codec_ctx != nullptr) avcodec_close(o_codec_ctx);
  
      if (o_codec_ctx != nullptr) avcodec_free_context(&o_codec_ctx);
  
  
  
      ofmt_ctx = nullptr;
  
      o_codec_ctx = nullptr;
  
  
  
      // Filter stuffs
  
      if (buffersrc_ctx != nullptr) avfilter_free(buffersrc_ctx);
  
      if (buffersink_ctx != nullptr) avfilter_free(buffersink_ctx);
  
      if (filter_graph != nullptr) avfilter_graph_free(&filter_graph);
  
  
  
      buffersrc_ctx = nullptr;
  
      buffersink_ctx = nullptr;
  
      filter_graph = nullptr;
  
  
  
      // Conversion image.
  
      if (picture_rgb24 != nullptr) av_frame_free(&picture_rgb24);
  
      picture_rgb24 = nullptr;
  
  }
  
  
  
  void px::GIF::FFMPEG::cancel()
  
  {
  
      // In-case of failure we must close ffmpeg and exit.
  
      av_write_trailer(ofmt_ctx);
  
  
  
      // Release and close all elements.
  
      release();
  
  
  
      // Delete the file on disk.
  
      if (remove(m_filename.c_str()) != 0)
  
          PX_LOG0(PX_LOGLEVEL_ERROR, "GIF::FFMPEG - On 'cancel' failed to remove the file.");
  
  }