Recherche avancée

Sur d’autres sites (12264)

• Building FFMPEG for Visual Studio development

  28 juillet 2016, par gboy

  I’m trying to use ffmpeg in Visual Studio 2013 C++ software (ultimately as part of an OpenCV project) - but right now I’m just trying to get basic FFMPEG functionality. In general, when building in Visual Studio, I build 64—bit software with Multi-threaded DLL runtime libraries. I have built ffmpeg using the general instructions for ’Native Windows compilation using ... MinGW-w64’ at http://ffmpeg.org/platform.html#Windows (I provide a more detailed set of steps I followed below...).

  After building the ffmpeg software on my system, I tried to create a simple ’hello world’ project in Visual Studio 2013. Specifically, I tried to implement the initial tutorial file presented at http://dranger.com/ffmpeg/tutorial01.html. Upon building the project, I get the error :

  c :\msys64\usr\local\ffmpeg\libavutil\common.h(45) : fatal error C1083 : Cannot
  open include file : ’libavutil/avconfig.h’ : No such file or directory

  The following are the detailed steps I took to build ffmpeg and create my basic Visual Studio project :

  ============ Building ffmpeg ===============

  1. Downloaded and intalled msys2-x86_64-20160205.exe from http://msys2.github.io
  2. Ran update-core to update the Msys2 install
  3. Ran pacman -Suu (twice) to complete the update (following the instructions about updating shortcuts, etc.)
  4. Then I quit out of the MSys2 shell and opened the MinGW-w64 Win64 Shell. In this new shell :
  5. Installed the following packages using pacman -S The list of packages I installed is : make, pkg-config, diffutils, mingw-w64-x86_64-yasm, mingw-w64-x86_64-gcc, mingw-w64-x86_64-SDL, git
  6. Then I cd’d into cd /usr/local
  7. Ran git clone https://git.ffmpeg.org/ffmpeg.git ffmpeg
  8. I wanted to build the ffmpeg library ’out-of-tree’ of this MSys64 folder. So, in the regular file system of my Windows machine I created a folder at C :\ffmpeg
  9. Back in the Win64 Shell, I cd’d to this new folder : cd /c/ffmpeg
  10. Then ran /usr/loca/ffmpeg/configure --enable-shared
  11. Then make -r
  12. And, finally make install

  Now, if I had to guess, my ’flaw’ was in the options I used when calling the ’configure’ script of ffmpeg. Do I need to use particular options so that I can take the ffmpeg libraries built here and use them as dynamic (DLL) libraries in Visual Studio ?

  ========== Configuring my Visual Studio Project ============

  Here’s how I created a simple hello world project in Visual Studio to see if ffmpeg is working.

  1. I created a new Visual C++ ’Empty Project’ in Visual Studio 2013

  2. I then configured the project properties as follows :

     a. In C/C++ => General => Additional Include Directories, I put

     C :\msys64\usr\local\ffmpeg

     b. In Linker=>General => Additional Library Directories, I pointed to each of the built library folders (basically I pointed at all of the libraries that were built to ensure I was not inadvertently missing the critical one). The list is as follows :

     • C :\ffmpeg\libavcodec
     • C :\ffmpeg\libavdevice
     • C :\ffmpeg\libavfilter
     • C :\ffmpeg\libavformat
     • C :\ffmpeg\libavutil
     • C :\ffmpeg\libswresample
     • C :\ffmpeg\libswscale
     • C :\ffmpeg

     c. In Linker=> Input => Additional Dependencies, I pointed to the particular libraries (again - I pointed to all of the ones present). The list is :

     • avcodec.lib
     • avdevice.lib
     • avfilter.lib
     • avformat.lib
     • avutil.lib
     • swresample.lib
     • swscale.lib

  3. I then created a new source file called ’tut01.c’ and copied/pasted the code from http://dranger.com/ffmpeg/tutorial01.c

  4. Then hit F7 and got the error specified above about not finding avconfig.h

  The above is my best guess as to the steps I need to follow to get this working in Windows (btw, it’s Windows 10, 64-bit) & Microsoft Visual Studio 2013. What should I change to get this basic program to build and run ?

• Encoded images into H264 video are skipped and/or missing ?

  25 juillet 2013, par Jona

  I'm trying to encode images into an H264 MP4 video. The issues I'm having is that some of the images are skipped or at the end of the video simply missing. I need the video to play every single image I encode since it is an animation.

  Any help setting the encoder properly would be greatly appreciated !

  Encoder settings :

  AVCodecContext *c;
  
  ...
  
  c->codec_id = AV_CODEC_ID_H264;
  
  c->bit_rate = mOutputWidth*mOutputHeight*4;//400000;
  
  /* Resolution must be a multiple of two. */
  
  c->width    = mOutputWidth;
  
  c->height   = mOutputHeight;
  
      /* timebase: This is the fundamental unit of time (in seconds) in terms
  
       * of which frame timestamps are represented. For fixed-fps content,
  
       * timebase should be 1/framerate and timestamp increments should be
  
       * identical to 1. */
  
  c->time_base.den = mFps;
  
  c->time_base.num = 1;
  
  c->gop_size      = 12; /* emit one intra frame every twelve frames at most */
  
  c->pix_fmt       = AV_PIX_FMT_YUV420P;
  
  ...
  
  av_dict_set(&pOptions, "preset", "medium", 0);
  
  av_dict_set(&pOptions, "tune", "animation", 0);
  
  
  
  /* open the codec */
  
  ret = avcodec_open2(c, codec, &pOptions);
  
  if (ret < 0) {
  
      LOGE("Could not open video codec: %s", av_err2str(ret));
  
      return -1;
  
  }

  Update 07/24/13 :
  I was able to achieve a better video by setting the gop_size=FPS and writing the last video frame repeatedly FPS+1 times seemed to resolve all issues. To me it seems odd to do that but might be something standard in the video encoding world ? Any tips feedback about this ?

• Parsing avconv/ffmpeg rawvideo output ?

  23 avril 2013, par DigitalMan

  I'm about to begin a project that will involve working with the output of avconv/ffmpeg, pixel-by-pixel, in rgb32 format. I intend to work with a raw byte stream, such as from the pipe protocol. Basic pointer arithmetic (C/C++) will be used to iterate over these pixels, and modify them in arbitrary manners in real-time.

  I've created a very small file using rawvideo format and codec, and opened it up in a hex editor. As expected, it's just a series of pixels, read right to left, top to bottom. No distinguishing between lines - no problem, if you know how wide the video is beforehand. No distinguishing between frames - no problem, if you also know how tall the video is. No file header for frame rate, or even what the encoding (rgb32, rgb24, yuv, etc.) is - again, as long as you already know, it can be worked with.

  The problem occurs when - for one reason or another - some bytes are missing. Maybe the stream isn't being examined from the beginning, which is likely be the case in my project, or maybe something just got lost. All the pre-existing knowledge in the world (besides maybe a byte count of what's been missed, not gonna happen) won't prevent it from happily chugging along, with an incorrect offset of line and frame.

  So, what I'm looking for is an option for rawvideo, or possibly some other format/codec, that will allow me to work with the resulting stream at the pixel level, in RGB, yet still have a clear definition of where a new frame begins, even if it happens to start "looking" in the middle of a frame. (Width, height, and framerate will indeed be known.)