Recherche avancée

Sur d’autres sites (9090)

• reading video mpegts stream from stdin using libavformat libavcodec

  1er février 2012, par D Starkweather

  I'm trying to read video frames from stdin using ffmpeg's libav* lib collection. Simply passing "pipe:0" or "pipe :" as the filename to avformat_open_input_file() in my program does not do the trick. (e.g. cat /dev/video0 | ./myprogram OR ./myprogram < /dev/video0 ; it also fails using file.avi in place of /dev/video0).

  Peeking into ffmpeg.c, I find the following bit of code using tcsetattr() to set the termios :

  struct termios tty;
  
  
  
  if (tcgetattr (0, &tty) == 0) {
  
  oldtty = tty;
  
  restore_tty = 1;
  
  atexit(term_exit);
  
  
  
  tty.c_iflag &= ~(IGNBRK|BRKINT|PARMRK|ISTRIP
  
                        |INLCR|IGNCR|ICRNL|IXON);
  
  tty.c_oflag |= OPOST;
  
  tty.c_lflag &= ~(ECHO|ECHONL|ICANON|IEXTEN);
  
  tty.c_cflag &= ~(CSIZE|PARENB);
  
  tty.c_cflag |= CS8;
  
  tty.c_cc[VMIN] = 1;
  
  tty.c_cc[VTIME] = 0;
  
  
  
  tcsetattr (0, TCSANOW, &tty);
  
  }
  
  signal(SIGQUIT, sigterm_handler); /* Quit (POSIX).  */
  
  }
  
  
  
  avformat_network_deinit();
  
  
  
  signal(SIGINT , sigterm_handler); /* Interrupt (ANSI).    */
  
  signal(SIGTERM, sigterm_handler); /* Termination (ANSI).  */
  
  signal(SIGXCPU, sigterm_handler);

  However, when I try this, it appears to have no effect. What settings do I need in the termios struct to continuously read video frames in ?

  Currently, all i am able to do is call av_read_frames a few times (return code 0) before it usually either segfaults or hangs waiting. The buffer usually contains just a blank screen (all black). With any luck, the video buffer will contain the first video frame it comes across, but does not update to the later frames.

  (I'm running this on Debian 6.0) Here's my version of ffmpeg :

  ffmpeg version N-36936-g4cf81d9 Copyright (c) 2000-2012 the FFmpeg
  developers built on Jan 19 2012 20:51:47 with gcc 4.4.5
  configuration : —enable-shared —enable-gray —enable-hardcoded-tables
  —enable-runtime-cpudetect —enable-libmp3lame —enable-libopenjpeg —enable-librtmp —enable-libtheora —enable-libvorbis —enable-libx264 —enable-libxvid —enable-zlib —enable-gpl libavutil 51. 34.101 / 51. 34.101 libavcodec 53. 57.100 /
  53. 57.100 libavformat 53. 30.100 / 53. 30.100 libavdevice 53. 4.100 / 53. 4.100 libavfilter 2. 59.101 / 2. 59.101 libswscale 2. 1.100 / 2. 1.100 libswresample 0. 6.100 /
  0. 6.100 libpostproc 52. 0.100 / 52. 0.100 Hyper fast Audio and Video encoder usage : ffmpeg [options] [[infile options] -i
  infile].

  Thank you very much for any help helpful tips. It is much appreciated.

  D. Grant Starkweather

• How to hack ffmpeg to consider I-Frames as key frames ?

  2 février 2012, par justanothercoder

  I'm trying to get ffmpeg to seek h264 interlaced videos, and i found that i can seek to any frame if i just force it.

  I already hacked the decoder to consider I - Frames as keyframes, and it works nicely with the videos i need it to work with. And there will NEVER be any videos encoded with different encoders.

  However, i'd like the seek to find me an I - Frame and not just any frame.

  What i'd need to do is to hack The AVIndexEntry creation so that it marks any frame that is an I-Frame to be a key frame.
  Or alternatively, hack the search thing to return I - Frames.

  The code does get a tad dfficult to follow at this point.

  Can someone please point me at the correct place in ffmpeg code which handles this ?

• FFMPEG for iPhone recorded video encoding

  16 décembre 2011, par Sarah

  Hi found lots and lots of links for video encoding through ffmpeg 1,2,3,4 etc but they all start with using terminal commands but when i try to implement any on terminal like :

  git clone git ://github.com/lajos/iFrameExtractor.gitit says that-bash : git : command not found.

  Also as per my knowledge it is not possible to use terminal command on iPhone. Can anybody point out how to encode a video recorded through ffmpeg in mp4 format and also to reduce the size of the video ?Thanks in advance.

  EDIT :
  I am already implementing this method to resize my video and it successfully takes place and I am able to send the video on server but then on server side it's giving problem in retrieving the data and to use it.

  - (void)imagePickerController:(UIImagePickerController *)picker 
  
  didFinishPickingMediaWithInfo:(NSDictionary *)info
  
  {
  
      [self convertVideoToLowQuailtyWithInputURL:videoURL1 outputURL:[NSURL fileURLWithPath:videoStoragePath] handler:^(AVAssetExportSession *exportSession)
  
       {
  
           if (exportSession.status == AVAssetExportSessionStatusCompleted)
  
           {
  
               NSLog(@"%@",exportSession.error);
  
               printf("completed\n");
  
           }
  
           else
  
           {
  
               NSLog(@"%@",exportSession.error);
  
               printf("error\n");
  
           }
  
       }];
  
  }
  
  
  
  - (void)convertVideoToLowQuailtyWithInputURL:(NSURL*)inputURL 
  
                                     outputURL:(NSURL*)outputURL 
  
                                       handler:(void (^)(AVAssetExportSession*))handler
  
  {
  
      [[NSFileManager defaultManager] removeItemAtURL:outputURL error:nil];
  
      AVURLAsset *asset = [AVURLAsset URLAssetWithURL:inputURL options:nil];
  
      AVAssetExportSession *exportSession = [[AVAssetExportSession alloc] initWithAsset:asset presetName:AVAssetExportPresetLowQuality];
  
      exportSession.outputURL = outputURL;
  
      exportSession.outputFileType = AVFileTypeQuickTimeMovie;
  
      [exportSession exportAsynchronouslyWithCompletionHandler:^(void) 
  
       {
  
           handler(exportSession);
  
           [exportSession release];
  
       }];
  
  }