Recherche avancée

Sur d’autres sites (8006)

• No such file or directory Error with FFMPEG screenshot and CarrierWave custom model method

  9 juillet 2013, par dodgerogers747

  I am using AWS CORS to upload videos to my site, all of which works as planned.

  I have the following model method which runs as an after_create callback (for speed) to take a screenshot from the video file on AWS. I plan to move this out into a delayed job but I don't think this will solve this particular issue. Please advise if mistaken.

  I use FFMPEG to take a screenshot from the AWS self.file location, I then send the file to CarrierWave by saving the file to self.screenshot where it is uploaded to AWS.

  Approx. 50% of the time it errors out with Errno::ENOENT - No such file or directory for the location of the screenshot image.

  How can I rectify my code to remove this error and how come it only occurs around 50% of the time ? If anyone needs more code just shout.

  video.rb

  after_create :take_screenshot
  
  
  
  mount_uploader :screenshot, ImageUploader
  
  
  
    def take_screenshot
  
      location = "#{Rails.root}/public/uploads/tmp/screenshots/#{unique}_#{File.basename(file)}.jpg"
  
      system `ffmpeg #{log_level} -i #{self.file} -ss 00:00:0#{time_frame} -vframes 1 #{location}`
  
      logger.debug "Trying to take screenshot from #{self.file}"
  
      #pass the actual file to CarrierWave to handle the image upload
  
      self.screenshot = File.open(location)
  
      self.save
  
      logger.debug "Deleting tmp file: #{location}: #{File.delete(location)}" if self.screenshot.present?
  
    end
  
  
  
  def unique
  
      (0..6).map{(65+rand(26)).chr}.join
  
    end
  
  
  
  def log_level
  
      "-loglevel panic"
  
    end
  
  
  
    def time_frame
  
      rand(0..3)
  
    end

  Stack trace :

  Started POST "/videos" for 127.0.0.1 at 2013-07-10 03:58:49 +0800
  
  Processing by VideosController#create as JS
  
    Parameters: {"utf8"=>"✓", "authenticity_token"=>"6M1Ia+Ag2E3HVKH2PO/p7jewxSpMPdWeVHGA933Bzjw=", "video"=>{"file"=>"http://bucketname.s3.amazonaws.com/uploads/video/file/671a87fb-91de-4eaf-a38a-1b25c51798e5/Good_7iron.m4v"}}
  
    User Load (0.3ms)  SELECT `users`.* FROM `users` WHERE `users`.`id` = 9 LIMIT 1
  
     (0.1ms)  BEGIN
  
    SQL (0.2ms)  INSERT INTO `videos` (`created_at`, `file`, `question_id`, `screenshot`, `updated_at`, `user_id`) VALUES ('2013-07-09 19:58:49', 'http://bucketname.s3.amazonaws.com/uploads/video/file/671a87fb-91de-4eaf-a38a-1b25c51798e5/Good_7iron.m4v', NULL, NULL, '2013-07-09 19:58:49', 9)
  
  Trying to take screenshot from http://bucketname.s3.amazonaws.com/uploads/video/file/671a87fb-91de-4eaf-a38a-1b25c51798e5/Good_7iron.m4v
  
     (0.8ms)  ROLLBACK
  
  Completed 500 Internal Server Error in 3550ms
  
  
  
  Errno::ENOENT - No such file or directory - /Users/me/rails/project/public/uploads/tmp/screenshots/WCACLIC_Good_7iron.m4v.jpg:
  
    app/models/video.rb:24:in `initialize'
  
    app/models/video.rb:24:in `open'
  
    app/models/video.rb:24:in `take_screenshot'

• Get frame from live video stream

  10 avril 2013, par rickster

  I am streaming live video from my camera on my android phone to my computer using the MediaRecorder class.

  recorder.setCamera(mCamera);
  
  recorder.setVideoSource(MediaRecorder.VideoSource.CAMERA);
  
  recorder.setOutputFormat(MediaRecorder.OutputFormat.THREE_GPP);          
  
  recorder.setOutputFile(uav_UDP_Client.pfd.getFileDescriptor());                     
  
  recorder.setVideoEncoder(MediaRecorder.VideoEncoder.H264);

  That's the basic idea. So I would like to show this stream in real time. My plan is to use FFMpeg to turn the latest frame into a .bmp and show the .bmp on my C# program every time there is a new frame.

  The problem is there is no header until I stop the recording. So I can not use FFMpeg unless there is a header. I've looked at spydroid and using RTP but I do not want to use this method for various reasons.

  Any ideas on how I can do this easily ?

• Why frame->pts increases by 20, rather than by 1 ?

  19 mars 2013, par user1914692

  Following the exmaples of ffmpeg : decoding_encoding.c and filtering_video.c, I process one video file taken by iPhone. The video file : .mov, video dimensions ; 480x272, video Codec : H.264/AVC, 30 frames per second, bitrate : 605 kbps.

  I first extract each frame, which is YUV.
  I convert YUV to RGB24, and process the RGB24, then write the RGB24 to a .ppm file. It shows the .ppm file is correct.

  Then I plan to encode processed RGB24 frames to a video file.
  Since MPEG does not support RGB24 picture format, I used AV_CODEC_ID_HUFFYUV.
  But the output video file (showing 18.5 MB) does not play. Movie Player on Ubuntu claims an error : Could not determine type of stream.
  I also tried it on VCL. It simply does not work, without any error information.

  My second questions is :
  For each extracted fram from the input video file, I get its pts as follows according to filtering_video.c :

  frame->pts = av_frame_get_best_effort_timestamp(frame);

  I print out each frame's pts, and find that it increases by 20, like below :

  pFrameRGB_count: 0,  frame->pts: 0
  
  pFrameRGB_count: 1,  frame->pts: 20
  
  pFrameRGB_count: 2,  frame->pts: 40
  
  pFrameRGB_count: 3,  frame->pts: 60

  Where frame is the extracted frame from the input video, and pFrameRGB_count is the count for processed frame in RGB24 form.

  Why are they wrong ?