Recherche avancée

Sur d’autres sites (10247)

• python cv2 video playing quality

  10 août 2017, par TheRutubeify

  Why it is so big difference in quality between MPC-CH Windows Media Player and cv2 cv2.VideoCapture(’*.mp4’) when playing the same file ! Codec H. 264 E

  How I can improve quality with cv2 ?

  Here is the screenshot Left(cv2) Rigth(MPC-CH) :
  

  cv2.namedWindow("Final", 0)
  
  cv2.resizeWindow("Final", 300, 200)
  
  cap = cv2.VideoCapture('1.mp4')
  
  while(cap.isOpened()):
  
      qwe, frame = cap.read()
  
      cv2.waitKey(30)
  
      gray = cv2.cvtColor(frame, cv2.COLOR_BGR2BGRA) #COLOR_RGB2RGBA
  
      cv2.imshow('Final',gray)

  Any suggestions ?

• Trying to concatenate 11 seconds of black video onto the end of a 5 min video. Getting non-monotonous DTS in output stream error

  14 janvier 2020, par TidusWulf

  I have a video shot from my smartphone that is 3840x2160@57.74. Originally, it was exactly 5 minutes long. I replaced the audio with a music file that is 5 minutes and 11 seconds using ffmpeg -i video.mp4 -i audio.mp3 -c copy -map 0:v:0 -map 1:a:0 output.mp4 I expected 11 seconds of black at the end, but instead the video output freezes on the last frame for 11 seconds. There’s definitely something funky going on because when I try to upload to youtube, it only sees the first 5 minutes. The last 10 seconds of audio get dropped.

  I tried making a second black clip with ffmpeg -f lavfi -i color=c=black:s=uhd2160:r=57.74 -t 11 -pix_fmt yuv420p blk2.mp4. When I try to concat the two files with ffmpeg -f concat -safe 0 -i list.txt -c copy teacup6.mp4 I get a huge list of errors such as [mp4 @ 00000166fda994c0] Non-monotonous DTS in output stream 0:0; previous: 27002920, current: 3728954; changing to 27002921. This may result in incorrect timestamps in the output file. for what appears to be about 11 seconds worth of frames, so basically the entire 11 second black clip. When I play it in VLC it goes black, but I suspect it’s not actually processing/playing 11 seconds of good black video because if I click along the track timeline, visual glitches start happening in the canvas.

  Here is what ffmpeg tells me about my inputs and outputs as far as framerate, pixel format, etc.

  [mov,mp4,m4a,3gp,3g2,mj2 @ 00000239e6cb2880] Auto-inserting h264_mp4toannexb bitstream filter
Input #0, concat, from 'list.txt':
 Duration: N/A, start: -0.023021, bitrate: 72244 kb/s
   Stream #0:0(eng): Video: h264 (High) (avc1 / 0x31637661), yuv420p(tv, bt709), 3840x2160, 71994 kb/s, SAR 1:1 DAR 16:9, 57.74 fps, 59 tbr, 90k tbn, 180k tbc
   Metadata:
     handler_name    : VideoHandle
   Stream #0:1(und): Audio: mp3 (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 250 kb/s
   Metadata:
     handler_name    : SoundHandler
Output #0, mp4, to 'teacup5.mp4':
 Metadata:
   encoder         : Lavf58.35.102
   Stream #0:0(eng): Video: h264 (High) (avc1 / 0x31637661), yuv420p(tv, bt709), 3840x2160 [SAR 1:1 DAR 16:9], q=2-31, 71994 kb/s, 57.74 fps, 59 tbr, 90k tbn, 90k tbc
   Metadata:
     handler_name    : VideoHandle
   Stream #0:1(und): Audio: mp3 (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 250 kb/s
   Metadata:
     handler_name    : SoundHandler
Stream mapping:
 Stream #0:0 -> #0:0 (copy)
 Stream #0:1 -> #0:1 (copy)
Press [q] to stop, [?] for help

  alternatively, I tried to make a black 3840x2160 png file and create an 11 second slideshow. It didn’t turn out any better. I tried a single image for 11 seconds @ 57.74fps, I also tried a looping slideshow of that one image, looping on every single frame, at 57.74 fps. I should note that wether a slideshow or a color generator, the filesize for the 11 seconds of black comes out to about 160KB. Strangely small in my opinion, but I was chalking it up to a good compression algorithm.

  I tried to do this in Da Vinci Resolve 16.1 instead and the video wouldn’t play. I also couldn’t find a way to maintain the same unusual 57.74 framerate with the free version, so it was undesirable anyway. I tried to re-process using Handbrake, but it threw an error. The output of the Handbrake processed video file also gets stuck on the last frame of the video, instead of going to black.

• Video encoder & segmenter for HLS VoD poor quality

  10 juillet 2017, par Murilo

  I am trying to encode and segment video for HLS on demand(VoD).
  I am using the following code for such :

  ffmpeg -i 20170706_174314.mp4 -c 24 \
  
          -vcodec libx264 -acodec aac -ac 1 -strict -2 -b:v 128k \
  
          -profile:v baseline -maxrate 400k -bufsize 1835k \
  
          -hls_time 10 -hls_playlist_type vod -vsync 1 \
  
          video_chunks/index1.m3u8 \
  
          -c 24 -vcodec libx264 -acodec aac -ac 1 -strict -2 -b:v 128k \
  
          -profile:v baseline -maxrate 700k -bufsize 1835k \
  
          -hls_time 10 -hls_playlist_type vod -vsync 1 \
  
          video_chunks/index2.m3u8

  I tried this other code also just for segmenting but had the same exactly problem :

  ffmpeg -i 20170706_174314.mp4 \
  
  -c:a libmp3lame -ar 48000 -ab 64k  -c:v libx264 -b:v 128k -flags \ 
  
  -global_header -map 0 -f segment \
  
  -segment_list video_chunks/test.m3u8 -segment_time 10 -segment_format mpegts \
  
  video_chunks/segment_%05d.ts

  Later on I create another playlist with bandwidth separators to call on the two other playlists generated with the code above.

  This code was working great on some videos but yesterday I recorded a video with my Samsung J7 Prime phone to test since the videos will be generated by phone and this video was poorly encoded. The quality sucks and some parts of the video turned Black&White.

  Another thing I noticed on this video is that the following message kept appearing in loop until the end of the encoding&segmenting process.

  Past duration X too large

  Where X is a decimal really close to

  0.675316

  The link to the video is below :

  Dropbox Link

  My FFmpeg version :

  ffmpeg --version
  
  ffmpeg version N-86482-gbc40674 Copyright (c) 2000-2017 the FFmpeg developers
  
    built with gcc 7.1.0 (GCC)
  
    configuration: --enable-gpl --enable-version3 --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-libmfx --enable-nvenc --enable-avisynth --enable-bzlib --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-lzma --enable-zlib
  
    libavutil      55. 66.100 / 55. 66.100
  
    libavcodec     57. 99.100 / 57. 99.100
  
    libavformat    57. 73.100 / 57. 73.100
  
    libavdevice    57.  7.100 / 57.  7.100
  
    libavfilter     6. 92.100 /  6. 92.100
  
    libswscale      4.  7.101 /  4.  7.101
  
    libswresample   2.  8.100 /  2.  8.100
  
    libpostproc    54.  6.100 / 54.  6.100

  SO : Windows 10

  EDIT1 : Link to the output
  If you see the output it might be worth saying I am also seeing the message

  VBV underflow(Frame X, -Y bits)