Recherche avancée

Sur d’autres sites (6724)

• Stack AVFrame side by side (libav/ffmpeg)

  22 février 2018, par dronemastersaga

  So I am trying to combine two H264 livestreams of 1920x1080 resolution side-by-side to a livestream of 3840x1080 resolution.

  For this, I can decode streams to AVFrames in libav/FFmpeg that I would like to combine into a bigger frame. The Input AVFrames : Two 1920x1080 frames in NV12 format (description : planar YUV 4:2:0, 12bpp, 1 plane for Y and 1 plane for the UV components, which are interleaved (first byte U and the following byte V))

  The way I have figured out is with colorspace conversion (YUV to BGR) in libav, then to change it to OpenCV Mat, then to use hconcat in OpenCV to stack together, then colorspace conversion (BGR to YUV) in AVFormat.

  Below is the method currently being used :

  //Prior code is too long: Basically it decodes 2 streams to AVFrames frame1 and frame2 in a loop
  
  sws_scale(swsContext, (const uint8_t *const *) frame1->data, frame1->linesize, 0, 1080, (uint8_t *const *) frameBGR1->data, frameBGR1->linesize);
  
  sws_scale(swsContext, (const uint8_t *const *) frame2->data, frame2->linesize, 0, 1080, (uint8_t *const *) frameBGR2->data, frameBGR2->linesize);
  
  Mat matFrame1(1080, 1920, CV_8UC3, frameBGR1->data[0], (size_t) frameBGR1->linesize[0]);
  
  Mat matFrame2(1080, 1920, CV_8UC3, frameBGR2->data[0], (size_t) frameBGR2->linesize[0]);
  
  Mat fullFrame;
  
  hconcat(matFrame1, matFrame2, fullFrame);
  
  const int stride[] = { static_cast<int>(fullFrame.step[0]) };
  
  sws_scale(modifyContext, (const uint8_t * const *)&amp;fullFrame.data, stride, 0, fullFrame.rows, newFrame->data, newFrame->linesize);
  
  //From here, newFrame is sent to the encoder
  
  </int>

  The resulting image is satisfactory but it does lose quality in colorspace conversion. However this method is too slow to use (I’m at 15 fps and I need 30). Is there a way to stack AVFrames directly without colorspace conversion ? Or is there any better way to do this ? I searched a lot about this and I couldn’t find any solution to this. Please advise.

• Bash script : automate ffmpeg encoding for mpeg-dash

  13 février 2018, par Massimo Vantaggio

  I’m writing a bash file to create video encoding and concatenation for dash live streaming use,
  Basically it read an input video folder, encodes all videos into three resolution formats, after that it concatening them to create three adaption sets.

  DIAGRAM :

  

  This script checks for fps conformance,

  force/scaling resolution if the input is not 1920 x 1080p,

  Insert the channel logo png,

  Cut the end of all videos input in order to make them finish with closed gop, this to ensure that there are not videos with audio and video track with different lenght.

  ISSUE :

  Actually I’m not sure that the concatenation process respects the closed GOP alignment as it after the encoding..
  I try to cut also the end of the concatenation result in order to make it finish without decimals on a closed gop too, but im unable to erase all decimals from the total duration :

  total duration in seconds: 826.795000
  
  total duration corrected in seconds: 826

  But the real duration measured by ffprobe is

  824.044000

  I try to check keyframes alignment with mp4box has they teach without any result :

  MP4Box -info TRACK_ID source1.mp4 2>&1 | grep GOP

  This is the first time that i work with "video scripts" and probably i don’t know what input to give for TRACK_ID

  BASH SCRIPT :

  #!/bin/bash
  
  #CANCAT 0.2
  
  
  
  cd input
  
  times=()
  
  fps=()
  
  for f in *.mp4; do
  
  
  
      _t=$(ffprobe -i "$f" -show_entries format=duration -v quiet -of csv="p=0")
  
      times+=("$_t")
  
  
  
      _f=$(ffmpeg -i "$f" 2>&1 | sed -n "s/.*, \\(.*\\) fp.*/\\1/p")
  
      fps+=("$_f")
  
  done
  
  #SUM ALL DURATIONS
  
  TOTALDURATION=$( echo "${times[@]}" | sed 's/ /+/g' | bc )
  
  #DELETE DECIMAL
  
  DURROUND=$(echo "$TOTALDURATION" | cut -d'.' -f1) 
  
  #GET REST OF DIVISION BY 2 AS GOP
  
  TOTDELTA="$((DURROUND%2))"
  
  #SUBTRACT DELTA FROM TOTAL DURATION
  
  TOTDUR="$(($DURROUND-$TOTDELTA))"
  
  
  
  #GET NUMBER OF ELEMENTS IN FPS ARRAY  
  
  tLen=${#fps[@]}
  
  #CHECK FPS EQUALITY     
  
  for tLen in "${fps[@]:1}"; do
  
      if [[ $tLen != ${fps[0]} ]]; then
  
          printf "WARNING: VIDEO’S FRAME-RATE ARE NOT EQUALS, THE PROCESS CAN’T START."
  
          printf "%s\\0" "${fps[@]}" |
  
              sort -zu |
  
              xargs -0 printf " %s"
  
          printf "\\n"
  
         exit 1
  
      fi
  
  done
  
  for f in *.mp4; do
  
  #GET DURATION OF EACH VIDEO
  
  DUR="$(ffprobe -i "$f" -show_entries format=duration -v quiet -of csv="p=0")"
  
  DUR=$(echo "$DUR" | cut -d'.' -f1) # DELETE DECIMAL
  
  #GET FPS OF EACH VIDEO
  
  FPS="$(ffmpeg -i "$f" 2>&1 | sed -n "s/.*, \(.*\) fp.*/\1/p")"
  
  #ROUND FPS OF EACH VIDEO
  
  FPSC=$( echo "($FPS+0.5)/1" | bc )
  
  #REMOVE EXTENSION FROM VIDEO FILE NAME
  
  NAME=$(echo "$f" | cut -d'.' -f1)
  
  
  
  #GET GOP
  
  GOP="$((FPSC*2))"
  
  DELTADUR="$((DUR%2))"
  
  DUR="$(($DUR-$DELTADUR))"
  
  
  
  #ENCODE 1080p
  
  ffmpeg -y -i "$f" -i ../logo/logo.png -c:a aac -b:a 384k -ar 48000 -ac 2 -async 1 -c:v libx264 -x264opts keyint=$GOP:min-keyint=$GOP:no-scenecut -bf 0 -r $FPSC -b:v 4800k -maxrate 4800k -bufsize 3000k -profile:v main -crf 22 -t $DUR -filter_complex "[0:v][1:v]overlay=main_w-overlay_w-10:10,scale=1920:1080,setsar=1" ../buffer/${NAME}-1080.mp4
  
  
  
  #ENCODE 720p
  
  ffmpeg -y -i ../buffer/${NAME}-1080.mp4 -c:a aac -b:a 256k -ar 48000 -ac 2 -async 1 -c:v libx264 -x264opts keyint=$GOP:min-keyint=$GOP:no-scenecut -bf 0 -s 1280x720 -r $FPSC -b:v 2400k -maxrate 2400k -bufsize 1500k -profile:v main -crf 22 -t $DUR ../buffer/${NAME}-720.mp4
  
  
  
  #ENCODE 360p
  
  ffmpeg -y -i ../buffer/${NAME}-720.mp4 -c:a aac -b:a 128k -ar 48000 -ac 2 -async 1 -c:v libx264 -x264opts keyint=$GOP:min-keyint=$GOP:no-scenecut -bf 0 -s 640x360 -r $FPSC -b:v 800k -maxrate 800k -bufsize 500k -profile:v main -crf 22 -t $DUR ../buffer/${NAME}-360.mp4
  
  done
  
  
  
  
  
  #enter in buffer
  
  cd ..
  
  cd buffer
  
  
  
  #CONCAT 1080 SET
  
  # with a bash for loop
  
  for f in ./*1080.mp4; do echo "file '$f'" >> 1080list.txt; done
  
  
  
  ffmpeg -f concat -safe 0 -i 1080list.txt -t $TOTDUR -c copy ../output/1080set.mp4
  
  
  
  #CONCAT 720 SET
  
  # with a bash for loop
  
  for f in ./*720.mp4; do echo "file '$f'" >> 720list.txt; done
  
  
  
  ffmpeg -f concat -safe 0 -i 720list.txt -t $TOTDUR -c copy ../output/720set.mp4
  
  
  
  #CONCAT 360 SET
  
  # with a bash for loop
  
  for f in ./*360.mp4; do echo "file '$f'" >> 360list.txt; done
  
  
  
  ffmpeg -f concat -safe 0 -i 360list.txt -t $TOTDUR -c copy ../output/360set.mp4
  
  
  
  #CLEAN BUFFER
  
  rm *.mp4
  
  rm *.txt
  
  
  
  echo "CONCAT COMPLETED:"
  
  echo "frame-rate: $fps"
  
  echo "total duration in seconds: $TOTALDURATION"
  
  echo "total duration corrected in seconds: $TOTDUR"

  The full file with relative folders :

  BASH SCRIPT WITH FOLDERS

  RESULT VIDEO

  There is someone who can help me to understand why I can not eliminate the decimals of the total duration during concat ?
  And how to check overall keyframes allignment ?
  Also any impovement that i ignore is welcome !

  Thanks a lot !

  Massimo

• Why is ffmpeg trying to produce both h.264 as well as h.265 ?

  7 mars 2018, par hydra3333

  I think I only ask for h.265 output, but the output loge below seems to indicate it is trying to produce 2 video streams as output.

  "C:\SOFTWARE\ffmpeg\0-homebuilt-x64\built_for_generic_opencl\x64_8bit\ffmpeg.exe" -hide_banner -v verbose  -threads 0 -i "G:\HDTV\0nvencc\test-mp4-03\ABC HD interlaced.aac.mp4" -t 15 -threads 0 -an  -sws_flags lanczos+accurate_rnd+full_chroma_int+full_chroma_inp -filter_complex "[0:v]yadif=0:0:0" -pixel_format yuv420p -pix_fmt yuv420p -strict -1 -f yuv4mpegpipe - 2> .\zzz1.h265.txt 
  
  | "C:\SOFTWARE\ffmpeg\0-homebuilt-x64\built_for_generic_opencl\x64_8bit\ffmpeg.exe" -strict -1 -hide_banner -v verbose  -threads 0 -i - -strict -1 -c:v:0 libx265 -crf 28 output.mp4 -an  -y .\zzz.h265.mp4 2> .\zzz2.h265.txt 
  
  
  
  -----------------------------
  
  <snip to="to" make="make" code="code" block="block" smaller="smaller">
  
      Stream mapping:
  
        Stream #0:0 (h264) -> yadif
  
        yadif -> Stream #0:0 (wrapped_avframe)
  
      Press [q] to stop, [?] for help
  
      [h264 @ 000001b78ca87c00] Reinit context to 1920x1088, pix_fmt: yuv420p
  
      [graph 0 input from stream 0:0 @ 000001b78ca264c0] w:1920 h:1080 pixfmt:yuv420p tb:1/90000 fr:25/1 sar:1/1 sws_param:flags=2
  
  </snip>

  Output file #0 (pipe:):
  
    Output stream #0:0 (video): 2 frames encoded; 2 packets muxed (1072 bytes); 
  
    Total: 2 packets (1072 bytes) muxed
  
  Conversion failed!
  
  -----------------------------
  
  Routing option strict to both codec and muxer layer
  
      Last message repeated 1 times
  
  Input #0, yuv4mpegpipe, from 'pipe:':
  
    Duration: N/A, start: 0.000000, bitrate: N/A
  
      Stream #0:0: Video: rawvideo, 1 reference frame (I420 / 0x30323449), yuv420p(progressive, left), 1920x1080, SAR 1:1 DAR 16:9, 25 fps, 25 tbr, 25 tbn, 25 tbc
  
  Stream mapping:
  
    Stream #0:0 -> #0:0 (rawvideo (native) -> hevc (libx265))
  
    Stream #0:0 -> #1:0 (rawvideo (native) -> h264 (libx264))
  
  [graph 0 input from stream 0:0 @ 0000027f5ebefa00] w:1920 h:1080 pixfmt:yuv420p tb:1/25 fr:25/1 sar:1/1 sws_param:flags=2
  
  x265 [info]: HEVC encoder version 2.7613d9f443769
  
  x265 [info]: build info [Windows][GCC 7.3.0][64 bit] 8bit
  
  x265 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX
  
  x265 [info]: Main profile, Level-4 (Main tier)
  
  x265 [info]: Thread pool created using 8 threads
  
  x265 [info]: Slices                              : 1
  
  x265 [info]: frame threads / pool features       : 3 / wpp(17 rows)
  
  -----------------------------