Recherche avancée

Sur d’autres sites (9777)

• ffmpeg - MXF with 4-channel audio : how to create proxies (high quality / small file size) and preserve the audio mapping

  8 juillet 2018, par WhatsYourFunction

  We’ve got MXF sources (h.264 video at UHD (3840x2160) with 4-channels of (4 - PCM S24 mono sources)
  We want Proxies — smallest file size at highest picture quality
  The compression applied to the video and audio essences can be anything,
  And the wrapper can be either MXF or QuickTime
  but we need to preserve the audio mapping (i.e. the Proxy must be 4-channel audio)

  How to do that with ffmpeg ?

  • EDIT Adding ffprobe :

  Metadata :

  uid : ***

  generation_uid : ***

  company_name : CANON

  product_name : EOS C300 Mark II

  product_version : 1.00

  product_uid : ***

  modification_date : 2018-06-28T08:29:24.000000Z

  material_package_umid : ***

  timecode : 02:50:31:17

  Duration : 00:06:35.40, start : 0.000000, bitrate : 395842 kb/s

  Stream #0:0 : Video : h264 (High 4:2:2 Intra), yuv422p10le(tv, progressive), 3840x2160, SAR 1:1 DAR 16:9, 23.98 fps, 23.98 tbr, 23.98 tbn, 47.95 tbc

  Metadata :

   file_package_umid: ***

  Stream #0:1 : Audio : pcm_s24le, 48000 Hz, 1 channels, s32 (24 bit), 1152 kb/s

  Metadata :

   file_package_umid: ***

  Stream #0:2 : Audio : pcm_s24le, 48000 Hz, 1 channels, s32 (24 bit), 1152 kb/s

  Metadata :

   file_package_umid: ***

  Stream #0:3 : Audio : pcm_s24le, 48000 Hz, 1 channels, s32 (24 bit), 1152 kb/s

  Metadata :

   file_package_umid: ***

  Stream #0:4 : Audio : pcm_s24le, 48000 Hz, 1 channels, s32 (24 bit), 1152 kb/s

  Metadata :

   file_package_umid: ***

• FFMpeg : Add background image during the process of video creation

  10 juillet 2018, par Killer

  In reference to my previous question FFMpeg : merge images with audio for specific duration I have successfully merged few images with audio for specific duration by using following command.

  ffmpeg \
  
  -y \
  
  -f concat \
  
  -safe 0 \
  
  -r 1/5 \
  
  -i concat.txt \
  
  -i audio.ogg \
  
  -c:v libx264 \
  
  -profile:v high \
  
  -crf 17 \
  
  -preset ultrafast \
  
  -strict experimental \
  
  -t 15 \
  
  output.mp4

  In order to add a background image, I tried to use https://superuser.com/a/876275/299733 and other solutions that exist on the web. But the given solution doesn’t overlay my images properly and I was getting black video throughout the duration. Therefore, I encode the resultant video from earlier command once again via :

  ffmpeg \
  
  -y \
  
  -loop 1 \
  
  -i bg.jpg \
  
  -i output.mp4 \
  
  -filter_complex "overlay=(W-w)/2:(H-h)/2:shortest=1,format=yuv420p" \
  
  -c:v libx264 \
  
  -profile:v high \
  
  -crf 17 \
  
  -preset ultrafast \
  
  -strict experimental \
  
  -t 15 \
  
  output2.mp4

  Now, I am able to get the desired result. Is there any way to merge both requests into a single pass ? Or a better way without loss of any performance ?

  Additional details :

  concat.txt

  file '/home/shubham/Desktop/FFMpeg/image_1.jpg'
  
  file '/home/shubham/Desktop/FFMpeg/image_2.jpg'
  
  file '/home/shubham/Desktop/FFMpeg/image_3.jpg'

  Based on @gyan response

  Updated concat.xml :

  file '/home/shubham/Desktop/FFMpeg/image_4.jpg'
  
  duration 5
  
  file '/home/shubham/Desktop/FFMpeg/image_5.jpg'
  
  duration 5
  
  file '/home/shubham/Desktop/FFMpeg/image_6.jpg'
  
  duration 5
  
  file '/home/shubham/Desktop/FFMpeg/image_6.jpg'

  Updated Command :

  ffmpeg \
  
  -y \
  
  -loop 1 \
  
  -i bg.jpg \
  
  -f concat \
  
  -safe 0 \
  
  -i concat.txt \
  
  -i audio.ogg \
  
  -filter_complex "[1]fps=25[v];[0][v]overlay=(W-w)/2:(H-h)/2:shortest=1,format=yuv420p" \
  
  -c:v libx264 \
  
  -profile:v high \
  
  -crf 17 \
  
  -preset ultrafast \
  
  -strict experimental \
  
  -t 15 \
  
  output.mp4

  The problem is that when images are of different resolution or even for the same resolution the images are skipped. And most of the times last image is shown. However, there are no criteria of which image is selected and which one is skipped.
  SAMPLE :
  https://drive.google.com/file/d/1JxCsV2eudKzdgWWuefXqdaWPaBf9Dzzd/view?usp=sharing

  However, if I repeatedly used the same image or copy the image and rename it. In both the cases, I get proper images on the background without any skipping.

  EDIT : 09 JULY 2018

  As @gyan stated in comments : Convert each image to same resolution and type.
  I check the info of images. via

   ffmpeg -i image_X.jpg

  And found out two images have different encoding :

  Image 1 : Stream #0:0 : Video : mjpeg, yuvj420p(pc, bt470bg/unknown/unknown), 640x480 [SAR 72:72 DAR 4:3], 25 tbr, 25 tbn, 25 tbc

  Image 2 : Stream #0:0 : Video : mjpeg, yuvj444p(pc, bt470bg/unknown/unknown), 640x480 [SAR 72:72 DAR 4:3], 25 tbr, 25 tbn, 25 tbc

  Which is the possible cause of the failure in merging.

• HLS-streaming on VLC - cookies alternative

  5 juillet 2018, par unicornslayer

  I’m trying to build a streaming web api which creates m3u8 playlists of .ts files in .net core.

  I’ve built the live streaming option and it works fine on vlc.

  For example, in VLC -> File -> Open Network put my link : http://localhost:39986/api/Stream/MyChannel/index.m3u8 and VLC gives me the live version of the channel.

  But I need to implement the option of starting a video from a fixed time : http://localhost:39986/api/Stream/MyChannel/index.m3u8?timeStr="2018-07-04T17:16:01". In this case, I would need to create a session to know from which point to create the playlist, right ?[1]

  If so, sessions require an ID stored as a Cookie (right ?[2]). But from what I’ve read and what I’ve tested, vlc doesn’t support cookies.

  So my question is : is there any workaround for this scenario ?[3]

  It’s an university project and I am very confused about this task.