Recherche avancée

Sur d’autres sites (9151)

• lavu/hwcontext_vulkan : Support write on drm frame

  11 juin 2024, par Haihao Xiang

  lavu/hwcontext_vulkan : Support write on drm frame
  Otherwise nothing is written into the destination when a write mapping
  
  is requested.
  For example, a vulkan frame mapped from a drm frame (which is wrapped as
  
  a vaapi frame in the example) is used as the output of scale_vulkan
  
  filter, it always gets a green screen without this patch.
  ffmpeg -init_hw_device vaapi=va -init_hw_device vulkan=vulkan@va
  
   filter_hw_device vulkan -f lavfi -i testsrc=size=352x288,format=nv12
  
   vf
  
  "hwupload,scale_vulkan,hwmap=derive_device=vaapi:reverse=1,format=vaapi,hwdownload,format=nv12"
  
   f nut - | ffplay -
  Signed-off-by : Haihao Xiang <haihao.xiang@intel.com>
  

  • [DH] libavutil/hwcontext_vulkan.c

• Oddly-named file "not found" when trying to move/copy via MacOS Terminal [closed]

  15 juin 2024, par Ben C

  This is a head scratcher.

  


  I need to move these three oddly-named files from an external 2.5" SSD to my desktop. The file names were intended to have variables replaced by date info, however, clearly it didn't work. So I'm left with these filenames that MacOS seems to want to interpret rather than treat as a string. I wish it was as easy as tossing in single quotes or escaping chars, but so far, that hasn't worked.

  


  Files: Test01-$(internal:date_y)-$(internal:date_m)-$(internal:date_d)-.mp4 Test01-$(internal:date_y)-$(internal:date_m)-$(internal:date_d)-0001.mp4 Test01-$(internal:date_y)-$(internal:date_m)-$(internal:date_d)-0002.mp4

  


  I'm on MacOS. In the Finder, the files are not visible. But they are not hidden files.

  


  In Terminal, however, when I navigate to "/Volumes/TestDrive", then run a quick "ls", I can see all three files no problem. Including permissions, size, owners, full filename, etc.

  


  However, when I attempt to move the files to my desktop, and rename in the process (even if I don't rename), Terminal tells me that "No such file or folder can be found" or something very close to that.

  


  I've tried using mv and cp commands to put the filename in single quotes so the filename is read literally. Yet, I'm still given feedback that the file cannot be transferred because it cannot be found or doesn't exist.

  


  mv 'Test01-$(internal:date_y)-$(internal:date_m)-$(internal:date_d)-.mp4' ~/Desktop/Test01-01.mp4' 'cp 'Test01-$(internal:date_y)-$(internal:date_m)-$(internal:date_d)-.mp4' ~/Desktop/Test01-01.mp4

  


  When I try to escape special characters instead of quotes, I am also told the file doesn't exist. But clearly it does when I list out the contents of the drive. And there's only 3x .mp4 files, two hiddne files .fseventsd and some spotlight file.

  


  mv Test01-\$\(internal\:date_y\)-\$\(internal\:date_m\)-\$\(internal\:date_d\)-.mp4 ~/Desktop/Test01-01.mp4

  


  I've tried copying by inode. No luck. I've tried pulling the videos into ffmpeg (CLI based media mgmt tool) to see if I can get some info on the files, and same thing, ffmpeg (or ffprobe) both will tell me the file doesn't exist...even though I can list the files and see that it does.

  


  I hope I'm missing something obvious, but but it seems all the obvious approaches are not yet working for me.

  


  So my question is, what do I need to do to make these files "exist" so that I can rename them and back them up ? Happy to go down any rabbit hole.

  


  Thanks in advance !

  


• the "non-monotonic DTS in output stream " error while concatenation even after reencoded input files

  19 juin 2024, par nish

  hi I am trying to write a Python program that is trying to first edit different videos separately under the heading "intro" "story" and "byte" with the help of FFmpeg and Python subprocess module and then concatenate them in the function named "bind". At first, I encountered this "Non-monotonic DTS in output stream" then i reencoded the input files before concatenating them but i am still getting this error

  


  # function to assemble all the videos 


  


  def bind() :

  


  bunch = ["final_intro.mp4","final_story.mp4","finalbite.mp4"]
new_bunch =[]
for video in bunch:
    name = f"re_{video}"
    re_command =[
        "ffmpeg", "-y",
        "-i", video,
        "-c:v", "libx264",
        "-c:a", "aac",
        "-strict", "experimental",
        "-b:a", "192k",
        name
    ]
    subprocess.run(re_command)
    new_bunch.append(name)
    

with open("concat_list_final.txt", "w") as f:
    for video in new_bunch:
        f.write(f"file './{video}'\n")
  
command_final = [ 
    "ffmpeg", "-y",
    "-f", "concat",
    "-safe", "0",
    "-i", "concat_list_final.txt",
    "-c", "copy",
    "done.mp4"
]
subprocess.run(command_final)
for video in new_bunch:
    os.remove(video)
os.remove("final.mp4")
os.remove("updated_final.mp4")
os.remove("concat_list_final.txt")


return "done.mp4"


  


  for reference this is the error messages :-

  


  [mov,mp4,m4a,3gp,3g2,mj2 @ 0x14cf16040] Auto-inserting h264_mp4toannexb bitstream filter [mp4 @ 0x12ce06880] Non-monotonic DTS in output stream 0:0 ; previous : 152933, current : 127760 ; changing to 152934. This may result in incorrect timestamps in the output file. [mp4 @ 0x12ce06880] Non-monotonic DTS in output stream 0:0 ; previous : 152934, current : 128272 ; changing to 152935. This may result in incorrect timestamps in the output file..........