Recherche avancée

Sur d’autres sites (4345)

• movenc : Allow writing a DASH sidx atom at the start of files

  21 octobre 2014, par Martin Storsjö

  movenc : Allow writing a DASH sidx atom at the start of files
  This is mapped to the faststart flag (which in this case
  
  perhaps should be called "shift and write index at the
  
  start of the file"), which for fragmented files will
  
  write a sidx index at the start.
  When segmenting DASH into files, there’s usually one sidx
  
  at the start of each segment (although it’s not clear to me
  
  whether that actually is necessary). When storing all of it
  
  in one file, the MPD doesn’t necessarily need to describe
  
  the individual segments, but the offsets of the fragments can be
  
  fetched from one large sidx atom at the start of the file. This
  
  allows creating files for the DASH ISO BMFF on-demand profile.
  Signed-off-by : Martin Storsjö <martin@martin.st>
  

  • [DH] libavformat/movenc.c
  • [DH] libavformat/movenc.h

• ffmpeg waits to close the pipe in order to start processing data

  15 juin 2022, par Mahmoud Ferig

  I'm struggling with this issue.
Here I am trying to stream video using ffmpeg from images and audio file this code below works fine but it is start streaming after stdin closed, I would expect ffmpeg to start processing the pipe input as it is received, and immediately output the result.

  


  p = subprocess.Popen(
    [
        'ffmpeg',
        '-y',
        '-re',
        '-f', 'image2pipe',
        '-vcodec', 'mjpeg',
        '-framerate', '15',
        '-analyzeduration', '2147483647',
        '-probesize', '2147483647',
        '-i', '-',
        '-vn', '-i', audio_path,
        '-c:v', 'libx264',
        '-pix_fmt', 'yuv420p',
        '-preset', 'ultrafast',
        '-c:a', 'aac',
        '-f', 'flv',
        rtmp_url
    ],
    stdin=subprocess.PIPE,
    stdout=subprocess.PIPE
)

time.sleep(2)

i = 0
while True:
    current_images = [name for name in os.listdir(img_dir) if os.path.isfile(os.path.join(img_dir, name))]
    current_images = sorted(current_images)
    if i > fnum-1:
        print("frame read failed")
        break
    try:
        img = cv2.imread(os.path.join(img_dir, current_images[i]))
        success, buffer = cv2.imencode('.jpg', img)
        frame = buffer.tobytes()
        p.stdin.write(frame)
        i += 1
    except:
        time.sleep(2)
        pass

p.stdin.close()
p.wait()


  


  Why is ffmpeg waiting to close the pipe to start processing ? Can it be configured to start a live transcoding of the received stream ?

  


  Do you know how can I convince ffmpeg to start producing output immediately ?

  


  Thank you !

  


• FFmpeg : Combine video files with different start times

  18 mai 2016, par Fabian

  I have two webm files with audio and video recordings of a video conference session. Both files only contain one side of the conversation. They are not of the same length (someone has joined before the other one), but I have the unix timestamp in milliseconds of the start time of each video file.

  On a timeline they look like this :

  webm 1:   -----------------------------------------------
  
  webm 2:                 -----------------------------

  or like this :

  webm 1:   -----------------------------------------------
  
  webm 2:                          -----------------------------

  I would like to combine these two video files into one file so that :

  1. They appear next to each other (using the hstack option), and
  2. That they are mixed with taking the time stamps of the start times
     into account. The final video should then look like this :

  Target result : --------------===========================----

  The beginning and the end of the new video would show a black placeholder for the video file that has no data at this time of the mixed stream.

  At the moment I use this command :

  ffmpeg -i 1463408731413703.webm -i 1463408880317860.webm -filter_complex \
  
  "[0:v][1:v]hstack=inputs=2[v]; \
  
   [0:a][1:a]amerge[a]" \
  
  -map "[v]" -map "[a]" -ac 2  -c:v libvpx output.webm

  This creates a video like this :

  Not good result : =====================------------------

  i.e. the conversation is out of sync.

  How can I combine two video streams with different length and different start times using ffmpeg so that I will end up with "Target result" above ?

  Thanks a lot !