Recherche avancée

Sur d’autres sites (8567)

• FFmpeg uses too much memory when repeating split, select, overlay

  13 novembre 2020, par finefoot

  I'm running

  


  ffmpeg -i input.mp4 -filter_complex_script script.txt output.mp4


  


  with the following minimal example script :

  


  split[tmp],
select='between(t,1,2)',
select='between(n,0,1)',
[tmp]overlay=enable='between(t,1,2)':eof_action=repeat,
split[tmp],
select='between(t,3,4)',
select='between(n,0,1)',
[tmp]overlay=enable='between(t,3,4)':eof_action=repeat


  


  What I want to do is to take 1 frame at a certain position and repeat it for a certain duration, basically "pausing" the video, while overwriting to keep the same output length. In the example, I'm doing this twice : I'm using split[tmp] to get a second input stream to work on, select the time at position 00:01 with select='between(t,1,2)', select the first frame from that position with select='between(n,0,1)' and finally overlay that frame over the input. This repeats for a second time at position 00:03. I have tested this and it does exactly what I'm looking for.

  


  However, in my real script, I'm repeating this about 1000 times for different positions in the stream (and for shorter durations than 1 second) which results in running out of memory. What am I doing wrong ? What can I do to optimize ?

  


• Pipe numpy array to virtual video device

  16 mars 2021, par sre

  I want to pipe images to a virtual video device (e.g. /dev/video0), the images are created inside a loop with the desired frame rate.

  


  In this minimal example i only two arrays which alternate in the cv2 window. Now i look for a good solution to pipe the arrays to the virtual device.

  


  I saw that ffmpeg-python can run asynchronous with ffmpeg.run_async(), but so far i could not make anything work with this package.

  


  example code without the ffmpeg stuff :

  


  #!/usr/bin/env python3

import cv2
import numpy as np
import time

window_name = 'virtual-camera'
cv2.namedWindow(window_name, cv2.WINDOW_GUI_EXPANDED)

img1 = np.random.uniform(0, 255, (1080, 1440, 3)).astype('uint8')
img2 = np.random.uniform(0, 255, (1080, 1440, 3)).astype('uint8')

for i in range(125):
    time.sleep(0.04)
    if i % 2:
        img = img1
    else:
        img = img2
    cv2.imshow(window_name, img)
    cv2.waitKey(1)
cv2.destroyAllWindows()


  


• avcodec/eatgv : Check remaining size after the keyframe header

  28 juillet 2019, par Michael Niedermayer

  avcodec/eatgv : Check remaining size after the keyframe header
  The minimal size which unpack() will not fail on is 5 bytes
  
  Fixes : Timeout (14sec -> 77ms) (testcase 15508)
  
  Fixes : 15508/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_EATGV_fuzzer-5700053513011200
  
  Fixes : 15996/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_EATGV_fuzzer-5751353223151616
  Found-by : continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
  
  Signed-off-by : Michael Niedermayer <michael@niedermayer.cc>
  

  • [DH] libavcodec/eatgv.c