Recherche avancée

Sur d’autres sites (7914)

• lavu/tx : improve 3-point fixed precision

  14 février 2020, par Lynne

  lavu/tx : improve 3-point fixed precision
  There's just no reason not to when its so easy (albeit messy) and its also
  
  reducing the precision of all non-power-of-two transforms that use it.
  

  • [DH] libavutil/tx_priv.h
  • [DH] libavutil/tx_template.c

• Error writing to file when saving matplotlib animation with a given codec

  31 mars 2017, par idesh

  I’ve had trouble saving an mp4 animation created with matplotlib with a given codec. However I should mention beforehand that I could save the mp4 animation without specifying codecs and that seems to work fine, except that I am not able to insert this mp4 animation into powerpoint for a presentation. Powerpoint says it cannot read the file because it is missing 64bit codec.

  When digging further I discovered that Windows media player which is used by power point for the animations, has already some video codecs installed.

  So I thought of saving the initial matplotlib animation with one of these codecs, which is also compatible with ffmpeg. The command ffmpeg -codecs on terminal lists the supported codecs and I could spot the codecs common with ffmpeg and windows media player, ex : MSS1, MSS2

  So I tried to save the animation with the argument codec set as follows.

  anim = animation.FuncAnimation(fig, animate, 158,interval=300, blit=True)
  
  writer = animation.writers['ffmpeg']
  
  anim.save('film_v5.mp4', codec='mss1')

  Nevertheless it leads to an error, no matter what type of codec argument I put.

  So I was wondering perhaps there was someone savvy, willing to help me troubleshoot this problem. Despite my attempts I could not find any solution in the stackoverflow forum.

  Thank you in advance for your attention.

• swscale/output : Altivec-optimize yuv2plane1_8

  17 novembre 2018, par Lauri Kasanen

  swscale/output : Altivec-optimize yuv2plane1_8
  ./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt yuv420p \
  
   f null -vframes 100 -v error -nostats -
  1158 UNITS in planar1,   65528 runs,      8 skips
   cpuflags 0
  19082 UNITS in planar1,   65533 runs,      3 skips
  16.48 speedup ratio. On x86, SSE2 is 7. Curiously, the Power C version
  
  takes as many cycles as the x86 SSE2 version, yikes it's fast.
  Note that this function uses VSX instructions, but is not marked so.
  
  This is because several existing functions also make that mistake.
  
  I'll submit a patch moving them once this is reviewed.
  Signed-off-by : Lauri Kasanen <cand@gmx.com>
  
  Signed-off-by : Michael Niedermayer <michael@niedermayer.cc>
  

  • [DH] libswscale/ppc/swscale_altivec.c