Recherche avancée

Sur d’autres sites (9830)

• lavc/pcm_tablegen : slight speedup of table generation

  31 décembre 2015, par Ganesh Ajjanagadde

  lavc/pcm_tablegen : slight speedup of table generation
  This gets rid of some branches to speed up table generation slightly
  
  (impact higher on mulaw than alaw). Tables are identical to before,
  
  tested with FATE.
  Sample benchmark (Haswell, GNU/Linux+gcc) :
  
  old :
  
   313494 decicycles in build_alaw_table,    4094 runs,      2 skips
  
   315959 decicycles in build_alaw_table,    8190 runs,      2 skips
   323599 decicycles in build_ulaw_table,    4095 runs,      1 skips
  
   318849 decicycles in build_ulaw_table,    8188 runs,      4 skips
  new :
  
   261902 decicycles in build_alaw_table,    4096 runs,      0 skips
  
   266519 decicycles in build_alaw_table,    8192 runs,      0 skips
   209657 decicycles in build_ulaw_table,    4096 runs,      0 skips
  
   232656 decicycles in build_ulaw_table,    8192 runs,      0 skips
  Reviewed-by : Michael Niedermayer <michael@niedermayer.cc>
  
  Signed-off-by : Ganesh Ajjanagadde <gajjanagadde@gmail.com>
  

  • [DH] libavcodec/pcm_tablegen.h

• ffmpeg Concatention - negative pad values

  15 décembre 2015, par mmmyum

  I’m relatively new to ffmpeg and having an issue concatenating some videos.

  Occasionally, I’m running into an issue with my command that is producing negative pad values, preventing the concat from completing. Check out this gist for details and the command output.

  Brief summary is that I’m using filter_complex to scale down and/or pad videos. You can see the full commands in the above links, but here are examples of the filter text :

  [$count:v]scale=if(gt(ih\,iw)\,-2\,$VID_WIDTH):if(gt(ih\,iw)\,$VID_HEIGHT\,-2),pad=$VID_WIDTH:$VID_HEIGHT:(ow-iw)/2:(oh-ih)/2,setsar=sar=1/1[v$count];

  where $VID_WIDTH and $VID_HEIGHT are the smallest dimensions of the inputs.

  I’m concatenating hundreds of videos (in different groups), the vast majority of which work without issue - but occasionally I run into this negative pad values issue. I assume it’s because the input width/height is larger than the output width/height, so it produces a negative value, but I’m not sure how to fix this - I need the videos to be centered.

  Other details :

  • most videos are shot on a cell phone, in a portrait orientation (some in landscape, but I’d like to stitch them in portrait anyways)

  • very rarely I run into a video rotated 90 degrees (larger width than height) - not sure how I should handle this, but one issue at a time

  • usually all videos have the same dimensions, but occasionally one or more videos in a group will have slightly different dimensions, sometimes just by a couple pixels (see attached gist for example)

  • scripting everything in bash right now

• aacenc : move the TNS search and filtering before PNS

  6 décembre 2015, par Rostislav Pehlivanov

  aacenc : move the TNS search and filtering before PNS
  The original plan was to have TNS use data from the PNS search to better
  
  tune itself to noise but this was never used nor necessary. This should
  
  slightly boost the PNS accuracy if TNS was used.
  Signed-off-by : Rostislav Pehlivanov <atomnuker@gmail.com>
  

  • [DH] libavcodec/aacenc.c