Recherche avancée

Sur d’autres sites (12251)

• avfilter/x86/vf_blend : add 16 bit version for BLEND_SIMPLE, phoenix, difference...

  17 février 2018, par Martin Vignali

  avfilter/x86/vf_blend : add 16 bit version for BLEND_SIMPLE, phoenix, difference for SSE and AVX2 (x86_64)
  

  • [DH] libavfilter/x86/vf_blend.asm
  • [DH] libavfilter/x86/vf_blend_init.c

• What's the difference with crf and qp in ffmpeg ?

  12 novembre 2024, par Nova

  I read https://trac.ffmpeg.org/wiki/Encode/H.264 about h264 encoding and discovered qp.

  


  Q1 : What are the differences with crf and qp ?
  
Q2 : Is it better to use qp over crf overall, or is it only if for using qp 0 for best lossless ?
  
Q3 : Does qp have a known sensible setting if it's preferred ? So far, I know crf has the default value of 23 while 18 is a sensible preferred increase in quality, although I don't understand why 18 wouldn't be default if better sensible lossless.
  
Q4 : Would changing either of them cause incompatibility with non-ffmpeg players or just qp ?

  


  I'm converting from webm to mp4.

  


  I was going to test crf 23 and 18 and pick which is best but I can't seem to find any concrete information on this comparison or about qp.

  


• What is the difference between different fadein/fadeout curves in ffmpeg ?

  16 août 2018, par siods333333

  Here is the list of possible curves for afade and acrossfade filters from here https://ffmpeg.org/ffmpeg-filters.html#afade-1

  tri
  
      select triangular, linear slope (default) 
  
  
  
  qsin
  
      select quarter of sine wave 
  
  
  
  hsin
  
      select half of sine wave 
  
  
  
  esin
  
      select exponential sine wave 
  
  
  
  log
  
      select logarithmic 
  
  
  
  ipar
  
      select inverted parabola 
  
  
  
  qua
  
      select quadratic 
  
  
  
  cub
  
      select cubic 
  
  
  
  squ
  
      select square root 
  
  
  
  cbr
  
      select cubic root 
  
  
  
  par
  
      select parabola 
  
  
  
  exp
  
      select exponential 
  
  
  
  iqsin
  
      select inverted quarter of sine wave 
  
  
  
  ihsin
  
      select inverted half of sine wave 
  
  
  
  dese
  
      select double-exponential seat 
  
  
  
  desi
  
      select double-exponential sigmoid 

  Here is the code for them, from libavfilter/af_afade.c :

  switch (curve) {
  
  case QSIN:
  
      gain = sin(gain * M_PI / 2.0);
  
      break;
  
  case IQSIN:
  
      /* 0.6... = 2 / M_PI */
  
      gain = 0.6366197723675814 * asin(gain);
  
      break;
  
  case ESIN:
  
      gain = 1.0 - cos(M_PI / 4.0 * (CUBE(2.0*gain - 1) + 1));
  
      break;
  
  case HSIN:
  
      gain = (1.0 - cos(gain * M_PI)) / 2.0;
  
      break;
  
  case IHSIN:
  
      /* 0.3... = 1 / M_PI */
  
      gain = 0.3183098861837907 * acos(1 - 2 * gain);
  
      break;
  
  case EXP:
  
      /* -11.5... = 5*ln(0.1) */
  
      gain = exp(-11.512925464970227 * (1 - gain));
  
      break;
  
  case LOG:
  
      gain = av_clipd(1 + 0.2 * log10(gain), 0, 1.0);
  
      break;
  
  case PAR:
  
      gain = 1 - sqrt(1 - gain);
  
      break;
  
  case IPAR:
  
      gain = (1 - (1 - gain) * (1 - gain));
  
      break;
  
  case QUA:
  
      gain *= gain;
  
      break;
  
  case CUB:
  
      gain = CUBE(gain);
  
      break;
  
  case SQU:
  
      gain = sqrt(gain);
  
      break;
  
  case CBR:
  
      gain = cbrt(gain);
  
      break;
  
  case DESE:
  
      gain = gain <= 0.5 ? cbrt(2 * gain) / 2: 1 - cbrt(2 * (1 - gain)) / 2;
  
      break;
  
  case DESI:
  
      gain = gain <= 0.5 ? CUBE(2 * gain) / 2: 1 - CUBE(2 * (1 - gain)) / 2;
  
      break;
  
  }

  How do they look like ? How do they sound like ? Which one is recommended for fadein+fadeout and crossfade ? Personally I’m just trying to avoid audio clicks, maybe crossfade is a bit of an overkill here.

  Related link : http://manual.audacityteam.org/man/fade_and_crossfade.html . Not sure how audacity names translate into ffmpeg names though.