On Wed, 7 Jan 2015 10:46:47 -0800 Philip Langdale <phil...@overt.org> wrote:
> > Honestly, I don't understand what the encoder is doing at this point. > The 1.02 seems to be the 'correct' scale factor for a PAL DVD, but is > incorrect for anything else. I used 1024 to avoid the floating point > problem, but you are right that pre-multiplication is a better way to > handle that. Right now I'm playing around with an NTSC DVD and I can't > find the right magic to stop it mangling the aspect ratio. > > At this point, I'm not convinced that darWidth and darHeight are even > representing the display presentation size at all. Hopefully Agatha > can understand what's going on here. Now, here's something weird. I set the darWidth and darHeight to '1' and ffprobe reports this: Stream #0:0(eng): Video: h264 (Main), yuv420p, 720x480 [SAR 15:22 DAR 45:44], SAR 186:157 DAR 279:157, 29.97 fps, 29.97 tbr, 1k tbn, 59.94 tbc Note how it somehow ends up with 45:44 for the DAR, which happens to be 1.02 - so that's where that comes from. I have no idea where the SAR comes from. Alright, so we now have a situation where we need to compensate for this 1.02 scale factor in our dar calculation. We need to scale up the width/height values enough that heigh * 1.02 is an integer - hence why the 1024 helped previously. So, for our NTSC dvd case, I might use darWidth=8520, darHeight=4896 and this will correctly compensate. ffprobe now reports: [SAR 1775:1496 DAR 5325:2992] which is close enough that mplayer (for example) will correctly derive 852x480 for the display size. This is crazy, but it's clearly what the encoder is doing. --phil _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
