On 2017-02-14 22:25, Mark Thompson wrote: > On 14/02/17 19:44, Daniel Oberhoff wrote: >> filter strictly “halves” the image efficiently, which is often exactly what >> is needed >> likely much faster than using scale > > Did you benchmark this? How? > > $ time ./ffmpeg -f lavfi -i allyuv -vf 'scale=iw/2:ih/2' -vframes 400 -f null > - > ... > frame= 400 fps= 26 q=-0.0 Lsize=N/A time=00:00:16.00 bitrate=N/A speed=1.05x > ... > real 0m15.365s > user 0m11.092s > sys 0m4.272s > > $ time ./ffmpeg -f lavfi -i allyuv -vf 'halve' -vframes 400 -f null - > ... > frame= 400 fps= 22 q=-0.0 Lsize=N/A time=00:00:16.00 bitrate=N/A speed=0.873x > ... > real 0m18.392s > user 0m46.280s > sys 0m3.656s > > So it uses four times as much CPU as swscale to be marginally slower?
I would be tempted to blame the lack of SIMD for the poor performance.
signature.asc
Description: OpenPGP digital signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
