Hello all, We do some performance critical processing with ffmpeg involving image scaling and warping. To support this we did implemented some new algorithms and parallelized one.
1. parallelize remap 2. implement remap_frac that interprets data as 13.3 fixed point (i.e. 3 bits for subpixel warping) 3. implement halve that makes the image exactly halve size and is faster than using scale we would like to contribute this back to ffmpeg as it may be useful for others. do you think so to? if so, should we send one big patch or split it up somehow? Best Daniel Oberhoff _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
