On 12/7/19, Tao Zhang <nowe...@gmail.com> wrote: > I'm sorry for the late reply. > > Paul B Mahol <one...@gmail.com> 于2019年12月7日周六 上午2:48写道： >> >> On 12/5/19, Paul B Mahol <one...@gmail.com> wrote: >> > On 12/5/19, Tao Zhang <nowe...@gmail.com> wrote: >> >> Hello everyone, >> >> Can I assume this patch is ok if no comments or objections? >> > >> > Yes, give some time for it to be applied. >> > I'm busy with other stuff right now. So this patch LGTM (Note to >> > committer to bump minor of libavfiter upon pushing). >> >> I do not get anywhere near output of this filter with either >> avgblur/gblur/boxblur? >> >> How should output of this filter look like? > Assume keeping planes and radius unchanged， > 1) the output is near avgblur/boxblur when sigma close to INT_MAX > 2) the output is near raw input when sigma close to one, so why not > zero? just keep code simple, I need to gurantee the > denominator larger than zero > 3) the output is local adaptive that means blur stronger on flat area > while blur weaker on edge otherwise.
Is there plan to make it faster with SIMD? _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
