On Sun, May 12, 2019 at 10:47 PM Paul B Mahol <[email protected]> wrote:
> On 5/8/19, Song, Ruiling <[email protected]> wrote: > > > > > >> -----Original Message----- > >> From: ffmpeg-devel [mailto:[email protected]] On Behalf > >> Of Dylan Fernando > >> Sent: Tuesday, May 7, 2019 8:27 AM > >> To: [email protected] > >> Subject: Re: [FFmpeg-devel] [PATCHv2] lavfi: add gblur_opencl filter > >> > >> Anyone have any comments/feedback? > > I think unsharp_opencl with a negative amount should do similar thing as > > this one. > > Not really. > > > What's the difference? Better quality? or better speed? > > This one can blur image with larger radius. > > But why step parameter was removed? > _______________________________________________ > ffmpeg-devel mailing list > [email protected] > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > [email protected] with subject "unsubscribe". This one doesn't use a Gaussian approximation, it uses a Gaussian kernel calculated with: matrix_horiz[i] = (1 / sqrt(2 * 3.14159*pow(s->sigma, 2)))*exp(-(pow(x, 2) / (2 * pow(s->sigma, 2)))) with kernel size 6 * sigma. Should there be a parameter for adjusting the kernel size? _______________________________________________ ffmpeg-devel mailing list [email protected] https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
