[FFmpeg-user] how is sws gauss filter implemented?
pnorthug at gmail.com
Thu Oct 27 14:01:38 CEST 2011
We have been using ffmpeg in our research to filter movies and it has
been a huge help. I was wondering if someone can explain the details
of the sws gauss option. If I am downscaling a movie from say 64x64
per frame to 32x32 using gauss, what happens? Is each frame convolved
with a 2d-gaussian filter and then subsampled? We mostly output each
frame to a png after applying the filter.
Any details on the number of taps used to design the filter or
anything else would be appreciated. I couldn't find much documentation
and was challenging for me to understand the source, for example, in
More information about the ffmpeg-user