Thank you, that makes perfect sense. I am working on a project where we use the ffmpeg libraries to perform cropping and scaling. When we start breaking stuff we want to be sure that we know what the data looks like.
________________________________________ From: [email protected] <[email protected]> on behalf of wm4 <[email protected]> Sent: 01 February 2015 15:53 To: [email protected] Subject: Re: [Libav-user] How is cropping and scaling implemented On Sun, 1 Feb 2015 14:44:11 +0000 "Kristensen, Odd-Arild" <[email protected]> wrote: > Fair point, cropping should not modify the pixels. But at some point you > would want to remove the pixels outside the selected region, or copy the > pixels inside the region to the destination frame. I dont’t see either of > these things happening in filter_frame() Am I looking in the wrong place? Cropping can be implemented by changing the pixel data pointer to point to the top left pixel. Nothing else needs to be done. Why are you even asking? It's completely transparent to the user. _______________________________________________ Libav-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/libav-user _______________________________________________ Libav-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/libav-user
