[Ffmpeg-devel] interlace and motion artifacts

Michael Niedermayer michaelni
Fri Mar 3 14:05:47 CET 2006


On Fri, Mar 03, 2006 at 01:25:32PM +0100, Luca Abeni wrote:
> Hi All,
> On Tue, 2006-02-21 at 09:24 +0100, Luca Abeni wrote:
> [...]
> > > what about making the generic C based scaling part LGPLed
> > > while leaving the asm code under GPL if someone ports the code to ffmpeg?
> > I was waiting for someone else to volunteer, but since noone is speaking
> > up...
> > I think I can try to have a look at porting the swscaler C code to
> > ffmpeg.
> Ok, I finally found some time, and I started by updating my old patch to
> current ffmpeg and libswscaler. It seems to work well.


> Now, I want to begin supporting pad & crop in my sws glue code (ffmpeg's
> img_resample_full_init() takes crop & pad parameters as input... I did
> not know how to handle them with libswscaler, so my current code falls
> back to ffmpeg resampling if padding or cropping is used). To do this, I
> am trying to understand how cropping and padding interact with
> rescaling...
> My first (maybe naive) impression is that cropping&padding could be
> handled by simply manipulating the input and output size. Is this
> wrong? 

no, its not wrong, that should work

> While trying to understand the code, I noticed that the owidth & oheight
> members of ImgReSampleContext do not seem to be used anywhere
> (pad_owidth & pad_oheight are used instead). Am I wrong? Can someone
> enlight me on this?

cvslog, no i dont remember at all ...



More information about the ffmpeg-devel mailing list