On Wed, Aug 26, 2015 at 11:28 AM, Paul B Mahol <[email protected]> wrote:
> On 8/26/15, Robert Krueger <[email protected]> wrote: > > Hi, > > > > I created two files using these command lines: > > > > ffmpeg -i fcp7_xdcam_ex_1080_25p_1s.mov -vf > > scale=in_color_matrix=bt709:out_color_matrix=bt601 -c:v ffv1 -an > > scale_test_bt601.nut > > > > ffmpeg -i fcp7_xdcam_ex_1080_25p_1s.mov -c:v ffv1 -an > > scale_test_unscaled.nut > > > > Using framemd5 I verified that the resulting output is identical. Is this > > some sort of optmization of swsscale (i.e. not doing anything when size > and > > pixel format remain unchanged)? Can it be forced to do the conversion > > anyway? I know that there is also vf_colormatrix but that is much more > > limited than swscale and it's not LGPL. > > Looking at code and if I'm not missing something it's just not implemented > to > be used like that, perhaps better to fill bug report. > Thanks. I created ticket 4805. This also happens when using swsscale directly using the api, which is where I ran into it. I just used the command line to verify this. So the ticket is more general. _______________________________________________ ffmpeg-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-user
