I not sure that this makes sense, but I have edited the lens_calibrate.py script to use sRGB as the input color profile, rerun the script and updated the lensfun database. Now if I import the raw flatfield images used to create the correction data into darktable, with the usual 'enhanced color matrix' input profile, the vignetting correction is excellent, being just a little under done in the extreme corners, but that is no bad thing.
Jim On Tue, Jan 8, 2019 at 8:50 AM Andreas Schneider <[email protected]> wrote: > On Tuesday, January 8, 2019 9:48:27 AM CET Jim Robinson wrote: > > I have tried your new version of the script and whilst the polynomial > > coeffients are slightly different the overall outcome is essentially the > > same, i.e very poor correction for raw file in darktable. This is not > > surprising as the changes shown on screen on changing the input color > > profile from 'enhanced color matrix' to rec2020 are barely detectable. > > Unfortunately I cannot get the old calibrate.py script to compare > outcomes > > with to work. > > Hmm, do we need to apply a neutral base curve to get an estimated > correction? > > > > > -- > Andreas Schneider [email protected] > GPG-ID: 8DFF53E18F2ABC8D8F3C92237EE0FC4DCC014E3D > > >
_______________________________________________ Lensfun-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/lensfun-users
