[argyllcms] Re: Faking the white point in camera calibration
- From: "Gerhard Fürnkranz" <nospam456@xxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 31 Aug 2011 12:02:04 +0200
-------- Original-Nachricht --------
> Datum: Wed, 31 Aug 2011 19:09:48 +1000
> Von: Graeme Gill <graeme@xxxxxxxxxxxxx>
> An: argyllcms@xxxxxxxxxxxxx
> Betreff: [argyllcms] Re: Faking the white point in camera calibration
> Stephen T wrote:
>
> > Now, rather than editing the wtpt tag, maybe I can edit the .ti3 file
> and then profile that? I
> > am thinking to edit the XYZ_X XYZ_Y XYZ_Z and RGB_R RGB_G RGB_B data for
> GS00 to make everything
> > neutral. I can make the reference Lab neutral by setting a and b to
> zero. Then I can convert
> > these numbers to XYZ. For the camera, I can use the average of the
> measured values and set
> > R=G=B.
> >
> > Is this safe or would it make problems for colprof and screw up the
> model fitting?
>
> That may work, as long as the change in the white patch is not too great.
I'd tend to add a faked white patch with RGB_R=RGB_G=RGB_B=100 and
"corresponding" XYZ numbers to the .ti3 file. It may be a bit problematic
though to estimate the XYZ numbers of this patch, as this involves
extrapolation and potentially even clipping (if a profile with CIELAB PCS is
used for estimating the XYZ numbers from the RGB numbers with icclu). And if Y
of the faked white patch happens to be greater than 100, then it may also be
necessary to downscale/darken all XYZ numbers in the .ti3 file correspondingly.
Regards,
Gerhard
--
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de
Other related posts:
