[argyllcms] Re: Colprof feature request: amount of blackpointhuecorrection control
- From: Nikolay Pokhilchenko <nikolay_po@xxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 13 May 2009 20:53:51 +0400
Gerhard FЭrnkranz wrote:
>
> -------- Original-Nachricht --------
> > Datum: Wed, 13 May 2009 09:48:10 +0400
> > Von: Nikolay Pokhilchenko <nikolay_po@xxxxxxx>
> > An: argyllcms@xxxxxxxxxxxxx
> > Betreff: [argyllcms] Re: Colprof feature request: amount of black
> > pointhuecorrection control
>
> > Gerhard, the xicclu uses "simple" gamut clipping by default.
> > When You try xicclu -b -s255 -fb -ir LSM2484.icc, the clipping
> > for relative intent will be by CAM Jab.
>
> You're right, but I still find it interesting that nearest distance clipping
> of "absolute black" in CIELAB versus CIECAM space makes so much difference
> for this particular printer profile.
>
> -Gerhard
>
> --
No, Gerhard, it's obvious if You'll take a look at profile gamut in both
spaces. Try this:
iccgamut -ir -pl LSM2484.icm
rename LSM2484.gam LSM2484-pl.gam
iccgamut -ir -pj LSM2484.icm
rename LSM2484.gam LSM2484-pj.gam
viewgam -cr -s -t0.5 LSM2484-pl.gam -cg -w -t0 LSM2484-pj.gam LAB_Jab.wrl
then see resulting Lab_Jab.wrl by VRML-viewer.
You'll see, that nearest to absolutely black point of Lab gamut will be at
R=G=B=0, but nearest to true black Jab gamut point isn't RGB zero (for higher
precision use -d4 or even -d3 key for iccgamut if possible).
Other related posts: