[argyllcms] Re: XYZ/LAB to RGB
- From: Gerhard Fuernkranz <nospam456@xxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Tue, 12 Jan 2010 14:23:27 +0100
Pascal de Bruijn wrote:
>
> Right, so you all made your point, on how pointless is to try to
> replicate a CC...
>
> But, I still like to know how the RGB values are approximated...
It's vice versa, the RGB numbers are the exact, prescribed device
numbers that will be sent to the printer when you print the target,
while the XYZ numbers in the .ti1 and .ti2 are just a preliminary
approximation for the resulting colors (and AFAIK they are only used for
rough sanity checks during chartread). After chartread, the .ti3 file
will contain the actually measured XYZ (or CIELAB or spectral) values of
the printed patches.
[I think, if a device profile is supplied to targen, it uses this
profile for computing the XYZ numbers in the .ti1 from the RGB numbers -
but I'm not absolutely sure]
> Say for example I'd like to include some self measured skin tones to a
> generic chart?
The RGB numbers in the .ti1 and .ti2 are in the device color space of
the printer on which you eventually print the target. So I suggest you
build a preliminary device profile for your printer and use this profile
to convert the desired XYZ/CIELAB numbers to RGB (e.g. using the icclu
or xicclu command).
Regards,
Gerhard
Other related posts:
