Hello,

Am 21.04.06, 15:32 +0200 schrieb BERGER Cyrille:

> > My suggestion is to Keep it Simple.
> >
> I am not sure what your point is, you mean we should use the same type
> for all channels in all colorspaces with the same range ?

Put every colour space in the same range and keep it there until a 
storage format need something different.

The keep simple suggestion is to put all integer data equaly 
into unsigned types, like unsigned char or unsigned short. The 
value range starts with zero and ends with the maximum of that type.  
For the CIE*a and CIE*b channels the neutral point is just the middle 
between zero and maximum.

Floating point numbers range normally from zero until 1.0. Of course over 
white and below black is allowed for that data type by its nature.

The decision to handle all data in one type, for instance float,  has 
some benefits as well. But this is not what I discussed in my former 
email. Probably you handle data types very generic in your code 
allready. So there is less need for any change.


kind regards
Kai-Uwe Behrmann
                                + development for color management 
                                + imaging / panoramas
                                + email: ku.b@gmx.de
                                + http://www.behrmann.name

_______________________________________________
kimageshop mailing list
kimageshop@kde.org
https://mail.kde.org/mailman/listinfo/kimageshop