"He who marches out of step hears another drum." ~ Ken Kesey
Tony,Underbasing is another really complex idea. And once again, the solutions out there are all over the place. Underbasing works best with the HSB model (surprise.) The difference between Brightness and Value is Brightness is based on emissivity and Value is based on reflectance. Think of emissivity as how much light is being emitted. So for a computer pixel 100% birghtness means 100% power to the pixel. So we want to put 100% color behind 100% brightness color values. If you want to see the radical differnce between Brightness and Value calculations use the color picker in Photoshop and compare hs B value with L ab for a color like 185 Red. That is the simple answer, but it gets more complex, naturally that rabbit hole againWhen you have a wet white under or over a color, you are desaturating toward white. In order to maintain saturation, you need variying degrees of dry white and wet with. So the transform is really two dimensional. The first dimension is to preserve the brightness and the second to preserve saturation. So two colors with 100% brightness will have dry white under the 100% saturation and wet whites behind the high brightness but decreasinging saturation colors. Again, this benefits from having the ability to create selective underbasing by color in the separation.As a guide in picking colors for your separtion. My experience has been the best gamut results come when the B value is greater than 60%.This type of underbase can be created in Photoshop if you use the HSB filter. It is getting harder and harder to find and to my knowledge is not available for 16 bit color calculation. That means it is only available for CS5 and lower. If anyone knows differently, I would really like to know about it.
I was just thinking I do not want to get to far away from my life raft the water is getting pretty deep around here. Looking forward to your post pierre
you give me too much credit, I just have a million questions and very little actual knowledge to contribute. . . Mark is doing the heavy lifting here and it is just AWESOME to get this kind of information!pierre
Actually and this my sound like it is out there... But because we are dealing with math I see us being very close to what you just outlined within a year. Not with every detail you have listed but very close. Because color values and densities in the digital space can be scanned based on numerical values which Jeff just started playing with that and has a path to making it work. So smart separations are on the horizon for sure. But all based around HSB color which in theory could be tied into pigments but I am not sure if will get it that far. But being able to scan the color and present smart options for the separation is possible.