The RGBW Color Space
by Reinhard Kietzmann, Berlin
A Novel Color Model for Color / White Tone Separation
Idea and MotivationEvery color space has its own strengths and weaknesses, there is justification for all of the basic color spaces as there are
RGB: The natural color space of digital cameras and monitors. Color and lightness information are tightly coupled. Thus it is well suited to adjust neutral tones and correct for color casts, brightness and mixed light situations.
CMYK: The printing color space. As the color information is separated from the black component it is best to use for adjusting shadow detail and subtle colors.
>Lab: The magic working color space. The complete separation of color and lightness information allows precise control of shadows and highlights as well as the opportunity to correct color casts and increase saturation with dramatic effect.
HSB: The rarely used relative of Lab. It separates hue, saturation and brightness, but its difficult to use for making adjustments to photographs. More suitable as source of masks.
RGB and CMYK come in various flavors, but this should be neglected for this discussion.
Though RGB is the most commonly used color space, many photo processing experts often like to use CMYK and Lab because of the separation of color and lightness components. No such thing exists for RGB – so far.
If appropriate, why not try to stay in RGB, but mimic the separation of the K component in CMYK. RGB is based on additive color mixing in contrast to the subtractive color mixing of CMYK. Thus the fourth component for an RGB color space which should contain the lightness information should be white (W) rather than black (K). Et voila – the RGBW color space is born!
RGBW: The working RGB color space with color/white tone separation. Color and lightness information are largely decoupled. This allows for more subtle highlight and brightness adjustments as well as color cast correction separate for saturated or gray picture parts and it allows easy saturation improvements.
How can I convert photos to RGBW?Photo processing programs like Photoshop, Paint, Gimp etc. have no integrated ability to handle additive color spaces with more than 3 components, at least not to my knowledge. Thus a little trick is needed to create and work with RGBW pictures. The white component of a picture is calculated and subtracted. The result is a regular RGB picture that only contains the color components. This forms the first layer. Then the white component is put into a second layer on top of the color component layer and added to the first layer. Done. The tricky part is the separation. The white component needs to be isolated using RGB-BW conversion functions and a saturation mask. Then it can be subtracted from the original RGB picture leaving only the color component remaining. A Photoshop macro that is doing the job for you can be downloaded here: RGBWmacro.
Note: In order to run macro you need to have the filter "Hsbhsl.8bf" installed, which can be found either on your Photoshop installation CD (until Photoshop CS3, search for the file name) or on the Adobe web site for Windows systems here and for Apple systems here , under "optional plug-ins".
See the examples below with some separated photos in the order of RGBW, Color component, White component.
How can I work with RGBW pictures?The ideal way is to select either the color or the white component layer and do curve adjustments. Curves applied to the different channels of the color layer can modify colored parts of a picture or increase the saturation. Curves applied to the channels of the white component layer can remove color casts from white or gray parts of the picture or modify the overall picture lightness and highlight details. Give it a try! The RGBW color space opens a new dimension in picture manipulation.
HSL separation in RGBIf you are looking for a complete separation of color and luminance the HSL color space is the one to work in. However, in Photoshop it is not really supported well and its hard to work in it. You can use the second macro called HSL separation. It is provided with the above given download file, too. It separates color (hue and saturation) and luminance in different layers and allows separate manipulation of these components while still in RGB color space.
Your feedback is welcomeYour comments, ideas, success stories and questions is always welcome. Feel free to write to .
© Dr. Reinhard Kietzmann, Berlin, October 2008. www.reinhard-kietzmann.de
zurück zum Seitenanfang> zurück zur Artikelübersicht