In computer graphics there is a long tradition to do rendering by computing images from geometric and illumination models. This can be done by various ways eg. raytracing.
Even though the illumination models have become increasingly complex and physically correct, the images still have some plastic feeling about them. How come?
When I started my PhD studies in 1991 I began to look into this. I soon became convinced that the fault was in the RGB calculations.
RGB stands for Red, Green and Blue primaries. It stems from the awareness that the eye uses a few primary sensor channels. By mixing just a few primary colors all other perceivable colors can be created. Thus by using red, green and blue light most of the colors the eye can see can be created.
The illumination models are based upon knowledge how light are reflected/transmitted/absorbed by physical surfaces/compounds. The calculations are approximations to what happens with real photons.
In computer graphics one have always made the assumption that illumination models are linear and orthogonal in the RGB space. Thus everything is correct. But it isn't!
What I put forth in my reports are quite revolutionary for the folks in the computer graphics area, but also for the traditional graphic arts, printing (more so in desktop publishing and digital press). What I have PROVEN is that it is ALL WRONG.
Please read my reports:By clearly writing "The text is Copyright (c) 1991 by Clas Tegenfeldt" you may use it, spread it, refer to it, or make citations from it. Do not make local copies on the web but please feel free to make links to this page.
Back to http://www.bemi.se
Last updated: