(Peter Wollenberg) wrote:
……….
Seeing what you actually want, the solution is quite simple:
RGB model, white Background
Layer 1: shape with 255 R 255 G 0 B; mode = Normal
Layer 2: shape with 255 R 0 G 0 B; mode = Difference
Layer 3: shape with 0 R 255 G 0 B; mode = Difference;
that’s all
BTW, IMHO you have a misconception of "adding" colors. You want the colors to subtract, not to add! CMY are the subtractive primaries and black represents zero luminosity.
Thank you for your interest. I know what you mean about colours adding, but for me school paints came first, so I normally think blue plus yellow equals green, etc.
To try to get a better idea of what was going on, I set up three experiments.
1. I set up a new file, white background and RGB colour mode. Then I created three overlapping circles on separate layers, set each layer to ‘Difference’, and coloured them respectively 100% red, green and blue. Rather to my surprise, each layer came up as the complementary colour, but these combined as I expected, with true black in the middle. However the individual colours appeared rather desaturated compared with the following experiments.
2. I set up another new file, as before, but this time I set each layer to ‘Multiply’, and coloured them 100% cyan, magenta and yellow. This looked much the same as the first diagram, except that the individual colours appeared more fully saturated, but the black in the middle was less saturated.
3. I set up a third new file using CMYK colour mode, and used the same arrangement of layers as in experiment 2. This time the black was fully black, and the reds greens and blues appeared stronger than in experiment two.
When I sampled the colours in each region of each image, I found that in experiment 1 the primary colours were each 100% of the opposite two colours (ie the nominally red region was 100% green and 100% blue), the secondary colours were 100% of a single colour, and the triple colour was true black.
In experiment 2 each region was the appropriate combination of C M & Y, but these were achieved by various combinations of R, G & B, but generally at significantly less than 100%. This made the single and double colours appear more saturated than in experiment 1, but the triple colour had quite large percentages of R, G., and B, so that it was really a dark grey.
In experiment 3 (with the CMYK colour scheme) the double and triple colours had significant components of K. In particular the triple colour had 90% K, and R,
G., and B were all zero.
When I used the techniques of experiment 2 for an actual test case, I found that it did not work nearly as well as I had hoped, because I tended to see the overlap region as a separate shape, so that was difficult to visualise the two shapes I was actually comparing. So now I am wondering about alternative ways of achieving a similar result. I suspect I might do better if I drew constant width outlines around each shape, in different colours, and only displayed the outlines.
Anyone know a good way to draw a constant width outline around an irregular shape?
James McNangle