"Do you agree that there are thousands of these combinations that can all be considered yellow?"
Agree.
What is in my mind and how to narrow the results:
- If I take a A4 (210 × 297 mm) paper and I make in Excel cells 5x5mm, then I might get (210/5) x (297/5) ~ 43x59=2537 variations of yellow. Maybe 4 of such pages with 10.000 variations of yellow is good enough for me.
- What does it mean variations? Up to what extent should we change R, G, B and to say that is still yellow?
The only “natural” way to describe it, and I have no idea now how to translate it in computers interpretation of the color as numbers RGB, is to think at the wide spectrum of the solar white light and then the use of a Newton’s prism experiment (
https://www.khanacademy.org/partner...ce/mit-k12-physics/v/newtons-prism-experiment ) to separate the colors, then we get the rainbow. Let’s think at the colors in the rainbow as individual colors and not variations of the yellow, even if using RGB system could reproduce some of them.
What is the purpose of it?
- I have in my hand a yellow paper and I would like to try to print “the same” yellow color on another A4 paper. By “the same” I understood the best approximation that I can get with my non-professional LaserJet printer. I just want to come as close as possible to the original yellow paper that I have in my hand, but it does not need to be a “industrial” grade with very high precision of matching.
How do I want to do it?
- I just want to try few pages of yellow color with many small spots, rectangles and compare with the original yellow paper. If few smartphone cameras + few scanners + few naked eyes cannot make the difference, then is good enough.
- The idea is to scan the original paper and the reference paper with many variations of yellow and to compare them. The scanning is done with the same devices, under the same conditions of light, distance. I do not need to know what color code that on the screen is or in reality and I do not care as long as they match close enough.
- I do not have a better description of what I want to achieve, because I do not have better possibilities to achieve better results.
If in Excel with VBA (or Photoshop) I run 0 to 255 on color variation on each cell and then play also with (hue or saturation or luminance) around the yellow RGB (255, 255,0) then I think would be a start. But I am just wondering if it could be perhaps more possibilities.
If you tell me that yellow means the spectrum of light with wavelengths between 570 and 590nm, then I do not care, because I have no means to measure it and it does not help either to achieve my purpose.
Maybe put it in other words, to describe better “the rabbit hole”: I am trying to implement an approximation of color management for potentially non-color-managed devices using a non-calibrated tool like a smart phone camera with a non-standard light source.
I try to rely on the fact that the same scanning device gives a systematic error, as an offset if you want, even if autofocus may introduce errors, but I have no other better way to do it except to use several devices to compare the results, given the fact that smartphones, scanner and PCs are everywhere nowadays.
The method above does not tell you what yellow color code is, it just tries to “clone” the original colored yellow paper.
If I may consider the error from these scanners constant and if several devices which compare the original scanned paper and the new printed yellow scanned paper declare the same yellow (even within small variation) and then those papers are also compared by few pairs of naked eyes in solar white light, but not direct sun light, and cannot make the difference, then is good enough for me, I achieve my purpose and I am out of the “rabbit hole”.
Now, you understand why I need such A4 paper with many variations of yellow.
The Excel file with VBA code provided below is a start, but how to get more variations by playing with R and G and not only with B, in such way that we do not get into other colors detected by our eye as one from rainbow spectrum for example?
We cannot have green or red or blue or black or white and say that is a variation of yellow, because even if for the computer is, for my eyes is not.
How to limit the RGB numbers variation and to stay within yellow nuances, which seen by our eyes is still declared yellow? I realize is a problem also from person to person how color is detected, but I just speak in general referring at the fact that I do not want to R and G and B numbers to be changed in 16 millions possibilities, which represent all the colors.
Perhaps a simple approach is to start with RGB(255, 255, 0) and run 10.000 variations of yellow around this point without breaking the rainbow rule.
Any idea how to implement it?
Attached is an Excel file with small variations of yellow done in VBA.
