I know multiple programs that calculate colours based on a percentage. I really love when colours are displayed this way as they seem to fade in a non-luminous dull way that is really indicative of a less bold style.

I am just wondering if anyone knows how these new faded colours are calculated based on a starting colour (RGB255 code) and how I might calculate these colours myself?

I am in particular looking for a good faded green that would look good when displayed as minor contours and I feel knowing this process may help me find one. (If you have a suggestion for a colour, please share!)

There are several ways. By mixing the colors, mathematically:

``````(1-a)*FGcolor+ a*BGcolor
``````

Would mix the colors as if fg color was alpha blended to bg color. You could blend to white black or any other color. the vairable a i between 0 to 1. Example in python:

``````from Tkinter import *

def ablend(a, fg, bg):
return ((1-a)*fg[0]+a*bg[0],
(1-a)*fg[1]+a*bg[1],
(1-a)*fg[2]+a*bg[2])

root = Tk()
num=8.0

fg=(100, 200, 170)
bg=(255,255,255)

for i in range(int(num)+1):
col = ablend(i/num, fg, bg)
e = Label(root, text=" "*70, background='#%02x%02x%02x' % col)
e.grid(row=i)

root.mainloop()
``````

This would result in a window that blends colors. You can freely change the bg color.

Image 1: result of code

You could also blend in other ways. like rgb/hsb/hsl interpolation. Sample using hsv interpolation:

``````from colorsys import rgb_to_hsv, hsv_to_rgb

# replace ablend with hsvblend
def hsvblend(a, fg, bg):
fgh = rgb_to_hsv(fg[0]/255.,fg[1]/255.,fg[2]/255.)
bgh = rgb_to_hsv(bg[0]/255.,bg[1]/255.,bg[2]/255.)
ret = hsv_to_rgb(
(1-a)*fgh[0]+a*bgh[0],
(1-a)*fgh[1]+a*bgh[1],
(1-a)*fgh[2]+a*bgh[2])
return ret[0]*255,ret[1]*255,ret[2]*255
``````

And so on…

Image 2: Using hsv blending

And here is a bonus image showing nonwhite bg that behaves a bit better than fully saturated white.

Image 3: Showing better of the differences between the modes (white and black is problematic for hsv)

PS: sorry for extremely messy code