Been posting about this in another thread but thought I should make a dedicated one.
I've got a vague idea that I could set up some stepper motors to depress syringes to dispense small amounts of primary coloured paints in small and hopefully accurate quantities to closely match a colour selected from an image on a computer. With the aim of making mixing colours for acrylic or oil painting a simpler process.
Going with cyan, magenta, yellow, black, and white since that should give the widest range of colours possible from a handful of primaries.

(these are blunt needes, not much risk of accidentally injecting myself)
So trying to match this target colour with 60% cyan, 20% magenta, 60% yellow, and 0% black by the CMYK colourspace.


Dispensed .60ml cyan, .20magenta, .60 yellow, and .40 white. Or as close as I can get.

It's somewhat close. To the right is adding 20% more black at a time to 60% to match the darker shade in that earlier image.
So there's a few things I sort of understand that I need to explain, or at least state that i'm aware of. I'm aware that the RGB display gamut doesn't entirely overlap with the gamut of CMYK, and i'm aware that displays are emissive where as paints are reflective. Or additive and subtractive. And that ambient light (as well as the colour balance between cameras and displays) will make it very difficult to get an exact match. Even jpeg compression shifted the shade of that colour swatch a bit.
The other thing is that CMYK in printing works by overlapping dots of somewhat tranlucent inks, and the white is from the background. That's why im working on the assumption that the amount of white to mix in will be whatever percentage is left over from only the highest percentage of the other colours.
What i'm having a hard time getting my head around is the idea that these volumes might need to be proportional to each other in some way. And perhaps different pigment levels in each paint might affect the ideal proportions. I think more tests are in order, but i'm thinking about how I could try get some actual data to calibrate stuff with.
Mechanically i'm also considering if there's a better way to dispense. The plastic syringes tend to flex a bit and trapped air bubbles up at the plunger end don't help either. Ideally i'd want something that'd really jet the paint out to minimize strings and droplets hanging on the tips of the nozzles.