# The Naked Scientists Forum

### Author Topic: Colors' Color  (Read 11538 times)

#### eric l

• Hero Member
• Posts: 514
##### Re: Colors' Color
« Reply #25 on: 30/08/2006 12:13:34 »
Actually all cones react to all wavelengths !  The difference is that the S-cones react more (not exclussvely)to shorter wavelengths, while M and L cones react more to longer wavelengths.
(See again http://en.wikipedia.org/wiki/Colour_perception)

The primary colors for absorption are not linked to a specific wavelenght.  If you would measure the spectrum refelected by the yellow, cyan or yellow ink in your printer, you would actually see a graph with a maximum, but with a multitude of small peaks and valleys.
The so called primary colors are defined by centuries of experience from painters, printers, photographers...  That is also the reason why the primary colours for your printer are not the same as the primary colours for your screen.  The light emmited by the red, green and blue "guns" is closer to monochromatic light than the light reflected by a printed or painted surface with only one of the primary colors.

#### lightarrow

• Neilep Level Member
• Posts: 4586
• Thanked: 7 times
##### Re: Colors' Color
« Reply #26 on: 30/08/2006 13:41:12 »
quote:
Originally posted by eric l

Actually all cones react to all wavelengths !  The difference is that the S-cones react more (not exclussvely)to shorter wavelengths, while M and L cones react more to longer wavelengths.
(See again http://en.wikipedia.org/wiki/Colour_perception)

The primary colors for absorption are not linked to a specific wavelenght.  If you would measure the spectrum refelected by the yellow, cyan or yellow ink in your printer, you would actually see a graph with a maximum, but with a multitude of small peaks and valleys.
The so called primary colors are defined by centuries of experience from painters, printers, photographers...  That is also the reason why the primary colours for your printer are not the same as the primary colours for your screen.  The light emmited by the red, green and blue "guns" is closer to monochromatic light than the light reflected by a printed or painted surface with only one of the primary colors.
Yes to everything, excepting to: "The so called primary colors are defined by centuries of experience from painters, printers, photographers...  That is also the reason why the primary colours for your printer are not the same as the primary colours for your screen".

The main reason for this difference is that with ink or paint or pigments, you make the others colours by "subtractive synthesys", while with lights by "additive synthesys" (I hope I have correctly translated from italian).

It means that, mixing two pigments, white light hitting on this new pigment looses those wavelenghts absorbed by both pigments, so you subtract them from the white light; mixing to coloured beams of light (or putting close together two tiny coloured points, as in the monitor's screen) you have to add the wavelenghts.

Example: mixing red light and green light produces a yellow light. If you look at a brilliant yellow spot on your screen through a powerful lens, you can see it's actually made of red and green pixels.

Mixing red and green pigments gives a completely different colour (dark greenish or reddish brown). So you cannot obtain yellow colour mixing pigments of others colours.

Because of these two different ways to obtain colours, it was found that the entire visible spectrum is better reproduced (better doesn't mean perfect!), using only three lights, by mixing the (primary) colours: red, green, blue; while instead, mixing pigments, with the three (primary) colours: magenta, yellow, cyan.

In this way, every colour, made with lights or with pigments, is unambiguously determined by a three-components vector: (x,y,z) where each component represent the intensity of every primary colour.

This is very useful, e.g., in digital elaborations of photos and pictures in general, because every coloured point on the monitor's screen, or every coloured dot printed on (or read from) paper, is identified by a triplet of numbers, and so, easily elaborated mathematically by softwares.

#### bostjan

• Sr. Member
• Posts: 107
##### Re: Colors' Color
« Reply #27 on: 31/08/2006 04:29:05 »
i beg to differ.  you can start from any three colors 120 degrees apart on the color wheel and make any other color additively or subtractively, just as well.  try it out if you don't believe me.

as far as the cones reacting to all wavelengths, this is not exactly clear.  they do react to a range of wavelengths.  obviously, they do not react to infrared or ultraviolet or anything beyond that.  what i was referring to, was the maximum of each kind of cone.  each one has a global maximum sensitivity correlated to a primary color.  one of them (i forget which- violet, or maybe green) acts goofy with a couple of local maxima as well.

#### lightarrow

• Neilep Level Member
• Posts: 4586
• Thanked: 7 times
##### Re: Colors' Color
« Reply #28 on: 31/08/2006 08:07:23 »
quote:
Originally posted by bostjan

i beg to differ.  you can start from any three colors 120 degrees apart on the color wheel and make any other color additively or subtractively, just as well.  try it out if you don't believe me.
Please, try to make bright yellow subtractively, of course without using orange-yellow mixed with yellow-green! It would be cheating! And in this case, you won't be able to obtain all the other colours (but very unsaturated ones) if these two are two of the primaries. For example, with these two and blue, you can't obtain red, ecc.
About trying it out, don't worry, I have done it a lot of times.
quote:
...as far as the cones reacting to all wavelengths, this is not exactly clear.  they do react to a range of wavelengths.  obviously, they do not react to infrared or ultraviolet or anything beyond that.  what i was referring to, was the maximum of each kind of cone.  each one has a global maximum sensitivity correlated to a primary color.  one of them (i forget which- violet, or maybe green) acts goofy with a couple of local maxima as well.
Yes but not completely exact. Yes because each retinal receptor (cone) has a sensitivity spectrum (curve) with a global maximum sensitivity correlated to a primary color. Anyway, those curves, even if with very low levels of sensitivity, have tails on the left and on the right, stretching in all the visible spectrum.
« Last Edit: 31/08/2006 14:06:40 by lightarrow »

#### bostjan

• Sr. Member
• Posts: 107
##### Re: Colors' Color
« Reply #29 on: 31/08/2006 19:48:33 »
You saying you cannot subtract to yield red?!  You have not tried enough, then.  Surely, Yellow and Magenta subtracted yield red.  To get any very bright colors subtractively is difficult, but try pale green and tiny bit of red.  If you got brown, mix in white to lighten it up.

There is nothing magical about any certain color, I assure you.

And I never said that the sensitivity curve did not span the visible range, merely that it did not span outside this range.  I stated this quite clearly in the first and second sentences you quoted.

#### lightarrow

• Neilep Level Member
• Posts: 4586
• Thanked: 7 times
##### Re: Colors' Color
« Reply #30 on: 01/09/2006 07:40:32 »
quote:
Originally posted by bostjan

You saying you cannot subtract to yield red?!  You have not tried enough, then.  Surely, Yellow and Magenta subtracted yield red.  To get any very bright colors subtractively is difficult, but try pale green and tiny bit of red.  If you got brown, mix in white to lighten it up.

There is nothing magical about any certain color, I assure you.

And I never said that the sensitivity curve did not span the visible range, merely that it did not span outside this range.  I stated this quite clearly in the first and second sentences you quoted.

I said it's not possible to obtain saturated red with orange-yellow, yellow-green and blue as primaries.
Magenta and yellow, of course can make red subtractively! They are primary colours of subtractive synthesys! Otherwise, how could a printer make red as well as all the other colours?

About a brown (or dark) pigment or ink, if you add white, you obtain of course a lighter colour, but not saturated anylonger (I am assuming you know very well what "saturated" means).

#### syhprum

• Neilep Level Member
• Posts: 3822
• Thanked: 19 times
##### Re: Colors' Color
« Reply #31 on: 01/09/2006 14:13:17 »
When I first joined the colour printing business the explanation I was given for subtractive colour generation was that white light passed through the printing ink removing some components, was reflected by the white paper and then passed out through the ink again and the normal additive process then took place

syhprum

#### lightarrow

• Neilep Level Member
• Posts: 4586
• Thanked: 7 times
##### Re: Colors' Color
« Reply #32 on: 01/09/2006 15:15:27 »
quote:
Originally posted by syhprum

When I first joined the colour printing business the explanation I was given for subtractive colour generation was that white light passed through the printing ink removing some components, was reflected by the white paper and then passed out through the ink again and the normal additive process then took place
Yes, but this happens only for those kinds of printing where pigment dots never overlaps one onto another. In that case it's the same as coloured light dots on the monitor's screen forming an image, and the syntesys is additive.

You can make this simple experiment: on a white paper, draw ~ 20 thin lines of red and green colour, alternatively, one by the other, without overlapping. You should do it with bright colours.

When you have finished, looking at this drawing from a short distance, you see a reticolus of red and green lines, but from some meters apart, you see a yellowish stain.

#### The Naked Scientists Forum

##### Re: Colors' Color
« Reply #32 on: 01/09/2006 15:15:27 »