How reliable are the colour blindness previews in Illustrator and Photoshop?

Question

Adobe Photoshop and Illustrator have options to preview what artwork will look like to people with red/green colour blindness (protanopia and duteranopia). How reliable are these previews?

For example, I produced the images below using Photoshop's colour blindness simulation. The red and green tiles are still somewhat distinguishable. Can I rely on people with protanopia or duteranopia being able to distinguish those colours?

I'd also like to know how much these conditions vary by person. Are these conditions that have a range of severities, where some people with protanopia are better able to distinguish red and green than others? Or are they more constant, where two people with protanopia would perceive colours the same way?

Full colour

Duteranopia

Protanopia

Answer 1

I was also quite convinced there were three 'pure' types of color-blindness, but apparently it can range quite a lot, not only because of the type of CB, but also because of environmental conditions, light amount of light.

The reason for color-blindness is 'a faulty eye cone'. There are three cone types that are used to perceive light colors, but for some people one type of cone perceives light slightly out of alignment. The three types of CB you see described in Adobe correspond to each of the three cones that can be faulty.

The effects of anomalous trichromatic vision can range from almost normal color perception to almost total absence of perception of the ‘faulty’ color.

(...)

About half of people with anomalous trichromacy will see the world in a similar way to those with dichromacy but their ability to perceive colours will improve in good light and deteriorate in poor light.

(...)

People with anomalous dichromacy can have either inherited colour blindness, in which case their ability to see colours will remain the same, or they can have acquired it, in which case their condition could get worse, or possibly improve over time.

_Source

There is general agreement that worldwide 8% of men and 0.5% of women have a colour vision deficiency. The 8% of colour blind men can be divided approximately into 1% deuteranopes, 1% protanopes, 1% protanomalous and 5% deuteranomalous. Approximately half of colour blind people will have a mild anomalous deficiency, the other 50% have moderate or severe anomalous conditions.

So it seems like there are indeed different degrees of color-blindness, that are not accounted for in this preview. It is, however, a very valuable tool, as whatever works for a full cone deficiency will work for a milder color-blindness.

Answer 2

There are many of types of color blindness, whereas red (protanopia, from prot+an+opia) and green (deuteranopia, from deuter+an+opia) color deficiencies are the most frequent.

The genes for green and red sensitivity are located in the X chromosome, this is why men are more prone to be color blind then women (women have two X chromosomes - one from mother and another from father, whereas men has only one X, thus the defects in one X received from mother are expressed. Why from mother? Because, father gave Y chromosome to a boy).

There are many simulators of color blindness. Let's check with the picture provided if the Ai gave similar images to simulator.

Look at the pictures -

As you can see Illustrator gave images sufficiently close to the simulator images in the case of established color blindness. Thus, we can stay that the simulator of Illustrator is close to the independent one.

Although, we don't know the exact calculations behind the scene from this small study we can assume that probably Ai simulator is sufficient for color blindness demonstration.

Protanomaly and deuteranomaly are mild color sensitivity defects and should be less relevant than established defects.

There are a lot of sources describing theories of color blindness and from these one could built its own simulator.

Common picture of confusion lines (from this excellent resource) -