Color Blindness Test based on Confusion Lines of the CIE 1931 Color Space

This is the third and last part of a series, where we have a closer look at a color blindness test which is based on the confusion lines of the CIE 1931 color space. In the first part I introduced the CIE 1931 Color Space and the second part was looking a bit deeper into the theory of Confusion Lines.

Most tests which check for color blindness are based in some way on confusion lines. But I would like to focus your attention on this test which is regenerated for every single trial based only on confusion lines.

The author of the test describes the design for his color vision test as follows:

  1. For each color vision deficiency (protanopia, deuteranopia or tritanopia) test, five confusion lines are selected randomly. Three points on each confusion line that can be distinguished by the normal observer are chosen.
  2. For each test, a confusion line is selected randomly. A point is selected randomly among the set of three. It is displayed in the upper test panel at a random luminance in the form of random dots. The three points are displayed in the lower three panels at a random luminance and also in the form of random dots.
  3. The subject is asked to check one of the lower panels which color matches that of the test panel.
Color Blindness Test - Example Screen
Color Blindness Test – Example Screen

As I described in part two of this series, colorblind people can distinguish only a handful of wavelengths of the color spectrum. Compared to normal vision it is around 10% to 20%. But if you are affected by color blindness you automatically start to interpret a difference in lightness as a different color. This makes you think to see more colors than you actually can distinguish.

If we look at the description of this color blindness test this is exactly the point of it. It plays with the lightness and adjusts the three choices to the same lightness, which makes it really hard to get the match if you are colorblind.

I took the test. It showed me cruelly – cruelly again, that I am very much colorblind. But let us have a look at my results. I took each color blindness test three times and completed each time 50 decisions. What is amazing to me is, that in all three tests the results didn’t differ a lot, or better said almost none. Let me start with the best results:

My Tritanopia test results

Run Correct Wrong Correct%
1 47 3 94%
2 49 1 98%
3 47 3 94%

Most of the time I could tell a difference between the three choices. But still I had to guess every single time. I suppose the problem is that I can’t name the colors I see, I just see different colors and nothing more. (→ Tritanopia Test)

My Deuteranopia test results

Run Correct Wrong Correct%
1 31 19 62%
2 30 20 60%
3 35 15 70%

The results are much worse compared to the first test. Although I still could distinguish usually one of the three colors from the other two it was more of a guessing game than really knowing anything. (→ Deuteranopia Test)

My Protanopia test results

Run Correct Wrong Correct%
1 26 24 52%
2 24 26 48%
3 25 25 50%

This color blindness test was pure guessing. I couldn’t tell a color from the other and I am very much amazed of the results. The lightness adjustment made it impossible to me to match them correctly. (→ Protanopia Test)

The results of this color blindness test in three parts approve the outcomes of other tests. I am protanopic or at least affected by a very strong protanomaly and as the doctor would say: “You are completely colorblind.”

More information about the color blindness test based on confusion lines and the online tests itself can be found at the Color Vision page of biyee.

This was the third and last part of this series. The other parts can be found at
Part 1: CIE 1931 Color Space
Part 2: Confusion Lines

27 responses on “Color Blindness Test based on Confusion Lines of the CIE 1931 Color Space

  1. Peter

    That is a cool test. Thanks for writing it up so thoroughly. Interesting to point out that you actually did significantly better than chance on the protanope test, as chance would be 33%. I’m a protanope too, so I’ll post my own results soon to compare.

    Also reminds me of this paper:

    Maybe I should email the author and ask for his code.

  2. Daniel Flück

    Peter, it would be nice to see your results as well. I definitely was astonished to see my actually good results in the deuteranopia and protanopia tests.

    Thanks for the link. Unfortunately it’s quite expensive for a single subscription. Would be great to here from you again (I haven’t forgotten about your page…).

  3. Garry King

    I would really like to do these tests (Tritanopia, Deuteranopia and Protanopia Confusion Lines), but the links are broken.

  4. Daniel Flueck Post author

    Garry: the server was for a few hours not available. But is up and running again. I hope you can enjoy the color blindness test know.

  5. Daniel Flueck Post author

    Jean, thanks a lot for posting your link. I’ve already seen your color blindness test and am planning to write an article about it in a future post.

  6. George

    I am completely guessing in the tritanopia test! how do you pass it?, are there any other tritanopia tests and do you have any suggestions as to why I can’t see yellow highlighters?- I am a tritanope by the way.

  7. Daniel Flueck Post author

    George, Tritanopia is very uncommon so there aren’t many tests available. As for online tests up to now I only of this test described here.

    Unfortunately I can’t tell you why you can’t see yellow highlighters. Maybe it’s because yellow looks more bright to your eyes. I just now that for us red-blind people, red looks much darker and can’t be distinguished from black very well sometimes.

  8. Rami

    I did your test. While I was doing the tritanopia test I had some difficulty and thought that I am going to have some false answers but to my surprise I got all the 41 answers true which means 100%. I did 10 more and still got 100% of them right. I than did the deuteranopia test and got 53 true answers and 11 false answers which makes 82% right. On the protanopia test I had 67 correct answers and 23 false ones which makes 74% right. While doing the protanopia test there was places where I had a lot of troubles and I am suspecting the fact that my intuition helped me. I don’t know what can I conclude. I am pretty sure that there is nothing wrong with my screen because I already did Ishihara online tests and have asked people to tell me what they see, they could see a lot of things I couldn’t and vice-versa. What I want to know is how can I interpret the results of these tests. Can I say that I have a protanomaly?

  9. Rami

    One last thing, how strong is my colourblindness considered? i.e in which of these category would you place me: mild, moderate, strong or absolute?

  10. Rami

    This moorning I did the test again without spending too much effort trying to concentrate, my results got worth. 39 correct v.s 20 incorrect on the Deuteranopia test which makes 66%. 33 correct v.s 24 on the protanopia test which makes 57%. The protanopia test seemed more strange to me. I don’t know how valid is this (see the middle of the page or a bit before in the paragraph or sub-paragraph entitled dichromacy) but I can tell that I don’t see at all the 37 no matter how hard I try, not even a red trait, the 49 is pretty clear though I see a 47, but the 56 is very clear. What do you think?

  11. Daniel Flueck Post author

    Rami, thanks for sharing all your results. My results also get worse if I don’t concentrate to much. As for the Wikipedia site, I can’t see the 37, I have only a very slight feeling of a 47 and I can see a little bit the 56. This just for you to compare with me. I suffer a strong protanomaly or protanopia and I also would like to know it for sure. Unfortunately this can only be done with a Anomaloscope test a certain eye specialist centers.

    I can’t tell what exactly you’re suffering from. But I would say it is some mild to strong from of protanomaly.

  12. Dr. T

    Jean, I checked your test and it looks very good to me. It confirmed the strong Protanomaly I had derived from the other tests, but giving a more clear result. Giving the correct results and the meaning of the wrong ones is a nice thing. Minor problems: Some of the translations in the German version did not work and the color-box questions would be easier with a fixed order of the colors in the proposed answers.

  13. Kate Dicey

    My test results came out at:

    Deuteranopia: 97%
    Protanopia : 99%
    Tritanopia: 97%

    What does this mean in terms of my colour vision?

    I am dyslexic. I find it easier to read dark blue lettering on a very pale blue background, or dark red or brown on a cream background, than black on a stark white background.

    I used to teach English language and Literature. Over the years I had a number of students with dyslexia and other reading disorders, and taught kids with blue, green orange, and even red spectacles.

    Have you ever found a correlation between learning disorders such as dyslexia and colour vision anomalies?

    Nowdays I work as a dress and costume maker, and use colour a lot in my work.

  14. Daniel Flück Post author

    Kate, this means you’re not colorblind (or only very very slightly) as you could match almost always the correct colors.

    Concerning dyslexia, I’ve never read of any relation between this and color vision. I also don’t believe that there is any, as this or some physically different disabilities.

  15. Kate Dicey

    Daniel, I didn’t think there was a correlation, but it’s great to have that confirmed. It’s also good to know that my colour vision is good! It’s bad enough being so short sighted I can only just see past the end of my nose! ;)

    I have seen research that says that dyslexia is more a ‘processing malfunction’, where bits of the brain are wired differently from a ‘normal’ brain, rather than a problem with the eyes themselves. I just wondered what correlation there might be between the two as so many dyslexics can be helped by coloured spectacles.

    Thank you for your time. Good luck with further research.

  16. Sean L.

    Thanx for the test. I’ve always been curious about my color blindness. Now I know suffer from Protanopia,40%,42%, 40%. Deuteranopia test was a bit better, 65%, 55%, 55% even though I was just guessing. My Tritanopia was better actually knowing some were right and not just guessing, 95%, 85%, 95%.

  17. Clive

    A very interesting test. My results were as follows:


    1st test: 46 correct, 4 incorrect, 92%
    2nd test: 43 correct, 7 incorrect, 86%
    3rd test: 47 correct, 3 incorrect, 94%


    1st test: 27 correct, 23 incorrect, 54%
    2nd test: 32 correct, 18 incorrect, 64%
    3rd test: 31 correct, 19 incorrect, 62%

    (or as I like to call it: guessing)

    1st test: 21 correct, 29 incorrect, 42%
    2nd test: 21 correct, 29 incorrect, 42%
    3rd test: 24 correct, 29 incorrect, 48%

    All in all the decline is consistent with my diagnosis, and the tests are remarkably consistent.

    My less than half score in protanopia is interesting as it is higher than I had expected since the I’d have estimated my score to be more random than that since I couldn’t tell the colours apart in that test.

    I shall be interested to repeat this experiment again to see if those numbers hold up.

  18. Clive

    I just took the test at Jean’s site and the results were:

    Deficiency likelihood: 73%
    random 0/24
    similar 0/24
    deuteranopia 0/19
    deuteranomoly 2/7
    protanopia 7/15
    protananomoly 1/7
    tritanopia 0/15
    trintananomoly 0/7

    deuteranomoly 29%
    protanopia 44%
    protananomoly 14%

  19. Clive

    I retook these tests. I altered one thing. I altered the gamma level from 2.2 to 5.0, as there seems to be a correlation between protanopia and brightness for distinguishing different hues (not colours) on that basis I wanted to see if increasing the gamma would make any significant impact on my scores.


    1st test: 46 correct, 4 incorrect, 92%
    2nd test: 50 correct, 0 incorrect, 100%
    3rd test: 49 correct, 1 incorrect, 98%


    1st test: 36 correct, 14 incorrect, 72%
    2nd test: 38 correct, 12 incorrect, 76%
    3rd test: 32 correct, 18 incorrect, 64%

    (or as I like to call it: guessing)

    1st test: 22 correct, 28 incorrect, 44%
    2nd test: 27 correct, 23 incorrect, 54%
    3rd test: 25 correct, 24 incorrect, 51%

    Increasing the gamma has seemed to have had a positive effect on my scores all round. I even scored above 50% on two of the protanopia tests!

  20. Josh

    Wow. People at school told me I should get checked because i mix up blues and greens and sometimes yellow and purple. So i looked for a test online and found this. Thanks alot.

    Tritanopia: 13 right, 20 wrong, 39%
    Deuteranopia: 27 right, 1 wrong, 96%
    Protanopia: 20 right, 2 wrong, 90%

  21. Andy

    My Results:

    Tritanopia – 86%
    Deuteranopia – 61%
    Protanopia – 49%

    thanks for the test. I was having trouble determining whether i am deuteranopic or protanopic.

  22. Nagesh

    My results
    Tritanopia – 93%
    Deuteranopia – 43%
    Protanopia – 67%

    Great test.
    Deuteranopia test was a big pain. Complete guess work. But still I can’t name a single color in all the three tests. Is it common for all color blind people.

    Any how gr8 site. Keep up the good work.

  23. Darryl R.

    I found all these tests confusing and was surprised to see how well I did with the Tritanopia when it seemed so difficult, but no where near as difficult as the others. It seems like there should be a “no match” option and an “all the results look the same” option. Plus sometimes the choice appeared to be missing. I’m assuming I just couldn’t see it.

    Tritanopia – 92%
    Deuteranopia – 21%
    Protanopia – 48%

  24. Ian

    Tritanopia 59 correct 3 incorrect – 95%
    Deuteranopia 44 correct 4 incorrect – 91%
    Protanopia 66 correct 4 incorrect – 94%

    The online D-15 (took mulitple times) says I have normal color vision; however, I fail the PIP test everytime and I recently was unable to take a certain job because I barely failed the FALANT test (Farnsworth Lantern).

  25. morri

    The test is strange , as it shows either the exact right colour or only different shades of a clour(aka bright gree dots and then a choice of normal gree, petrol green and forest green dots.. )
    or blue dots and then a tan coloured dot a purple dot and a green dot , and then again it shows exactly the rigth dots..