For those who want to test their perception of colour, I made a little game called "What's My JND"

https://www.keithcirkel.co.uk/whats-my-jnd/?r=ARUjKP__-ve-

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus

What's My JND? 0.0066
Can you beat it? https://www.keithcirkel.co.uk/whats-my-jnd/?r=ApQfKP_97_vA

My opinions are legally binding!

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@piggz @hp annoyingly good. The best score I've seen someone get is 0.00018.

@keithamus @piggz I assume that the quality of your monitor also plays a role. Both having a really good and a really bad monitor might help? A really bad one might exaggerate color differences through clipping?

I do have >=150% vision. But I think that doesn't necessarily translate to color vision. But I'm not an expert.

@hp @keithamus I have to wear glasses pretty much all the time, so I guess that doesnt translate to colour perception. My monitor is just a midrange HP laptop, some of them we're really touch and took some pondering time.
@hp @piggz The display you're using definitely matters. I suspect having an old school TN panel and looking it at it from a really sharp angle might distort enough to see the difference, but also those panels could barely show all of sRGB so might not cut it.
@keithamus @piggz well I guess then I'm sorry about my annoyingly good score? 😅
@hp @keithamus How does the number of rounds work, i see people doing different numbers?
@piggz @hp There are 6 "rounds"/steps. If you're at round 6 and get one wrong, it'll back off to round 5, to gain confidence that you didn't just fluke it. Each round lowers the threshold. But there's a cap of 40 trials.
@keithamus I got 0.0012. I have asus vy279 monitor.
@hp @keithamus I got 0.0098, which seems worse than most people's scores. I was using an M1 Macbook Air.
@astrashe @hp @keithamus I wouldn't worry. You also have to factor in that people who have a worse result are also less likely to publicly share it.
@astrashe @hp @keithamus happy to share my 0.0098 also.
@hp @keithamus
I got 0.0029 ... towards the end there seemed a mixture of pretty easy and downright impossible! 😂
@keithamus Also great for suddenly seeing how dirty your screen is 🤢
@joface this is the one consistent piece of feedback. Maybe we should start a support group.
@keithamus @joface I pulled out a cloth and wiped my screen down to get the optimal score I could 😂 glad it's not just me
@Lukew @keithamus @joface I was complaining about my dirty screen protector… when I realized the problem was really my glasses. 😳😬

@a
Adding my own for posterity, using a Viewsonic 27" monitor. WIll have to try it at home on my other monitors for comparison.

What's My JND? 0.0039
Can you beat it?

https://www.keithcirkel.co.uk/whats-my-jnd/?r=AYUgKP__7b8y

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus What's My JND? 0.0032
@johannes it's a great score! The best I've seen is 0.0018.

@keithamus Switched screens, and managed to get it down to 0.0026. Maybe I need to clean the screen.

Addictive 😅

@keithamus On an iPad Pro with Nanotexture Display + Paperlike protective film (which looked really dirty on some colors and created some "noise"). Touch input using Apple Pencil, so sometimes some px off.
@keithamus https://www.keithcirkel.co.uk/whats-my-jnd/?r=AVQgJv__-znc
Damn, it made me realize how much uneven my main work screen is, lol.
EDIT: oops, wrong url
What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus What's My JND? 0.0038
Can you beat it? https://www.keithcirkel.co.uk/whats-my-jnd/?r=AXgfKP___DbL

I guess me being a graphics programmer pays off. Though at some point the dust on my monitor has more contrast than the colors :D

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@lisyarus @keithamus I found that when the response flashed at me it was like a glare stain and took forever for my eyes to try and adjust (it’s nighttime here). I have glare sensitivity and have to wear shaded lenses even indoors depending on the light. I did ok, I think, for an old lady.

@keithamus Some time past half I had to clean my screen as it was hard to tell if it was an edge or just a smudge. Path 2/3 I swear some colours were like one degree away in a single channel, ain't no way anyone can see that.

https://www.keithcirkel.co.uk/whats-my-jnd/?r=ALAjKP__-P_e

@keithamus What's My JND? 0.0028
Can you beat it? https://www.keithcirkel.co.uk/whats-my-jnd/?r=ARUiKP__93yv
Now, I suspect the monitor you are using can have an impact. I was using my phone, which is a Samsung Galaxy S24, and I increased brightness to about 85% in a not very bright room. Also, in one of my correct answers, I was bet unsure, and my guess was quite off, but counted as correct, so probably some luck there.
What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus I'm kind of confused as to what you're meant to input when you notice the difference?
@emkingma you're supposed to click the area where you think the background transitions from one colour to the next. It could be clearer, but I am not sure how best to make it clearer.

@keithamus thanks, that's all I needed to know.

Mine's 0.0062
https://www.keithcirkel.co.uk/whats-my-jnd/?r=AmghKP__7l-8

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus @emkingma I feel like it needs a little instruction "Click on the line between colours" displayed before your first click. I was confused until I'd got the first one wrong.
@harry_wood @emkingma Good feedback. I've updated the intro wording.

@keithamus What's My JND? 0.0047
Can you beat it? https://www.keithcirkel.co.uk/whats-my-jnd/?r=AdUfKP__7zSz

Not a true representation as i realised i could move the wondow around my screen and the colours changed... lol!! Might be time for a new monitor...

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus got 0.0018, but I'm also on an HDR display and was wondering if that helps with getting a lower value there.
@karolherbst they're all in sRGB so HDR _shouldn't_ matter, but having a good display which can represent all of sRGB certainly will matter.
@keithamus @karolherbst on some systems, SDR is stretched to fit HDR, which can increase the gap between two colours.

@keithamus so it seems like there is an issue with the test. There are certain rounds that only have a single color across the entire page, e.g. "background: linear-gradient(to right, rgb(149, 126, 211) 24.8914vw, rgb(149, 126, 211) 25.1914vw); display: flex;"

I verified with a color picker on my desktop and it also displays the same color across the entire range.

Though I'm aware that there is just a limited range for colors, but was wondering if that's intentional behavior.

@karolherbst on one hand, I could sound really clever and say these rounds are designed to be a false test to trip up the cheaters. On the other hand I could admit that my failure as a programmer and say that these are likely rounding issues...

I've not yet decided which reply to go with.

@keithamus both would be very credible, but having a "I don't see any difference, it's the same color" button might spice this up a little...
@keithamus This was fun. My partner got 0.0018!

@keithamus Well I guess it's a draw.

The dirt streaks on my screen kept throwing me off while searching for the disturbance in the force.

@keithamus
Light colors fooled me, especially greens. I'm marking this down to my display not being able to display these nuances, never ever my eyes (note that this sentence contains sarcasm)

What's My JND? 0.0032
Can you beat it? https://www.keithcirkel.co.uk/whats-my-jnd/?r=AUAgKP__7zj1

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@keithamus

damn, got a very low score somehow. 0.0012. i can't figure out how to play it, a tutorial would be nice. does the click need to be pixel-perfect? even if i click dead-on, sometimes it says "Precise" but sometimes it's just "Correct".

https://www.keithcirkel.co.uk/whats-my-jnd/?r=AHogKP___ftI

Edit: or maybe it counts time. it took me almost a minute on some screens.

What's My JND?

Find your Just Noticeable Difference in colour perception. How small a colour difference can you actually see?

@orsinium The messages are just random flavour text, and there's a fudge factor - you don't have to get it exact. Time isn't a factor, I didn't want people to be stressed out by it.
@keithamus Some messages were so flavoured that I clicked wrong because of laughing. 😉 @orsinium
@keithamus cool utility! 0.0037 on a pixel 9 pro in a sunny room. Do you know if age is a factor for color perception? Would have to think sex is, if xkcd has learned me anything
@tal age must be a factor, anecdotally, sex seems to be. Also the quality of display definitely matters, but worse displays might be better to an extent!
@keithamus @tal I always thought that comic was about how m/f genders are _taught_ to name colours, rather than the physical perception of them
@Rubelicious I think it can be both, of be open to the idea of neuroplasticity in color perception.
@tal this is one of the most fascinating threads I've read in a long time 😊
@keithamus I'd willingly share those details if you ever want to add the option
@tal ah it's mostly for fun, I'm not tracking anything (other than visits to my site). I'm not what I'd do with the data if I had it!