Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Color Test

xrite.com

Post reply on HN
Print view View original

31–40 of 48 posts

Re: Color Test

#31

Post by onli »

sudont wrote: post #22
onli wrote: post #20

That is a bad designed test, for all the reasons you learn in psychology when talking about experimental test design. Most importantly: How is this thing scored? Imagine you have everything but one block in the right order. Now every element following that false positioned block is one apart from the right position. Are they now wrong as well? Or are they scored as correct as they are in the correct order in itself?…

Most likely you could score with a levenshtein distance. http://en.wikipedia.org/wiki/Levenshtein_distance

Good idea.

Well, it all depends what you want to test. Normally, such an experiment should try to detect which colours when compared can still be distinguished, detecting the smallest notable difference/patterns which don't work for subgroups. I don't see how that could work well with the measurement.

If it really is about the whole row of colours, then the levenshtein distance could indeed work well.

Re: Color Test

#32

Post by OMBUG »

bnegreve wrote: post #24

From what I understand the score should be in the range [0, 99], but I got that on the result page: > Best score for your gender and age range: -160 > Highest score for your gender and age range: 444445389 A bug, I assume.

Just clicking through without doing any reordering got me 1042 so I don't think it's in the [0,99] interval. I'm assuming 99 is just an arbitrary "you have terrible eyesight" cutoff.

Re: Color Test

#33

Post by JoelSutherland »

zeckalpha wrote: post #27
gavinpc wrote: post #23

Earlier quoted context omitted.

Also remember to turn off f.lux!

That shouldn't make a difference. If anything, it should help as it would better match your screen to the ambient light, allowing you to make better choices. I've heard a lot of people in graphics who use f.lux or redshift only to turn it off when they do graphics work, but I find more accurate results if it is left on. This calls for another experiment!

It will make a difference. Your eyes can see a much wider color space than your monitor can display. By shifting colors, f.lux leaves some of the monitors color space unused, compressing the remaining colors so that some are closer together (or the same even).

Re: Color Test

#35

Post by zipppy »

gueno wrote: post #17

When inspecting element I saw IDs like "patch_ROW_COLUMN" in container, just put them in order of IDs and you should get a perfect score. I got 8 instead of 0, bug ?

I scored a 0, so it is possible; there may still be a bug (perhaps my order was wrong but it still gave me a perfect score for some reason), but it is possible.

Re: Color Test

#36

Post by Dogamondo »

Last time I did one of these I nearly went deaf. (If anyone remembers the screamer at the end of the famous flash based one about 10 years ago..)

Re: Color Test

#38

Post by bestest »

sorry, could not resist, it's _hacker_ news after-all isn't it:

http://www.colormunki.com/game/huetest_kiosk (that's the source of the test iframe), open console, paste, enter, submit score:

var rows = document.getElementsByClassName('drag_container'); for(var i = 0, l = rows.length; i < l; i++){ var row = rows[i]; var items = row.getElementsByClassName('drag_patch'); items = Array.prototype.slice.call(items); items.sort(function(a, b){ return a.id.match(/\d+$/)[0] - b.id.match(/\d+$/)[0]; }); row.innerHTML = items.map(function(a){ return a.outerHTML }).join(''); }

Re: Color Test

#39

Post by nilkn »

sudont wrote: post #22
onli wrote: post #20

That is a bad designed test, for all the reasons you learn in psychology when talking about experimental test design. Most importantly: How is this thing scored? Imagine you have everything but one block in the right order. Now every element following that false positioned block is one apart from the right position. Are they now wrong as well? Or are they scored as correct as they are in the correct order in itself?…

Most likely you could score with a levenshtein distance. http://en.wikipedia.org/wiki/Levenshtein_distance

It seems this would not fix the issue of having one block off and thereby pushing a bunch of other blocks one off from their right position. You'd still have to move all those blocks back, which means the Levenshtein distance would not be short.

Re: Color Test

#40

Post by StavrosK »

nilkn wrote: post #39
sudont wrote: post #22

Earlier quoted context omitted.

Most likely you could score with a levenshtein distance. http://en.wikipedia.org/wiki/Levenshtein_distance

It seems this would not fix the issue of having one block off and thereby pushing a bunch of other blocks one off from their right position. You'd still have to move all those blocks back, which means the Levenshtein distance would not be short.

The levenshtein distance of "abcdefghijklmnopqrstuvwxyz" and "acdefghijklmnopqrstuvwxyz" is 1.
Post reply on HN