JBSchroeds wrote:nou wrote:A quick look at the spreadsheet linked above and comparing Citadel range with Vallejo game colour shows it's useless for any practical application - it returns Vallejo Blue Grey as closest match for both Celestra grey and Rakhart Flesh, two vastly different paints, and Charred Brown as closest match for both Dryad Bark and... Abaddon Black. It also indicates that
GW Dark Reaper is a layer version of both Mechanicus Standard Gray and Incubi Darkness base...
Reading methodology description and assumptions made provides a clear explanation why it's the case - website color swatches do not represent paint colors nowhere near accurately and using them as a basis for comparison makes this spreadsheet a rather pointless excercise.
For all of your examples:
GW Celestra Grey vs Vallejo
GC Heavy Blue Grey =
d4.79 "mediocre match"
GW Rakhart Flesh vs Vallejo
GC Heavy Blue Grey =
d8.52 "not recognizable as the same color"
GW Dryad Bark vs Vallejo
GC Charred Brown =
d8.42 "not recognizable as the same color"
GW Abaddon Black vs Vallejo
GC Charred Brown = d7.82 "bad match"
GW Dark Reaper vs
GW Mechanicus Standard Gray = d3.22 "sufficient match"
GW Dark Reaper vs Incubi Darkness = d9.12 "not recognizable as the same color"
So only Dark Reaper and Mechanicus Standard Grey are close, but still visually distinguishable. All the rest are poor to bad. What the sheet is saying in these cases is that there is no close match in the two ranges. So you gave three cases of double "matches" but for most of those they aren't actually matches, just as close as the ranges could get.
That said, there may well be a big problem with how it gathered its color data. But you'd hope companies would give swatches on their sites that were accurate representations of the colors they are selling. As the producers of these colors they should already know what the sRGB value of the color is and then actually use that value in their advertising. And before you hold a pot of paint up to your monitor to take a picture saying "see see, totally different!", ask yourself this: has my monitor been professionally color calibrated to have an average delta E below 2 and cover 98%+ of the sRGB color space? Most aren't, and unless you're using something built for professional artists then yours probably hasn't been. And that's goes towards the point I was making to begin with: us being able to accurately identify the color based on a picture isn't all that realistic.
Thing is: I am a proffesional graphic designer, I do use calibrated monitors and other assorted color matching tools for print production daily. Even more so, I own both Charred Brown and Dryad Bark and those are pretty damn close paints, both based on burnt umber pigment with only slight difference in green component and Dryad Bark is slightly darker - you cannot use them interchangeably in a single army, but you can substutute one with the other for following any tutorial or paint scheme. So if Charred Brown vs Dryad Bark has nearly the same
dE as Rakhart Flesh vs Heavy Blue Grey (vastly different colors) and Dark Reaper vs Mechanicus grey has
dE twice better, then the method of establishing
dE is flawed beyond utility. According to this methodology Charred Brown has better
dE to pure fething black than to paint of same intensity and based on the same physical pigment which you have to actually paint one over another to see the difference. And there is a perfect match in layer range for Mechanicus - Eshin grey, literally same pigment in different medium.
As to companies providing accurate swatches - it is pretty much pointless exactly because popular home monitors and perception of colour. Most mass produced monitors on factory settings have a heavy blue tint, yes, sRGB ones too, because daylight brightness requirements and perception of black. When designing for web nobody really bothers with exact color matching because of that and you only bother with relative distinguishability and stability of colors and eyeballing. You even have such disclaimer in Reaper tool
FAQ.
Sidenote: in practice sRGB compliance tells you only that you can distinguish separated grey, blue, green and red gradients to sufficient degree in optimal lighting conditions. Coupled with intended brightness way above print production makes most sRGB monitors useless for any paint color matching applications and sRGB mode is always way, way off from even manually calibrated. Moreover - you cannot have usefull „factory calibration” because calibration has to take ambient room light colour and brightness into account, if you want to have both daylight and calibrated monitor you have to use perpetual adjustment mode of hardware calibrators.
To sum up: the only valid way to do paint compatibility charts is with actual physical samples, either by eyeballing or with colorimeter (if you have samples side by side you can utilize digital camera with
RAW mode as a colorimeter). All other tools are very limited novelty.
And as to available novelty tools: Citadel Paint app has a photo input matching mode built in.
