Ridgway's Color Scheme to RGB?

Factual Questions
1

In my study of Latin, I’ve often been irked by terms for color, which are often vaguely applied in ancient sources, or at least the intended differences are only somewhat untangleable, and are differently ascribed among various modern Latin websites, including the Latin Wikipedia, which has the problem of being woefully incomplete. I suppose in reading ancient documents its only necessary to know broadly what the difference is, but I study Latin as a modern language, and the lack of specificity bugs me. However…

William T. Stearns’ book Botanical Latin describes color nomenclature for use in the naming of species. Stearns organizes the names according to a system devised by Ridgeway, which is available for perusal on Google Books, with some of the color plates actually scanned in color. I have scanned in the black-and-white chart from Stearns: Part 1, Part 2:

Having no previous experience in color theory, I am baffled at the numbering and lettering scheme here. Before I go beating my head against this system, I thought I’d check in here and see if it seems obvious to anybody how one converts the Ridgeway encoding to the hexidecimal RGB coding scheme that my computer knows how to interpret.

Any idea? Thanks in advance.

2

There isn’t going to be such a direct translation for a number of reasons.

Your screen is not capable of displaying most of the colour space the human eye can perceive.

The perception of colour is critically dependant upon the viewing conditions, most importantly the luminant (light source) but also second order effects including absolute intensity and the background light.

Even allowing for your screens inability to display the colours, screens have varying colours when driven with raw RGB levels. The screens are not linear in intensity, and the colours used for red green and blue vary across screen technologies. Two screens driven with raw RGB of the same levels can display different colours.

What you need is a translation of the Ridgway scene to the ICC colour space. That would then enable you to translate to colour on your screen within limits. It won’t be the exact colour of the painted plates you got in a book, but is close where it can be.

You might find this site interesting, it includes a set of the Ridway colours rendered. I don’t know how carefully the translation has been done. Given it is just into HTML colour space, and not something more accurate it suffers from the problems above.

3

How is it that the painted plates such as in Ridgeway’s book can avoid these problems? Surely they are also subject to lighting conditions.

The chart you link to does give some information, though the numbers it gives don’t appear to map onto the chart given by Stearns, which seems to have Roman numerals to represent shades of darkness within which letters and numbers represent, um, hue and tint?

This Color Terminology in Biology seems to be using the names that Stearns maps onto Ridgeway, but has not only the problem that the indexer mentioned (so many colors falling under the same name make it a problematic naming system) but also has a problem perhaps intrinsic to attempts at such specificity that a lot of things that would generally be called ‘blue’ (caeruleus) fall outside the specific options given for that term.

4

They are. This is why colour is so hard. Every step of the way you need to know exactly what it is you are defining or measuring. We can probably assume that the books were intended to be viewed under daylight. Exactly which daylight might be harder to know - one assumes that is specified in the book.

I think the numbers are taken from the NBS/ISCC Dictionary of Color Names, which includes Ridgway’s colours. Also link to on the original page ref.

There was an implied bit of sneakiness available to you in the chart of colours. Since they are rendered in your browser, they have HTML RGB colour values, and simply viewing the source HTML code would get you your initial request.
I was really trying to indicate why the whole subject is so difficult, and unsatisfactory.
In controlled conditions - probably in a museum - the colour patches could be viewed next to a natural subject, and matched. Same illuminant and conditions, and you can get a standard that allows people to agree on what they see.

But the moment you have different conditions, different illuminant, or representation of the colours that appear on systems that don’t have the same colour gamut, it gets really hard. Worse, for many natural life colours, the colour isn’t due to pigments. There will be colours that we simply don’t have useful technology to reproduce in a manner that is useful to an end user.

But much depends upon what you are trying to achieve. If you want to be able to reliably print a colour that will be indistinguishable from the colour of a bird, you will have trouble. If you also want that same colour to be reproduced on a computer screen, and to be indistinguishable from the bird or the printed colour, you are in deep deep trouble.

If you just want to work out what the hell the various colours in the reference book look like, it is probably totally satisfactory to use the RGB values from the scanned book. It may be that there is more information available about the initial scanning that might allow proper colour mapping to be found.

The trouble with the number schemes used is that they just chose arbitrary ways of dividing the colours, without knowledge of colour theory, and the steps are unlikely to be consistent, or even linear cross the multidimensional space. They are all spot colours, and the only translation from them to a proper colour space is as a lookup table with as many entries as discrete colours.

5

Vae mihi. It begins to look like Stearns’ latinization of Ridgway’s scheme wouldn’t answer even if I understood it properly. And that’s in addition to the problems with attempting to capture what may be perceived in the real world using the RGB system. Here’s one example of the sort of thing I assume you’re talking about – you cannot show on a computer the difference between three terms for Black, which Vicipaedia deals with by simply getting it wrong:

niger is the basic term for black, but in a stricter sense, it’s a shiny black, as of polished obsidian.
āter is a non-shiny black, which I believe is called matte in English.
As Döderlein puts it, niger is a color, but āter is a lack of color.
pullus is a duller black, less pure or striking, but nonetheless too dark to call grey. It’s the black of soil, or as I like to think of it, the black of a laundered shirt that stands out as lighter when put next to a new shirt.
Vicipaedia tries to reduce these differences to mere shades, but it’s a lot more complicated than that, in ways that can’t be reproduced on a computer, but could probably be discriminated in real life with a fair degree of consensus. And this isn’t even to deal with such terms as coracīnus the black of ravens, piceus the black of pitch (which does not necessarily entail the absoluteness of the equivalent description in English), mōrulus the purplish black of mulberries, anthracinus the black of coal, or caccabātus the sooty black of cooking pots. I’m not very perceptive on these matters, but I can see why these terms all describe different qualities of black that cannot be reproduced on a computer screen.

Still, it would be useful to be able to assign general color values to all the terms which may appear in a Latin dictionary which admit to such assignment, though you seem to be suggesting that colors in general are at best forced into one scheme or another. The name I might assign to an RGB value has a problematic relation to the best equivalent I could find in a box of crayons.

And of course, different languages categorize differently. There is no one Latin term that will answer for everything we sweep under the English term ‘brown’, except the recent coinage brunneus, which is a latinization of a german word. Also, like in many languages, a term for ‘orange’ has had to be invented after the discovery of the New World, aurantius.

Here are some possible goals that will make this complicated issue managable, I hope:

  1. Group all the color terms I can find in general categories
  2. Sort out which are the most basic terms within each category, suitable to be names for the categories they are members of
  3. Comb through assigning some visual aids, including examples of items that are described by each term, RGB values, and where possible the equivalent in colors from the Crayola line.
  4. Argh.
6

Funnily enough, a near-mint copy of Ridgeway’s Color Standards and Color Nomenclature was featured on last week’s Antiques Roadshow, and the appraiser noted that Ridgeway’s system never caught on. The book was worth rather a lot, though.

7

Brown is a difficult colour, since it doesn’t actually exist. All browns are dark orange or orange reds, and part of the perception of their colour is in contrast to their surrounding colours. Which is another difficulty in colour perception anyway.

Argh indeed. You are looking at something that a lot of work has been done on. But it is also difficult and frustrating. That initial site I linked to is a very interesting effort, and worth reading right through. A lot of colour theory has come from the printing industry, and most of the rest from the film and television industry. The limiting factor with both of these is that they are constrained by their reproduction technology. The standard colour gamuts that they use are limited by either the range of inks, dyes, or the limitations of phosphors (for CRT and plasma displays) or filter spectra (for LCD). Also complicated by the lighting and reflectance characteristics. Once you use anything with a line spectra (florescent lights, white LEDS, LCD backlights, for instance) it gets really very very messy.

If you want to start somewhere, this is good. There is lot to cover.

However, as I noted above, if you want to match colours of natural things, there will be perceived colours that cannot be reproduced. That is short of building a box that mixes spectrally pure colours of light.