Wikipedia talk:WikiProject Color/Archive 4

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Archive 2

Archive 3

→

Articles for the Wikipedia 1.0 project

Hi, I'm a member of the Wikipedia:Version_1.0_Editorial_Team, which is looking to identify quality articles in Wikipedia for future publication on CD or paper. We recently began assessing using these criteria, and we are looking for A-class, B-class, and Good articles, with no POV or copyright problems. Can you recommend any suitable articles? Do you have a set of basic colors that you would consider publishable? Please post your suggestions here. Cheers, Walkerma 03:28, 7 March 2006 (UTC)

Personal opinion (this seems to be a dormant project because I've never been able to achieve any discussion of these articles): I believe all of the articles on colors have a serious flaw. They do not contain any references; the colors could be made up for all the reader knows. The infobox is also full of context-free numbers. I could not therefore recommend any of them. Some of the articles about color science are good, but they lack consistency. Notinasnaid 08:28, 7 March 2006 (UTC)

Thanks a lot for that - I'll note that on our listings. Maybe a simple goal for the project could be getting a basic set of colours up to A-Class? Hopefully in time things will change. I think with minor re-organisation of refs the article Color would be A-Class (i.e., it could be considered as a possible FAC), do you agree? Thanks, Walkerma 16:58, 9 March 2006 (UTC)

Quiddity has started doing assessments, see below. Rfrisbie^talk 14:46, 24 August 2006 (UTC)

Assessment and list - help with template?

I've tried to fix our Template:WikiProject Color banner, to include the assessment scale (class) variable, and importance variable, but cannot get it to display correctly. (I tried to use {{genre}} as an example. (I also got horribly overwhelmed by info at Work via Wikiprojects))

If someone else can get that done, I have the urge to go find and paste the banner on all color-related pages, and give them a rough assessment/roundup.... --Quiddity·(talk) 22:05, 23 August 2006 (UTC)

I mixed the two templates a bit, threw in some color images and came up with this. [1] It seems to work for class=FA, A, B, etc. I don't know all the details you want, so I didn't replace the current template. Rfrisbie^talk 00:39, 24 August 2006 (UTC)

Perfect! Thank you.

I'm using these 4 articles as a test batch - Talk:Peach (color), Talk:Color tool, Talk:List of colors and Talk:Color. I'm going to take it slow, so that I don't make any mass-mistakes :) Anyone experienced is welcome to plunge ahead. (I'll have to figure out the bot-automation page-creation (eg Wikipedia:Version_1.0_Editorial_Team/WPScience#Wikipedia:WikiProject_Color) component of it all later.) --Quiddity·(talk) 01:10, 24 August 2006 (UTC)

Whoo-wee! That 1.0 project does have some details! I started a few cats related to the template at Category:Color articles by quality using chemistry as a model. Hopefully, that's enough for your test article assessments. I also linked to Wikipedia:Version_1.0_Editorial_Team/WPScience#Wikipedia:WikiProject_Color. See the Index for examples. Rfrisbie^talk 02:35, 24 August 2006 (UTC)

I noticed your assessments were changing the display but not the cat. When I changed them to inital caps - "Start" "Stub" - the cats changed. I went back to the template and added lower case cases. It seems to work either way now. Rfrisbie^talk 03:22, 24 August 2006 (UTC)

The assessment page linked on the template is good to go. Rfrisbie^talk 04:01, 24 August 2006 (UTC)

I added the importance ratings to the template using Template:Chemistry, along with the related cats. This update automatically shows "not yet" rated messages and allows for "class=NA" on non-article pages. Rfrisbie^talk 15:58, 24 August 2006 (UTC)

As I was surfing the color cats, I came across two red ones that orphan pages and/or cats - Category:WikiProject Color and Category:Non-article color pages. Is there any interest in creating these cats? Rfrisbie^talk 16:28, 24 August 2006 (UTC)

Apparently! I looked at chemistry again and modeled the classifications more-or-less after them. Rfrisbie^talk 17:42, 24 August 2006 (UTC)

You rock! Great work all over :) --Quiddity·(talk) 19:16, 24 August 2006 (UTC)

I've been through Category:Color stubs, and am going through Category:Color now. The following is a list of the articles I couldnt determine whether we covered and/or whether our WP template tag should be applied. (Hopefully I'm doing everything else right... I'm erring on the side of over-inclusion, so that we can clean up the categories from a central location. or something.) (I'm not an expert, just a wikignome - feel free to change any of my assessments) -Quiddity 20:44, 25 August 2006 (UTC)

Right, everything in the subcats of Category:Color have been tagged, except these:

Color metaphors for race
Cerebral achromatopsia (I included color-blindness already..?)
Color legibility (stub, needs merge)
Electromagnetic spectroscopy
List of flags by number of colors
Mordant red 19
Indiglo
Brown ribbon
Purple ribbon
Xanthochromic
Sodium nitrite
and nothing in Category:Horse coat colors.

as I wasn't sure whether we (want to) cover them. (?) (I'd guess not..) --Quiddity 07:26, 26 August 2006 (UTC)

Thanks to Longhair's tips, Template:WikiProject Color now also helps the bot track comments. The template places them into Category:Color articles with comments and the bot (hopefully :-) adds them to the summary tables Index · Statistics · Log. The project template can be even fancier! See Template:WP Australia and Template:WPBiography for examples. Rfrisbie^talk 19:29, 26 August 2006 (UTC)

Importance of color shades articles?

Is it safe to say the "shades" of Category:Colors are of "low" importance, with the named color for each shade subcategory "high" for primaries, or "mid" for everything else? Lots of importance ratings could be added if there's a basic triage rule of thumb that can be followed. Rfrisbie^talk 03:35, 26 August 2006 (UTC)

That was basically the assumption i was working under. (I've almost finished tagging them all with the template. Will do the last of them now..) --Quiddity 06:06, 26 August 2006 (UTC)

Cool. I'll just sit back and watch the bot tables fill up! :-) Rfrisbie^talk 06:10, 26 August 2006 (UTC)

Done. And now my hands hurt! *mutter*

I've been erring on the side of under-rating importance.

It'll update with that last batch at 3am UTC tomorrow. There're lots to be rated, classed, and re-evaluated in the bottom-half still ;) --Quiddity 07:45, 26 August 2006 (UTC)

Progress and regress

Recently there seems to be a movement towards removing entries on colours and it might even be gainisng speed. Since it seems contrary to this project and on suggestion from [[Johan the Ghost _seance ]] I am bringing it up here.

The best example is the case of the entry for ecru which had its own entry but now redirects to beige and before that to yellow, see Talk:Yellow#Ecru for more information.

The deleting proposals can be found in the [history file] and can be briefly summarised as 152 other color pages doesn;t show notability, shows we need to do cleanup... this fails notability by only possibly being Wiktionary entry, which already exists followed by no possibility of being anything beyond substub, this is what Wiktionary is for. I cannot find any discussion leading up to wipe and redirect. This suggests another 150 colours are about to be deleted. I for one think this would decrease the usefulness of Wikipedia. --14:48, 12 May 2006 (UTC)

I definitely support your position, that all colour stubs are useful (but need Color-coordinates/swatch infoboxes), and will be slowly expanded, and should not be deleted or merged.

I don't know if anyone else is watchlisting this page, but it does seem the right place to discuss this issue. -Quiddity 17:36, 12 May 2006 (UTC)

I do not support including non-notable colors. I especially do not support making up swatches for them. I support only including colors that are part of a published and non-proprietary standard. I fully support deleting all entries for nonstandardised and non-notable colors. Notinasnaid 11:34, 13 May 2006 (UTC)

The notability of a colour is what this is going to center on. What criteria are you suggesting we use?

"Published and non-proprietary" sounds sensible, but that includes hundreds of colours. see List of colors, which is only a tiny selection. Just to pick two randomly: I like the box at the bottom of Prussian blue, and think Camouflage green covers three names with a good start for a stub.

There is a song about the colour Ecru (which is now a redirect, but imo shouldnt be) by Ken Nordine, for example, and it's not a new word:[2].

They're not fan-cruft or similar, so I can't see removing any of those as being beneficial.

However, for an example of a stub that might not belong, see Gray-Tea Green. This doesnt appear to be one of the X11 color names (most of which i would obviously Not advocate creating an article for), and i don't know what colour space it does belong to. -Quiddity 18:57, 13 May 2006 (UTC)

I am brand new to this project (and with good timing from the looks of it!). I read through a lot of the discussion above, about the proliferation of color pages. The suggestion about having a standard, limited set of "major" colors like red and blue was a good one. They could have the color range swatches that were proposed, and a list of common shades, even including descriptions if appropriate. For example, Beige might list Ecru as a shade, and point out that Ecru tends to be a light, yellow-ish, off-white shade of beige. (Although I'm not sure that I'd put Ecru with Beige so maybe that's a bad example.)

I'm not totally stuck on any particular set of starter colors; maybe going through an 8-pack of crayons and then adding a few other popular ones would be a good start. Then, if we find that a particular color is getting a lot of text with it, we might consider breaking it out. For example, Teal could be included as a shade of (blue? green? what about the color Blue-Green?). If we find that it ends up with three or four paragraphs of description, then it might be worth breaking it out as its own page.

Lastly, on the topic of color swatches. On the one hand, I agree that it seems a little presumptuous to define one definitive "Red". On the other hand, I have come here in the past looking for particular colors, and it was very helpful to see a swatch. Hearing that Puce is "generally considered to be dark red to brownish-purple" was very helpful to me; to see an actual swatch would have been even better. (There is already a swatch there but it looks more like mauve to me, unless that's just my browser.) Standard problems with browser and monitor differences - this is why we used to work only with a few dozen colors on web pages.

Now that I've rambled long enough, I'll end by saying that I think this is a very worthwhile project and one that I'd like to contribute to. I'm going to take a look at the article for Color next, and see about organizing some of the major colors. This project seems to have been dormant for a while; anyone for starting it back up? --Laura S 00:44, 17 May 2006 (UTC)

I should clarify what I mean by "standard", since you used it in your reply in a completely different way than I intended. Easily done. To my mind a standard is the result of a standardization process, a recognised list of colors, an "official list" if you like. For example, the HTML/CSS list of colors in web colors is based on a standard; I think the color part of this is good, useful, and encyclopedic. Unfortunately, there has been nothing to stop anyone in Wikipedia deciding that SkyBluePink is a color, creating an article, and putting an info box with some made up colors in it. So I advocate removing all color boxes that do not contain standard values, and removing all non-standard values added for standard colors (for example, CMYK values for colors that were standardised as RGB). We also have to address the issue that multiple important standards might use the same name for different colors. See green as an illustration of this problem. For this reason it is vital that color infoboxes identify the standard that applies. A crucial decision to make is what is considered an acceptable standard. Merely being on the web is not enough; I have found many pages with new made-up lists. Notinasnaid 09:57, 19 May 2006 (UTC)

More: I have to say that the example of a pack of crayons is the exact opposite of what I think should be happening. You and I have no way to measure those colors accurately, nor any reason to declare that this blue crayon is a standard for blue. I do see a conflict here, and it is important to resolve this. There are really two outlooks being taken onto the color project: scientific, based on color science, measurement, standards; and artistic, based on a desire to work with colors. Both are valuable, and as someone whose head is more into color science I recognise that the artistic outlook is important and sincere. A lot of work has been done, too, and I respect that but at the same time, Wikipedia's policies have to be considered. I have several times seen a statement like "it's important that Wikipedia include some value and sample for each color." The implication being that some color is more important than accuracy. But Wikipedia has an absolute rule of verifiability: making up some color, or picking a personal favorite between several is the dreaded original research. So I couldn't agree with this statement less. It is very important Wikipedia does not include some kind of value for a color unless it is in the context of a well defined (and well explained) standard. No articles that I have seen meet this criterion. I don't think a "head in the sand" approach can work here; as Wikipedia matures its policies are being more rigorously enforced, and if the color boxes are to survive in any form, they need to change to a verifiable approach. I'd like to add that I'm very glad to see this project come back from the dead; various attempts have been made to deal with these issues without raising any interest at all. Notinasnaid 10:07, 19 May 2006 (UTC)

I've been giving some thought as to how to bridge the artistic/scientific gap. While Red is included in many color standards, and could have an infobox, my favorite illustration of red is #Red (color) above. This clearly shows a range of reds. Now, what about ecru? My initial research throws up a dictionary definition but no standard. Clearly on that basis (my arguments above would suggest that) we can't have an infobox because there is no standardised color. But how about a photograph of something that is in the fabric ecru? Indeed, ecru is by its dictionary definition first a fabric, second its color. There is also I believe a standardised cotton color. This is no good for info boxes, unless the makers provide an absolute color reference for it, but we could photograph a hank of the cotton. Notinasnaid 10:59, 19 May 2006 (UTC)

Wow, for a project that didn't raise much interest for so long, we seem to have generated quite a bit of it now. (For the record, I count that as a good thing.) We seem to be having several discussions in one thread, which is a little confusing. I have a feeling it's also making it seem like we all disagree more than we actually do. For the purposes of clarity, I'm going to attempt to break some of these discussions out so we can give them the focus they deserve. -- Laura S | talk to me 22:47, 19 May 2006 (UTC)

I'm mostly just watching and commenting, no expertise just curiousity. Take all my comments with a "weakly believe" prefix :) I prefer doing cleanup work in the background mostly. -Quiddity 18:20, 25 May 2006 (UTC)

Scientific vs. artistic

There seems to be a rift between defining and describing color from a scientific point of view and an artistic one. As Notinasnaid said, "There are really two outlooks being taken onto the color project: scientific, based on color science, measurement, standards; and artistic, based on a desire to work with colors. Both are valuable..."

As a matter of disclosing any bias I might have, I come more from the artistic standpoint, but find the scientific view interesting as well. In considering readers, there will be some looking for each type of information. I think our challenge (and value) lies in representing both of these views adequately and in a complementary fashion. Having said all that, I'm not sure what the solution is; only that I'm up for trying to find one. -- Laura S | talk to me 23:04, 19 May 2006 (UTC)

Absolutely. Inclusionism is generally better. -Quiddity 18:07, 25 May 2006 (UTC)

Inclusion of colors

The next topic is whether to include all colors we can lay hands on, or only certain colors. The forming consensus seems to be to have articles for only a set of "major" colors (how to select those is another question).

My personal addition to that would be rather that eliminating all the other color articles and stubs, which could lose a lot of good information, we should sift through the color articles. Any color that can be verified (from an art book, etc.) and is non-commercial could be added to the appropriate main color article. Ecru above was probably a bad example for a few reasons, but take the example of maroon. Defining precisely what maroon is might be difficult (unless there's an established standard), but - and I do say this without verification so please don't vilify me - it's generally accepted to be a dark shade of red. Therefore, again assuming we can verify this, I would support including "maroon" in an article for red, possibly in a "shades of red" section. This would retain the valuable information without cluttering Wikipedia with dozens of stubs and non-notable information.

So my questions would be, is my first paragraph accurate, and what does everyone think of the second part? -- Laura S | talk to me 23:15, 19 May 2006 (UTC)

I'm an inclusionist(mergist if necessary to avoid deletion, but stubs are fine by me). So i say all actual colours deserve their own page. So ecru, maroon, puce, indigo, mauve, aquamarine, etc all get their own page.

But obscure things like crayola colours (electric lime, purple pizzazz) are all on a single page. --Quiddity 17:50, 25 May 2006 (UTC)

Defining notable colors and standards

This is the other big one. If we are agreed to create or improve a limited set of color articles, which ones do we pick? My crayon suggestion was mainly just illustrative (the Crayola 64-pack was my favorite "toy" as a toddler so I have a slight crayon bias); I have seen suggestions for using spectral color. Or perhaps primary and secondary colors (which gets us to the same place essentially). I would add black, white, and probably brown and grey to the list as well.

Then there is the whole X11, HTML, web-safe colors, etc. discussion. I think those are important to discuss, and maybe even list the definitions of the colors of them, but each one certainly does not deserve its own article. -- Laura S | talk to me 23:31, 19 May 2006 (UTC)

Proposed Wikipedia policy/guideline on color/colour articles and swatches

For some time I have felt that there are problems with color articles in Wikipedia. It seems anyone can decide that a color exists, create a swatch, convert to dubious CMYK values using any method they please, and define an article. I see this as a problem, because of Wikipedia's requirements for verifyability. Others may not. However, I do believe that many things in the color articles are open to challenge under Wikipedia:Verifiability and that as this is enforced more strictly, most articles are currently open to deletion.

To deal with this I have created a proposal. It is in formal language because that's how proposals have to be to work, but I am not trying to forestall discussion. I don't know the mechanisms of Wikipedia policies either, so if there is any consensus to adopt this (or another) policy some guidance will be needed on what happens next.

To avoid accusations of hiding stuff in formality, some of the side effects of accepting these proposals would be: deleting non-notable color articles; removing color values that aren't from an authoritative standard; and removing CMYK values in swatches.

Proposal and following two sections posted by Notinasnaid 19:48, 1 June 2006 (UTC).

Proposal

1. Notability and verifiability. Wikipedia may include articles on color names only where there is sufficient notable and verifiable information available to write a non-stub article on aspects of the use of the color name. Verifiability of the use of a color name must go beyond finding a web site that lists or uses the name.

2. Use of color disclaimer. Wikipedia may include illustrations such as photographs of objects to illustrate the meaning of a color name; in all cases these shall link to a color disclaimer page which emphasises that exact color reproduction cannot be guaranteed and that neither the monitor nor a printout should be used as a reference. A similar disclaimer shall be used in any illustration which shows a color or ranges of colors in such a way that they might be taken as being authoritative.

3. Color values only if in a standard. In some cases a color name will be included in an authoritive standard which gives a definition or conversion in terms of specific color models or color spaces. It is proper to include this information provided the color name standard is not proprietary (see item 6). Conversions to other color spaces that do not form part of the standard or source shall not appear.

4. Strict limits on swatches. In limited situations, Wikipedia may include swatches of colors which attempt to directly show colors in the browser via HTML. Such swatches may only appear where

(a) the color is defined in an authoritative standard of color names

(b) the color name standard identifies a color in either sRGB color or as a simple RGB color mixture without calibration

(c) the swatch indicates whether it is sRGB or or uncalibrated RGB

(d) the swatch does not include any other color spaces unless they form part of the same standard

(e) the swatch identifies and links the color standard used.

5. Pre-vetting of acceptable color sources. Wikiproject color shall maintain a list of authoritative and acceptable sources for color names; adding of sources needs to be subject to a consensus procedure.

6. Excluding proprietary color spaces. To protect Wikipedia from legal action, color names and values from proprietary color spaces shall not be systematically listed, and color values shall not be included at all, unless the color space owner specifically releases this information under a license suitable for use in Wikipedia.

Commentary from proposer

1. Notability and verifiability. The idea here is to make color entries follow normal Wikipedia standards. For example, if writing about a rock star or web site, there are notability standards, and every article must be verifiable through sources. (Many articles lack sources, but this is a deficiency, and one this project should be keep to remedy). Note that this item does not say anything about whether the article has a swatch; there may be colors with no formal definition, but with important cultural, artistic, political or other connections that are worthy of an article. The author of such an article should record the connections but should not attempt to formalise the definition, because that is original research. Note also that these standards don't propose what the notability standards are, just that we need to have some; the details would be a major debate in themselves, so I'd recommend at this stage just debating whether to have standars, and only move on to what is in them if there is consensus on having some.

2. Use of color disclaimer. Wikipedia has disclaimers about medical and legal advice. While colors in pictures aren't perhaps so dangerous, it is very important, if Wikipedia is to have any claim to definitive information, that it is explained that the colors you see on your monitor or printout are not controlled and may not match the original intention. People can and do use the information they see on screen to make expensive decisions, like colors in printing or decoration, in ignorance that their monitor is not a reference standard (a few monitors are calibrated but most are not). We can have a general disclaimer page, and it should be linked to from every example of color (in picture or swatch).

3. Color values only if in a standard. This is an attempt to apply Wikipedia's rules for verifiability to articles about colors. I am convinced that many color articles include RGB values that have been found on a web site, or simply picked by the editor. The latter must surely count as original research, and picking an unchecked value of some web site is no better. This is why I propose that specific color values only be included if they are in a standard. For example the HTML/CSS standard. By "standard" I mean to imply something formal, published; countless web sites make things up. Also some web sites assign RGB values to particular standard colors which aren't part of the standard (such as, I suspect, the colors of brands of crayon). This project should decide what standards are acceptable. In addition, only the color values from the standard should be used, and the color space made clear. For example, RGB and sRGB are very different in their meaning. If a standard gives a value in a different space such as Lab or YUV then this is the only data that should appear in the article, not a conversion (see next item). Note also that the same color may have radically different colors in different standards (see, for example, Green); this can simply be presented, it isn't Wikipedia's job to pick one over another.

4. Strict limits on swatches. Unavoidably, we get a bit into color science here. The gist of the argument is that Wikipedia should not be making up conversions. This is inevitable in the case of where an RGB value is converted to CMYK; these are not absolute color spaces and the RGB colors/CMYK inks make a radical difference to the actual color, as does the choice of what to do about the many RGB colors that cannot be accurately reproduced in CMYK ("out of gamut"). When accurate RGB to CMYK conversion is done there are many choices to be made (e.g. "target profile" and "rendering intent"). How can Wikipedia pick just one and say "this is standard CMYK"? That isn't the job of Wikipedia. The CMYK article contains a widely used formula. But it is not an accurate formula, or a standard. I propose that only sRGB and (uncalibrated) RGB colors in color spaces get swatches. sRGB because that's what browsers show, and RGB because it isn't a specific color, therefore treating it as sRGB is no more wrong than anything else. Some argue that Wikipedia should show some CMYK value because people will come here looking for one. I argue that's exactly why it should not show a value that cannot be accurate except in very precisely chosen situations, and isn't actually very useful.

5. Pre-vetting of acceptable color sources. I can't see another way to make this all work.

6. Excluding proprietary color spaces. It would without a doubt be useful for Wikipedia to include colors from industry standards such as Pantone, but intellectual property of this nature may be vigorously defended. This is not a battle Wikipedia would want to fight.

Further commentary and discussion

Seeing as this is a thorough proposal with many parts, would it make more sense to move this proposal and its associated discussion to a subpage, like Wikiproject Color/Color Proposal (or somesuch)? -- Laura S | talk to me 20:40, 1 June 2006 (UTC)

Notinasnaid, thanks for this proposal. It's obvious you put a lot of thought into this, and hopefully it will go a long way toward putting us on the track to making some serious improvements in the color articles. My comments follow.

Notability and verifiability. I agree with this, with some expansion. I want to make sure that we don't go deleting all stub articles, because some may be useful. That is, just because an article is a stub doesn't necessarily mean the color is not notable. The AFD process would probably cover this issue. Also, some colors may merit a paragraph or so within another article (this is the mergist in me talking); we'll need to differentiate how much mention a color warrants.
Use of color disclaimer. This one is very important. However, instead of linking the image to the disclaimer, I would handle this in the caption. So each illustrative image or swatch would have something like this appended to its existing caption: "Note: Colors vary depending on output device." And that note would then link to a page with fuller information.
Color values only if in a standard. I agree fully. A great example of this is International Klein Blue. This color was developed as a special pigment of paint. The article's infobox gives the RGB value as though it's absolute but something tells me Yves Klein did not provide this value. It makes no sense to have anything other than perhaps an approximate swatch (with a big disclaimer that it's not exact), just to give readers an idea of the color, and maybe a photo of the big square of IKB that hangs in MOMA. Color space values, just like any other information, should only be provided where relevant, and where verifiable.
Strict limits on swatches. Need to clarify what we mean by swatches. If we're talking about a big box of a single color defined in the HTML, then I agree completely. These swatches need to be provided only where it makes sense to show an RGB or sRGB swatch, and need to be accompanied by information about what they are. However, occasionally we've used the term "swatch" to talk about a photograph, illustration, or collage containing examples of a color. I think these are very helpful to readers and should not be abolished. As long as we provide a disclaimer stating that these are not intended to show exact color values, but rather a range of colors that are "blue" or "green", I strongly support inclusion of this latter type of swatch.
Pre-vetting of acceptable color sources. No real opinion on this one just yet. It sounds like a good idea. I just want to make sure we don't end up exluding important colors that are not formally part of a standard, but still merit inclusion (such as IKB). However, I suppose it wouldn't be a big problem to drop a note on a talk page to gain concensus for addition of the color.
Excluding proprietary color spaces. Absolutely agree, no arguments here. Question though: this is purely hypothetical, as I can't think of any examples right now, but what if there is a color from a proprietary standard (say Pantone) that becomes highly notable for some reason (maybe a company uses it for all their advertisements, etc.)? Obviously it would still be inappropriate to list color values, but would it be inappropriate to list it at all? It seems this would be akin to any other product, such as Kleenex, provided we didn't include information on how to create the color. Of course, IP law gets tricky.

-- Laura S | talk to me 14:50, 4 June 2006 (UTC)

This is all great! Glad we seem to agree so well. I favour any mergist strategy too.

Re. the proposal's placement, it seems clear and short enough that, we could just add the 6point synopsis to our main wikiproject page, with a link to the commentary here? (multiple subpages make it harder for new participants to watchlist everything appropriate.) I'll move the deprecated (by this) templates to the bottom of the main page. -Quiddity 19:01, 4 June 2006 (UTC)

It almost seems too easy, after all that, that we reach agreement so quickly :) I've been looking around, and this will involve *lots* of rework. I think if we can determine a few "major" colors to start with (such as red, green, etc.) we can not only clean up those articles if they need it, but also come up with better examples instead of swatches, merge in any "minor" colors that are notable enough to be mentioned but not enough so to have their own article, etc. Coming up with that list may be problematic. How do we determine such a list? Some seem obvious (black and white) but what about cyan?

Many of the templates will go away or at least change significantly, as you've already demonstrated. We may be able to use the infobox still, in a less universal fashion (such as including only the relevant color space(s)).

After all this discussion, I'm itching to start working, but not sure where to start. I'd still like to get Color up to FA, and it's already a great article, but needs references which may be tough. Also it's highly scientific. Someone mentioned on the talk page that the whole article doesn't mention things like hue and tint; there's really no artistic treatment at all. Got any ideas?

Lastly, I think Notinasnaid is taking a brief wikibreak until next weekend, so we may not hear anything from him/her until then. -- Laura S | talk to me 01:23, 5 June 2006 (UTC)

I've added these as "guiding principles" to the main project page, with the text of them (plus summaries of subsequent discussion) at Wikipedia:WikiProject_Color/Principles. This way once the discussion is in some archive, we'll still have a clean copy to work from. - Laura S | talk to me 01:43, 20 June 2006 (UTC)

I have some concerns about this proposal. It seems to me that it is tryint to over-enforce criteria on color descriptions. As an example, an institution may use a specific color as part of their trademark. They may define the value of that color on their website. Yet, if the institution is not in the day-to-day business of defining colors, then they would certainly not appear in a list of pre-approved sources. However, their definition of the color in their logo would certainly be verifiable. Johntex\^talk 03:57, 6 July 2006 (UTC)

This was pretty much my only concern with the proposal. However, the purpose of those aspects of it, as I understood it, was not so much to exclude anything not specifically in a named standard, but rather to ensure that the only colors we write about are notable and verifiable. We are also trying to avoid the legal issues of dealing with a proprietary color. I think if there is a color that doesn't fit the letter of these guidelines but still fits the spirit, it would not meet resistance from the project. Besides, it doesn't hurt anything to raise the question of a certain color on a talk page to let people comment on it. -- Laura S | talk to me 17:41, 6 July 2006 (UTC)

Spectral color

i have added some material to spectral color on the physics of spectral color and integration of the spectral colors in visible light. i have some further ideas on wavelength discussion, alternative light sources for visible light, etc, if a consensus indicates more such material would be desirable. i am open to editing or moving text i created just now to another article. im a newcomer to "color" editing so i am open to suggestion. Anlace 15:32, 20 May 2006 (UTC)

Swatches

This one seems to cause a lot of grief, which is why I'm breaking it out separately. Do we or do we not include swatches of colors, and if so, what should they be like? I'm of the opinion that the current swatches aren't really doing any of the colors justice, especially because of differences between screen and print. At the same time, I think it is essential that we illustrate at least the major colors with examples.

I move for a multiple example solution - like the collage picture for Red. The article for green shows a couple photographic examples of "things that are green", and I think that's really excellent. This way people understand we're not trying to represent a perfect green, but rather a range. If we say "leaves are green", they can go outside and see some leaves and go "ah, that's green". Also, and this is purely a matter of aesthetics, but the photos are much nicer to look at than a big chunk of some randomly chosen shade of green. -- Laura S | talk to me 23:37, 19 May 2006 (UTC)

I agree. Most people who come to wikipedia I'm betting are there to look at more generalized aspects of a color in layman's terms. The things that colors have come to represent, uses for colors and emotions it evokes and is associated with are far more interesting and encyclopedic than swatches and color bars.MiracleMat 03:54, 22 May 2006 (UTC)

Agree. Multiple examples are inherently better than single. -Quiddity 18:26, 22 May 2006 (UTC)

So it sounds like we have a consensus then? Once we figure out what colors to include, we can provide some multi-color swatches and examples of "things that are x color". Now to figure out which colors to include. The other discussions have all been painfully quiet lately... -- Laura S | talk to me 15:22, 25 May 2006 (UTC)

I do have a major proposal in preparation which impacts on this; can I crave your indulgence until it is ready, hopefully soon. Notinasnaid 15:24, 25 May 2006 (UTC)

Fine by me, we can't really do much about the swatches anyway until we decide which colors to include. Does your proposal touch on that as well? -- Laura S | talk to me 18:10, 25 May 2006 (UTC)

Shades of Blue Template

Hello, Color WikiProject. I'm not really all that good at making templates, or coding or whatnot, so I just thought I would throw this out there for anyone who would be more capable than I am to fix this issue. The Shades of Blue chart along the bottom of all of the blue-based color pages seems to have an error when linking to the color "prussian blue". Instead of going to Prussian blue (color), it goes to the article on Prussian blue, the anti-semitic pop duo. So anyone who is capable, please fix this when you have a moment? :) Thanks a lot.

Thanks for bringing this up. It's likely that the "shades of blue" template will probably go away as a result of the above discussion, but I've fixed it for now. Also, please remember to sign your posts with ~~~~ which will add your username and a timestamp. Thanks! -- Laura S | talk to me 01:14, 5 June 2006 (UTC)

Sorry about that. I should've remembered the signature tag, since I'm used to talk pages by now. And thanks for the quick fix to the template, even if it is probably going away. Just a good "principle of the matter" thing considering what it was linking to before.-Resident Lune 03:40, 5 June 2006 (UTC)

Updated swatch template proposal?

PB29 - #003BAF

Ultramarine Blue

Hi, I wanted to propose an addition/modification to the swatch template (or possibly a new template with a different name). I've never created a template before, though I wouldn't mind learning the syntax if people think it would be a worthwhile template. I created the code for the pigment article to show swatches that estimate the appearance of various historical pigments, based on CIELAB measurements converted to sRGB color space. Like all swatches, its an approximation only, but I think they are useful for giving context. In this case, I've mirrored the appearance of the image thumbnail, to keep a consistent appearance. Likewise, use of the css classes means that it scales and changes gracefully with changes to skin, and future modifications to mediawiki software.

Parameters could be float (left, right or center), width, height, swatch color, text color, color name, swatch text (w/ sRGB value + other relevant color space info/pigment identifier. Let me know what you think. Phidauex 17:03, 11 August 2006 (UTC)


<div class="thumb tleft">
<div style="width:120px;">
<div style="width:112px; height:50px; background-color:#003BAF; color:#FFFFFF;">PB29 - #003BAF</div>
<div class="thumbcaption">Ultramarine Blue</div>
</div></div></div>

It's definitely preferable to our current {{Swatch}} template, so I'd say go ahead and replace that one (check the few places it's being used, before and after for breakages, and otherwise everything should be simple).

There was some recent discussion on the use of swatches -- see point #4 in the summary at Wikipedia:WikiProject Color/Principles, (or the original comments, just above). Thanks :) --Quiddity·(talk) 19:26, 24 August 2006 (UTC)

#003BAF

Ultramarine^[1]
#003BAF

I've created Template:Color swatch. Based on your suggestion and the German template de:Vorlage:Farbmuster. Example at right. --Quiddity 00:18, 9 October 2006 (UTC)

Looks nice. :-) Rfrisbie^talk 00:33, 9 October 2006 (UTC)

Color articles getting long??

One color article that appears to be getting longer gradually and might end up over 50KB (it is already around 40KB) is Blue. I can't find what the best section to put into its own article is. Georgia guy 19:37, 24 August 2006 (UTC)

CMYK / colorspaces

I think it is in this day and age to if we include color values, we need to also standardise color profiles for the CMYK values. For RGB, an assumption of sRGB is probably the wisest choice for the RGB side for three reasons: 1. As far as rgb profiles go, it is the most common, being the default for almost any scanner or camera. 2. Microsoft windows assumes everything is sRGB, unless otherwise specified, meaning it will be approximately right for any uncalibrated windows setup. 3. sRGB has a normal RGB gamut, and will match on uncalibrated monitors much better than the second most common standard, Adobe RGB, which has both advantages and disadvantages due to it's wide gamut nature.

In short, sRGB works best with uncalibrated gear, and works best with windows. This means almost anyone on wikipedia.

In terms of a standard CMYK profile, SWOP is the way to go. Most common by far. The SWOP primary colours are all close enough to inkjet printer inks to work adequately. All good.

Now to convert RGB values to CMYK values, we shouldn't use the calculations of 1-r = c type. These assume that CMYK inks are the exact colour opposite of RGB phosphors. This is nowhere even remotely close to the truth. Calculations of this kind (even though they're throughout the textbooks such as FVDF&H and FVDFH&P) are useless, except as an abstract exercise in matrix algebra. Unfortunately the only tool I know that will calculate reliably with both profiles is the ColorSync Utility (part of MacOS X). (Screenshot, and suggested settings below). Also, one final drawback is that you'd need to use another app to divide the 0-255 values by 255 (not... as is instinctive to divide by 256). Anyway, this I see as the best way of standardising the CMYK swatches, and making them useful. using the swatch CMYK values then should resonably resemble the screen, except for out of gamut colours (like the pictured crt magenta) which will resemble things as close as possible.

link to explanatory image (not directly included because it is not free)

I don't understand the problem with including this image on the page. The use on this page certainly falls legally under fair use. --jacobolus (t) 10:08, 15 May 2007 (UTC)

One more important thing: THIS PROCESS NEEDS TO BE UNDERTAKEN FOR AT LEAST THE PROCESS INK COLOUR ARTICLES

Magenta (printing) is defined in RGB as 255,0,255, which is nothing like Process Magenta which is most like 236,0,140. It's a HUGE difference, and needs to be fixed, because the way things are defined in that article, even the sample is outright wrong.

If people agree to this conversion process (which should actually be able to be replicated with the photoshop colour picker IFF the profile setup is changed to match, but I need to check), then I will volunteer to recalculate all the infobox CMYK values, and fix up the process colour articles Cyan, Magenta, and Yellow to include both the screen and process colours. What do you all think? It's a starndardised, useful solution to the whole mess? — Preceding unsigned comment added by Bb3cxv (talk • contribs)

I thought that the proposed policy, already agreed, or so I thought, would have the effect of almost completely eliminating the CMYK values included in articles in any case? Notinasnaid 21:16, 29 August 2006 (UTC)

But taking up the gauntlet anyway (can't resist): why SWOP, a North American standard, rather than Euroscale or a Japanese standard? Which SWOP? Which rendering intent? Which CMM? All of these affect the result, precisely why I argued that CMYK doesn't belong. (Which is not to say that the discussion above wouldn't produce more accurate values). Notinasnaid 21:19, 29 August 2006 (UTC)

Having reread my replies, I realise it comes over as rather, well, negative. I apologise to Bb3cxv if this was the perception. This is spilling over from a very long standing frustration with the awful state of the color articles, about which this project is at last doing what I have neither the time nor patience to do. Notinasnaid 13:41, 30 August 2006 (UTC)

The rational for removing CMYK values as far as I've seen is because the conversions currently used are quite frankly worse than useless, in that when printed don't even vaguely approximate their RGB screen counterparts. In short it's because of the current conversion procedure. With a valid conversion these CMYK values would be more of use.

Why SWOP? (Well SWOP v2)? First, SWOP v2 is the most common CMYK profile. Second, inkjet printers tend to work on an assumed SWOP profile, as their inks tend towards the US ink colours. Third, lack of ambiguity; SWOP is US web coated, euroscale has coated/uncoated variants. Japan standard is also unambiguous, but is somewhat uncommon. Admittedly dot gain is the weakness of the swop system, but because most hw assumes SWOP or at least supports it, SWOP is the path of least resistance (excepting the expression of pointless anti-american sentiment). Besides, noone uses imagesetters these days, it's all ctp... As for CMM? Any ICC derived system. It's why we have standards. Rendering intent should be absolute, after all these are solid colours, not raster images. Warping the in-gamut colours would be stupid. Colours significantly out of gamut should have a note. --Bb3cxv 06:18, 3 September 2006 (UTC)

More use, indeed, but the problem as I see it is that people expect information in an encyclopedia to be authoritative. But there cannot be an authoritative conversion of RGB to CMYK, unless arbitrary decisions are made about color spaces. You will find that CMM does indeed make a significant difference to the answers that you get, another nail in the coffin of an authoritative answer. The standards define the format of ICC profiles and their broad semantics, but not algorithms. The CMMs give different answers; that's why we have different CMMs! Notinasnaid 12:42, 7 September 2006 (UTC)

Essay on WP's direction

For anyone not watchlisting the /Principles page, check out Wikipedia talk:WikiProject Color/Principles for John Reid's very interesting comments on color, and the state and direction of our articles. --Quiddity 04:37, 31 August 2006 (UTC)

Misc remarks

I've just added myself to the participants' list. While you may have seen my name in some of the color articles' histories, I don't expect to become a very busy contributor. BTW, I'm affected by red-green color blindness and use WhatColor to make sense of some color articles.

I also added an interwiki to the (sort of) sister project on de:, de:Wikipedia:WikiProjekt Farbe.

You may want to have a look at , which (using template de:Vorlage:Farbdarstellung) is a disclaimer and rough check concerning the color reproduction of computer monitors. It's far from actually having a color calibrated monitor but it directs the attention to the problem.

Speaking of templates, I'm of the impression that some color articles suffer from navigation template overkill, e.g. Orange (colour). I especially dislike Template:EMSpectrum. --Pjacobi 07:25, 31 August 2006 (UTC)

Shades of... Subsections

I've noticed that many of the colour articles are sprouting huge list of colours in there shades subsection. Specifically: Azure (color). Lavender (color), Magenta, Red-violet, and Violet (color). I am wondering if this is the right approach consider the existence of shades navigation templates. PaleAqua 20:59, 18 September 2006 (UTC)

Please put such subsections only in ones that are the main colors for their categories; Red, Orange, Yellow, Green, Cyan, Blue, Violet, Brown, Pink, and Gray. Georgia guy 21:18, 18 September 2006 (UTC)

It's not PaleAqua, it's User:Keraunos, who has been doing a ton of work on color articles the last few weeks. I'll invite him to join the color project, and respond to this thread. --Quiddity 22:15, 18 September 2006 (UTC)

As noted further up the page, and in our recently written project principles, we intend to remove the vast majority of color swatches and templates, as there is no official/legitimate/singular definition for any color shade.

As pertaining to these "Shades of ..." sections in particular, they are inappropriate and should be removed. The values are only relevant at web color related articles. --Quiddity 22:44, 18 September 2006 (UTC)

My idea in putting the shades in was to have shades for each color on the color wheel. This morning I did the shades for Azure. I think it is very interesting for people to be able to see all the different shades of sky blue. The flag of Sweden, the color Azure itself, the UN flag, the Ukraine flag, various web colors, and Crayola sky blue, all represent "sky blue" but the colors are all slightly different and this way people can compare them. Keraunos 12:03, 19 September 2006 (UTC) This evening I added a color box to the article Rose and the color bands below. I figured out this morning mathematically how to distinguish the Rose group of colors from Red and Magenta. I got the idea of putting the color box in the Rose article from whoever put the color box in the Azure article. They formulated the formula for Azure such that it was exactly halfway between Blue and Cyan. So I did the same for Rose, formulating the formula for Rose so that it is exactly halfway between Red and Magenta. Pale Aqua said in a discussion that he thought Rose (meaning the color between Red and Magenta) was one of the tertiary colors, so I put in a color box for it. Keraunos 12:09, 19 September 2006 (UTC) I put in the article on the color Rose how to figure out mathematically whether a color is in the Red, Rose, or Magenta group. I am very interested in color. I have spent years studying many different color wheels and color systems. The Rose group of colors has been neglected over the years and it needs to be emphasized. Keraunos 12:14, 19 September 2006 (UTC) I have a question. Once you have the hex code and the RGB values, how do you calculate the CMYK values? Keraunos 12:21, 19 September 2006 (UTC) Is that color calculator pictured above on the Macintosh System 10 (which I have), or does it have to be downloaded for a fee? Keraunos 12:25, 19 September 2006 (UTC)</includeonly>

Please do not attempt to calculate CMYK values! See discussion and proposals above. Also, I think that calculating RGB values in this way is not desirable, though I don't want to discourage improvements to the calculators. I believe the only RGB colors used and reported are those in well defined international standards (again, see above). Notinasnaid 12:41, 19 September 2006 (UTC)

I don't recall saying anything about the color rose. I understand the temptation to put in color coordinates. I put in some swatches that I have since removed myself. BTW to give an example of of approximate colors are the system I'm working at right now has two different manually calibrated LCD monitors, but when I drag a one window to the other monitor I can see immediate color shifts. The utility is located in /Applications/Utilities/ under Mac OS X. Though using it to calculate CMYK values likely violates WP:OR. PaleAqua 16:10, 19 September 2006 (UTC)

Infobox Suggestion

About these coordinates Color with only RGB
— Color coordinates —
Hex triplet		#ff0090
RGB^B	(r, g, b)	(255, 0, 144)
Source		Unsourced.
B: Normalized to [0–255] (byte)

About these coordinates Registration black

— Color coordinates —
CMYK^H	(c, m, y, k)	(100, 100, 100, 100)
Source		Unsourced.
H: Normalized to [0–100] (hundred)

About these coordinates Color with only HSV
— Color coordinates —
HSV	(h, s, v)	(220°, 240%, 120%)
Source		Unsourced.

About these coordinates Color with all
— Color coordinates —
Hex triplet		#ff0090
RGB^B	(r, g, b)	(255, 0, 144)
CMYK^H	(c, m, y, k)	(0, 75, 0, 0)
HSV	(h, s, v)	(220°, 240%, 120%)
Source		Unsourced.
B: Normalized to [0–255] (byte) H: Normalized to [0–100] (hundred)

About these coordinates Color with all & source
— Color coordinates —
Hex triplet		#ff0090
RGB^B	(r, g, b)	(255, 0, 144)
CMYK^H	(c, m, y, k)	(0, 75, 0, 0)
HSV	(h, s, v)	(220°, 240%, 120%)
Source		X11 color names
Color space		sRGB
B: Normalized to [0–255] (byte) H: Normalized to [0–100] (hundred)

About these coordinates Color w/source & Caveat
— Color coordinates —
Hex triplet		#ff0090
RGB^B	(r, g, b)	(255, 0, 144)
CMYK^H	(c, m, y, k)	(0, 75, 0, 0)
HSV	(h, s, v)	(220°, 240%, 120%)
Source		X11 color names
B: Normalized to [0–255] (byte) H: Normalized to [0–100] (hundred) What ever I want down here.

About these coordinates Green

— Color coordinates —
Hex triplet		#00ff00
RGB^B	(r, g, b)	(0, 255, 0)
Source		Unsourced.
B: Normalized to [0–255] (byte)

I've started designing what could be a replacement for Template:Infobox color. This draft template only shows the Hex triplet/RGB, CMYK, and HSV lines if they are provided when calling the template with those parameters. It also adds a space for linking to the source of the coordinates and a replaceable caveat at the bottom. The default caveat will probably need to be change to link to some place in the main article space. PaleAqua 01:11, 20 September 2006 (UTC)

Some of this design was based on discussion at this archived talk page. The draft template is at User:PaleAqua/Sandbox1. I also intentionally left out some of the category stuff from the draft so that it wouldn't pollute category pages. PaleAqua 01:22, 20 September 2006 (UTC)

This is good stuff. I think, though, that the box still isn't quite complete. It needs to give the color space in which the color values are defined, or (in rare cases) indicate that a standard didn't give one. For instance, for all web colors defined by W3C, the colors are in sRGB color space. This probably needs to be a separarte line e.g.
Reference color space: sRGB.
I also feel that we should not be having any colors without sources, so having info boxes that don't work without a source could be a good way to achieve this, since people will always want to be adding their favorite. Notinasnaid 08:33, 20 September 2006 (UTC)

I'll see about adding an optional color space line. I can also make the source field required, though my intent was to make this template be a drop in replacement for the existing one. I wonder if it could be adjusted so that it would automatically add color infoboxes without sources to a unsourced category so that they could be easily found and fixed. PaleAqua 08:52, 20 September 2006 (UTC)

Also: in the final case, perhaps we should be separating into two boxes the concept of "greenness", illustrated by this excellent picture, and which does not imply any particular color coordinates; and specific color co-ordinates. The case of green is an important one: observe that the article already includes two different color co-ordinates, from different sources. Linking on "source" to an explanatory page giving acceptable sources (as approved byu this project) would also be good. Notinasnaid 08:35, 20 September 2006 (UTC)

I see your point. It might make sense just to use separate thumbnail for the picture the exemplify the concept of a color. PaleAqua 08:52, 20 September 2006 (UTC)

I'll probably need better contents for the missing source text. I also added the colorspace field. If we can get an acceptable source list together I'll link source to it. PaleAqua 09:06, 20 September 2006 (UTC)

Colors on the Color Wheel

Responding to a message sent to me by PaleAqua that I was trying to define some colors too exactly, what I was thinking about was the fact that the color wheel is a circle with twelve points. In the Hue code (h) in the hsv section of the color box (which is derived from the Munsell color system), on the color wheel 0 degrees = red, 30 degrees = orange, 60 degrees = yellow, 90 degrees = web color chartreuse, 120 degrees = green, 150 degrees = spring green, 180 degrees = cyan, 210 degrees = azure, 240 degrees = blue, 270 degrees = violet, 300 degrees = magenta, and 330 degrees = rose. All of those colors are calibrated with formulas that match those coordinates (when I inserted the color box for rose a few days ago I engineered it so the RGB values would match the color at 330 degrees) except orange, which is calibrated somewhat too far toward the yellow instead of halfway between red and yellow. To have a scientific color wheel, it seems to me all the colors should be calculated to match the twelve points representing the twelve major colors on the color wheel, three primary (red, green, and blue), three secondary (yellow, cyan and magenta), and six tertiary (orange, chartreuse, spring green, azure, violet, and rose). Keraunos 09:54, 22 September 2006 (UTC)

I'm afraid I have to say, despite your excellent intentions and hard work, that this seems like original research, and as such can have no place here. Can you convince us otherwise? Notinasnaid 10:50, 22 September 2006 (UTC)

In addition, you are adding definitions for new varieties of color, such as in Green. An objective of this project is to make sure colors use sources as Wikipedia needs, including the values in color swatches. Can you provide authoritative sources (e.g. an international standard) for your colors and swatches? Notinasnaid 16:07, 22 September 2006 (UTC)

For the colors I added in Green, I simply used a Digital Color Meter (which is software for the Macintosh System 10 computer) to measure the RGB and accompanying hex values for the colors of the Irish flag and the flag of the association of Islamic nations, and I termed those colors Shamrock Green and Islamic Green, respectively. For the pigment green, I used the color meter on the green in the picture of the subtractive primaries in the primary colors article. As for the color wheel, as I said those degree figures I used are not original research but taken from the Munsell Color System and they are the degrees that are used in the hsv codes in the color boxes. For example, if you look at the hsv codes for the colors in the green article, you will see that h is equal to 120 degrees, which means the position where green is on the color wheel. "h" means "hue" which means the various colors on the color wheel. Keraunos 16:24, 22 September 2006 (UTC)

I was staying quiet in the hope to get more discussion, but no luck? I feel that measuring colors is original research, and should be removed. Also, coining color names. The Irish color flag uses a green already defined in Wikipedia as a Pantone color; that color can be mentioned, but in my view should not be translated in any way. Is there really nobody else who wants to contribute, and perhaps suggest a way in which Keraunos' wealth of experience and patience can be used in a way that is compatible with Wikipedia's policies? Or am I wrong: is measuring colors a perfectly acceptable way to create articles? Notinasnaid 18:13, 29 September 2006 (UTC)

I strongly agree that coining color names is Original Research. Any names that are used should be properly referenced/etymologised.

Measuring color is even more problematic, as the colors portrayed are not necessarily accurate.

Sorting out the color-shade articles is this wikiproject's most complicated task, and needs to be discussed at length soon. There is a lot that should be merged or deleted. --Quiddity 19:21, 29 September 2006 (UTC)

Project directory

Hello. The WikiProject Council has recently updated the Wikipedia:WikiProject Council/Directory. This new directory includes a variety of categories and subcategories which will, with luck, potentially draw new members to the projects who are interested in those specific subjects. Please review the directory and make any changes to the entries for your project that you see fit. There is also a directory of portals, at User:B2T2/Portal, listing all the existing portals. Feel free to add any of them to the portals or comments section of your entries in the directory. The three columns regarding assessment, peer review, and collaboration are included in the directory for both the use of the projects themselves and for that of others. Having such departments will allow a project to more quickly and easily identify its most important articles and its articles in greatest need of improvement. If you have not already done so, please consider whether your project would benefit from having departments which deal in these matters. It is my hope that all the changes to the directory can be finished by the first of next month. Please feel free to make any changes you see fit to the entries for your project before then. If you should have any questions regarding this matter, please do not hesitate to contact me. Thank you. B2T2 23:27, 25 October 2006 (UTC)

Many of the colour articles need to be wikified

Hi, I've spent a bit of time organizing and re-wording colour related articles. Many of the articles I have come across:

Need to be better organized:
- Variations of a colour should be in a section; ideally this could be:

==Variations of [colour name]==
===Variation one===
•••
===Variation two===
•••

Section heading should use "Sentence case" for headings; not "Title Case"
Use capitalization where needed; colour names should not have a title case, so please use lowercase (unless it begins a sentence)
A single colour infobox should appear at the top to show what a typical version of the colour looks like. When in doubt, choose either the "official" variation, or the web-safe
Try to not discuss too much about colour theory in each article, as there are excellent articles already that describe these (which are well linked to each colour article)

+mwtoews 15:43, 27 October 2006 (UTC)

One stub of many

This article could use some work as we'd like to use it as a reference to solve an edit war. Also, I'm fairly sure there are plenty of other stubs to be taken care of so you might want to make that a goal as well. I'd also appreciate if maybe some of you could help with the consensus on that talk page as well ^^ GrandMasterGalvatron 16:02, 27 October 2006 (UTC)

It's good that it doesn't try to define RGB colors, but it needs sources. No article without sources can be thought of as a good model for another. Notinasnaid 16:04, 27 October 2006 (UTC)

I've found this http://www.fancifulsinc.com/mall/Metalic_Paints.asp for the color. There was also a good page about horse coats I came accross here: http://www.thundervalleywalkers.com/puffchampagnes.html GrandMasterGalvatron 16:19, 27 October 2006 (UTC)

Intro paragraphs

Please take extra care that the intro section, above the first section, is useful to the casual browser and well-written. The intro for purple has been pretty awful for a long time; particularly so for more than a month. +sj + 05:35, 13 November 2006 (UTC)

Spelling: Color vs. Colour

I recently realized that there's a problem in most of, if not all, the templates for this project. Wikipedia's policy on spelling says that although Wikipedians should not try to standardize the entire encyclopedia to either American or British spelling, the spelling should be consistent within any given article. This means that the Wikiproject Color templates don't comply with this - when used on pages like Orange (colour), they conflict with the standard spelling of the article by using "color" instead of "colour". I'm going to, in the near future, go through the templates to add a parameter to each that allows the spelling to be changed from "color" to "colour", with the default being "color" (the original, and corresponds to the name of this project) if no option is selected. Since this option is so important for this project, I hope that use of the soon-to-be-created spelling parameter spreads beyond the Orange (colour) article, which I will correct myself. Nihiltres 17:11, 13 November 2006 (UTC)

I ran into this problem with the Additive Color article. The article title uses the American spelling, while the spelling throughout was the British spelling. I went through and changed the spellings to be consistent with the article's title, but the inter-linked article titles on the page are still a problem. 68.44.28.9 20:46, 15 April 2007 (UTC)

Articles for deletion

I thought I should let the project know about an article I nominated for deletion, since it purports to be about a color. Lesley (color), Wikipedia:Articles for deletion/Lesley (color). Notinasnaid 19:53, 18 November 2006 (UTC)

New nominations for deletion: Byoblue‎, Gray-Tea Green Notinasnaid 18:03, 15 December 2006 (UTC)

inappropriate use of color coordinates

Use of sRGB as a connection space neutral ground for RGB triplets is laughable. When specifying colors that relate to the physical world, there are radiometric and photometric values one can use, that have been designed specifically for that. Photometric values such as those standardized by the CIE standards body are the ones that should be used. They won't be bound to the specific gamut of display color spaces such as sRGB. Many natural pigments fall completely outside the gamut of sRGB, and henceforth, your entire SRGB-normalized reference system collapses like a deck of cards. Being aware of the CIECAM02 model of color appearance would help convert the CIE coordinates into practical RGB values for users who wish to simulate naturally-occuring colors under specific constraints of illumination, and avoid some surprises with metameric behavior.

As it currently is, the system used in Wikipedia is completely inadequate.

There is plenty of reference available on the subject, from books by Billmeyer & Salzman for the less technically enclined, to more recent works. An authoritative source on the topic is definitely Prof. Mark Fairchild of the Munsell Color Science Laboratory in Rochester, NY.

If you refer to previous proposals, on which there has been sadly little movement, the proposed policy is that the only color values to appear would be those defined in standards, and the only values shown would be the value in the standard. For example in the case of the CSS standard, all colors are predefined to be in sRGB. It would be as wrong to convert these to XYZ or Lab as it is to convert them to CMYK. I consider all color conversions and color measurements to be original research (and hence fobidden) though this has not been tested. Welcome to the project! Some more people who understand color are definitely needed. Notinasnaid 13:20, 21 November 2006 (UTC)

Thanks for welcoming me. Converting sRGB triplets to XYZ would be ridiculous indeed. But if there is a ban on original research, where do these incredible color coordinates come from ? Who measured Prussian Blue to yield this sRGB triplet, and under what conditions ? Prussian Blue will look totally different under daylight and under tungsten illumination, as you know. These values simply cannot be trusted by anyone. There are attempts at standardizing (read, open source) color names and naturally-occuring color phenomena. With these standards the original color measurement methodology is standardized and published. Without some sort of standardization nothing prevents one from introducing any fancy variation of "pink" and specify values which will absolutely never yield the naive inventor's color appearance.

Stablepedia

Beginning cross-post.

See Wikipedia talk:Version 1.0 Editorial Team#Stablepedia. If you wish to comment, please comment there. ★MESSED ROCKER★ 03:29, 26 November 2006 (UTC)

End cross-post. Please do not comment more in this section.

Alternative approach - greater use of photographs?

I realise I am walking into a large project anew, but it occurs to me that with a subject as inherently complex as colour we run the risk of building an enormous scientific resource with plenty of original research and scientific contention, which is not what WP is about. Many colour articles at the moment seem ridiculously complex and largely unhelpful to the casual visitor who needs to clarify what is meant by a certain term.

Would it be worthwhile to consider making photographs a much more integral part of colour entries? If a visitor wants to know what is meant by taupe, for example, it strikes me that it would be most effective to see a dozen or so thumbnail photographs of taupe items, along with perhaps a basic explanation of the common interpretation of the word. Sure, scientific data could be included, but I contend that it shouldn't form the essence of the article. Multiple photographs would have the advantage of 'averaging out' lighting conditions and monitor inaccuracy, and it would allow the visitor to see various different shades that may all be considered appropriate to the colour in question, in 'the real world'.

I don't feel it is necessary or even desirable to delete entries that may not be 'significant' enough, particularly as there are many unofficial or local colour names. Surely it is enough to cite a source of an early use of the word to describe a colour, to allay the creation of frivolous and self-serving fictional colours?

Brewabeer 00:33, 13 December 2006 (UTC)

Beer color help request

On Wikipedia talk:WikiProject Beer there's been some discussion about beer color recently, as we work on expanding Beer style.

Beer color is graded using three related systems: Standard Reference Method, degrees Lovibond, and EBC. We know what the SRM ranges are for any particular beer style. What we'd like to do is take those numbers and turn them into little boxes (HTML cells, in my ideal world) filled with the appropriate color, to enhance the beer style articles.

What's a simple way to go about this? --Stlemur 20:10, 28 December 2006 (UTC)

Wikipedia Day Awards

Hello, all. It was initially my hope to try to have this done as part of Esperanza's proposal for an appreciation week to end on Wikipedia Day, January 15. However, several people have once again proposed the entirety of Esperanza for deletion, so that might not work. It was the intention of the Appreciation Week proposal to set aside a given time when the various individuals who have made significant, valuable contributions to the encyclopedia would be recognized and honored. I believe that, with some effort, this could still be done. My proposal is to, with luck, try to organize the various WikiProjects and other entities of wikipedia to take part in a larger celebrartion of its contributors to take place in January, probably beginning January 15, 2007. I have created yet another new subpage for myself (a weakness of mine, I'm afraid) at User talk:Badbilltucker/Appreciation Week where I would greatly appreciate any indications from the members of this project as to whether and how they might be willing and/or able to assist in recognizing the contributions of our editors. Thank you for your attention. Badbilltucker 16:46, 30 December 2006 (UTC)

color article style guide?

there's alot of variance in structure and content among the different color articles... is anyone working on any semblance of a style guide in this WikiProject? Blueaster 23:21, 2 January 2007 (UTC)

Proposed rename of RGB color space.

Unlike articles like sRGB color space, the RGB color space article is not about a color space, whose name is RGB. Rather, it is about a number of color spaces, each from the RGB color model.

I believe the title of this article may help to reinforce the mistaken idea that there is a single color space called RGB. I therefore propose renaming it in the plural, "RGB color spaces" (with redirects sorted out - RGB color space would remain a redirect - and minor rewording). Seeing the plural in lists of color spaces should emphasise the point.

Comments? Notinasnaid 12:55, 12 January 2007 (UTC)

Sounds good. —Quiddity 11:58, 16 January 2007 (UTC)

Done. Notinasnaid 12:19, 16 January 2007 (UTC)

Color category names

I have been tidying up category membership for Category:Color and its subcategories. I find there is an additional Category:Colors. This is reasonably tidy, but it took a while to figure out the intended difference: the former is about colors in general, the latter about specific colors identified by name. I propose to change the second category to be called "Named colors". (Note that articles are not supposed to be in both a category and its parent category, so articles should be in only one of these). Comments? Notinasnaid 11:11, 16 January 2007 (UTC)

Wikipedia's category scheme(s) generally terrify me; it's an infinitely recursive ontological free-for-all! So any order you can bring to a small segment would be welcome :) —Quiddity 11:55, 16 January 2007 (UTC)

Infobox Suggestion Revisited

I just was a little bold and made the infobox changes live. I'm also started going through the web color pages and adding sources to the infoboxes. PaleAqua 02:01, 4 April 2007 (UTC)

Finished most of the core 16 Web colors, save for Black and White. I've been thinking of making sRGB a standard option of the template however possibly with a sRGB=1 parameter instead of how I've currently done it. PaleAqua 06:18, 4 April 2007 (UTC)

Various proposals pertaining to the Shades of ... templates & categories

Browsing through the talk pages of the Shades of ... templates & categories I found a couple of proposals put forth. I've also put a few extras forth myself. I'll summarise them here. J i m p 17:14, 5 April 2007 (UTC)

Move to Variations/Varieties of ...

(Discussion)

It had been suggested that Category:Shades of orange be moved to Category:Variations of orange. Were such a move to go through all Shades of ... categories along with their associated templates should follow. I don't know that this is necessary but suggested that Varieties of ... would be better. J i m p 17:14, 5 April 2007 (UTC)

To me "Shades of ..." sounds better than "Variations of ..." or "Varieties of ...". I do understand the concern with meaning of shade, but think that it covers the current usage. Merriam-Webster online has these definitions for shade 8 a : a color produced by a pigment or dye mixture having some black in it b : a color slightly different from the one under consideration PaleAqua 19:15, 5 April 2007 (UTC)

That has pretty much been my take on it too. As I'd noted, Cambridge Dictionary (and Wiktionary) also acknowledge this meaning of shade (m-w's b), though ADH doesn't mention it. I say leave them & save (a lot of) work. J i m p 19:50, 5 April 2007 (UTC)

Moving/splitting to Shades of purple

(Discussion)

I have proposed that the Shades of violet template & category be moved to Shades of purple, the latter being the broader term. Otherwise we should probably split Shades of purple out from Shades of violet. J i m p 17:14, 5 April 2007 (UTC)

New layout for shades without swatches

(Discussion)

I have proposed a new layout for the Shades of violet template whereby those colours without swatches are integrated into the main list with the words no swatch written where the swatch would be. J i m p 17:14, 5 April 2007 (UTC)

What to include as a shade

(Discussion)

It had been noted that some of the shades featured on the Shades of yellow template did not look yellow. I agreed and proposed that the template be trimmed. Such a precident might affect other templates also. J i m p 17:14, 5 April 2007 (UTC)

I've been worried myself that we are ending up with a large amount of unsourced colors in both the templates and in the articles themselves. I've seen a few articles which have more than a few screenful of color bars. Also there are color pages that will have 3 or 4 "Shades of ..." navigational templates at the bottom of the page. I think a lot of this needs to be trimmed. PaleAqua 19:19, 5 April 2007 (UTC)

I can understand there being a few articles with two templates but to find a reasonable justification more than this would be rare. I had done a little trimming of blue (article & template) but it may be time to put the nail scissors away and unsheath the sword (in fact I'm already brandishing it at the Yellow). J i m p 20:00, 5 April 2007 (UTC)

Organisation of shades

(Discussion)

It had been suggested that the shades featured on templates be organised by colour rather than by alphabetical order. It was suggested that they go from light/saturated to dark/unsaturated. I like the idea in principle, of course, this would be no easy task & may indeed prove impossible. However, I suggested that light/saturated to dark/unsaturated go along the horizontal with hue (i.e. magenta to red to orange to green to cyan to blue to indigo to violet to magenta to red) going along the vertical. J i m p 17:14, 5 April 2007 (UTC)

I would personally recommend sorting by hue, then saturation, then value. Although not perfect for every occasion, we would have a less arbitrary and consistent way of organizing the vast multitude of colours.--88.110.211.3 23:26, 1 August 2007 (UTC)

Comments

Hi, just a quick note. As mentioned in a few places on this talkpage, there was a general consensus that we need to move away from having the "shades of" templates at all, and work to merge and/or delete the majority of the color-stub articles. Primarily because most of them are only applicable to web color, and others are unlikely to ever be expanded.

So, quite seriously, put away the sword and unsheathe the lightsaber. We need a serious, stern, thorough cleanup of the color stubs and color templates. Something like 75% of the Category:Stub-Class color articles should be deleted (or merged if appropriate), and quite a few of the Start-Class articles, too. Hope that helps. --Quiddity 18:57, 6 April 2007 (UTC)

Merging is probably a better idea whenever sources can be found. Articles can be merged to a single "Shades of" article, and perhaps the heraldry colours would be best merged into a single heraldry colours article as well. Otherwise, there are some ridiculous colours out there, without proper sources: Girlsnberry anyone? Nihiltres 13:22, 10 April 2007 (UTC)

Opinion on removing the "Shades of...." color charts from an article?

Does it seem generally acceptable to remove the "Shades of..." charts from an individual color article? Magenta has three of them, for red, pink, and violet, including about a hundred colors. The way they're used in the article is just hanging at the bottom of the page, with no section header or explanatory text. I proposed here in magenta's talk section keeping magenta listed in the categories, but removing the actual charts at the bottom of the page. A couple people agreed, but someone suggested discussing it in a broader color article area, otherwise they would likely be replaced for consistency between color articles. Thoughts? -Agyle 21:20, 21 September 2007 (UTC)

I'm replying here as we try to move discussion from talk:magenta to this page. Someone in t:m said "The Wikipedia user needs them to find their way to the other colors of the same type. You need to think of the convenience of the Wikipedia user." I agree that they're helpful and informative in navigating colors, but an article's See Also section can link to the "shades of" articles to provide the same access. -Agyle 22:11, 22 September 2007 (UTC)

Wikipedia:WikiProject Color/Sources for Color Coordinates

I've started a page to serve as a place holder for color coordinate situations. PaleAqua 06:33, 9 April 2007 (UTC)

I am delighted to see the new info boxes, which show real progress in the articles. But I would like to initiate discussion of something I see as a real problem: the choice of sources. There is much enthusiasm for adding sources, but all sorts of things are being added as sources, including articles in other Wikipedias and forum posts, which don't seem to meet Wikipedia:Attribution#Reliable sources.

I think the project needs to go further than the policy page and decide here and now what is an acceptable source. This seems to be what the link above is about, but what are the controls on it?

Should the project only accept colour values from a list of approved sources? If so, apart from the HTML/CSS specifications, what if anything shoould be approved. The issue here is that countless people have spent time creating names for colours; there is no obstacle to publishing contradictory lists. X11 is a particular problem because while it is widespread and important, the article itself acknowledges there is no standard table.

What do you mean, there is no standard table of X11 colors? The Wikipdedia X11 color list clearly indicates the hex code and RGB value for each X11 color. I tested all the X11 colors with a Digital Color Meter and all the codes are rendered accurately. Keraunos 14:07, 11 April 2007 (UTC)

There is no problem in principle with multiple sources giving different values for a colour. If we consider the sources acceptable, we simply show them separately, as in cases like the HTML and X11 colours which have the same name but different values.

I would propose that once the project has ruled on acceptable sources, that all unsourced colour values be removed. Not necessarily, the colour boxes or articles, which could for example use photos or description.

Comments? Notinasnaid 11:54, 11 April 2007 (UTC)

Supplement. I think we also need to be decisive about whether or not it is legitimate to show a color value based on measuring color digitally in images, pictures or scans of printed color. It may be that excellent sources like the two encyclopedias of color cannot be used to derive color values. Notinasnaid 12:01, 11 April 2007 (UTC)

You asked "What do you mean, there is no standard table of X11 colors? The Wikipdedia X11 color list clearly indicates the hex code and RGB value for each X11 color." Well, Wikipedia is not a valid source. What would be needed was a "standard" X11 color table. The X11 color table exists in X11 as the rgb.txt file (see X11 color names). X11 color names specifically says "These charts are not a standard set of colors to be found on any X11 system". The reference to "a typical rgb.txt file" suggests that the file itself is nonstandardised.

So long as we deal with CSS colors, we don't have a problem, they can be traced to a W3C standard. Do you know of a true standard set of X11 colors: not a particular set, a set that is somehow blessed by a standards organisation? Notinasnaid 21:23, 15 April 2007 (UTC)

For the X11 colors I would suggest the rgb.txt from X.Org would be the ones to use as they are the defacto stewarts of the X11 standard. However the rgb.txt file does not specify a RGB color space which is a problem.PaleAqua 22:20, 15 April 2007 (UTC)

As to using digital colorimeters to get values there are several problems. These are just a few of them 1) Not all colors can be represented in a given RGB color space, so many colors are displayed as an approximation. 2) Verify few websites actually include color spaces in there design, so browsers end up making a guess. Many choosing something like sRGB. 3) Web browsers will often adjust colors based on calibration settings. It's possible that the same web page will show different RGB values in a color meter depending on the browser and calibrations.PaleAqua 22:20, 15 April 2007 (UTC)

I have noticed that a lot of the color info boxes cite a forum post on a Battlefield 2 website. This is utterly unacceptable!! Where did people get the idea that a source is a source even if it is not the original source?? All the colors listed on that forum post are simply the W3C standard for HTML colors. I propose that all the links to the BF2S website should be redirected to the listing of HTML colors so as not to confuse. —Preceding unsigned comment added by 216.174.23.10 (talk) 17:33, 6 September 2007 (UTC)

Please do. If you could check them for correctness while you're at it, that would be very helpful. Dicklyon 20:16, 6 September 2007 (UTC)

Spectral Colors

I started a framework to add information to color articles for spectral colors. While I understand the wavelengths and frequency ranges are approximate, I have seen several articles using different ranges. For example: Orange (colour) gives the wavelength as ~585-620 nm, Color gives 590-625 nm, and Visible spectrum gives 590-620 nm. Any suggestions on how to handle this? PaleAqua 00:50, 20 April 2007 (UTC)

Just find sources, and preferring sources with wider ranges is probably appropriate, since color interpretation plays into range. Nihiltres^(t.c.s) 01:01, 20 April 2007 (UTC)

I suggest trying to find a reputable book which lists such spectral ranges. The color names don't really mean much of anything; they are arbitrary and highly culture- and observer-dependant. --jacobolus (t) 11:12, 4 May 2007 (UTC)

Lack of Spectral / Photometric color definitions, Lack of information of color naming schemes

I'm sorry to burst your bubble, but color naming is a science. It's as old as language, and A.H. Munsell has produced since 1905 the first rational color naming model. There are standards usable for notating color values, which are not at all represented in the WikiProject Color body.

Furthermore, there is little in the articles, besides chemistry, that relates colors to physical phenomena - spectral information is the only truly reliable information we can have of color, as it is independent of the illuminant and the reproduction system. When scientists measure pigments and dyes, as well as other forms of color phenomena (self-luminous, etc..), spectral analysis is the approach. Instruments like spectroradiometers, tele-spectroradiometers, monochromators are used precisely for that. Then, there are color equations one can use to transform spectral curves into various simulations of the original color phenomenon. Any conversion of a spectral function into the shorthand that tri-stimulus systems like XYZ, Lab, xyY will have specific shortcomings, not least which is metamerism. Further than that, linear photometric spaces such as XYZ have to be converted into output-specific color spaces such as sRGB or AdobeRGB, qualified under a certain white point and gamma function, in order to allow reproduction (simulation) with target devices (soft copy devices such as LCD diplays, or hardcopy devices such as inkjet printers). By the time all thes conversions have taken place, the limited dynamic range and gamut of output devices make it nearly impossible to truly reproduce the natural phenomenon that was characterized with a spectral function. In order to make practical reproduction possible, the viewing conditions have to be standardized as well. When a color is characterized in one of the articles here, the specific viewing conditions under which the approximation can be trusted absolutely has to be specified, otherwise naive users will instantly be misled. Someone who wishes to emulate Cerulean Blue in a Photoshop painting will seek the information at Wikipedia, and dial in, credulously the values in his airbrush color mixer... meanwhile her screen isn't calibrated, or is calibrated at a gamma of 1.8 and AdobeRGB color space, she is working strictly in a bright room with windows facing north, and proofs her colors with a consumer grade Epson inkjet printer. The results will have nothing to do with Cerulean Blue oil paint on a canvas. Users should be taught that color definitions contained herein can only be used under strict evaluation conditions, which should be enumerated. There are international standards for these evaluation conditions, depending whether one works in the electronic (soft copy) world, television, or in the press. Use of these standards would lend Wikipedia some credibility in this sector.

The strong emphasis seen here on using the very limited sRGB color space is ill-advised. sRGB is a display color space, not a connection space or a perceptual space. You are right in assuming (if that is the root assumption) that the world being roughly divided into 96% Windows PCs and the rest, Wikipedia users are likelier to have their displays unknowingly preset to an sRGB color space at a gamma of 2.2. But for 99% of these people, the screen is uncalibrated anyway. sRGB was not intended to specify colors in a way that is device-independent. There are color models designed specifically for that. When you use RGB, HSV, Hex triplets and so forth, you have to state very explicitly what conversion function were used, what rendering intent was used, and what the viewing conditions should be to trust the color appearance to within a specified degree a accuracy (less than 5 delta-e ?).

If you're not a color scientist and are interested in the topic, please have a look at publications from the world authorities on the subject, the Munsell Color Science Laboratory, and the Commission International de l'Éclairage. --Ppanzini, 21 November 2006

I don't dispute your science, but I content it is not relevant when we report what is in a verifiable and important standard like CSS. You can read the CSS standard, and it says that in their specific context, the word 'Red' has a meaning in terms of sRGB. Wikipedia should not say any more than that, and not seek to define what 'red' means, but it would be wrong to exclude important standards and what they say. Wikipedia's job is to report what verifiable sources say. That's why articles on creationism and evolution can coexist. Notinasnaid 13:56, 21 November 2006 (UTC)

But instead of taking an argumentative position on this, let's try and find common ground. What would you propose to see in an article called "red"? I do know that I don't want to see what is there at present. Notinasnaid 14:00, 21 November 2006 (UTC)

I have to agree with the original poster here. What the CSS standard says is of very little encyclopedic value, and while there are scads of articles with made up hex triplets created by this wikiproject, all of the articles about color science, and all of the articles about historical color systems (when they exist at all), are woefully inadequate. The organization of colors into "shades of" {blue, brown, cyan, gray, green, orange, pink, red, violet, white, yellow} is at best unscientific and misleading. This wikiproject seems to be more about inventing hex triplets for arbitrary color names, than about explaining color theory in any encyclopedic manner. In short, most of the activity of this wikiproject seems like a waste of time, both of editors and readers.

I realize this sounds like harsh criticism, and I don't mean to offend, but it's not like online resources about legitimate color theory are impossible to find. I myself am no expert, but was able to find dozens of useful articles online.

I've been trying to improve the Munsell color system article (here's a diff of changes), about the system on which modern color science was founded, and I added an image and cleaned up some of the prose on the Lab color space article. But really all of the articles about color systems (RGB/CYMK/HSV/HSL/Lab/Luv/XYZ color model/space, color temperature, additive/subtractive color, hue, saturation, gamut, etc. etc.) are currently full of vague, unscientific, and misleading claims, with an over-emphasis on formulae (which nearly no readers will be interested in or find uses for), and a heavy bias towards RGB/CYMK, even though these models have the least to do with color science. It would be great to have some other wikipedians try to help fix up the color-related articles to be presentable.

What is needed is sourced, verifiable, and precise historical and scientific explanations. All too much of the current information is unreferenced, vague, inaccurate, and misleading.

—jacobolus (t) 11:05, 4 May 2007 (UTC)

Also of note, John Reid's comments on Wikipedia talk:WikiProject Color/Principles have been around since August, but seemingly no action has been taken in such spirit. Specifically, I couldn't agree more with the conclusions at the bottom:

“Almost all of the hundreds of individual color articles are pointless.…”
“… We have not the capability to duplicate or subtitute for these tools; the attempt deceives the public and arouses the contempt of experts.”
“…Merge and redirect…”
“Banish ugly, bulky, misleading {{Infobox Color}}. Readers who want bullshit numerical values for colors can get them from List of colors, which should be linked from every color article. Templates such as {{Shades of orange}} are acceptable. I do think they need work, though. For one thing, most of the shades shown shouldn't even have articles. They should all be derivatives of one standard template and that should be made more compact and attractive.”

(—John Reid at Wikipedia talk:WikiProject Color/Principles)

—jacobolus (t) 11:31, 4 May 2007 (UTC)

While I think I dislike many articles as much as you do for many of the same reasons I have to take issue with your statement " What the CSS standard says is of very little encyclopedic value". It may have very little scientific value, but this isn't a work on science, but on, well, everything. Color articles include information on linguistics, history, art, politics, sport... every one of these topics is encyclopedic, though of course much in the color articles is not.

Web design is certainly enclopedic, and so is what the CSS standard says. Now, what the CSS standard defines as "red" is just what the CSS picked a name for. It isn't a better or more absolute value of red than any other, but it is sourced, verifiable, and very important (to people working in that field). No article should, however, simply say "this is red". But I have no problem with an infobox explaining that "this is the CSS web color red". It is also, interestingly, the only color definition we can be sure of displaying correctly in a browser, since it is defined by how it displays in a browser. Different people see different "CSS reds" but they are all seeing exactly what the CSS standard specified. Notinasnaid 21:41, 8 May 2007 (UTC)

Taking issue with something you quote: "Readers who want bullshit numerical values for colors can get them from List of colors, which should be linked from every color article." No! An enclopedia should not have little corners of bullshit, out of site and attention of people who care (though Wikipedia has many such places today). If made up color combinations aren't encyclopedic, they aren't encyclopedic in any article.

However, perhaps there is some merit in showing what some source has identified as being a named color; there are a number of reference works which have done this. I have a problem, however, with assigning any numeric code which was not in the reference. A photo, which is clearly understood not to be an exact representation of a color may be the answer. Maybe even (going out on a limb), a shade in a box which displays but has no hex value shown. Notinasnaid 21:45, 8 May 2007 (UTC)

I don't think there's a problem with saying, somewhere on the page, that what css calls "red" is actually such-and-such an sRGB triplet, which translates to such-and-such values in L*a*b* or L*u*v* or whatever under such-and-such illuminant. But that's not what the currant articles do. They say: "Red: #ff0000", which is just completely arbitrary. It would be better to say, “Red is a hue corresponding to the spectral colors of wavelength ~625-740 nm. This is approximately L*a*b* hue angle

h_{ab}

-2°–2°” (or whatever; I'm making that number up). Then we can show some slices of the L*a*b* space (converted to sRGB for display) of those hue angles, and users can see what sort of colors are being discussed. --jacobolus (t) 22:12, 8 May 2007 (UTC)

Moreover, in response to “No! An enclopedia should not have little corners of bullshit, out of site and attention of people who care.” the point is that these colors are only encyclopedic in the context of CSS colors. Designers don't go looking through Wikipedia's color articles to find the hex values and names of css colors. That would be silly: there are better resources for that. But if wikipedia does want to provide such a resource, it should be on one page, where it's actually of use to someone. Not spread out among the various hundreds of color articles, where it only serves to mislead readers. --jacobolus (t) 22:21, 8 May 2007 (UTC)

Couldn't have said it better myself Wikidan829 22:50, 8 May 2007 (UTC)

RGB, HSV, etc

(reposted from the red talk page) Are these things really necessary in the color infoboxes? Are they accurate enough? Sure, while #FF0000 might be red, so is #AA0000,and even #110000, just not as bright(meaning not as many photons of that wavelength are being emitted). It is still red regardless. I'm not really for deleting things, but is there a good reason for having this? This is an article for the color "Red", I don't see why the computer application of colors seem to take a sort of importance in them. Wikidan829 16:45, 8 May 2007 (UTC)

No, they're misleading. Color names don't have any precise particular definition as understood by most people. If someone wants to add information about the "css named color called red", that would be fine, but pretending that sRGB's FF0000 is at all special, or deserves a place in the top of the "red" article is a mistake. I really liked the idea of just using pictures to represent color names, instead of swatches or numbers. If we want a technical number, a range of wavelengths for the hue would be the only appropriate choice, IMO. --jacobolus (t) 17:35, 8 May 2007 (UTC)

I like the idea of using photos instead of swatches as well. Not to say that the HTML codes and other color spaces shouldn't be in the article. My disagreement is with the lack of real importance to the color that this information provides. It's relevant, not important. Does anyone else think differently? Wikidan829 17:42, 8 May 2007 (UTC)

There are a few good reasons to not support this proposal:

Brown is a variant of orange in approximately the same way that #AA0000 is a variant of red (#FF0000), so you create ambiguity by not distinguishing.
Colours as named are and should be defined as their absolute - that a colour is a range is generally understood. No one will doubt that #FF0001 is, essentially, red, but the absolute red, by which red is defined, is #FF0000. This is a caveat that applies to all colours, and should be on the infobox help page linked from the as I suggested on that help page's talk.
For colours which can be defined as wavelength or frequency ranges, there's a spectral colour attribute which can and should be filled in, which generally solves the problem presented.

Given the above reasons, to start, I think that what exists is reasonable. Nihiltres^(t.c.s) 18:25, 8 May 2007 (UTC)

Unfortunately, I disagree. Again, we're getting back to "what the color means to a computer". In the red apple, my red shirt, the red box that my new hard drive came in, in the entire universe of colors, what makes a computer representation(who says the computer recreates "absolute red" anyway?) so important, and so relevant, that it deserves to be in the infobox? In the natural world of color, why are RGB, HSV, and HTML(which is RGB) noteworthy in the infobox, the ultimate summarizing table of the article? Wikidan829 18:35, 8 May 2007 (UTC)

Because that's how they've been defined. I understand the concept of the range, and understand that no monitor is perfect, but colors are generally scientifically defined. For hex values, they're defined in HTML, CSS, X11, or SVG, all accepted standards, and should be sourced as such. If you want an image to define the color, add an image to the infobox in question, and it will display the image appropriately, in addition to the usual information. The natural world is limited to subjectivity - the carpet near me right now looks beige, but in reality it is made up of a plethora of knots in beige-like colors, rather than any standard beige. Since color standards do exist, we should implement them as valid information within the article -although standards are arbitrary, they are standards, and standards are appropriate. Nihiltres^(t.c.s) 19:10, 8 May 2007 (UTC)

This is the problem. They are accepted standards - in the computer world. I cannot go to the library, pull out a book on colors(before color monitors were even distributed) and find hex codes and rgb values from 0-255 to show "what the color should look like". Computers and software codes are not the ultimate authority on color. Of course a photo is subjective, but it is generally the best way to represent it. It is much more realistic than the "pure red" that a computer monitor displays.

The picture wasn't the original point anyway(see above). The point is that these values are entirely specific to the computing community, and unfamiliar (and more importantly, useless) to the general public who may be reading this article. It's great that you understand, I understand, and jacobolus understands what they are, but a lot of people who go to the article, will not. Wikidan829 19:26, 8 May 2007 (UTC)

Who is going to come to this article, look at the RGB or HSV values and say "wow, now I know what red is?" Wikidan829 19:48, 8 May 2007 (UTC)

I think that infoboxes as currently used give a false idea of what color is. The coordinates have the irony of both being over precise by giving exact values and under-percise because of calibration issues, that fact that not all colors fall neatly into any one particular RGB space. I do think, however, think that colors that are defined in a standard should be included. It is just that color boxes give too much weight currently to sRGB values. I think pictures, CIE values (or equations see Selective yellow and wavelength/frequencies for spectral colors should be the key items in an information box. Additionally information like is it a spectral vs extra-spectral may be good also depending on the case. Also information from the artistic POV information may also be appropriate. Is it considered additive, secondary, etc? PaleAqua 20:07, 8 May 2007 (UTC)

Exactly, going back to the note in my first post. For every color space, you will need to specify a range(this could include slight variations in green as well since the upper range of red can be close to orangish). The values are still specific to computers. As an artist, you wouldn't look up the RGB value to find the paint you want. I do not think that we should take the RGB and HSV stuff out of the article completely, I just think that the infobox is the wrong spot for it. It is not that important. Wavelength and frequencies are general enough, they do not relate to only computers, they are real, natural world numbers. These have tons more importance. Wikidan829 20:16, 8 May 2007 (UTC)

Yeah. I think if something more specific than just hue wavelength is required, some approximate range of hue angles, lightnesses, and colorfulnesses in a standard model like CIECAM would be appropriate. But hex triplets really don't belong there. --jacobolus (t) 21:27, 8 May 2007 (UTC)

"absolute red" is a completely meaningless term. There are much more chromatic "reds" than can be represented by sRGB.
And when you say "Since color standards do exist, we should implement them as valid information within the article -although standards are arbitrary, they are standards, and standards are appropriate.", what does that actually mean? There is no official "color naming" body. If you want, you can call that "css named red" (incidentally something like X = 0.44, Y = 0.22, Z = 0.014), but to say that is "red" is just misleading. Real color standards don't use names (unless we're talking crayola crayons here).
And one last thing. Nihiltres: tell me, which is orange and which is brown. ;)

--jacobolus (t) 21:29, 8 May 2007 (UTC)

more discussion about color values

I'd like to add to this discussion a point "Hey, we went over this". I produced a detailed proposal, which never seems to have been either rejected or implemented. See #Proposal above. It does cover a lot of this ground. Notinasnaid 21:47, 8 May 2007 (UTC)

I like points 1, 2, 4, and 6. Points 3 and 5 are pretty questionable. --jacobolus (t) 21:55, 8 May 2007 (UTC)

I agree. Why are you so tough on keeping these CSS and computer colors anyway? Are you a web designer? This information is useless to the general public. I don't want to accuse you of bias, but I want to know why you are so persistent over this. Why don't we list all the countries that have red in their flag while we're at it? Because it's unimportant, just like RGB and HSV values. Wikidan829 21:57, 8 May 2007 (UTC)

I see it fitting nicely into a "Red in computing" subsection. Wikidan829 21:59, 8 May 2007 (UTC)

Beyond that, do you have any thoughts about John Reid's comentary here? --jacobolus (t) 22:17, 8 May 2007 (UTC)

Most of this makes sense. I must concede most of the points raised. This doesn't mean, however, that we should abandon what we have so far. I think that as widely used definitions of particular colours, definitions in particular color systems are information relevant to an article, regardless of the fact that the average computer monitor is completely uncalibrated, and that this doesn't apply to much outside of digital devices - while we cannot guarantee that an end-user's monitor will display any particular colour accurately, we can assume that an ideal monitor would display the colour, and most monitors will display a reasonable facsimile of the colour, assuming that lighting is not incredibly bad and that the gamma and gamut of the monitor are both acceptable. As mathematical and scientific idealizations or standard values for a colour (rgb, HSV, spectral, for that matter all of the current infobox's values), such definitions are probably OK when taken in the context that they are standardized and often official definitions of a colour, based on the idea that a certain density of a wave of photons of a particular wavelength or mix of wavelengths will generally produce the same sensation in any given human brain attached to a normal eye, and not that the definition will necessarily produce that exact colour in all contexts. It is also worth noting that some information may be somewhat esoteric for the average reader, but is nonetheless reasonable - Wikidan829's comment that the information regarding computer standards "is useless to the general public" seems unfounded, since information is not necessarily useful to any particular person - I don't think that the infobox information in Mars, the featured article earlier in my time zone's day, is useful to me, but it is clearly important to the subject of Mars. A picture to give a general idea of what such a colour would look like is a good idea, but it's limited. Part of the problem is that colour is very subjective - It's worth noting the linguistic challenges of translating the idea of the perception a colour, notably in the case of green and blue, which share words in some languages, not to mention that optical illusions can alter the perception - a brown colour can look orange when surrounded by dark objects, because of the brain's tendency to correct for the effects of low light in colour, as jacobolus noted. This doesn't mean that we should merge the Brown and Orange articles, however - the idea seems somewhat absurd, since brown and orange are, to the average person, "obviously" different. What I propose is that mathematical, scientific, and digital definitions of color be kept, since as definitions only, wouldn't they be acceptable? The only alternative I see to allowing idealized color definitions is suggesting that the only definition of a colour we can have is through an image, which may, for the same reasons, be as corrupt a definition when displayed on a computer monitor, or, worse, a mobile browser. If there's anything that needs to be added to the colour infobox according to this discussion, all I can think of are:

Standard inclusion of the easy-to-use "image" parameter with an image giving a wider sense of what the colour looks like.
An explicit disclaimer on the infobox (not just a ) with a link to a more detailed discussion (ideally an article) on colour variability across media, devices and ambient lighting, showing that the colour definitions present are idealized, and ideally suggesting that colours may extend over particular ranges.
Notes of some sort indicating the use of a color definition - web colour definitions are notable, but implying in any way that they impose upon reality is, as some have pointed out, ridiculous.

I don't want to see colour definitions go simply because they're fallible - virtually any strict definition of colour is, so including the ideas behind the standards seems to be a good idea in principle - otherwise, why would the RGB, HSV, colour wheel, YEV, CMYK, Cie, and other color systems have been invented in the first place? I generally support the idea behind Notinasnaid's proposal (though I personally and irrationally like swatches), as long as the idea that fallible standards are regardless important can be kept in mind. Whew, that took a while to get out. Nihiltres^(t.c.s) 00:45, 9 May 2007 (UTC)

Thank you for your input :) I agree, we should have pictures. I'm not so much concerned about actually displaying a color on the screen and calibration, etc. I'm talking specifically about the numbers that you need to type into MSPaint or an HTML page to reproduce "what the color can look like". They are relevant, yes. Important enough for the info box? I don't think so.

The numeric values should not be there. If they are to be there, then they need to represent a range. Maybe for Red, with a 0-255 range, we might say 180-255 is red. This is necessary to be accurate. I'm not trying to be facetious. If this is simply infeasible, then it should be removed completely and put in a subsection of the page. If someone wants to make swatches, they need to also represent that range as well. The swatch would have to represent different intensities of red(all the way down to brown), go from point to point towards Violet or Orange, and also represent color saturation(going from almost gray to almost pink). Again, if it's not feasible, then it shouldn't be there. It is inaccurate to show a single color (#FF0000) and say "this is red". More accurately, "this is considered a shade of red".

As far as what I meant by "general public", maybe Mars wasn't the best example :) It has size, rotation speed, how many satellites it has, volume, mass, tilt, etc. I'm sure most people who made it through high school(I mean that jokingly) would understand what those meant. Representing a color in a 3 byte base 16 number? Not so much high school material. Knowing how much red, green and blue to apply to "create the illusion"(which is what it is) of that color, also very deceiving. Remember, when you see yellow on your computer, it's merely an optical trick, and it's not really yellow.

Again, thanks for the input. I feel we are getting somewhere. Wikidan829 01:00, 9 May 2007 (UTC)

----

I think you're missing the point. When you talk about the “…context that they are standardized and often official definitions of a colour,” you are assuming that there is some standardized or official definition of color. That is a wrong assumption, which survives from the 19th century and before, but has little relevance to modern color theory.

Scientific color models (CIELAB, etc.) are very careful not to assign arbitrary names to particular colors, because they recognize the inherent problems with that. When you say:

“I don't want to see colour definitions go simply because they're fallible - virtually any strict definition of colour is, so including the ideas behind the standards seems to be a good idea in principle.”

you are missing that there are no “strict definitions” of color which are based on names. The “strict” definitions (in, e.g., any of the CIE models) are based on spectra, not names.

“otherwise, why would the RGB, HSV, colour wheel, YEV, CMYK, Cie, and other color systems have been invented in the first place?”

RGB was invented because it is a convenient way to numerically represent colors which can be displayed on a screen. It doesn't attempt to be useful from a human perspective; its values are completely arbitrary, based on the mechanics of displays, not on human vision.
HSV is a derivative of RGB, and is nearly as useless. It is based on 15th-19th century ideas about color.
CYMK is obviously used for printing on offset presses, etc. and also has little to do with human vision. It uses arbitrary ink colors, and is device dependent.
CIE XYZ was invented in order to enable matching multiple spectral power distributions of the same perceived color, so that colors could be specified in 3 dimensions instead of infinitely many.
Derivatives such as L*u*v* and L*a*b* were transformations of XYZ data so that it would be more perceptually uniform, as for instance the Munsell color system is. They are basically hacks layered on top of XYZ to get the space to look more like Munsell. But if we use any kind of color values, they should be based on Munsell/XYZ/L*a*b* values, which are actually a standard. sRGB is simply too limited, and too irrelevant to color thoery.

None of these systems has any kind of “strict definitions” of color names, or indeed has anything to do with color names whatsoever. To associate “red” or “blue” with a specific color is just misleading and counterproductive (it basically perpetuates pre-20th century ideas).

--jacobolus (t) 01:38, 9 May 2007 (UTC)

Is there a reason that Xiong's image of red things (at right) isn't used on the red page? --jacobolus (t) 01:34, 9 May 2007 (UTC)

jacobolus's main idea to take away from this discussion

I'm not actually all that interested in getting rid of the color articles that exist. I don't think they are particularly informative or encyclopedic, but they don't do all that much harm. What I am worried about though is that no one in this Wikiproject seems to be very interested in improving the content of articles about color science or color theory history. What I would really like to see is good articles on such subjects, including information about the motivation and history of color models such as:

And of course fixing up more general articles such as color space, color model, color solid, gamut, opponent process, primary color, additive color, subtractive color, white point, color temperature, hue, saturation, lightness, etc. etc. etc.

I want Wikipedia to have well researched sections on the RGB and CMYK articles explaining how those spaces are not perceptually uniform, and pointing readers in the direction of more useful color models. I want the current emphasis on highly-technical descriptions and formulae in the color space articles to be reduced, in favor of clear explanations in layman's terms about the benefits, history, intentions, and uses of various color models are.

I would not agree that the more perceptually uniform spaces are necessarily "more useful". The RGB and CMYK spaces probably account for 99+% of all color representations in use, since they are what monitors and printers use. Nevertheless, those spaces have their uses and should be well explained. Dicklyon 15:03, 15 May 2007 (UTC)

And this WikiProject doesn't seem to share those goals. Instead of striving to explain, it seems bent on documenting trivialities like the hex triplets for colors described in the CSS spec, which are of very little general use, except to web designers, who already know where to find that information. Or a long list of color names, each of which apparently deserves its own stub article, with no content other than an info box and some vague hand-wavy description.

Mostly, I want wikipedia to stop perpetuating unreferenced unscientific ideas about colors—many of which come straight out of uninformed debates held in the mid-19th century—and convincing readers they must be accurate through an overabundance of jargon and technical formulae, and a lack of clear prose.

—jacobolus (t) 02:13, 9 May 2007 (UTC)

I can help with some of that. I agree the color science articles are important, and it's something I know a bit about. Dicklyon 04:36, 15 May 2007 (UTC)

I wish I could help with the content of these pages but worry my knowledge currently is at the enough to be dangerous point. I agree that they should very much be a focus, however I do worry though that some of the other color pages are actually a problem and have a WP:NPOV problem in that they imply even indirectly that 24-bit true-color RGB is the end-all of color definition, &c. PaleAqua 06:02, 15 May 2007 (UTC)

For all that I agree with much of the above discussion, jacobolus, how do we incorporate such information into the Wikipedia entry, especially if those colour systems have no definitions of any particular colour? The troubling thing I keep hearing is that we don't have any definition of colours by the systems which are more scientific and reliable. The theoretically convenient RGB, HSV, and, to some extent, CMYK defined values will, at least, generally produce a colour which is perceived as somewhere in the range covered linguistically as a particular colour, despite their inaccuracies. I like the idea of adding or converting to information (whether a range, a definition, or otherwise) covered by these reliable systems, but if there is nothing to which to convert, it seems (from my point of view, at least, please clarify) that all you are suggesting is that relevant information defined by reliable sources (the current definitions for those articles which have sourced infoboxes) shouldn't be included in the article. Please explain how we can incorporate this more accurate picture of colour standards into the colour articles (e.g. Blue, Red, Orange (colour)) - without a practical application, this discussion is meaningless. Nihiltres^(t.c.s) 14:24, 15 May 2007 (UTC)

I think it would be useful to have an example or two of certain colors that are specified in sRGB being converted to some other spaces. But mostly, color coordinates of particular colors are not very useful, while understanding the spaces and conversion methods is quite useful. For color coordinates, we should stick to what is sourced reliably, which means probably just the sRGB coordinates, 24-bit, of web colors, and a few things like that. It doesn't help anything to give "approximate" RGB or CMYK colors in unspecified spaces as a way to "approximately" define a color. A picture is good enough; in which case sRGB is the required space, and the pictures will be approximate when the numbers can't be sourced. Dicklyon 14:35, 15 May 2007 (UTC)

Agreed - I don't mean to imply that unspecified RGB spaces are appropriate: sRGB should be standard for RGB definitions except under special circumstances. Nihiltres^(t.c.s) 15:07, 15 May 2007 (UTC)

slices through L*a*b* space showing colors in sRGB which could be called "orange"

I think first, each article can have several photographic images showing objects considered to be of that color. We can try to find the best examples we can, that show off a range of colors considered to fall under each particular word. Beyond that, one possibility would be to show some slices through L*a*b* space (or some Munsell chips like at Template:Munsell-5PB-5Y, or whatever) for the hues like "red", "yellow", "purple", etc. Alternately, we could give color coordinates in some absolute space (or even better a spectral power distribution or similar) for particular pigments or light sources. So we could say that "cobalt blue", in such-and-such conditions, has such-and-such a color, and then we could show a picture showing the closest color which can be reproduced in sRGB, with a disclaimer if the color is out of gamut. Of course, all of this takes some amount of work and study. But many of the current color strips shown on those articles are inaccurate, and of dubious utility. --jacobolus (t) 19:03, 15 May 2007 (UTC)

I like this. This is more or less what I was getting at. No single value, no monochromatic swatches. This is a way more accurate depiction of a color. Good work jacobolus. Wikidan829 19:14, 15 May 2007 (UTC)

I really like the idea of slices through L*a*b* space and the spectral power distribution. Especially as regards the spectral power distribution: if we can find sources and have one for each colour in some standardized look, it would be an ideal addition to the infobox, which should probably take some overhaul as a result of this discussion. (By the way, I still recommend the inclusion of the old values in the infobox - barring nonstandard color space entries - so long as we add a "web colour" or such header to that section, and strictly require sources.) Nihiltres^(t.c.s) 02:39, 16 May 2007 (UTC)

As J noted above, spectral power distributions are applicable to particular pigments or illuminants. They are NOT generally applicable to COLORs, however, since colors live in a 3D subspace of spectral space, so their spectra are not at all unique. By the way, I did add some reflectance spectra of yellow pigments to yellow a while back; I think they're interesting, but certainly not definitive like color-space coordinates are. You can't get to Lab from pigment spectra or vice versa; if you multiply a pigment spectrum by an illuminant spectrum, then you can get to Lab space. Dicklyon 04:49, 16 May 2007 (UTC)

Yeah, I realize that. Many colors, though, are associated with particular families of pigments - cobalt blue, for example, being one where it's clear that particular pigments would be appropriate for the purpose. Might not entirely work for all colours, but I like the idea - as long as we can have something standardized, appropriate, and sourced, I'm happy. The point I'd like to make regards application - I think that these good ideas should be spread throughout the articles as part of the infobox, and if we reach a consensus of what to add, what to move, and how, the articles will be improved by the effort. Nihiltres^(t.c.s) 11:45, 16 May 2007 (UTC)

Check out this nifty chart by Bruce MacEvoy. --jacobolus (t) 18:47, 16 May 2007 (UTC)

Color comparison strip charts

User:Keraunos has added a bunch of comparison strip charts to color articles such as violet (color), magenta, yellow, blue, etc. The one place I made a little survey, violet, the consensus was to get rid of it. Do we need to do that survey in each article? Or can we get some opinions here, please? Dicklyon 04:36, 15 May 2007 (UTC)

While it is apparent that a lot of effort went into making these charts, I don't believe they belong in the articles. I agree with User:Notinasnaid on the Talk:Violet (color) page that about the issue of unsourced and original research for color names and their coordinates. RGB coordinates really only apply to computer and video output, and many colors fall outside of the gamut. Color names are not precise. For example a large range of hues or shades could be considered Iris (color). PaleAqua 05:51, 15 May 2007 (UTC)

I personally don't like them. But I think showing a range of colors in some form is useful. I particularly think that putting photographs of objects in each color in each article is a good idea, even up to 10 or 15 per article, or as many as can fit comfortably without spilling past the end of the text of the page. --jacobolus (t) 19:14, 15 May 2007 (UTC)

Actually, for colors which fall out of gamut, it might be useful to show a vertical slice through L*a*b* or Munsell space with grey for out-of-gamut colors, and then outline a range on that chart where the color in question is commonly understood to fit, as a way of giving users some understanding of how much brighter (or not) it might be than the color they are able to see on their monitor. Sort of use such slices in the same way as articles about places use the little geo-locator maps. --jacobolus (t) 19:17, 15 May 2007 (UTC)

What in the world are colors like "blanched almond" (#FFEBCD) "papaya whip" (#FFEFD5), "banana mania" (#FBE7B2), and "macaroni and cheese" (#FFB79B)? I'm really not sure who the expected user of such information is. And either way, the information is mostly arbitrary; it's not like Crayola, etc. specifications have anything to do with sRGB. And in either case, just because some marker/crayon/software producers use arbitrary color names, doesn't mean it's a good idea (for them, for users, or for us). It would be reasonable to have one page on "web colors" with information about web colors (and fortunately it already exists ;). And the info about crayola colors should be scrapped as inaccurate and un-encyclopedic. So I say just scrap the strip charts altogether. --jacobolus (t) 01:33, 16 May 2007 (UTC)

I agree. After that we can review the "shades of" templates, and make sure they include only sourced colors. Besides web colors, there may well be other useful reliable sources; I haven't checked. We might even could have an article on Crayola colors if they're verifiable. I informed Keraunos of this discussion, and we should give him and others a chance to comment before we take action, yes? Dicklyon 01:45, 16 May 2007 (UTC)

(moved from talk:magenta)

The strips of color in the comparison chart, which generally duplicate the colors provided at the bottom via the templates, are found now only in magenta, cyan, and indigo of all the common color names. I propose we remove it from here, as it's ugly, strange, and duplicative. Any objections or support? Dicklyon 04:29, 12 May 2007 (UTC)

I strongly object to removing these charts. I created all of them and I think they are beautiful. They are not duplicative--the Color Comparison Chart displays the shades of a particular color in approximate order of their shades (from the lightest at the top to most saturated in the middle to the darkest at the bottom) rather than in alphabetical order as in the Shades Template at the bottom of the article. The purpose of these Color Comparison Charts is to enable the Wikipedia user to more easily pick out a particular color which they may need for a particular use. For example, if someone is going to design a website, repaint a room, paint their house, or purchase a new automobile, they can look at the Color Comparison Charts and choose which color is best for or is closest to the color they need. It is much easier to do this when the colors are arranged in order of their shade instead of being arranged in alphabetical order. In addition, they display colors such as Crayola colors which may not be in the regular color articles and thus allow the user a greater selection of colors to choose from. I am restoring all of them with a short explanation as to their purpose and use. Keraunos 07:42, 14 May 2007 (UTC)

If a user wants to pick a color scheme for a house or similar, they should first use a color space such as L*a*b*, in which colors can be precisely specified. But second and more importantly, they should examine and compare actual paint chips. If they are going to design a website, they can use something like Adobe's kuler or similar. If they're going to purchase an automobile, they should examine cars in the possible colors, as no image on a computer screen is going to be an adequate representation. Wikipedia is an encyclopedia, not a list of arbitrary proprietary color names. --jacobolus (t) 01:51, 16 May 2007 (UTC)

I'm moving this bit of discussion to Wikipedia talk:WikiProject Color. Anyway, what does "order of their shade" mean, exactly? Color is not one-dimensional. It seems to me that they are roughly ordered by lightness, with colorfulness and hue ordered arbitrarily. I'm not sure why this is a particularly logical ordering. --jacobolus (t) 02:11, 16 May 2007 (UTC)

Keraunos seems to have opted out of the discussion, and there's no other objection, so I'll go ahead and take the strip charts out again. Dicklyon 00:52, 19 May 2007 (UTC)

Project Spelling Style Consistency

There are many articles on Wikipedia which use colour instead of color. Shouldn't all articles within the scope of a project use a consistent spelling style? 65.197.192.130 21:39, 17 May 2007 (UTC)

See WP:MOS#National varieties of English for details. Wikipedia policy is to keep uniform spelling/grammar/etc. within each article, but only to prefer a particular spelling when the article deals with a topic relevant to a particular variety of English: i.e. an article about a town in Australia would get Australian spelling/grammar conventions, etc. I don't think it's encouraged to standardize spellings across larger sets of articles, as to quote from MOS: “There are many more productive and enjoyable ways to participate than worrying and fighting about which version of English to use on any particular page.”--jacobolus (t) 21:57, 17 May 2007 (UTC)

I agree. It should be isolated to within the article, and the spelling of words other than color are to be taken into consideration. Basically whichever variation beat it to the punch.Wikidan829 22:01, 17 May 2007 (UTC)

Unsourced trivia and such

User:Keraunos continues to be the most active contributor to color articles. Unfortunately, most of his contributions are either unsourced trivia or color infoboxes representing original research (measuring color coordinates out of books and web sites). I'm tempted to add a fact tag to every trivia item (e.g. in the so-called popular culture sections) and then delete them in a month if no source is provided. What do others feel is the appropriate approach here? Dicklyon 17:10, 26 July 2007 (UTC)

I personally don't care so much about the articles about specific colors; I don't find them particularly useful, and I don't have the time to clean them up—it's too big a job, for not much benefit IMO. But ideally (for me) the color articles would all be filled with photographs of items so-colored, to give the idea that color is not discrete, instead of infoboxes w/ swatches. --jacobolus (t) 23:49, 26 July 2007 (UTC)

I'm debating going after the Green article. I've got some experience dealing with OR-laden articles (see Long hair). After doing a study of Sir Gawain and the Green Knight, it has become apparent to me that color is very important symbolically. The color articles just aren't done right. They're largely lists, not prose; trivia, not real research. I'm going to do some scholarly research in the future and drastically reorganize this particular article with the goal of GA status. Wrad 21:22, 11 September 2007 (UTC)

Go for it! Having even one good article about a color would be nice as an example. --jacobolus (t) 21:33, 11 September 2007 (UTC)

That's my experience, too. One good article in a set somehow makes the others that much more bearable. I'll probably make a rough draft in my sandbox, first, so I'll have to keep you posted if anyone wants to help out. Wrad 21:40, 11 September 2007 (UTC)

I have been making drastic changes to the Green article on the way to GA status. Things are looking good. I'm going to go through a serious fact-checking process next, as well as further cleanup. Any of you color experts who wish to contribute are welcome to it. I'm more of a culture guru. Wrad 05:31, 17 September 2007 (UTC)

Wow, that's really massively better. Great work! It would indeed be nice to figure out some standardized types of art/art history/psychology/color science information to put on color articles... the current infoboxes, etc. that were standardized-upon are pretty weak. A little ways up this page I put an image of a few slices of orange hues in the L*a*b* space. Do you think such a thing would be generally useful for colors? If so, I could probably create some number of uniformly-spaced hue slices, that could be mixed and matched for various articles. --jacobolus (t) 09:40, 17 September 2007 (UTC)

Color infobox request

Could someone fix the color infobox so that an optional image can be included? The green article would definitely be more pleasing if I could combine the picture and the infobox. Thanks, Wrad 00:05, 12 September 2007 (UTC)

I frankly think we should only put a picture in the info box, falling back on a swatch if no suitable image can be found. Anyway, be bold: create a new info box template in whatever manner you like, and if others dislike it, things can be hashed out at the wikiproject color talk. --jacobolus (t) 09:46, 17 September 2007 (UTC)

See also the 2 threads above: #Infobox Suggestion and #Infobox Suggestion Revisited. --Quiddity 17:24, 17 September 2007 (UTC)

Speaking of info boxes, shouldn't we have sources to go with info? Most of them have numbers in RGB and other forms, in many cases with no source. I just removed the source from the green one, which had been copied from the red one, and was a source only for the wavelength range (the CRC book source, that is), which the green one didn't even have. Some colors have a webcolor, html, css, etc. source for RGB numbers, but many have nothing. Isn't it better to omit the info than provide unsourced info? And don't we need a separate source for each separate data item? How about just using footnotes on sourced items, and removing the others? Dicklyon 21:44, 17 September 2007 (UTC)

I'm not sure I understand, but whatever we do, it does need to cite the information. The best solution is to add sourced info, not so much delete unsourced info, although that is sometimes necessary if it's totally false (but how would you know that unless you had found a source?). Why do we even have a source parameter in these infoboxes? Shouldn't the source just go with each individual fact? Wrad 23:12, 17 September 2007 (UTC)

Yes, I think a ref on each fact is the way to go. But as the standard for inclusion/removal, it's about verifiability, not about whether it's true or false, since we have no way to determine the latter in general. Before we can call the green article "good", I think we have to take out all the unsournced stuff. Of course, we ought to allow plenty of time for people to work on finding sources first, so the thing to do is to tag all the factoids as needing sources. Dicklyon 00:14, 18 September 2007 (UTC)

That's my plan. I'm going to do a fact check. I'm just worried someone might delete everything before I can do this. Even if things aren't sourced, I have found that they can point me in the right direction to find good sources. It's a lot easier than a blank slate. Also, by the time this gets to GAC, all unsourced info will be deleted, so don't worry. (Sorry if my last post was misleading on this point.) As it is, i think we're doing pretty good. The article is steadily getting better. Wrad 00:34, 18 September 2007 (UTC)

Agreed. Thanks for the good work. Dicklyon 01:02, 18 September 2007 (UTC)

This wikiproject: ideas

Wow. I just read through this project's talk page and good grief! Seems like a whole lot of arguing about hex and RGB and whatnot that makes my head spin. Meanwhile, the color articles for blue, red, green, and yellow are arguably the worst articles on wikipedia in proportion to their importance to wikipedia. All four are among wikipedia's "Vital Articles", limited in number to one thousand total on the most basic and vital subjects of mankind. All four are absolutely, sickeningly pathetic. Full of trivia. They read like disambig pages. I only say this because I know most of you agree with me, so I don't think I'm going to offend anyone. (To be honest the Color article is pretty dang good.) I think that this project needs to get more organized. Fact is, color articles are difficult to write. They are so prevalent in our lives that to write a comprehensive article on even one color would take nothing less than an Omniscient, someone who knew everything about everything (joke). Luckily, among all of us we may be able to come close. Perhaps, as in the Dinosaur Wikiproject, we should categorize ourselves. Color articles need people who are good with the sciences, and people who are good with language and human culture. Members could be organized into these groups so that we know better who to ask for help when we need it. We may also need to do a roll call to determine active members. Once we get one color to GA (Right now I'm working on Green, but will need help.) We can develop a basic guideline for all major color pages to follow, and work within that format. Thoughts? Wrad 02:23, 18 September 2007 (UTC)

Your work on green may inspire us to do more like that. I hope so, because I agree that most color articles are just sinks for unsourced trivia so far. Dicklyon 02:34, 18 September 2007 (UTC)

Not just red, yellow, green, and blue (by the way we should make an article for unique hue), but also white, black and grey, along with basically every other color article (purple/violet/magenta, pink, orange, brown, teal/cyan/aqua, etc.) are just terrible. --jacobolus (t) 04:35, 18 September 2007 (UTC)

Merging minor colors together

I think the best course of action, if there is actually enough man-power to do it, is to figure out a smallish number of broad color categories (red, green, brown, etc.) and put their trivia lists into prose and fill out the missing information, then merge other minor color variations ("olive green", "chartreuse", "azure", "crimson", etc.) into those articles (or sub-articles “variants of brown”, etc.), and only split them off on their own if they have enough material to fill out an article (and in some cases—“wheat”?—just remove them). Otherwise, there are just too many minor color articles to keep track of, and their quality will remain abysmal. --jacobolus (t) 04:35, 18 September 2007 (UTC)

As an example, I don't think that blue-green, teal, aqua, cyan, turquoise, cerulean, aquamarine, bondi blue, and category:shades of_cyan have enough useful information between them to fill out one decent article. I think they should be consolidated into one article at blue-green, where any differences can be explained, and where a common set of information about the history of blue-green pigments, etc. can be placed, where someone looking for information can actually find it, instead of having to hop through 10 articles about essentially the same thing. (I'm not quite bold enough to make such drastic decisions without some consensus beforehand though :)) --jacobolus (t) 04:55, 18 September 2007 (UTC)

Sounds goods to me. A page like Minor shades of green or something like that. White and Black are obviously also important articles that should be fixed. Wrad 05:09, 18 September 2007 (UTC)

At least for the blue-greens, I think they can just be merged into one article. For some colors with larger numbers of arbitrary names associated to them, it might be necessary to break the list into a sub-article. --jacobolus (t) 05:38, 18 September 2007 (UTC)

I just looked back over the discussion on this page, and only about 5-6 have actually participated in the discussion in months. After another week or so of waiting for input, I'm just going to merge the blue-greens together, and redirect them all to blue-green, and then next perhaps create a "Named variations of green" article into which all the green stubs can be dumped. Frankly, most of them should just be deleted, but that seems more controversial than merging them, so even though no compelling arguments for keeping them (in the 2 years of discussion in this wikiproject) were ever made, I'll stick with the merge to avoid revert wars. --jacobolus (t) 16:43, 19 September 2007 (UTC)

I'm interested in seeing how it turns out. Go for it. Wrad 16:49, 19 September 2007 (UTC)

Actually, I started taking a look at the minor color articles, and now think many of them should be deleted outright. I think even the more inclusionist editors around here would generally agree with me for articles like thistle (color), which basically says "thistle is the color of a thistle" as its only content, with no sources for its swatch, and no other information. --jacobolus (t) 22:06, 19 September 2007 (UTC)

I agree many of those articles probably should be pruned. Category:WikiProject Color articles needing infobox sources is probably a good place to start finding such articles. PaleAqua 21:39, 21 September 2007 (UTC)

Appeal regarding Azure (color) and Azure (heraldry)

There has been much discussion of the relative importance of "azure" as the generic name for a color or "azure" as one of the standard tinctures in heraldry. I wonder if some of the participants here would care to participate in the discussion and help break the impasse one way or another. I also think the article(s) in question could benefit from some improvement, perhaps along the lines of what User:PaleAqua describes above. --Dystopos 21:03, 3 October 2007 (UTC)

ambiguous color terms

In searching through online color books at google, I found the following passage in a book from 1913 (quoting a pamphlet by Milton Bradley):

“The list of words now employed to express qualities or degrees of color is very small, in fact half a dozen comprise the more common terms, and these are pressed into service on all occasions, and in such varied relations that they not only fail to express anything definite but constantly contradict themselves… Tint, Hue, and Shade are emmployed so loosely by the public generally, even by those people who claim to use English correctly, that neither word has a very definite meaning, although each is capable of being as accurately used as any other word in our every day vocabulary…”

I found this amusing, because I have found this still to be the case, at least on Wikipedia color articles. Thought I would share with the class. :) --jacobolus (t) 10:58, 8 October 2007 (UTC)

JPEG and GIF

This has probably been mentioned been mentioned before, but I've been seeing a lot of JPEG (and a couple GIFs) used as illustrations in color-related articles. Since JPEG is lossy, this is probably not a good idea. Is the viewer seeing the effect you wanted to illustrate, or is the viewer seeing the effect the lossy compression added? Worse, if the image is thumbnailed, it will be recompressed to a different JPEG, and there's no way to know exactly what will happen. Thus, I recommend we use PNG, with possible exception for huge images. Even if you're trying to illustrate JPEG or GIF themselves, you can convert to PNG as the last step before upload to ensure thumbnailing doesn't recompress the image. Superm401 - Talk 20:27, 11 November 2007 (UTC)

GIF would certainly be a bad idea, as it has a limited number of colors. JPEGs, on the other hand, can show millions of colors pretty accurately; many more, and with more precision, than a human can distinguish. The artifacts of JPEG compression are usually only visible as subtle high-frequency defects near edges, not as color errors. Dicklyon 20:40, 11 November 2007 (UTC)

Green

I recently nominated this article for GA status. Input and aid are desired. Wrad 04:25, 1 December 2007 (UTC)

Template:Infobox color Symbolism

Per a discussion in Talk:Green#another infobox?, a field for including symbolism has been added to the Template:Infobox color. I also broke out the spectral colors into an optional section at the same time. PaleAqua 07:06, 3 December 2007 (UTC)

^ Ultramarine

[1] Ultramarine

[1]