Talk:Pigment/Archive 1

This is an archive of past discussions about Pigment. 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

Heat absorption

Does pigment affect the amount of heat absorbed? — Preceding unsigned comment added by 67.62.163.140 (talk) 21:21, 26 April 2005 (UTC)

Pigment vs. dye

The discussion of pigments vs. dyes is very confusing to someone like me who's unfamiliar with the distinction. First it says what the distinction is; then it says there's a very clear distinction, and repeats what that distinction is; then it says that some things are both, which makes me doubt the statement that there's a very clear distinction. I suspect this is all accurate, it's just not phrased in a way that's clear to people who don't already understand it. Could someone clean up that paragraph? Thanks, --Elysdir 17:59, 16 January 2006 (UTC)

WP:COTW vote

Pigment (5 votes, stays until July 22)

Nominated July 8, 2006; needs at least 6 votes by July 22, 2006

Comments

Needs such things as images, industry, ecology, biological benefits, etc. Basically just a list. Davodd 23:23, 8 July 2006 (UTC)

Use of the word organic versus biological

For the sense in which the word "organic" was being used here, let us instead write "biological". To a chemist, an organic compound is a compound with carbon based chemistry. Important modern pigments such as Phthalo Blue and Phthalo Green, (which are both symmetric, aromatic compounds in which a three-dimensional lattice of carbon rings holds in place a single copper molecule) are the product of organic chemistry. Thus, we need to use the physical scientists' definition of the word "organic". --Metzenberg 10:21, 16 July 2006 (UTC)

Perhaps somebody else can figure out how to make the image monarch_butterfly.jpg show up on this page. --Metzenberg 11:53, 16 July 2006 (UTC)

Links to Sinopia and Handprint - External links about fine artist pigments

Before you delete external links, why don't you look at the sites and see how informative they are:

Sinopia is a commercial site, but it offers extraordinary information. For example the website demonstrates fresco techniques, how to make one's own oil paints, and so forth. The selection of pigments is astonishing. The company has both synthetic modern versions of pigments and actual historic natural versions of the colors that one can compare side-by-side. I like to visit Sinopia whenever I am in that neighborhood in San Francisco, just to look and touch. The tiny basement store on 22nd Street is like a museum. I would characterize the entire business (Sinopia) as a hobby business or a labor of love for the owner. There is no way that the tiny market that this business serves merits such a well-designed and informative website on a financial basis. Sinopia is a company that created a great website because the owner just loves pigments, and it is one of the most informative sources on the Internet for fine artists. It is of geat interest, for example, to Art History students.

Handprint is a completely non-commercial site by one person (whom I have never met in person). He lives somewhere around Santa Rosa, California, and has no commercial affiliations whatsover in the art materials industry. He just likes to paint. His descriptions of watercolor paints, of how to mix and use paints, and the pigments that are available are very highly respected. This is an extremely high quality non-commercial link.

Until a few minutes ago, I found no web page at this address, which is why I deleted it. I did look. Outriggr 06:07, 17 July 2006 (UTC)

Kremer Pigments I did decide to leave deleted. The old Kremer pigments store in lower Manhattan was as good as Sinopia. Kremer has recently moved to a new location, and I haven't had a chance to get to the new store yet. I hope the new store will retain some of that "below 14th Street" character, that the move uptown does not diminish the opportunity to see and examine hundreds of pigments, often to compare different versions of the same pigment from different manufacturers. Both Sinopia and Kremer are like museums to artists. Kremer hasn't really developed a website of Sinopia's caliber (although it is still an informative one). --Metzenberg 05:34, 17 July 2006 (UTC)

Scientific and Technical Issues

I have added an outline for scientific and technical issues to be covered. This is not an exhaustive list, however. These are some of the practical properties of pigments that govern their use in particular applications.

Another issue that should be touched on is the reactivity of pigments with each other. For example, lead carbonate (a white) blackens with exposure to sulfur, as it is transformed into lead sulfide. --Metzenberg 06:13, 17 July 2006 (UTC)

Pigments by color

I'd like it if someone could make a list of pigments by color, e.g. Purple Pigments: then a list with links to those particular chlorins or whatever they may be. Another list for green pigments, etc.--Ashi — Preceding undated comment added 23:39, 25 January 2006 (UTC)

I have added a several paragraphs here about Colour Index International, which is the authoritative information source for pigments by color. Since there are typically dozens of pigments for each color, maintaining an exhaustive list on Wikipedia does not seem feasible or appropriate. --Metzenberg 00:12, 18 July 2006 (UTC)

Style issue - Capitalization in this article.

In the color industry, it seems to be fairly conventional to capitalize the names of pigment colors, even generic names, while chemical names are lowercase. I have followed those conventions here, and I would suggest that other editors see these examples before changing capitalization.

Phthalo Blue, but phthalocyanine blue
Indian Yellow and Indian Yellow Hue
Vermillion and Vermilion Hue, but vermilion colors (to indiciate generically, colors that are similar to Vermilion the pigment.

This kind of departure from normal capitalization conventions is common in techical writing and makes it much easier to understand technical materials. --Metzenberg 02:50, 18 July 2006 (UTC)

Industrial standards

Manufacture and classification of pigments is subject to an extensive list of international (ISO) standards. Is somebody that is expert in manufacturing able to write about them? Here is a partial list:

ISO-787: General methods of test for pigments and extenders.
ISO-8781: Pigments and extenders - Methods of assessment of dispersion characteristics

Other ISO standards govern specific pigment groups, such as cadmiums, iron oxides, fluorescent pigments, and titanium dioxide. --Metzenberg 20:43, 16 July 2006 (UTC)

I've created a bare outline for the section on standards. I'd like to see somebody else work on this. How about somebody that understands color models and color measurement? Another issue is certainly how the color attributes of pigments are measured and compared. --Metzenberg 03:43, 18 July 2006 (UTC)

Albinism

On the biological note, the sentence in the intro about albinism is not strictly correct. Albinism is a result of a lack of melanin only (not pigment in general). Many albino reptiles have a yellowish tinge to them because they still generate yellow pteridine pigments in their xanthophores. And, of course, the characteristic red eyes found in albinos are due to haemoglobin, which is itself a pigment. Rockpocket 17:54, 18 July 2006 (UTC)

Separate article on Biological Pigments

Biological pigments really merit a separate article. It is not feasible to create and maintain a single article that covers both subjects. They are very diverse. A separate article about biological pigmentation needs to be created, and a lot of links on other pages will need to be fixed. This page should focus on pigments used in industrial, decorative, and artistic applications. Any comments? --Metzenberg 00:12, 18 July 2006 (UTC)

I think you should wait until this article becomes too long before splitting it up. I also think Pigment, biology is a bad title. If indeed there is a need for it, it should be called Biological pigment or Pigments in biology or something. —Keenan Pepper 00:28, 19 July 2006 (UTC)

I created the article called Pigment, biology in part because I realized that it was a huge topic in and of itself. It was becoming obvious that there would be no way to create a well-structured and readable article that covered both topics. They are really too dissimilar. The two topics really had nothing in common other than the definition of a pigment. Furthermore, the biology article needs to focus on areas such as evolution (how pigmentation evolves, assuming you are not a "creation scientist") and diseases (both hyperpigmentation and hypopigmentation). If you are interested in pigment in biology, then feel free to rename that article and by all means edit it. --Metzenberg 10:39, 19 July 2006 (UTC)

Touch on Color Models

I would like to see a paragraph added to the beginning of the Manufacturing and industrial standards section that says more or less that color chemistry has developed into a modern technical field, and that scientific and omjective standards now exist for measuring color. Can someone explain the models for measuring color, and the ones that are used with pigments? --Metzenberg 20:49, 19 July 2006 (UTC)

Vermeer's The Milkmaid

The Wikimedia commons has three different scans of this painting by Vermeer. Unfortunately, the scans look very different and render the color very differently. On my crummy flat panel monitor, I cannot possibly judge which one is the best scan. This painting is a good one for how well is represents the artist's lavish use of the expensive colors Indian Yellow and lapis lazuli. I would consider it general knowledge that Vermeer made lavish use of expensive colors. It is one of these things that every Art History 101 class probably talks about. Is there a good reference on Vermeer that we can cite? Does somebody own a book about him? --Metzenberg 21:18, 19 July 2006 (UTC)

This version seems to be the best. It is a little less 'sharp' than the others, but the colors are more lively and realistic. It was clearly taken by a professional photographer, and looks far better than the other two on my fairly-nice Hiatchi SuperScan CRT. However, I'm hesitant to add it to the article, which is now very image heavy. It is an article about pigment, of course, so one would expect a lot of images, but we are starting to crowd out the text. I'd suggest waiting until we have more text before adding another image. Phidauex 03:31, 20 July 2006 (UTC)

Girl with a Pearl Earring, a novel by Tracy Chevalier, is a fictional account of one of Vermeer's most famous paintings. In Chevalier's novel, and in the film based upon it, the artist uses lapis to paint the headscarf on a young servant girl. Vermeer (played by Colin Firth in the film version) admonishes the servant girl Griet (played by Scarlett Johansson) to keep this secret from his wife, knowing that his wife will be jealous.[6]

and later

In Girl with a Pearl Earring, Vermeer's patron remarks that Vermeer used "cow piss" to paint his wife. Since mango leaves are nutritionally inadequate for cattle, the practice of harvesting Indian Yellow was eventually declared to be inhumane. Modern Indian Yellow Hue is a mixture of synthetic pigments.

why is this trivia about a novel and movie being included in an article about pigments? It's a good movie, yes, and shows the paint mixing process, but it's hardly a primary source of information and it's very unprofessional. — Preceding unsigned comment added by 63.27.156.13 (talk) 00:31, 24 November 2006 (UTC)

Swatches

I've added a section of swatches. I've been wracking my brain all evening trying to figure out the best way to display these, and the best values to use. No one seems to have any good CIELAB measurements of historical pigments, or if they do, they aren't sharing. I'm currently using information from the The Dictionary of Color Names, which simply maps color names to one of the 267 NBS/ISCC "centroid" colors that use the standard color language description. Its not as accurate as it could be, but it is consistent, and gets us into the ballpark. I'm going to continue to look for more accurate measurement data, but this will give us a good idea in the meantime. Here is the code to create a swatch:

#143CB4

Ultramarine Blue

<div class="thumb tleft">
<div style="width:120px;">
<div style="width:112px; height:50px; background-color:#143CB4; color:#FFFFFF;">#143CB4</div>
<div class="thumbcaption">Ultramarine Blue</div>
</div></div></div>

The code is based on the output of the Image tags, so it has a consistent appearance, and the CSS classes will mean that its appearance will change alongside the appearance of image thumbnails, should the main CSS document change. The size of the swatch is arbitrary, and could be changed. Likewise, the inclusion of the Hex data is arbitrary. We can include as much, or as little, data as we feel is necessary. Once we get a good idea of what we want displayed and how, I'll make a template out of it for ease of use here and in other pages.

If anyone has a suggestion for datasets of raw measurements of any of these pigments, I'd be happy to convert them to the correct colorspace for web-viewing. I just need the data. Phidauex 05:53, 21 July 2006 (UTC)

Hi Phidauex. I am wondering how the average reader is to interpret the difference in hex color definitions between say, Ultramarine Blue (#007CAD) in this article, and the Ultramarine article (#120a8f)? Or Vermilion = #FF4D00; and in this article, #C10020. Could the explanation be built into the article (if my question isn't misguided). Thanks. Outriggr 06:05, 21 July 2006 (UTC)

There are three basic reasons why they don't look the same... The first reason is that there are many different pigments that you can call Vermillion. Even different batches of the same pigment will look slightly different, and different preparation methods and sources will change the color as well. Historical vermillion probably had a very wide color range, depending on who was making it. Even modern vermillion has a range, depending on who is manufacturing the pigment, and who is mixing it into paints. Second, there are different ways to measure the color. Different lighting, different specrophotometers, and different paint swatches will result in different measurements. And finally, there are many ways to convert the raw measurement into a 'colorspace' to define it, and then converting that color space into the sRGB color space that your computer display uses. There are a number of colors that your monitor cannot reproduce (called "out of gamut" colors), because of limitations inherent to any 'three primary color' system. So the conversion to sRGB is invariably an approximation, and different people will generate different approximations. However, I think I'll try to unify the different pigment sub-articles, so they use the same data I'm using on the swatches. The color values can be whatever we want, as long as they are backed up by real measurements, and good conversions to sRGB. If possible, the Vermillion article should probably contain a number of swatches, each corresponding to a certain pigment source or manufacturer. I hope to have some of this explanation in the pigment article as we work on it. Phidauex 20:39, 21 July 2006 (UTC)

Another suggestion I have ... incorporate the color index numbers. Have a look at my ecommerce site Dick Blick Art Materials and see our swatches. Although I have done some of them myself, the quality of the color match varies greatly. I have gone through a long series of young graphic designers and web developers, all of whom have worked on the swatches on the site. Some were very good at it, and some were not. One observation I have made to them is that professional artists understand how to see a dark and untinted pigment, and how it will brighten as it is tinted. So I have recommended to them that our swatches be dark. I tried to build a system where we would work on a colo9r corrected Macintosh and I built a small program for gamma shifting the Mac values to PC values. Unfortunately, new people would constantly not understand it and do things wrong. For example, they would instead use a non-color corrected PC, and yet enter the values into the color correction program. What a headache. --Metzenberg 06:27, 21 July 2006 (UTC)

The problem is that we know too much! This would be much simpler if we didn't have any understanding of color measurement, perceptual matching, gamma, gamut, color spaces, etc... I just got my hands onto some CIELAB measurements of raw pigments, which I can use to make 'better' swatches. I'll probably normalize the raw L*a*b values to a sort of 'uniform lightness', on the dark side (which makes sense to artists, as Metzenberg noted), then convert to sRGB color space at a gamma of 2.2 for display. I'll try to tweak the template to show both the original LAB measurements, the RGB values, and the original pigment identification (PB29, for instance). I'm also trying to get a book from the library that someone recommended to me regarding measurements of historic pigments. I'll do some more work on it this afternoon, expect it to be in a state of flux for the moment. ;) Phidauex 15:31, 21 July 2006 (UTC)

Ok, a few of the swatches have been updated with actual measurement information from Bruce at handprint.com! The measurements are of watercolor pigment swatches that he prepares in a very standardized way, and are the average value from a few brands and lots of the same pigment (which vary slightly). I've done my best to convert the measurements to sRGB, and I've added the pigment number (PB29) where appropriate/known. Thoughts? Phidauex 17:58, 21 July 2006 (UTC)

I would recommend Bruce's measurements! Tell him you are working on this, and see if he is interested in working on this article. --Metzenberg 21:37, 21 July 2006 (UTC)

There is a programmer on Wiki who has done a lot of color templates. He tends to get the programming right, but the colors he makes mistakes on. I wouldn't fault him for that, because those old CRT computer monitors ruin your vision. Would you like me to see if I can interest him creating a template system that can be used for all pigments, and can be used on a lot of different pages? --Metzenberg 21:52, 21 July 2006 (UTC)

I've noticed some problems in the swatches section. First, how were the measurements made. Pigments are metameric and therefore have to be measured with respect to a standard illuminant, hopefully the same as the one implied in your sRGB color space, or you'll have errors induced by the chromatic adaptation matrix. To measure pigments accurately, you need pretty high-end spectroradiometers or spectrophotometers - you won't do it with a Macbeth Eyeone or something of that caliber. Pigments seen under a source close to D50 or under a fluorescent source will look radically different. Your swatches do not specify what the white point of the illuminant is. I presume it is D65, and not D50 which is common in the graphic arts. Also, the mention that gamma will affect hue is technically incorrect. However, this may happen on computer displays due to non-linearities in the actual electronic transfer functions of the RGB channels, especially in the dard shades when using 10bit look-up tables or less. Pure, mathematical gamma is the slope of Luminance and has absolutely no bearing on Hue. In newer color appearance models such as CIECAM 02, there are more complex interactions at play which cannot be summarized by Hue, Saturation and Lightness alone. Do not confuse Lightness and Luminance, and Brightness - they relate to different things. Brightness is a radiometric quantity, Lightness is a photometric quantity, and Luminance relates to self-luminous objects such as computer displays. The gamma function of a computer display cannot be perfecly true to theoretical, and to the Lightness quasi-power function of the human visual system. For the swatches, you should specify: - the measurement apparatus - the illuminant - the RGB color space (as you do) - the display gamma (it is normally implicit in sRGB, but it's good to mention it explicitly) - the rendering intent used in the gamut mapping from XYZ or Lab to RGB. You'll have to choose whether you want to preserve individual hues more than saturation, or if you're interested more into maintaining the perceptual relationships between different pigments under such and such illumination. By all means avoid the ridiculous CMYK conversions seen elsewhere. Hex values should be specified as full gamut sRGB and non web-safe. If your goal is not just to show swatches on the Wikipedia web page but to provide artists with a good starting point to emulate pigments digitally, you are not bound by the small gamut of sRGB. You can specify values in wider RGB spaces such as BetaRGB, BestRGB, AdobeRGb, PhotoRGB or ColorMatchRGB, as long as you mention that the display won't be able to reproduce the exact appearance. If your measurement instrument gave you XYZ and L*ab values, it would be good to include those too. Be careful also how you average your samples - the median will be more representative than the mean. Ppanzini 04:49, 23 November 2006 (UTC)

Thanks for your comments, and I appreciate your clarification within the article. The measurements weren't taken by me, however I know that they are the median of a series of measurements taken under a D65 source. I took the D65 LAB color measurements, converted to sRGB (accepting the loss of gamut), and created the template.

Unfortunately, you are right, with the number of conversions done, the state of most people's monitor calibration, etc. the swatches are just a vague approximation. These swatches aren't really meant to be a resource for graphic artists, and I was afraid that too much detail could swamp the swatch, and give the impression that the view is more real that it actually is. The intent is merely to give readers a general idea of the appearance of some classic pigments, without being recklessly inaccurate with the color space conversions.

I created them because I was reading the article, and thinking to myself, "What does vermillion even look like?? What about crimson?? Tyrian purple?? How are Prussian Blue and Ultramarine different?" I couldn't imagine them, and it was bugging me, so I made an attempt to create reasonable approximations to orient the reader's mind.

If you have the ability to take measurements, or do conversions to other RGB spaces (I'm not familiar enough with them to reliably convert myself), then that would be a welcome addition. However, that would be a pretty labor intensive process, and I think that as long as we are clear about the nature of the swatch, and their intent as a general guide to the reader, we will be OK. Phidauex 19:13, 25 November 2006 (UTC)

Physical Basis

I like the figure showing light reflecting off of the pigment along with the effect on the light spectrum--but it's a bit incorrect. The figure shows A - B = C, where A is the solar spectrum, and C is the final spectrum. But B seems to be a transmittance spectrum. This isn't what is subtracted from A but rather what is allowed from A to pass through. It seems this could be fixed one of two ways:

Change B to an absorbance spectrum
Change the minus sign to a times sign.

I think the second option might actually be the more logical (and easy) way to correct it. Red Herring 11:39, 29 November 2006 (UTC)

I made the second correction described above. Red Herring 15:46, 4 December 2006 (UTC)

Solid Form?

"It must be stable in solid form at ambient temperatures." I don't see where this comes from, I'm not an expert but I assume it's still a pigment if it's in liquid form in a cell?...Ashi Starshade 20:29, 24 June 2007 (UTC)

Simple reflectance image >_<

I dislike the wording and the picture. It may induce wrong simplifications when getting acquainted with the physics of the process.

If it's a wide variety of colors, then three is not sufficient.

If it's just the three colors mixed, then it's a) not a wide variety, b) unnatural.

The second picture, "complex reflectance" shows that a pretty solid spectrum is being partly absorbed and the result is quite not monochromatic, which is correct. Pigments and dyes never scour the spectrum of all "wrong"-colored wavelengths, they subtract certain bands, and the rest of the spectrum causes an impression of certain color. "Simple reflectance" suggests that the thing left is actually the certain colored light, which isn't quite the case.

Basically, a person would dismiss the second, complex picture and imagine that a pigment is a bit weird sort of a fluorescent substance. As if it absorbed everything and then emitted certain color. Or absorbed everything except a narrow wavelength interval (like, you know, the color "blue" has a wavelength attributed to it).

"Simple reflectance" takes red, green and blue combination as the reference. This can't be a simple coincidence with additive color synthesis of the RGB-model. That model is nice and all, but it deals with spectral colors. Where 760 nm is purple and 400 nm is violet. All the more reason for the whole stuff to be horribly digested into, I repeat, a theory of pigment absorbing every wavelength except those which correspond to the color observed.

"Complex reflectance" will make even less sense then: there are plenty of blue waves left, sure. But there are quite much of purple, green, yellow, orange and red left. If someone has got the grip on how spectral colors look — he checks and finds that the only color that was wiped out (~600 nm) was actually yellow. *blink*-*blink*

The problem is that pigments do not work via additive color scheme. They subtract waves. When pigments eat up just the yellow spectral constituent of the white solar light, the resulting color will be the complementary to the yellow spectral color (which is blue, heh). Honestly, without any boasting on my part, I do not know, if you're taught this in colleges. I find that english wikipedia doesn't have anything about complementary colors.

The basics are like this: visible spectrum can be cut into colored segments based on the color sensed from the respective monochromatic wave sources. If you shine with the light of the segment waves you see a spectral color for that segment. But if you shine with all other segments' waves, you see another color — the complementary one. Example: white mixture bereft of green waves segment (which are ~500—560 nm) will appear red. Complementary for the spectral green is red. Same way the complementary for spectral orange (~595-605 nm) is greenish-blue. End of example.

I've set up a page in ru-wiki (native to me), here: ru:Спектральные и дополнительные цвета.

With all this stuff in mind, I'll reiterate again. "Simple reflectance" image makes false simplifications by bringing in the additive mechanisms into universe of subtraction. And if someone tries to imagine the playing of brown-colored pigment with RGB-light falling on it, that someone risks getting a trauma trying, too. Signing off. ^^ Легат Ская 91.78.151.205 (talk) 21:23, 30 June 2008 (UTC)

Oh, I've found the Complementary color, but it looks pretty incoherent, if you ask me. >_< It's not the color names or appearance that define complementarity, it's the certain spectral deduction that nets you the definition of complementary color. Or should it actually be "secondary color" in English? Легат Ская 91.78.151.205 (talk) 21:52, 30 June 2008 (UTC)