labeling, lightfastness & toxicity
Paint labeling is probably the least interesting topic on watercolor paints ... like reading the fire tag on a new mattress. Unfortunately, boredom leads to apathy, and apathy leads to ignorance, and it is this ignorance that paint manufacturers exploit through marketing hype.
Paint labels tell you what your are getting for your money, provided you know the difference between pigments, paints and "colors". If you don't, then marketing gimmicks will take control.
Most paints sold today are required to carry health warnings, but you'll discover why these warnings often overstate the risks of using watercolor paints.
pigments, paints & "colors"
The most important point to understand when selecting watercolor paints is the difference between pigments, paints, and "colors" or marketing names. This distinction is essential to understand if you want to get the right art materials in the watercolor paints you buy.
Because most artists have been trained under the "color theory" dogma that paints are just "colors", even knowledgeable artists or authors such as Michael Wilcox, Charles Reid, Susanna Spann, Betty Edwards, Jim Kosvanec or the late Zoltan Szabo do not always keep the distinction clear between pigments (colored powders), paints (mixtures of pigments and liquid vehicle) and "colors", the product names or marketing monikers assigned by each manufacturer to their paints. This results in frequent inaccuracies and outdated information in art instruction books, as illustrated in my book reviews.
The editors at publishing houses such as Watson-Guptill, North Light or Watercolor Magazine share in the responsibility after all, packaging and distributing information is their business. This confusion is an entrenched habit, abetted by the marketing techniques of art materials manufacturers, but informed choices by artists, authors and publishers can put it in the past.
What's in a Name? The diagram below shows that the problem comes down to a single question: what exactly is in this paint?
What you buy online (or in a store) is, of course, the paint the physical runny substance packed into a little tube. Each paint is given a marketing name that allows you to identify it in the paint rack or catalog page. But what you really pay for is the pigment in the paint, because pigments create the color.
Unfortunately, paint marketing names commonly do not inform you about the pigment that is actually in the paint. This is primarily because many watercolor paints contain more than one pigment.
the problem with paint marketing names
The single pigment paints contain only one pigment, with vehicle and additives, period. A paint with the marketing name cobalt blue would ideally contain only one pigment ingredient: cobalt blue (PB28). Single pigments are always the preferable formulation for artist's paints. You can directly evaluate the quality of the pigment in the color, saturation and tinting strength of the paint, different brands of paint made with the same single pigment will usually have a similar color and handling attributes, and the lightfastness of the pigment is as good as it gets. You know what you are working with.
Of course, artists must mix paints to get specific colors that no single pigment matches exactly. This is true for any nuance of hue, especially violet, green and orange, as well as for subdued or grayed colors (tones) and dark valued or near black colors (shades).
Paint makers often premix and package common color recipes as convenience mixtures of two or more pigments. A paint with the marketing name cyanine blue might actually contain the pigments ultramarine blue (PB29) and phthalo blue (PB15). Buying premixed paints is not usually a quality advantage over buying the pigments as individual paints, because premixed paints can hide flaws in the pigment ingredients, and because the convenience paints are often modified by the artist anyway. You can just as effectively mix the color you want from scratch, using single pigment paints. (There are two exceptions: the earth pigments are single pigment tones or shades of yellow, orange and red hues, and convenience greens provide a handy "home base" green that can be modified by mixture with any other paint.)
Paint manufacturers also use mixtures of relatively inexpensive pigments to match the color appearance of more expensive pigments (usually marketed as "student" paints); or they may use a mixture of reliable modern pigments to match the color of traditionally popular but fugitive paints. So a paint with the marketing name cobalt blue might actually be made with less expensive ultramarine blue and phthalo blue; a paint named alizarin crimson might actually be made with the more permanent quinacridone violet.
Many of the traditional "earth" pigments (yellow ochre, raw sienna, raw umber, burnt sienna, burnt umber, PBr7), as well as fugitive pigments (carmine NR4, alizarin crimson PR83, and rose madder NR9), polluting pigments (manganese blue, PB33), expensive pigments (cobalt blue, cadmium yellow, PY35, cadmium red, PR108) and highly toxic historical pigments (vermilion, naples yellow, emerald green) are imitated in this way. These hue mixtures can be quite acceptable color substitutes for the original pigments, but they may suffer the same quality and lightfastness problems as convenience mixtures.
By the way, good labeling practice requires the word hue in the name, so that you know cobalt blue hue is not really made with cobalt blue pigment. But marketers routinely ignore that practice, because they assume you already know that marketing names do not inform you about what is in a paint.
Finally, as a separate complication, the generic chemical names of modern synthetic organic pigments can make your eyes bug out. Phthalocyanine, diarylide, isoindolinone, benzimidazolone, anthraquinone, pyrazoloquinazolone, anthrapyrimidine, quinophthalone ... these are a chemist's technical terms that paint manufacturers almost always translate into something less forbidding for artists.
In each case single pigment, convenience mixture, hue mixture or chemically arcane pigment the paint manufacturer is free to give the paint whatever marketing name it deems appropriate. There is no regulatory limit or industry common practice to inhibit the creativity of the marketer.
the names of historical, fugitive or toxic pigments that are no longer used (mauve, carmine, rose madder, vermilion, bright red, dragon's blood, naples yellow, gamboge, indian yellow, chrome yellow, sap green, emerald green, van dyke brown, sepia, indigo, etc.). Sometimes these modern paints resemble the color of the traditional pigment, but they do not match it.
proprietary pigment trade names hansa yellow, irgazine red, monastal blue which can designate a specific pigment, although it can also be used to label a different pigment of similar color.
outdated labeling conventions peculiar to the art materials market (spectrum yellow, primary yellow or permanent yellow). "Permanent" is especially misleading: it was originally a 19th century paint marketing label that meant "a synthetic organic pigment that is not derived from aniline"; it has never identified "a paint that won't fade".
proprietary paint manufacturer names or pseudo technical pigment nicknames (winsor red, blockx red, scheveningen red, australian red, thalo red, azo red)
nonspecific color poetry (brilliant orange, vivid orange, translucent orange, warm orange, coral orange).
In all cases, again, these names tell you nothing about what is actually in the paint.
The crux is that paint manufacturers can name a paint anything they want. The result? If the marketing name is all you rely on, it is impossible to tell if two paints with the same name ("cobalt blue") contain the same ingredients, or if two paints with different names ("cobalt blue" and "cyanine blue") contain different ingredients and in every case, you don't know what those ingredients really are.
the marketing romance
There is another layer of marketing that defines the paint manufacturer's trade image or brand style a statement of the company's goals, its ingredient choices and paint manufacturing methods and of course its relationship to you, the purchasing artist. I call this the marketing romance, and it exerts an amazing power over the many compulsive collectors of colored gum among watercolor painters.
Nowhere else are exotic marketing names and picturesque stories used more consciously and skillfully than in the Daniel Smith mail catalog or web site. Take for example their "PrimaTek® pure, authentic mineral pigment" paints, made (as they claim) from crushed regional rocks. (What exactly is an impure, inauthentic, nonmineral pigment made from crushed rocks?) Here is the tout from the Daniel Smith Summer 2004 mail catalog for a paint made of powdered turquoise:
"Many cultures thought an amulet [of turquoise] worn on a horse's bridle protected the horse and rider from a fall. The legends are many. The magic is yours to own. ... The mystic beauty of turquoise has been felt by every culture and its use has crossed national and cultural boundaries. Paint with DANIEL SMITH Natural Sleeping Beauty Turquoise Genuine and catch a piece of its rich and colorful history for yourself. Prepare to be captivated."
Set aside the suggestion that you might be painting while riding a horse, and focus on the proposition that you catch a piece of rich and colorful history by purchasing a tube of paint. And ask yourself ... how big is the piece of rich and colorful history that fits into a tube of paint? (And if you get more colorful history, does that mean you get less paint?)
Laughably, this kind of marketing gibberish actually works. People are coaxed to buy these paints because they covet protection, legend, magic, mysticism, tradition, colorful history or captivation ... in short, because they seek a consumer experience. For these amateur artists, buying an amulet in an ersatz Indian souvenir shop, and buying a piece of colorful history, are pretty much the same thing.
Most marketers do not attempt to seduce you outright, as Daniel Smith does, but rather to distract your attention or mislead through adjectives that deliver empty claims. These hackneyed romance stories make liberal use of nonsense buzzwords such as genuine, pure, natural, authentic, made by hand, craftsmen, rare, classic, workshop, historical, time honored, original recipe, founded in ("founded in 1664"), handed down for generations, used for centuries and, of course, trusted by professional artists. The nonsense here is that none of these terms have any legal force, nor do they refer to any industry standards or regulated practices. They claim much, but guarantee nothing.
You defang this kind of marketing bite by reversing the qualifiers in each case (for example, "impure, inauthentic, nonmineral") or by looking into the technical documentation. Here for example is the "quality guarantee" used for many years by Blockx:
"We never use non lightfast pigments and never made second quality range. We want artists confortable with quality question so we make only the best one. We still use stone mills the only ones to reproduce an handmade job. All our colors are made in our workshop in Belgium by specialised craftsmen working under the leading of Jacques Blockx himself" [sic throughout!].
So ask: what is an unspecialised craftsman? ("Paintmaker" does not sound so romantic.) What is a handmade job? All commercial watercolor paints are manufactured in the same way: with ingredients measured by hand into large premixing tubs, thoroughly mixed on three roller milling machines fed and controlled by hand, packed in tubes with hand operated crimping machines, and identified with self adhesive labels applied by hand or with paint information preprinted on the tube. No paint is made by blending pigment and vehicle by hand with a mortar and pestle. And how specifically is paint milled with stone rollers superior to paint milled with the standard metal rollers (beyond the mere claim that "old ways are better")? And how does the fact that Blockx doesn't make a student (second quality) range of paints create any assurance about the quality of the paints they do make?
Finally, as I explain elsewhere, during the many years this guarantee appeared in the Blockx marketing literature, Blockx did in fact use impermanent pigments in several paints (see for example PY1, PR3, PR83 and PR106). Isn't it a more reliable guide to quality that a paint company lies or does not lie to you in its marketing materials?
As an example of factual investigation, let's drop by those diligent artisans at Old Holland and see what they have to say about their line of 168 watercolors:
"One of the unique characteristics of these watercolours is the unparalleled colour strength (maximum pigmentation). And while this high colour strength requires a slightly different approach on the part of the aquarellist (so little paint is needed from the tube or cup for the desired colour effect that you have to get used to the ratio of paint to diluent), the advantages are clear.
It is well-known that Old Holland attaches a hand-painted colour strip to its tubes of oil paint and acrylic paint, showing the paint in the tube in question. A different solution was chosen for the watercolours. The labels of the tubes and cups of watercolour do not show the full tone of the paint, but the undertone (a logical choice following on from the technique of watercolour painting). The colours on the labels are screen printed with the watercolour in this undertone, using a screen printing technique developed specially for Old Holland."
The claim that so little paint is needed from the tube or cup is an indirect justification of the high price Old Holland charges for their products. The unparalleled colour strength referred to can be assessed with a tinting test. I have personally inspected the comparative tinting tests of several watercolor paint brands conducted by a USA watercolor paint manufacturer, and while Old Holland watercolor paints are not noticeably weaker than other brands, they are certainly not "unparalleled".
It's relevant to examine the paint ingredient information to find out what pigments we are getting at unparalleled colour strength. To do this we must consult the Old Holland web site, and click on individual color swatches for ingredient information. A tedious process, but we discover that their two cobalt violet paints are actually hue mixtures of three (cobalt violet dark) or four (cobalt violet light) pigments, none of them genuine cobalt violet (PV14 or PV49); that sap green lake extra is formulated with black pigment (PBk7) plus four (!) other pigments; and that a staggering 28 paints (!) contain white pigment (PW4) and that's excluding another 6 varieties of white or gray paint!
Pause for a minute. Why would watercolor paints be formulated with white or black paint? Added white paint is an acceptable ingredient in oil paints, but is generally something to avoid in watercolors. And why are four or five pigments necessary in a hue mixture? Watercolor paints generally become less lightfast as more pigments are used to make them, because several pigments will respond differently under exposure to light.
The realization comes if we browse over to the oil paint section:
"Oil paint has a special place at Old Holland. It is the first product we manufactured. ... In 1985, following extensive research, Old Holland presented a revolutionary range of 168 oil paints, each with the highest degree of lightness [sic]."
Hmmm ... Old Holland makes 168 oil colours ... and 168 watercolors ... it's obvious what has happened. They purchased pigments and formulated pigment mixtures to create their inaugural line of oil colors, and when it came time to add watercolors to their brand, they just mixed the oil color pigment formulations with some gum arabic and glycerine unparalleled strength of white pigment and all!
As a last bit of detective work, we're alerted that something is amiss by the fact that most of the description of the watercolor paints is actually focused on the watercolor paint labels the missing color swatches, in fact. I worked for several years in the printing industry, and I can affirm that there is no screen printing technique "developed specially" for any client. Modern printing technology is highly standardized. But why print color samples that would be much easier to create with fast drying watercolors than with sticky and slow drying acrylic or oil colors? My conjecture is: the painted labels would fade in the retail display rack.
Here is the sleeve magic from Sennelier:
"Only the purest pigments are chosen and in the time honored tradition. The pigments are soaked in purified water for 24 hours before cold grinding for maximum consistency and luminosity. The binding solution is composed of hand-picked gum arabic from Senegal and honey from the Alps. The result is intensely deep colors that offer delicate transparencies and the distinctive "satin luminosity" unique to French watercolors."
In fact, these are standard manufacturing details described to make them sound special. All modern industrial pigments are shipped in "pure" form (or in impure form, since they usually contain additive chemicals to resist mold, clumping or degradation), all paint manufacturers premix the paint ingredients before milling; all genuine gum arabic is harvested by hand, etc. A few euphemisms (grinding means milling, binding solution means vehicle) and nonsense claims honey from the Alps is the same as honey from anywhere else, there is no distinctive "satin luminosity" unique to French watercolors are thrown in for poetic effect.
It's quite a shock, then, to read the brief and factually accurate product claims in a 2005 Utrecht catalog:
Made in our Brooklyn plant, Utrecht brand professional Artists' Watercolors rank among the world's finest for their transparency, lightfastness and working properties all at a friendly Utrecht price.
Your only defense against marketing romance, refund aside, is to apply common sense to the marketing claims. The easiest way is to put the claim in a different context. Handmade may sound as desirable as homemade, but if you imagine that handmade means some guy with an industrial blender in his garage, and homemade means granny drooled in the batter, then the magic evaporates. Does the fact that Old Holland started making paints in 1664 guarantee the quality of products made by the company today? Answer: no, it doesn't. Is honey from the Alps really a better humectant than honey from Spain or France? Answer: no, it's not. Are traditional manufacturing methods really better than modern methods? Answer: it depends, but typically no, they are not.
If you can't explain what the marketing claim means, or explain why the marketing claim makes a difference to the quality of the gummy stuff in the tube, then you can see the emptiness of the marketing gibberish and you've disarmed the marketing magic. The spell is broken, and you're a free spirit again. Now that's magical!
paint ingredient information
So how do you find out about the pigment in the paint? You find it listed on the paint packaging. After a long period of rampant labeling and marketing abuse, the art materials industry has voluntary adopted the ASTM standards for commercial paint labeling. These require manufacturers to list the paint pigments on the packaging, both as the pigment common name (such as "Dioxazine Violet" or "Cadmium Red") and as the pigment color index generic name, a code that identifies each pigment as a generic chemical compound.
(A third identifier, the five to seven digit color index constitution number, is a numerical code assigned to pigments to indicate a specific chemical formulation. This number is occasionally withheld to protect proprietary pigment recipes, and is not part of the ASTM standards.)
Color Index Name. The color index naming system is standardized, regulated and disseminated by the Society of Dyers and Colourists, London (UK), in collaboration with the American Association of Textile Chemists and Colorists (USA). The current naming codes for pigments and dyes are available to subscribers as the massive Colour Index International, currently in its 4th (2001) edition.
The pigment common name is often a technical chemical name ... something about benzimidazolone, or quinacridone, or quinophthalone, or thioindigoid ... we're no longer in that warm and fuzzy realm of poetic paint names that marketers use to move that product.
No matter: the color index name lets you identify pigments without the chemical jargon. It's a simple code, that consists of:
The letter P to denote a pigment (rather than a dye, D, or a basic dye, B); you will occasionally see N to refer to natural pigments such as cochineal, rose madder, gamboge or lapis lazuli
A letter to denote one of ten basic color categories: R for red, O for orange, Y for yellow, G for green, B for blue, V for violet, Br for brown, W for white, Bk for black and M for metallic
A number referring to a standard list of pigments within each color category. This number is assigned as a pigment is introduced for commercial use, and may be withdrawn or deleted if the pigment is no longer manufactured. (The symbol "N/A" for "not available" is used in the rare cases when a pigment is not included in the pigment list to protect a proprietary formulation.)
So PY40 refers to the 40th entry in the list of yellow pigments (aureolin), PO20 to the 20th pigment in the list of oranges (cadmium orange), NR4 refers to the fourth red listed in the natural pigment list (cochineal), and so on. (Numbering is not consecutive, as some pigments or dyes have been deleted over time.)
Now, the color index name or constitution number is the most reliable way to identify paint ingredients. To help you learn this color naming system which is very easy to use once you get familiar with it all paints in the guide to watercolor pigments are grouped by color index name.
The Color Index and "Color". Unfortunately, the color index name does not identify a consistent paint color, because some chemically equivalent pigments can exist in several forms, each with a different hue. Some of the most important examples: cadmium yellow (PY35) can be anything from a lemon yellow to a near orange, cadmium red (PR108) can range from a scarlet red to a dark maroon, natural iron oxide (PBr7) can be anything from a dull yellow or orange to a brownish black, quinacridone violet (PV19) can be a bright red, dull carmine, bright rose, or dark reddish violet, cobalt turquoise (PB36) can be a cerulean blue or a dull turquoise green, and cobalt titanate green (PG50) can range from a pale bright turquoise to a dull yellowish green.
The color index name and number refer primarily to the chemical composition of the pigment. So color varieties within the same CI name are produced through adustments in the manufacturing methods (especially in the amount of time the pigment is calcinated and in the extent or method of grinding into fine particles) or through variations in the the proportions of pigment ingredients or in the structure of the pigment crystal. Thus, the iron oxide color variations are produced by differences in the amount of added manganese, in the length of the calcination, and in the particle size; the cobalt and cadmium variations are produced by the particle sizes and the proportion of added secondary metals (tin or aluminum for cobalts, selenium for cadmiums); the quinacridone variations by the crystal form and particle size, and so on.
Even with these important exceptions, the color index naming system is very useful. A look at some commercially available green paints will show you immediately how the color index names are much superior to the manufacturers' marketing names if you want to find out the actual ingredients and paint colors involved in your paint selection:
The first four greens are made of exactly the same pigment, and will look almost identical on the paper, even though they have completely different marketing names. None of these manufacturer labels refers to the pigment's common name (phthalocyanine green), so the labels don't explain what is really in the paints, which might tell you whether or how much their colors or handling attributes differ.
The next three greens are mixtures of two pigments: the first two paints have different ingredients even though the marketing names and actual paint colors are the same; the last two paints are made with exactly the same ingredients (in different proportions), yet the marketing names (and paint colors) are different. Again, without the color index names, it would be impossible to sort this out.
The last two examples show two paints with the same name. Using the color index name, you find that one is made of three ingredients, the other of a completely different single pigment that almost every other paint manufacturer calls by its common name, viridian. (Vert émeraude is the common name for viridian in France.) You also can verify that neither paint contains a speck of the historical pigment emerald green, PG21, the poisonous copper acetoarsenite.
"Hue" Paints. To conform to the ASTM labeling standards, manufacturers must also use the designation hue for a paint that is named with a pigment common name but that does not actually contain the pigment. For example, a paint named Manganese Blue must contain the pigment "Manganese Blue, PB33," and if it does not, it must be called Manganese Blue Hue. Every paint manufacturer I know of does not respect the letter of this standard, especially for historical color names carmine, madder, sepia, indigo, van dyke brown, gamboge, emerald green, and any color name with sienna, umber or earth in it. No matter: even in these cases, the color index information will clear up most of the confusion.
Finally, the ASTM standards recommend that the packaging show the paint's ASTM or manufacturer lightfastness rating, and federal or state (California) law requires a health warning for pigments that may poison or cause allergic reactions (as described below).
Although paint manufacturers have become more forthright, you cannot depend on paint manufacturers to name paints accurately. With the color index name you cut through the marketing clutter and actually see what you're getting for your money. (Well, assuming the paint manufacturer is honest. A simple dissolving test shows, for example, that more than one paint sold as pure viridian also contains some phthalocyanine green to brighten up the color.) Deceptions aside, the color index name is the most reliable way to find out what you are buying, and you should only buy paints that show the pigment common name and color index name on the tube: "Dioxazine Violet, PV23," "Cadmium Red, PR108," and so on. Pigments listed in the guide to watercolor pigments, in the complete palette, or mentioned elsewhere in this site, always refer to the color index name so you know which pigment is being talked about.
Lightfastness refers to the chemical stability of a pigment under long exposure to light. As a source of energy, light can cause color and chemical changes in many pigments. These changes can cause the color to whiten, gray, darken, change hue, fade or completely disappear.
Permanence or fastness refers to the chemical stability of the pigment in relation to any chemical or environmental factor, including light, heat, water, acids, alkalis, or mold. For example, ultramarine blue is extremely lightfast, but it will fade if brushed with a dilute acid.
It may seem obvious that paint pigments should be extremely durable, yet surprisingly the chemical stability of watercolor pigments has been addressed in the USA through independent testing and standardized labeling only since 1984.
A Brief History of Lightfastness. Guides to the choice and use of pigments, including a discussion of their relative permanency, have been available to artists at least since the early 15th century, when Cennino Cennini described permanency problems and recommended stable pigments in his Il Libro Dell'Arte (Book of the Arts, c.1400). Even before that time, painters relied on craft tradition, workshop training and generations of experience with pigment acquisition and manufacture to create lasting paintings in fresco or tempera. By the 16th century, with the advent of oil painting in Italy, pigments were being imported or manufactured expressly for artists; by reputation the finest pigments were sold in Venice and Florence. Painters became less adept at making pigments themselves, but they still had to know how to make paints and how to use them for best results in easel or fresco paintings.
The traditional range of artist's pigments slowly began to expand toward the end of the 18th century, as the new science of chemistry and the new industry of chemical manufacture began discovering and producing new synthetic inorganic pigments of iron, copper and cobalt. In addition, new and more complex methods for purifying and laking organic colorants such as madder and carmine created an impressive number of new pigments. At the same time, commercial colourmen appeared as pigment providers and then paint manufacturers, and artists began buying premade paints. It was in this era that painters largely lost the practice of making paints, and with it a clear understanding of the chemistry of the pigments they used. They began buying manufactured paints, often as new and untested colors.
The English chemist George Field published the first scientific study of artist's pigments, Chromatography: A Treatise on Colours and Pigments for the Use of Artists (1835). Field documented that many pigments then in use, especially those made from vegetable matter, were fugitive. In part because of Field's influence, several Victorian artists (including Alfred Hunt and William Holman Hunt) began performing their own permanency tests before using any manufactured paint.
After about 1860, new industrial methods were invented for synthesizing organic (carbon based) dyes from coal tar or petroleum, and organic chemistry developed the infrastructure to discover and manufacture hundreds of new pigments. These new colorants stimulated a late 19th century fashion for bright colors in fabrics, clothing, furnishings, and artworks. Artists wanted brilliant pigments in order to appease the rising demand for dramatically colored paintings, so paint manufacturers began selling paints made with the new chemicals.
Unfortunately the lightfastness of these new dyes and pigments was often not scrutinized, and many artists denied or downplayed the permenancy issue. J.M.W. Turner was notoriously unconcerned about the permanence of his materials and methods. One anecdote has it that the paint manufacturer William Winsor chided Turner for buying so many pigments known to be fugitive to which Turner replied: "It's your business to make paints, it's my business to use them." Sir Joshua Reynolds, Vincent Van Gogh and James McNeil Whistler were also maddeningly slipshod in their use of painting materials: several of the Reynolds portraits I have seen in Europe now have a ghastly gray blue appearance because his laked carmine and yellow pigments have entirely faded.
The lightfastness issue came to a head in England in the early 1860's, when the art critic John Ruskin was asked by the British National Gallery to develop methods to display and conserve the items in the Turner Bequest. Soon knowledgeable collectors such as Ruskin and the museum curator J.C. Robinson were publicly voicing concerns about the permenancy of many commonly used art materials. Their claim that some watercolor pigments faded if improperly or even routinely displayed was at first hotly disputed. But the debate was put to rest by the Report ... on the Action of Light on Water Colours (1888) by the chemist Walter Russell and the amateur painter Capt. William de W. Abney. Like Field before them, they carefully documented that many watercolors did indeed fade under even moderate exposure to sunlight.
Lightfastness Standards. The first attempts to establish objective lightfastness standards appear toward the end of the 1880's, in the report by Abney and Russell, in standards developed by the German Society for Rational Painting Techniques, and in simple paint recipes and lightfastness evaluations published by Winsor & Newton in 1892. In 1907, a committee of artists and manufacturers met in Germany to set industry lightfastness standards, and adopted alizarin crimson (PR83) as a minimum lightfastness standard. Today it is considered to be far below the lowest standard for artist's materials.
Paint manufacturers continued to operate largely independently of these voluntary industry standards in large part because most artists continued to ignore the problem. Some paint brands began using the term "permanent" to convince buyers that their pigments were not as fugitive as the usual coal tar derivatives, but this label was (and still is) a marketing moniker rather than a guarantee.
In the USA, the unregulated state of the industry was addressed with paint manufacturing standards proposed in 1938 by Massachusetts artists working in the WPA (a Roosevelt era program of government funded art projects in public buildings across the USA). These were published as voluntary guidelines, for oil paints only, by the new National Bureau of Standards (NBS) in 1942 (revised in 1962). This Paint Standard remained the industry guide for many years.
The ASTM and Lightfastness Ratings. The introduction of acrylic binders, and the discovery of many new synthetic organic pigments, led the Artists Equity Association in 1976 to lobby the NBS for new standards. The NBS delegated this task to the American Society for Testing and Materials (the ASTM), the largest independent standards writing body in the world, which sponsored the actual work by a large group of European and American artists and paint manufacturers. This ASTM subcommittee developed testing methods and a classification according to the reaction of the pigment to a standard exposure to light, called a lightfastness test, and published its standards and reports in 1984. These define industry minimum standards for commercial paints the best quality paints should easily exceed them.
The Standard Test Methods for Lightfastness of Pigments Used in Artists' Paints (D4303-03) describes the industry accepted standards for conducting lightfastness tests with sunlight, fluorescent lights, a xenon arc fadeometer, cool white fluorescent lamps or fluorescent UV lamps. The Standard Specification for Artists' Watercolor Paints (D5067-99) describes the specific methods for testing diluted watercolor paints on filter papers, and lists the lightfastness test results for several dozen common watercolor pigments.
Only tests actually done on watercolors can guide your pigment selections. Lightfastness ratings of oil or acrylic paints are not a reliable guide to the permanence of watercolors made with the same pigment, because pigments last longer inside the protective coatings of oil or acrylic vehicles than when left bare on paper with an irregular coat of gum arabic. Keep in mind that some watercolor paint manufacturers simply quote the lightfastness test results provided by the pigment manufacturers, who commonly test their pigments in an oil or acrylic dispersion.
Blue Wool Test. Most lightfastness tests expose paint samples to abnormally intense light radiation, because this more quickly produces fading or discoloration in impermanent paints. Thus, samples of artist's pigments may be tested by exposure to direct sunlight, even though paintings are normally displayed under much weaker indoor lighting. (Thanks to low latitude sunlight intensity and largely cloudless climate, it is generally accepted that the most rigorous outdoor pigment testing is done at commercial sites in Florida or Arizona, USA.)
How much light exposure are we talking about? Direct outdoor sunlight or indirect sunlight in a bright room yields anywhere from 5,000 to 100,000 lux of illumination. This can produce a cumulative radiance exposure (depending on weather) of up to 25 megalux hours annually, which is 30 to 50 times higher than the brightest art gallery illumination.
The problem then is to measure the amount of light exposure the test samples receive. The current solution is the blue wool textile fading card or blue wool scale (at right). This consists of eight strips of wool mounted side by side on a small card; each strip or reference is colored with a blue dye that fades after exposure to a known amount of light. The dyes have been chosen so that each reference takes about two to three times longer to begin fading as the next lower reference in the scale. (Under normal solar testing conditions, reference 1, the least permanent, will begin to fade in 3 hours to 3 days, depending on geographic location, season, cloud cover and humidity; reference 3 will fade in 5 days to 2 weeks; reference 6 in 6 to 16 weeks; and reference 8, the most permanent, in 6 to 15 months.) These scales are used for paint lightfastness testing under international standard ISO 105-B, and are also used by gallery curators to measure the accumulated amount of light received by museum displays of paintings, textiles or photographic prints.
To perform the lightfastness test, a blue wool scale is exposed alongside the paint samples. When a reference strip in the scale begins to discolor, all paints that have also begun to discolor at that point but not before are rated as having that level of lightfastness. This procedure provides the blue wool ratings for paint lightfastness commonly cited by paint manufacturers.
The problem with solar radiation is that you are at the mercy of weather, season and geographic location. So lightfastness is also measured by exposing paint samples to prolonged artificial light, which produces more accurate and reliable results than sunlight testing. The instrument of choice is a xenon arc fadeometer or weatherometer (brand names vary), which can measure precisely the amount of illuminant energy the sample has received and can be programmed to cycle through alternating periods of light and dark, usually for a total exposure equal to 1000 hours of sunlight exposure. (Filters are used to adjust the xenon light to more closely match the sun's visible spectrum.) The xenon arc method provides an intense light that produces quicker results and with less heat exposure than solar testing. UV fluorescent or halogen lamps, which provide more UV radiation than incandescent lighting, can also be used, and in those tests a blue wool scale is again necessary to measure light exposure.
Through accumulated testing experience and scientific study, the ISO exposure levels are used to estimate the amount of time that a paint will remain visibly unchanged under exposure to natural light under normal display conditions that is, away from a window, under indirect sunlight or moderate incandescent light, and properly framed behind a UV protective glass or acrylic cover.
This table gives one version of the eight blue wool lightfastness levels, from 1 (fugitive) to 8 (extremely lightfast), with the amount of light exposure required to produce a color change at each level and the approximate match between the eight blue wool and five ASTM lightfastness categories. (For different and more stringent ASTM standards, see the discussion under doing your own lightfastness tests.)
Other rating systems have been defined on the blue wool levels, including the schemes using three categories developed by Robert Feller and Karen Colby: these lump blue wool 7-8 into a (I) "durable" or "excellent" category, 4-6 into a (II) "intermediate" or "marginal" category, and 1-3 into a (III) "fugitive" or "sensitive" category.
These lightfastness levels apply to artworks "under normal conditions of display" in art galleries that is, hung under controlled and reduced light conditions. The typical museum policy permits the display watercolors considered "lightfast" for no more than three months in a period of three years at illumination levels from 150 to as little as 50 lux; "fugitive" historical watercolors may not be displayed at all. When not on display, paintings are stored inside portfolios in dark archive cabinets and are shown only briefly by appointment with a museum curator.
Home display is another matter, and here is what the ASTM says on that topic:
In a normal home environment these times [required for paints to fade] can be expected to be shorter, especially if the artwork is located near a window, or in direct sunlight or fluorescent illumination, or is located in tropical or subtropical climates. When [the standard ASTM lightfastness test] was conducted in different locations and at different times of year, "Fugitive" materials took from a few days to 2 months to fade, while materials rated "Inferior" and "Fair" took from approximately three months to eighteen months to change color. Materials rated as "Good" showed no color change when Reference 6 faded but some of these showed a color change before Blue Wool Reference 7 faded. [This passage refers to the lightfastness categories defined in ASTM D5383-02, ¶9.1.1, which differ from the table above.]
However, paints classified in the blue wool category for "fair" lightfastness (III) in my 2004 lightfastness tests began to fade after six weeks of direct daily sunlight exposure during northern California summer and fall months, and all "fugitive" samples faded within two weeks. The ASTM projections seem to me charitable and optimistic.
All evidence considered, then, the consensus seems to be that it is best to use pigments rated BWS 7 or 8. In the guide to watercolor pigments I recommend that you do not use any materials rated BWS 6 or lower.
a blue wool scale
(top) unexposed; (bottom) exposed to sunlight for 800+ hours
Using lightfast paints does not mean that you can ignore the proper mounting and display of watercolor paintings. Mark Gottsegen's painter's manual provides extensive information in the chapter on "Picture Protection," and complete information on lightfastness testing and the ASTM paint standards. You can also order the technical documents I've cited directly from the ASTM web site. The most recent ASTM lightfastness standards for pigments used in watercolors were published in Standard Specification for Artists' Watercolor Paints (D5067-99); that report includes lightfastness ratings for many common watercolor pigments, though it also omits several common pigments and includes a few that no paint company currently uses.
lightfastness with a grain of salt
After this long discussion of paint lightfastness, here is the letdown. As an artist and consumer, you should be skeptical of published lightfastness ratings, whether of the permanency of a generic pigment, or the permanency of a specific brand of watercolor paint.
Both manufacturers and some independent paint gurus publish paint lightfastness tests to guide or reassure consumers. But these ratings can be seriously misleading. They imply that paint lightfastness is a fixed quality, constant over time, that can be measured accurately by a single test of a single tube of paint. This is never the case.
The first problem is that pigments do not have a specific, unchanging lightfastness. That is, the various pigment manufacturers who make pigments classified under the same color index name differ significantly in the quality of their raw materials, the efficiency of their manufacturing plant, the quality controls applied to their manufacturing methods, the chemical purity, particle size and laking substrate of their pigment product, and the additives used to stabilize the pigment for storage and transport. To make paints, the different paint manufacturers use these pigments in different concentrations, and add different amounts of extenders or brighteners. All these factors can affect the lightfastness of the paint. This means the same generic pigment can receive very different paint lightfastness ratings, depending on who manufactured the pigment and who manufactured the paint. This variation is especially large across the many synthetic organic pigments.
Second, different lightfastness tests can yield conflicting test results, a point the ASTM makes very clear in its testing literature. The effects of moisture or heat are not measured at all by the ASTM tests, but more importantly, concentrated and diluted paint layers will discolor at different rates, as I document for some synthetic iron oxide pigments.
And third, large scale lightfastness testing is costly and time consuming. In fact, no paint manufacturer tests all their paints: some manufacturers test some of their paints, but most do not test any paints at all. I believe Daniel Smith is the only manufacturer actually to test their pigment stock, but I do not know how frequently this is done. In most cases, the lightfastness rating on the tube of paint is actually the ASTM rating of the generic pigment, or the pigment manufacturer's lightfastness rating (which may be based on chemical reasoning rather than an actual light exposure test, or a test of the pigment in an acrylic dispersion), or a rating inferred from "common knowledge" or the chemical structure of the generic pigment.
When you realize that all these factors varied pigments, varied tests, different manufacturers, different quality control issues, incomplete testing, "conceptual" pigment ratings are at work in the marketplace, you can appreciate the problem with relying on a single published lightfastness rating. And this problem is significant because paint manufacturers have never tested the specific paints you buy in the store. No manufacturer tests all their paints, some manufacturers test some of their paints; and many never test any of their paints at all.
Do the "consumer reports" paint guides, or my guide to watercolor pigments, offer any help? Not really. The current (2001-02) Wilcox guide simply parrots the ASTM or manufacturer ratings, without testing any paints for lightfastness. The Page guide is based on actual paint lightfastness tests (so far as I can determine), though these are not ASTM standard tests and the tests were done years ago. The evaluations in all guides are based on a single tube of paint, although paint manufacturers make hundreds or thousands of tubes of each "color" in a year.
All these uncertainties make the published lightfastness ratings in marketing brochures or "consumer reports" paint guides hard to trust. But if you conclude that lightfastness tests are therefore worthless, you're wrong. The tests are important ... but the answer is to do your own lightfastness tests, rather than trust the ratings offered by the manufacturers or paint guides.
At this point, anyone reading this who actually works in the art materials industry or an art retail business is rolling on the floor laughing ... What, artists test paints? Ha ha ha, you're dreaming! Artists can't be bothered to test paints!
Well, that may be true, but if so it invites a look at the current paint selection practices among watercolor artists, and what effect those may have on art purchasers and on the market price and exhibition reputation of watercolor paintings.
Michael Wilcox has often ridiculed the ethics of paint manufacturers because they market impermanent pigments such as alizarin crimson (PR83), rose madder genuine (NR9) or aureolin (PY40). But what about the artists who buy these paints? After two centuries of accumulated experience, published guidance and painstaking scientific documentation of the issue, artists cannot claim to be ignorant of the problem of paint lightfastness, or claim to be unaware of the many substitutes for fugitive pigments (in particular the quinacridones) available from modern industrial chemistry. As the creators of works that fade, the responsibility for permanency lies entirely with artists.
There are publishers who collude in misleading students about these issues. North Light Books, Watson-Guptill and Watercolor magazine regularly publish painting tutorials or artist profiles that tout the use of indisputably impermanent pigments by professional artists such as Paul Jackson, Jeanne Dobie, Charles Reid, Al Stine, Linda Stevens Moyer, Steve Hills or Mel Stabin. Yet even here, as the authors of books or articles that mislead, the primary responsibility rests with the artists.
So why do many contemporary painters still use fugitive materials?
One problem is, unfortunately, that some paint manufacturers mislead their customers about paint permenancy. As explained above and in my review of Michael Wilcox's paint guide, some paint manufacturers simply quote the pigment manufacturer's assurances, or rely on technical information about the generic pigment, or quote ASTM test results performed on a competitor's product many years ago. Whether this practice is cynical or just negligent is not at issue: the point is that you cannot trust the lightfastness assurances of some watercolor paint manufacturers. And you don't know which manufacturers those are, until you test paints for yourself.
Another aspect of this issue is generational. Any survey of watercolor exhibitions, art instructional publications or retail consumer behavior shows it very clearly: older artists tend to prefer impermanent pigments. Perhaps because these pigments all date from the 19th century and are, therefore, "traditional," this preference is inexplicably turned into a badge of refined artistic standards. Here, for example, is the American watercolorist Eliot O'Hara, writing in 1946: "[Alizarin crimson], although somewhat too violet for a perfect spectrum red, is so good a mixer and so transparent that all but "sticklers" and perfectionists excuse the fact that it is not as completely permanent as several other less useful reds." This was written a decade before the many quinacridone and perylene substitutes became available, but the same glib arrogance is imitated even today. Product demand from these "old masters" sustains the manufacture and marketing of impermanent paints, and models a misguided indifference posing as refinement that younger artists are encouraged to emulate.
Shouldn't painters have the right to use whatever materials they choose? Well, obviously! Paint with food dye on latex, if that is your pastime, and enjoy! But artistic preferences become a public concern as soon as artists receive payment for their work. Then it is no longer a question of "exercising artistic freedom," but a question of business ethics. Selling a painting that will fade within a few years, with no warning to the buyer of that fact, can be considered a form of fraud, pure and simple.
Artists could easily advise buyers at the time of sale, "This painting [print] is made with one or more pigments that have been found by independent testing to fade after moderate exposure to light," and let buyers decide for themselves, at least forewarned they must mount and hang the work appropriately. Yet most artists who use impermanent pigments say nothing about it to their buyers.
Why not? Because the artists deny there is any problem. This denial takes many forms, but the two excuses I've heard most often are I have never seen any problems myself, and the paints I use are lightfast enough. (Lightfast enough for what is never explained.) Usually, these excuses are strung together: I have never seen any problems, so the paints must be lightfast enough. The first claim is certainly naive, if not cynical (one year framed in the studio or gallery is not ten years on a buyer's wall, and you know that). The second is muddled thinking: see for example Jeanne Dobie's comments on rose madder genuine (NR9).
In fact, I know a few watercolor painters who have developed the opinion that I'm not going to be intimidated by the lightfastness police, and one or two others who have told me flat out: once I sell the thing, it's not my worry. A few, candidly, feel they are not getting paid enough anyway for the labor they put in, so the buyer can have no complaints. And I wonder whether these artists are not the most honest, and speak for many others.
Which brings me to the primary reason for my concern with this issue: continued use of fugitive pigments by some watercolor painters depresses the price all can command for a fine painting. It poisons market confidence in watercolors and reinforces the entrenched belief among informed buyers and professional curators that watercolor paintings will fade. Not just some paintings, or possibly will fade, or will fade after extreme exposure to light: they fade, dude. This prejudice, in turn, justifies the practice: paintings are gonna fade sooner or later anyway, so why not use whatever paints I want?
Amazingly, despite common prejudice and beliefs, watercolors can be more permanent than oil paintings if the artist uses today's lightfast pigments and archival papers. I look forward to the day when a watercolor artist can command $2,000 rather than $200 for a superb full sheet painting, and when watercolors are considered the equal of oils or acrylics in the gallery and museum marketplace. That won't happen as long as talented, ambitious, otherwise responsible artists continue to assert that alizarin crimson or rose madder genuine are "lightfast enough" to clinch that sale.
health & environmental issues
Some artists want to attend to the health or environmental impact of the materials they use, and in a few cases watercolors do present some problems.
Environmental Impact. Environmental impact occurs through mining or raw materials manufacture and the disposal of manufacturing wastes. Unfortunately, though many high quality pigments are manufactured in Europe, Japan and the USA, the environmental consequences of industrial pigment manufacture are increasingly being exported to the Third World (China and India in particular), whose environmental laws and enforcement are of a different kind. This is one reason why pigments can be made there so cheaply.
Synthetic inorganic pigments made with cadmium, cobalt, chromium or manganese can generate severe toxic wastes in unregulated manufacture. In most cases, these can be remediated through waste treatment, though these procedures can make pigments such as manganese blue uneconomical to produce. (Paints made with mercury or lead are no longer available in watercolors, though some lead may be present in extremely small amounts as trace impurities in cadmium, zinc or nickel paints.)
I know of no way at present to find out whether pigments are manufactured in an environmentally friendly way, because (for proprietary reasons) paint manufacturers do not disclose where they buy their pigments, and there is no easy way to obtain trustworthy environmental impact statements from specific pigment manufacturers in, say, China. Painters have no effective control over these upstream environmental impacts.
Your control begins at the time you purchase and use the paint. When disposed of by washing down the sink, paints made with cadmium, chromium, copper, cobalt or nickel (nickel azo yellow, PY150, and nickel dioxine yellow, PY153) are hazards in untreated waste water for example, the leach field to your septic tank. The cumulative impact depends on how much paint you use, and for how long, but the State of California assures me there has never been a site ruled to be polluted because of the use of watercolor paints.
Toxicity Warnings. The same gang of synthetic inorganic paints raises health issues, too. But these need to be put in perspective. An artist's life in olden times was genuinely risky. Pigments such as mercuric sulfide, lead oxide or emerald green (originally made with arsenic) were toxic to manufacture and toxic to use and toxic meant it could kill you. Most painters lived through these dangers, however, so even starving artists were sufficiently forewarned not to eat their paints for dinner.
Today, paint health labeling standards for products sold in the United States follow the guidelines set down by the ASTM: "Conforms to ASTM D4236" means the paint packaging or label provides health information as recommended by the ASTM technical report Standard Practice for Labeling Art Materials for Chronic Health Hazards (D4236-94(2001)). (Some states, such as California, may require additional labeling.)
These ASTM guidelines require paints to be labeled toxic, or potential health hazards, if "in the opinion of a toxicologist" the pigment in the paint might produce "a chronic adverse health effect" as a result of any "reasonable foreseeable use or misuse" of the paints. For one interpretation of these standards applied to artist's pigments, see the excellent summary posted by the City of Tucson Environmental Management Division.
These guidelines do not specify what a "chronic health effect" is, or how serious it must be to qualify as "adverse", nor what a "foreseeable misuse" might be. They do not specify the type of proof necessary to form the toxicologist's opinion, nor the methods used to determine toxicity, nor the level and duration of exposure necessary to produce a toxic effect, nor in what proportion of the total population the toxic effect would occur, nor how serious the potential toxic effect would be.
In practice, then, the chain of definitions and logic connecting "opinion," "adverse," "forseeable misuse" and "toxic exposure" means paints may be slapped with a toxic label if they can produce any potential adverse health effects whatsoever as a result of any imaginable use by any ignorant painter or unsupervised child. That is a very broad brush.
In fact, with very few exceptions, art materials manufacturers have abandoned the use of extremely toxic pigments, notably all pigments containing lead, strontium or mercury, and have curtailed the use of most remotely toxic pigments including heavy metals such as nickel or cadmium.
Risk Assessment. The toxic effects of watercolor pigments arise from inhaling the pigment as a powder, dust, mist or vapor (when the paint is applied with a spray gun or heated on a stove); by swallowing the paint, or by skin contact with the paint through ignorance or sloppy working habits. In other words, the toxic effects of pigments depend almost entirely on your painting methods.
Risk from Inhaling. Pigments are most often inhaled when used in pastel chalks, or when the artist is working with raw pigment to prepare watercolor paints by hand, or when the paint is heated over a stove or as it dries with a high temperature air gun.
Nearly all artists buy their paints prepackaged in tubes or pans, so the hand manufacture of paints is not a significant source of risk. If you do use raw pigment for any reason, you should wear a face mask or filtering respirator mask while you handle the pigment and while you clean up after work.
Finally, watercolor paints are rarely sprayed on. Overall, then, the inhalation hazards from painting in watercolors with a brush or similar implements are close to nonexistent.
Risk from Ingestion. The principal risk with watercolor paints arises from swallowing them, and this most often happens when brushes are "pointed" or shaped with the lips or mouth while still wet with a toxic pigment. (You can also get an accidental mouthful from a burst tube while trying to unscrew a stuck cap with your teeth.) You should never put a brush in your mouth for any reason: use a paper or cloth towel to blot away excess water or shape the bristles. (I just snap the wet brush over the floor, which does both at once instantly.)
To be toxic in "foreseeable use or misuse," the exposure must extend over months or years, and must be to the most toxic (water soluble cadmium, lead chromate or cobalt arsenite) pigments. You can also be poisoned by eating the paints, but you literally must consume one or more tubes of paint to merit a doctor's attention.
The other risk is that children may be led by the lovely color of the paints to eat them, or pets may try them for a snack. Even when ingested, the cadmium or cobalt paints might make a child sick after a single dose, but they are not toxic enough to permanently harm or kill the victim.
Risk from Skin Contact. A third "hazard" is that the pigment may cause skin irritation in some people. I have never encountered anyone to whom this has happened, or even anyone who knew of someone to whom it happened, but it is possible.
Unless your reaction is immediate and strong, you may develop a skin condition slowly and not notice it at first. If any redness or tenderness develops on your skin within a day or two after using paints, consult a dermatologist for diagnosis. If you are sensitive to the copper, nickel, manganese or cobalt in paints, then you should take special precautions not to get the pigments on your hands or choose alternative pigments for the same color.
I personally know two artists, both female, who suffered acute and painful skin reactions on their hands through the use of acrylic or oil paints, but in both cases the reaction seems to have been caused by the paint resin vehicle and solvents, not by the pigments.
Used and disposed of properly, nearly all pigments used in modern watercolor paints have no toxic consequences whatsoever. For example, since the 1950's there has never been a recorded death due to ingestion of the cadmium pigment used in watercolor paints. Gottsegen puts it this way: "The hazards of cadmium pigments have been overstated in some published accounts, but do not ignore them." As a category of consumer products, watercolor paints are far safer than house paints, soaps, oven cleaners, solvents, bleaches, fuels, new carpets, pesticides or some foods (such as shellfish) not to mention oil paint solvents and acrylic resins.
That said, watercolors should always be handled with reasonable care and applied with appropriate techniques. They should be kept out of the reach or unsupervised use of children.
Watch that cat, too.