Tests

Bleeding
A pigment's solubility in a given diluent or binder is a good indication of its tendency to bleed. Solubility may be quickly determined in the following way. Place one-half teaspoonful of dry pigment in a jar or test tube with a few ounces of turpentine. Shake well for a few minutes. Then allow the colored mixture to stand quietly for half an hour. If the pigment is not soluble in turpentine, the pigment and liquid should separate from each other, leaving the liquid clear. If it remains discolored by the pigment, this is evidence of the pigment's solubility in turpentine. The test may be carried out in the same fashion with oils, alcohol, or water to determine the solubility of the pigment in each. It can be assumed that a pigment will bleed or migrate when it is employed in a technique that uses a liquid in which the pigment is soluble. [p. 26]

Color Stability of the Pigment when Combined with its Binder
A given paint, produced by a pigment plus a binder, may be checked for color change by exposing a sample painted on a nonyellowing ground, such as a gessoed panel. The sample should be hung on the studio wall in strong light [south light if possible] and observed periodically at intervals of a month or more. Some of the same paint should be kept in reserve in the tube and at a later date, when a comparison is to be made, should be painted on an identical ground. After drying well [about one week] it can be compared with the test sample and any darkening or fading noted.

The paint should be tested for its color stability when it is mixed with white. One volume of the paint being tested should be mixed with 20 volumes of zinc white paint and applied to two test panels. After they have dried, one sample should be exposed to a window facing south so as to receive strong sunlight for several months. Half of the panel should be covered tightly with cardboard or stiff paper backed with aluminum foil to exclude light, while the other half should be exposed to the sun. The second panel should be kept inside the room, away from direct sunlight, as a control. After several months the samples can be compared to see if the exposed section of the panel in the window has altered in comparison to either the shielded section or to the panel kept in diffused light.

Colors that appear identical or very similar and are similarly labeled by different manufacturers, such as the various brands of underpainting white, can easily be compared by painting out swatches of each product. These swatches should all be of equal size and thickness of application, painted without the use of a diluent on a nonyellowing ground, such as a gesso panel or a good-quality canvas. They should be hung together in normal light and compared periodically to determine which sample changes color the most. I have samples of a widely distributed, nationally advertised underpainting white that were set out on a panel with several other white paints. While the others [p. 27] remained white, in varying degrees of purity, within two years this paint changed color so strongly that its hue was closer to Naples yellow than to white.

Such samples, simply labeled, dated , and kept on a studio wall, can be prepared at practically no cost of labor, money, or time and, if retained for a few years, are extremely informative. [pp. 27-28]

Fading or Color Change Caused by Action of Light
The pigment may be checked for light-fast qualities by grinding it in a non-yellowing binder, such as gum-arabic solution, and applying the paint to a piece of good quality white rag paper, dimensions 2 by 6 inches. When this sample is thoroughly dry, cut it in two and place one of the halves on a window sill or wall where it will be in strong light. The other half should be placed in an envelope in a drawer where it will remain in the dark. After two months compare the two samples. If there is no apparent color difference, the pigment can be assumed to be reasonably light-proof.

Since in most painting procedures the pigment frequently will be diluted with white, test samples of the pigment mixed with a substantial amount of zinc white pigment [e.g., 1 part of color to 20 parts by volume of zinc white] should also be made. There are pigments that fade considerably when diluted, although they are light-fast when used full strength. [Kay, Reed. The Painters Guide to Studio Methods and Materials. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1983. p. 27]

Flexibility and Film-forming Properties of a Pigment or Paint
The character of a film formed by a given pigment can be compared to some extent with that of other pigments by painting samples of each on swatches of good-quality primed canvas. After these have dried well for a few months, a simple scratch test will often show the difference between an extremely soft or crumbly film and a tough durable one. Flexibility may be roughly evaluated by rolling the two dried samples over cylinders, or between the fingers, and noting the speed of cracking and the character of the crack. Although such tests sound extremely primitive, they are sufficiently sensitive to indicate to a knowledgeable painter at least the extremes in differences of quality. [p. 28]

And - It should be understood, however, that it is dangerous to generalize too much from the results of tests such as those mentioned unless the same results are obtained repeatedly with other batches of the same materials. Thus one should be cautious about leaping to the conclusion that a new pigment, X, bleeds in oils, just because one sample does so. Perhaps the tested sample was "improved" by the addition of a dye material. On the other hand, if several samples obtained from different sources should bleed, it would surely be wise to question the color. Scientific accuracy is possible only under carefully controlled conditions, and complicated problems or fine distinctions of quality and paint behavior probably cannot be investigated by most painters. However, even the gross results of the tests mentioned here can be extremely helpful to the practicing artist . [p. 28]