Watercolour lightfastness tests

One remarkable source of information on the Internet for watercolour painters is the Handprint Web site. It contains a wealth of detail about colour palettes, paints and pigments -- much of it of a technical or scientific nature. (But don't let that put you off.)

Handprint makes a strong case for testing paints in the way they respond to sunlight: the property of lightfastness. It is clear that some pigments are notorious for the way they fade or change colour after a certain amount of exposure to daylight. (These have been recorded as ones to avoid.) The recommendation is that a painter should test any new or unknown paints to ensure that they will not cause disappointment (or worse) after a couple of years.

Should this worry a beginner? I am inclined to say no, but it really depends on what happens to your initial paintings. (Are they for your eyes only -- to be kept filed away -- or will you share them with other people?) It looks as though some form of testing would be a good idea, if only to determine how resistant your first paints are to fading, and whether you need different paints in future.

Lightfastness testing is not as daunting as it sounds. The basic idea is that a piece of watercolour paper (archival and acid-free) is painted with a colour swatch, and this is exposed to natural daylight for at least six months. In order to assess the effects of daylight, either a part of the swatch is masked-off, or a separate piece of paper is painted at the same time and kept in a dark location. At the end of the test, the daylight strip and the masked/separate strip are compared for any changes. As is to be expected, the Handprint site gets quite scientific about the amount of light exposure, but for UK purposes a year should be sufficient to show up any issues.

I have scanned the outcome of my first test, which ran from March 2014 to March 2015. (Thankfully we were blessed with a fairly good summer in south central England in 2014!) I used Cotman 90 lb CP paper from a gummed pad, trimmed to 5 x 7 inches. The paints are predominantly W & N Cotman, although there is at least one W & N Artists' paint. I made masks out of strips of aluminium foil to cover the painted strips before fitting the sheet into the photo frame. The frame was placed in a south-facing (double-glazed) window for maximum light exposure.

The results initially looked a little disappointing. Eight paints (more than half) showed some evidence of fading. Only the three earth colours, Payne's Gray and Intense Green appeared totally lightfast.

However, things are not as bad as they seem. The two green paints in the leftmost column (column 1) were expected to be poor. The top one is the now discontinued W & N Artists' Sap Green: my tube must be close to 25 years old. It is obvious why W & N replaced it with Permanent Sap Green! The bottom paint in the column is Cotman Hooker's Green Dark of a similar vintage, and suffers from similar issues. I now know that I need to keep these tubes away from any paintings that I may later value...

Moving to the right, column 2 shows that both Cadmium Yellow Hue and Hooker's Green Dark have faded a little. This is more apparent in the more diluted pigments. The fading in the yellow may be due to the use of pigment PY97, which Handprint says can suffer a substantial loss of color in tints in some brands. Similarly, convenience greens made with pigment PG7 depend on the lightfastness of the yellow or orange pigments, in this case PY139 -- which is reported to have very slight fading in tints.

Column 3 shows fading in both the Cadmium Red Pale Hue and the Sap Green. Pigments PY65 and PR255 are used in the red, and Handprint says both have excellent lightfastness -- so this is unexpected. The Sap Green uses pigment PY139, so the comments above about Hooker's Green Dark apply.

In column 4 we have Permanent Rose (top) and Yellow Ochre (bottom). The Permanent Rose displays some slight fading. This is pigment PV19 (gamma quinacridone). Handprint notes that some brands of paint are rated very good rather than excellent, which can be down to the pigment suppliers. No fading was apparent in the Yellow Ochre.

The Ultramarine and Burnt Sienna in column 5 showed no obvious fading, nor did the Intense Green and Payne's Gray in column 7.

Finally to column 6. The Burnt Umber on the bottom is fine. The surprise is the Cerulean Blue Hue paint on the top, which showed some slight fading. W & N's Website claims it is pigment PB15 only, but compared next to a half pan of Intense Blue (also PB15 only), they are different animals! There seems to be an additional white pigment in the Cerulean Blue Hue, and whatever it is, it appears to be causing a degree of fading in sunlight. (Could the mystery white pigment become more opaque after prolonged exposure to sunlight, leading to apparent fading?)

I think this test has been useful. It has raised a couple of questions about the pigments that go into Cotman paint, and indeed whether the pigment labelling is as accurate as it should be? Perhaps it serves to illustrate that a budget range does keep costs down by using less expensive pigments? (But a side-by-side test would be useful here.) Reassuringly though, the modern Cotman paints have not faded into oblivion. Phew!