Tech Talk: Lightfastness tests – 5 months

full view of my test card –

sun-exposed half on left

center panel compares full-strength color –

sun-exposed paints to left of center

center panel compares tints (diluted washes) –

sun-exposed paints to right of center

Back in January of this year, I began a lightfastness test of these Mission Gold watercolors after receiving some samples of the paints. Because they are new to the market, there were no published lightfastness tests, and many of the pigments used in the manufacturing of these paints were poorly rated in the resources I have.  (See my previous posts here (introduction), here (at 2 weeks) and here (at 6 weeks).)  I put half of the sample sheet in a folder made of acid-free matboard and stashed it in my flat files.  My sun-exposed half of the test sheet started out with several hours of winter sun each day, but at this point, the sun is blocked most of the time by trees.  I've now moved the sample to a spot that gets sun for about 4-5 hours most mornings and will continue the test for about six more weeks.  However, there are such clear changes at this point that I would be remiss in not sharing the results.

I took these photos in full sun yesterday and have done a bit of digital cutting and pasting on the second and third images so you can more easily see the comparison of the full-strength pigments and the tints. (Note that the sun-exposed pigments are on the left side of the full-strength sample and on the right side of the tint sample.)

I should point out that pigments ranked as lightfast on some lists may indeed be lightfast in other media, but unreliable in watercolor paints. As I understand it, this is due to the binding agents used -- oil and acrylic paint bases being more protective of the pigments than the gum arabic used for watercolor.

As noted in earlier posts, the full-strength Permanent Yellow Light (PY17) has darkened with exposure and it has also faded significantly in the diluted wash. Quite a predictable outcome, since the pigment is known to be fugitive.

Yellow Orange (PY65 and PO13) is composed of a very reliable yellow pigment, but also contains the same unreliable orange pigment as Orange (PO13). Both show darkening at full strength and fading as a diluted wash, although it's more pronounced in the Orange sample.

The Permanent Red (PR112) is only slightly changed at either strength, but Permanent Rose (PR122 and PR209) is quite faded both at full strength and as a tint. A bit of a surprise, actually, since Permanent Rose is a mix of two supposedly reliable quinacridone pigments. [added note: There is some disagreement on the lightfastness of PR122, with Michael Wilcox labelling it as unreliable.] The Rose Madder (PR83:1) is actually alizarin crimson, a pigment which has been replaced in most professional-quality product lines by a more lightfast substitute. This paint changed only slightly at full strength, but is noticeably faded as a tint.

Permanent Violet (PV3:1) is anything but. It began showing signs of fading by the second week of sun exposure and is now seriously changed both at full strength and in the tint. This comes as no surprise, because the pigment is a known bad actor.

All three blue paints in my sample -- Ultramarine Deep (PB29 and PV12), Prussian Blue (PB27), and Peacock Blue (PB15:3 and PG7) -- seem quite reliable, despite the inclusion of a mystery violet pigment in the Ultramarine Deep. I can find no information in any of my resources on it, but it doesn't seem to affect the color stability.

Viridian (PG7) is actually phthalo green, which makes it very stable, but also a very strongly staining paint. Sap Green (PG36, PBr25, and PY17) includes the same yellow pigment as Permanent Yellow Light, which means it shares the same problems -- darkening at full strength and very noticeable fading in the tint.

Burnt Sienna (PR101) was a definite suprise -- fading badly across the board. Not a true burnt sienna, it is composed of a red pigment that is supposed to be very lightfast, but this sample didn't live up to the reputation of the pigment. 

VanDyke Brown (PBr9) is made of a pigment that does not appear in any of my resources, but is similar to one that is rated fugitive/unreliable. It faded particularly badly as a tint -- in both samples!! I had to go back to check my photos at 6 weeks to verify that there had been a noticeable difference in the exposed and non-exposed tints at that time. I therefore have to conclude that simply exposing this paint to air will cause it to deteriorate at tint strength.

So I'm sorry to say, but my advice would be to avoid Mission Gold paints as they are currently formulated -- unless you're planning to channel Picasso's Blue Period or keep your work in a drawer.

Tech Talk: Pigments - Week 6

Lightfastness Test Strips

(bottom half is the sun-exposed section)

Back in January, I painted these test strips of Mission Gold watercolors because I was concerned about the lightfastness of the pigments used in manufacturing these paints.  As I've reported before, I cut the sheet down the center and put half in a drawer in my flat files and half on an enclosed porch that gets a few hours of strong sun on clear days.  Despite the limited sun exposure during our rainy winter months, there are definite changes to many of these pigments.

The sun-exposed yellow has become darker in the full-strength strip, but other colors have either become slightly dull or are changing color slightly.  At full strength, permanent rose, permanent violet, and burnt sienna are faded and somewhat duller now and prussian blue is slightly faded and also leaning towards green. 

Changes are more noticeable in the tints, which is often the case with impermanent pigments.  While orange and peacock blue had only slight fading, there is noticeable fading to the permanent rose, rose madder, permanent violet, prussian blue, and burnt sienna.  Sap green and Van Dyke brown both showed moderate fading. 

Out of curiosity, I changed the white background of my photo to a 60% grey tone in Photoshop.  The neutral grey really makes the colors pop and makes it easier to see the changes to the sun-exposed pigments.

I'll be leaving the test section in place on the porch for the forseeable future.  Even with many hours of full sunlight, it can take several months to see the full extent of changes to pigments.  I plan to report back here in a month.

Tech Talk: Pigments – Week 2

 Lightfastness -- comparing full strength paint

(bottom half is sun exposed strip)

 

Lightfastness -- comparing tints

(top half is the sun exposed strip)

Two weeks ago, I began testing these Mission Gold watercolors. (See the original post here.) After painting the test strips, I cut the page in half and put one piece in my flat files and the other in a spot that would get direct sun. Following my plan to compare the sheets every two weeks, I put the two pieces together today, first comparing the full-strength strips and then the lighter washes.

There are a few changes at this point, although nothing drastic. As might be expected, the tints (light washes) were generally affected more than the full-strength colors.  Prussian blue faded slightly in the tint, as did burnt sienna -- both a bit of a surprise, since the pigments listed on the tubes are supposedly lightfast.  There was also a slight darkening of the yellow, particularly noticeable at full strength. The tint of Van Dyke brown faded slightly.  Neither of these was unexpected -- the yellow pigment was rated as fugitive by both Michael Wilcox and Hilary Page and the brown is unlisted, but its closest relatives are considered fugitive.

It's worth noting that the sun-exposed strip has had only a few hours of sun a day on the best days and we had 4 or 5 days of deeply overcast weather in the past two weeks. I'll continue to monitor the test strips and will post again when there is more information.

Tech Talk: Pigments

Watercolor test strip

I'm going to do a number of posts on the technical aspects of painting with watercolors. To start things off, I want to talk about paints. I was inspired to begin this series after receiving a couple of sample sets of a new line of watercolors, Mission Gold, a few months ago. I'll talk more later about the color test strip shown above, but first let's talk about how you can check the pigments used in your paints.

The first thing I do with an unfamiliar tube of watercolor is to look on each tube for the pigment information.  These days, all reputable brands of watercolor list the pigment number or numbers on each tube.  I've circled the numbers to make them easy to find on this assortment of watercolors from various manufacturers.

Pigment numbers (circled)

Pigment numbers are in the form PY## for yellows, PO## for oranges, PR## for reds, etc.  Once you have the pigment number (or numbers in the case of a mixed color), you can check several sources to learn more about them.  For years, I have relied on The Wilcox Guide to the Best Watercolor Paints and Hilary Page's Guide to Watercolor Paints.

You can check the raw pigments as well as the ratings for individual manufacturer's paints.  Hilary Page has tested all the paints herself and gives personal ratings, pointing out her favorites in each color, but Wilcox gives the most comprehensive information on the raw pigments -- pointing out pigments that may be suitable for oil or acrylic paints, but not for watercolors. 

There is also an online resource by Bruce MacEvoy.  He has created an exhaustive catalog of information related to watercolor, with a very thorough section on pigment information. By clicking on the color links at the top of that page, you can see every pigment by number along with detailed information about each one. There is an overwhelming amount of material, but for the purposes of this post, the critical thing is the lightfastness rating (Lf), the column at the rightmost edge of the rating tables. Lightfastness is rated from 1-8, worst to best.

 Now back to my color tests.  After looking up the pigment numbers on the paints in my sample sets, I was very concerned because some of them were ranked as fugitive or unreliable by both Wilcox and Page.  However, since this is a new line, neither Wilcox nor Page has tested the actual paints, so I decided to do my own lightfastness tests.  I had 14 unique colors in the sample sets, so I took an 11x14-inch piece of watercolor paper and divided it into 14 equal spaces.  I printed out the color name and pigment numbers in the center of each space and then painted a full-strength stripe across the middle of the paper, with identical lighter stripes on each side of the center. You'll notice I blotted out the paint at the center of the darker colors in order to see the names and pigment information. (I also lifted a stripe down the right side of each color, but that is unrelated to my test for lightfastness. I merely wanted to see how staining each color was.)

I cut the paper down the middle, labelled the backs, and taped one to a piece of acid-free matboard that I set where it will get as much full sun as possible. The other half is sandwiched between two pieces of acid-free matboard and tucked into a flat file drawer.  At present, the daily dose of sun is not great -- a few hours at best -- but my plan is to compare the two halves of the test sheet every two weeks for at least 6 months.

Stay tuned.