Comparing Blues

I got some new (to me) blue colors in oil recently, and I wanted to compare them with ones I have been regularly using.

• The top row is thick application with a palette knife and a "drawdown" to thinner at the bottom.

• The 2nd & 3rd rows are tints (tube blue + titanium white)

• 4th row is pure paint applied with a brush

• Bottom row is each blue with just enough white to bring the color out. Many blues (and other colors) are so dark straight out of the tube that they appear back, or colorless.

I tweaked this for color accuracy as much as possible, but it still lacks something compared to viewing in person

My conclusions? Hmmmm..... Well, they all have their uses and tendencies.

Straight out of the tube (no tinting):  Cerulean and Cobalt show the most pronounced color. Those two are lighter in value (without he addition of white) and my eye can register the hue better. This is Williamsburg "genuine" Cerulean (a more pricey color) my issue with it is that it's too pasty and thick. Which can be remedied, of course.

Cobalt seems like the least biased blue. Not strongly greenish or bluish to my eye. Williamsburg also makes a "Cobalt Deep" which I may try when this runs out. Some projects for making color wheels suggest Cobalt as a primary blue.

The Transparency of Ultramarine is nicely evident in the top drawdown. It's really the only one that evinces much transparency here.

I'd like to bring Prussian Blue back into some paintings again. The third row, far right tint is so much like what one sees in skies. Prussian seems like it occupies a spot in between Phthalo and Cobalt on the color wheel.

I think one could certainly use 3 out of these 5 in a painting to utilize the best features of each, depending on your painting's needs.

I am leaning toward using Ultramarine and Prussian for my warm and cool blues, respectively. I have been recently painting with a split-compliment palette using Ultramarine and Phthalo.

Decoding Pigment Codes

limited palette painting demo in chromium oxide green and pyrrole red

     In the process of working on materials for my limited palette workshop at Evanston Art Center, I came across a number of interesting things in the world of colors and pigments and oil paints.

Now, I always assumed that a paint's pigment code (such as PB29 for Ultramarine Blue) was absolute. Not so much, really. It is pretty specific yes, but 2 paints from 2 different manufacturers can vary quite a bit even if they have the same pigment code. Also, 2 different-looking paint colors can be made from the same pigment, by varying the grind and including additives.

This whole labyrinth came to my attention when I realized that Cadmium Lemon and Cadmium Yellow Deep oil paints (both from Williamsburg) are each composed solely of PY35 pigment. Now, Cad Lemon is a lighter, cooler yellow and Cad Yellow Deep is more like an orangey yellow and darker in value. Pretty different while both obviously still being "yellow." How can they both be the same pigment? I emailed both Williamsburg and Gamblin these sort of questions and got helpful responses from both.

From Scott Bennett at Williamsburg/Golden:

"There are many examples of paint colors that have the same pigment ID number but exhibit different colors. Iron Oxides in general have a wide variety of colors due to little tweaks in trace elements and pigment size. Phthalo Blues and Quinacridone Reds can have the same number but with a color and a sub number indicating a variation. How the pigment is ground during paint making will change the color. PR 108,..the various Cadmium Reds, is similar but the differences have to do with the addition of selenium.

Differing hues of Cadmium reds, orange and yellow are made by increasing the percentage of selenium, zinc or sulfur during one stage of production. The range of cadmium pigments, yellow, orange, red are basically cadmium yellow (cadmium sulfide) with some selenium added in place of sulfur (cadmium selenide). Therefore cadmium sulfide can be made in various shades ranging from yellow, orange to red. Mineral pigment produced from cadmium sulphide when heated with selenium becomes red.

So it is not just one variable but a range of variables depending on the pigment. And the same pigment in different binders can look very different."


Another thing I found is that in researching colors, you really have to go to the manufacturer's site's to get accurate information on pigment content. It can seem handy to just look on the Blick site (or some other retailer), but they tend to just assume pigment content based on the name and I have found numerous errors.

Below, see Blick's pigment info on the left, and Michael Harding's on the R:

If you want to read info on specific pigments read up on:
The Color of Art Pigment Database

Two-color Painting Explorations | Confounding Expectations pt. 1

I recently taught a workshop called "Limited Palette Painting" at Evanston Art Center. Teaching a class (my first time) was a wonderful experience and I hope to do it again soon.

The goal of the class was trying to concern ourselves with pigment and paint in a deeper way by pushing a small number of colors to do as much as possible.

To work along with my students, I did a painting and mixing grid for Burnt Umber (Blick) and Indian Yellow (Liquitex Heavy Body) acrylic paint.  Working somewhat backwards (it's more ideal to do the mixing grid first), I first made a painting (using a black & white photograph for reference) using just these 2-colors:

I was rather surprised to find green tones appearing, especially in the upper right. (BTW, I tweaked the photos to look as they do in real life as much as possible. Some subtleties and colors will inevitably be off.)

I wasn't expecting any greens from 2 warm colors like these. Possibly if I had used raw umber (a cooler color), then maybe yes.

So to prove that I wasn't seeing things, I then made a matrix of possible mixtures of burnt umber and indian yellow and tinted them:

If you look at the center square and some of the ones around it, sure enough, it appears greenish.

And because I had originally started using Indian Yellow in oils, I made this comparison between the 2 versions of Indian Yellow I own in oil and acrylic:

It may be hard to see here, but in person, the Williamsburg oil version (left column) maintains a richer color as it gets tinted. Out of the tube it is more orangey and Liquitex  is more like yellow ochre.

The real point here is not so much why unexpected results can occur with pigments, but that you should explore and get to know what your pigments can do!

Pigment codes only tell part of the story, but it is important to be aware of them.

The acrylic colors I used were:

Blick Raw Umber PBr7

Liquitex Indian Yellow PY139

And my oil versions:

Williamsburg Burnt Umber PBr7

Williamsburg Indian Yellow PY83 (Diarylide Yellow)

Given that the Williamsburg Indian Yellow uses a different pigment, you would expect a slightly different result. (BTW, originally Indian Yellow supposedly came from feeding cows a diet of nothing but mango leaves and then collecting and drying their urine to get the pigment.)

Well, at least Burnt Umber is consistent. Or is it? I also have a tube of Michael Harding Burnt Umber which I haven't used yet, and it lists the pigment as PBr6. I would think that a common earth tone like that would be consistent.

Be aware.