The challenges of painting clouds with softness and subtle color variations

Catherine Creek Cloudy Day

 I was at Catherine Creek State Park with a friend one March morning a few years ago when clouds were moving across the sky, blocking and then revealing the sun. We even got drizzled on for a bit, but had a lovely 2-hour hike through the basalt bluffs and weathered towers, with spring wildflowers blooming throughout the meadows and the oak trees still bare. I did one painting soon after that of Wankers Columns, but had several other photos I wanted to paint, including this one. When I took the photo, looking into the sun, I was really happy with the shapes and density of the clouds, framed below by this small rain-shaped gully bordered by rough basalt ridges and thin stands of white oaks.

But the photos themselves were complete duds, no color whatsoever in the clouds and the landscape completely silhouetted against the brilliant sky and almost-revealed sun. I loved the composition but hated the loss of color. For the years since, I knew I didn't have the skills to recreate the scene, with the brilliant spring grass and the subtle hues of the backlit clouds. In the last year, however, I've been practicing on clouds, working to develop a technique that would give me the look I wanted. I don't know how many times you have to paint a subject before you know it well enough to create it from your imagination, but that is my goal with my subjects, like the landscape and clouds here.

Acrylics have a lot of challenges in the best of circumstances, and more for slow painters like me. I have neither the experience nor the confidence to paint a new subject boldly and quickly, before the texture of the paint becomes either too dry, or too wet from trying to keep a mixed color from drying out over a period of several days--or weeks--while I work out the details of the composition. I almost never want to use the straight tube colors--I end up blending them, adjusting the value, and mixing colors freely with both their complements and their neighbors on the color wheel--whatever I have to do to get just the colors needed to work together.

I also like to layer colors to mix them. On this painting I used more than a dozen different mixes of gray, white, black, two blues, two purples, yellow, orange, and green to get the hues I wanted. And I frequently misjudge values, even with the reference photo on my iPad for the best color, and end up having to lighten or darken whole areas after I've painted them, to get the scene right. Sometimes it feels like I've painted the whole painting three different times, slowing bringing everything into the right balance.

My latest discovery has been that once I have the basic pattern of the painting done but the colors just aren't right, and they need just small adjustments, I can tweak them by picking a very small amount of paint on the right-sized brush, then wiping the brush gently on a piece of paper towel, so there's very little paint on the brush. With that, I can put the paint onto the canvas very carefully without it globbing or running, or me leaving so much that I have to wipe it off (which is also a useful technique.)Then, of course, I have to let that bit of paint dry so I don't smear it with additional strokes. That way, I can adjust the value or hue by small amounts. And when you do finally have the value relationships right, you can get amazing depth and subtle textures of colors.

Wrapping it all up

I spent most of the last week working over the rocks,  a day on the grasses and minimizing the outlines of the columns, then two more on the sky and the tree. I'm happy with the overall dark-light pattern of the loose rocks and how it angles up and to the right, curving slightly around this end of the butte.

The most interesting thing that happened was a disagreement between me and the bottom left corner of the sky. I wanted it to be dark clouds, to match the upper right, but overnight I realized that made the whole left side look as heavy as the right, and dragged down the energy of the whole painting. The next idea I had was to turn those clouds into lighter, more bluish shadows—but as I worked on those, the whole small area turned itself into light sky instead. As soon as that happened, it made the white clouds look even brighter, punched up the green of the grasses, and for some reason made the face of the columns stand out more, both in brighter values and also in a dimensional sense.

I was so taken with that brightening effect that I invoked my artist's license to delete the bare tree growing out of the butte face on the left side (sketched in, in previous versions) so it wouldn't detract from the vertical lines of the butte. My excuse is that 100 years ago, that tree would not have been there, but the butte and the rocks, and probably the grasses too, would have looked the same as they do now.

And I got just the spring color combination I wanted!

Balancing rocks for depth

Working on the foreground rocks has been interesting because they need to attract the eye and lead it  into the painting, but they can't attract the eye back down once it gets to the butte face. They have to be detailed enough to be interesting, but not enough to be more interesting than the butte. They need to be relevant to the design, but not stand out.

They're not finished yet, but I believe their shapes are defined enough now. I started by using dark gray washes to fill in and darken the whole foreground masses. Once they were sufficiently dark overall, I switched to a lighter gray and carefully painted the rock faces I wanted to highlight. After I had those highlights where I wanted—where they would establish direction in the overall design—I went back with the darkest dark yet to reinforce the shapes of the individual chunks of basalt.

At the same time I darkened the background scree slope on the right where it's in shadow.

All this is to try to show depth, and to focus attention on the butte face.

Main cliff done

I've put in a lot of detail on the main cliff face and the fractured columns in the upper right; I think it's time to stop working on this area. The colors and values are pretty much where I want them for now, and from 5' away, it looks like what I remember.

I'm surprised how rough and sloppy the edges of the individual columns look in close-up, when they look so great from a short distance away. Not going to change them now.

More green

The stones are almost all defined, and I've painted in a few of the highlights on the cliff face. I've added a lot of green tint on the columns, and darkened some greens in the grass. Not sure how much of the blue I'll be keeping in the next layer, which will be the red-brownish gray.

I've noticed one thing about this painting—the smaller the image is, the better it looks. Up close, it still looks like a rough watercolor. I'm still having fun with it.

Light to dark

I was working on darks today. I needed to identify the shadowed surfaces on the butte face so I can get the right colors in the right places, and I worked on the mid- and foreground rocks to shape them and bring them closer to their final color. I worked simultaneously with six different hues: a neutral gray, a blue gray, a red-brown gray, a more neutral brown gray, an ochre gray, and an olive gray—every hue I can identify in the reference photo. For the most part, I put them on pretty dark, a luxury that working in that acrylics gives me. Whenever I need to lighten an area, I can mix in white to do that. Which answers the question I was asking myself yesterday—I'll start using opaque colors as soon as I have to lighten something.

The work I've done so far really reminds me that the first art instruction I ever got was in watercolor, and in that training I learned to work light to dark because we weren't allowed to use white. When I finish the darks as much as I can, I'll start putting in the lighter bits—the lichens, lighter surface scale, and more reflective areas. To me this looks like a watercolor now, still having a high degree of translucency, the white surface of the canvas contributing to the highlights. I always hate to lose that translucency, probably because of that watercolor training.

Romancing a color scheme

I'm starting a new painting of yet another place in the Columbia Gorge—this is back in Catherine Creek State Park in WA, at a tall basalt butte called Wankers Column, favored by the rock climbing crowd for its crevices & columns. I took the reference photo on a partly rainy day in March while hiking with a friend. For those of you who have been there, this is the south face. There were no climbers out that day.

I'm in the middle of defining the dark areas in the rocks, boulders, and columns. I decided to use ultramarine blue with brown as the base color for all the stone (except that one red strip), partly to give the feel of a cold spring day. My plan is to continue using transparent washes on the stone surfaces. I'll be working on them with olives and pale gray for the lichen, burnt orange for the oxidized areas, and violet- and brown-grays for the rest. I'm curious to see when I'll have to switch to opaque colors (other than when I screw something up.) Usually I'll block in the first colors with opaque hues, then use washes on top of them. Just a slight difference in approach.

I had a big argument with my projector (the Tenker) about the photo I used; it wouldn't accept the format of this particular photo, while it did fine with others. I ended up having to convert the jpeg to a  tiff, and then to convert the tiff to a new jpeg before it would accept it. Still not sure what the problem was, but it wasted over two hours.

A post-sketch painting of Catherine Creek

Memaloose Island

After doing the sketches of the gorge I wanted to see if they made it any easier for me to make a painting of it, so I basically copied the image of the second one, a combination of the two photos—the basalt outcrop at Catherine Creek, and the gap just east of there. And no big surprise, it did make it a bunch easier, enough that I finished it in just over a week, working around a bit of stomach flu. It did take me a few hours' work coaxing the complex hills and bluffs into a close approximation of the way they actually look.

I'm still not very familiar with the geography there, but I think the near bluff on the right is Rowena Crest, and the more distant land on the left is around Lyle. (I added some buildings to the painting after I took this photo.)

The most fun part was the foreground, playing with the big color range from the black basalt through the dried and the green grasses, to the wildflowers, blue Camas and a pink thing that looks like a clover flower that I can never remember—Rosy Plectritus. I used 3 yellows, two greens, brown, orange, white, and black just for the grasses! Fun playtime!

I wanted to make a center of interest around the right cluster of flowers to bring attention to that area, so I put some soft highlights in the trees just above them and then darkened the ground just below them, and that gave them just a little boost.

Detail

It was nice to finish it so quickly—last year a painting like this would have taken at least 2 weeks of work and a lot more worry; I credit all the sketching & quickie exercises I've done this year for the speed improvement. I used to get into so many situations where it just seemed like there was no way out, and that doesn't seem to be happening any more. I'm getting into a habit of just trying something, anything, when I get stuck now, and once I stop worrying about the outcome, it's amazing how many times those wild-guess experiments just work.

Looking East

Looking East

I was actually afraid last night and this morning that I wouldn't be able to do another gorge sketch today. But I started the same way and used the same strokes, and they really look a lot alike. I learned something new, though—that with the Blick Professional Gesso I can scrub paint off with a wet brush  when I make a mistake and get almost all of it off without visible harm to the gesso. That's very handy. I first drew the rock outcropping at the wrong angle, and was able to fix it.

I did more work on the foreground in this one, wanting to see more of the grasses and do more to balance the foreground and background. I also mixed a lighter color for the distant bluffs.

Sketching the Gorge with a very used brush

Rowena Plateau

I decided it's time to start working from some of my photos of the landscapes around Catherine Creek State Park, east of White Salmon, WA, on the Columbia River. I sifted through them yesterday and grabbed a couple to start with. I was ready to start a canvas panel this morning when my muse whispered, "Do sketches." I spent the last two days doing color experiments so I was already in "quick" mode.

I mixed up a nice china blue from Pthalo blue and Dioxazine purple, and picked up a superbly beat-up and damaged small flat brush and started scribbling in the dark shapes. The bristles on one side are bent back 360º and they were great for scratching on trees, branches and other rough shapes, while the other side was still straight enough to sort of draw lines and put the paint on flatly. I did use a different tint on the river, a mix of Pthalo and Cobalt blue. I was going to color it in fully, but decided to stop with a value study.

I'm surprised how traditionally "sketchy" it looks, and yet how fresh and "real", as if I'd done it right there. It seems like a great example of brush-stroke texture as detail, and I'm liking how many different looks I got depending on the value and thick-or-thinness of the paint and how much time I spent on that bit. No real painter would ever throw a brush away, unless maybe the handle breaks off, and clearly this is why. What a gold mine!

A big loose-ish landscape

I was thinking I would try doing a photo-based landscape in the same brushstroke style I used on Wild Iris, but as I was painting in the sky and tree foliage, I decided to reduce the amount of texture by having fewer brushstrokes and more larger, smoother areas of color. Most of the water is pretty smoothly mottled, but I tried to delineate most everything else.

The location is the Catherine Creek State Park in WA, northeast of Hood River. It's a great wildflower area I go to with friends most years in May. As far as I know, only the elves use the stones to cross—there's a wood plank bridge for humans. It's where I took the photo from.

It was a pretty straight-forward effort, using a lot of glazing to get the myriad of greens and blues that are in here, and a guess at how many highlights to put in. They could change. This one took me a week, the longest painting I've done all year. I would personally call it "impressionistic" because it really is all about the light.

Why paint THIS picture?

I feel as though I'm finally getting closer to simplifying the picture enough to just capture what I want to, in order to answer the question, "Why this picture?"

Little waterfalls at Catherine Creek

Backyard trees, early morning

P.S. - I've finally spun off a blog that's just about my garden. It's at outinthegardendancing.blogspot.com.