Lately I’ve been writing a series of posts about the night, inspired, in part, by my fascination (obsession?) with bats. I’m thinking of developing this into a new illustrated essay, eventually. But the illustration part has posed some challenges.
If all the pictures in this essay will be night scenes, then they will have dark backgrounds and will probably be pretty low-key. Spare highlights over dense, deep shadows is the idea here. I’m not entirely sure how to do that.
For the most part, I’ve worked in ink wash over the last few years. Ink wash almost always involves black ink on white paper. There are white inks, (Note: Windsor & Newton and Liquitex make decent acrylic inks in white. ) but I’ve yet to find any that are truly opaque. (Note: This is kind of by nature: ink needs to be fluid and viscous, so it’s usually made with either dye or nano-particle pigment in a water-based binder, along with maybe some shellac or lubricant. It’s not as opaque as paint simply because it’s so much thinner; there’s not as much pigment or dye per volume. Will Bailey has an excellent video that goes much deeper into different types of ink and how they work. ) My white ink, undiluted, usually ends up gray when applied over a black ground, even after several coats.
As another approach, I’ve used digital coloring and compositing to create darker pictures with ink wash before. I’ll draw various parts of the image separately, scan each piece, then composite it all together in layers and apply color floods around the black ink lines. This is a pretty similar process to animation or comic books, and it works well. (Note: I’ll write more about how I do this later. Over the years I’ve found ways to handle masking and color that have been really helpful. ) But for this essay, I think I’d like to try a different approach. I want something that feels a little more organic, textured, and layered. A little more loose and gestural. And that probably means less computer.
Like watercolor but not
After some research and experimenting, I decided to give gouache a try. It’s a medium I haven’t really worked with before. Gouache is a water-soluble paint that uses gum arabic as its binder, just like watercolor. In fact gouache basically is watercolor with more pigment and, sometimes, a thickening agent (back in the day, they added chalk). It dries to a matte finish but colors tend to be bright and saturated. And, thanks to its denser pigment and thicker binder, it’s opaque. White gouache over a dark ground stays white. (Note: At least for an opaque color like Titanium White. Some whites are more transparent. )
Gouache behaves quite differently from ink and it’s taken a lot of getting used to. It’s hard for me to break old habits. With ink, I mostly have to use the white of the paper for highlights (except when I “cheat” and do it digitally) and, so, I work light-to-dark. I start with very light, dilute washes and build up darker values slowly, layer by layer. Once I’ve laid down some undiluted black, I’m committed to that area being a dark value. With gouache, on the other hand, I’m free to work dark-to-light. I can paint a black flood, and then paint over it with white later. That’s one of the main reasons I chose this medium and it’s wonderfully freeing. But I still catch myself trying to “protect” bright areas, dancing around them with a pointy brush, when I should have just steamrolled it all with black and added my lighter values later.
Gouache is also more difficult to layer. It’s not waterproof, so dry paint can be “reactivated” with a wet brush. That’s helpful in some ways but it often means that I accidentally pick up paint from a previous layer when I add a new one, or create blends where I don’t want them. Pigment inks are usually water-resistant once dry, so it’s much easier to add a layer, let it dry, and then work freely over top of it. Acrylic paint is also like this; it dries quickly to a flexible “plastic” film that is completely waterproof.
After some frustration I took to the internet and found some help there. In particular, this video from Ogygia Art on Youtube saved me a lot of gray hair. James Gurney, as always, has lots of good advice on his blog and YouTube channel on gouache’s unique character and quirks.
Trial and error
As a test, I thought I’d try to paint some moonflower blooms from our yard last fall. I took a photo one night when the flowers opened.
But gouache is a very different beast from ink, and I’ve struggled to learn how to handle it. I’m getting there, but I still don’t feel nearly as confident with it as I do in more familiar media. I ended up painting many versions of the same flower picture, trying another approach or possible solution in each. There were many problems to figure out.
First problem: finding the right ground
I wasn't sure what ground to use — this is, what I should paint on. If I’m going to use a dark background and I want to work dark-to-light, I figured it might make sense to use a toned paper instead of the bright-white, hot-press watercolor paper I usually work on. I love the look of charcoal and crayon on toned paper, and I’d seen some nice examples here and there of gouache on toned paper as well. Using a limited palette — a bright white and a near-black — maybe I could work up from the paper’s tone to get a loose, light-touch kind of effect. I did quick tests on some black and gray paper I have in my cabinet.
Unfortunately, it turns out my toned paper only supports dry media. As soon as I applied paint, it crumpled like a wet napkin. It did better if I applied paint in thick, dry coats, but then I didn’t have much control over tone. That made the subtle shadows and veins in the flower petals hard.
But I did like the idea of starting from a mid-tone ground and then building up (and down) from there. I went back to my favorite Fabriano watercolor paper, but used an underpainting to establish my non-white ground. After a few experiments, this approach felt promising.
Next problem: layers
I needed to get across the bright “glow” of the flowers, but also the deep black night in the background. I eventually realized the flowers need a brighter undertone than the background, or the whole scene would feel too dim and cold. I tried a few different combinations of underpainting and layered floods. A layer of Cyprus Green under Pyrene Black made a rich, velvety black. After I watched some more tutorial videos, I discovered that a brighter watercolor underpainting — Windsor & Newton’s Green Gold was pretty nice — helped the white highlights on the flowers, lending them a warmer tone that was striking against the darker, cooler surroundings.
But I kept striking out. Everything either got too muddy, turning into a baby-puke mess, or too bright, or too dim. Predicting the kind of texture and contrast I’d get with a given set of layers was really hard. I eventually arrived at something that feels okay, but I’m not really happy with it. The color and contrast are close but still don’t feel quite right.
One more problem: reproduction
A wonderful thing about ink is that it’s easy to scan and manipulate digitally. Black ink on white paper translates beautifully to an alpha channel on a computer. But layers of colored paint are much, much harder to work with. The colors I’ve been using don’t seem to scan well, and I’m also chafing a little at the lack of control I have to tweak (or entirely change) color digitally. Working in a more analog way means that I have to commit to color decisions much earlier, and I’m essentially stuck with whatever colors, tones, and textures that I’ve put on the paper. I end up losing a lot of the fine-grained precision that digital color spaces offer, and I miss having the chance to do that.
As an experiment, I scanned in my last version and then converted it to grayscale, as though it were an ink wash drawing. Then I applied some gradient maps and other color effects to digitally re-create the color palette I was going for. What I ended up with felt better, but still not quite there. The gold and green tones in the paint don’t quite translate to the gray values I would want as a starting point for digital color.
What I think I’ll try next time is I’ll paint only in gray, black, and white. Then I’ll go back to handling color entirely in the computer. I can build up values and texture with the grayscale paint. When I scan it in, I can create alpha channels from the painting, just as I would with ink, and then color it. That might give me the best of both worlds: the organic textures, layers, and gestures of analog media, with the precision and flexibility of digital tools.
I can’t help but be disappointed that my first attempts with gouache didn’t land where I hoped they would. But these things take time. It takes time to learn to handle a medium. And it takes time to figure out the right process.
In some ways, I feel like I’ve spent most of the last 20 years of my career trying to figure out process. How do I execute an idea from start to finish? What’s the right combination of tools and approaches to end up where I want? When should I use what I know and when is it worth exploring unfamiliar ground?
I’ve learned a lot about gouache and it’s also been a good chance to reflect on where I’m trying to go with my illustration work. Where ink lends itself to precision and detail, gouache gives me space to be a little looser and messier. In this way ink and gouache work well together in counterpoint. Broad gesture and detail.
Gouache insists on economy. This is good for me. I have a bad habit of overworking things, getting too fussy, worrying my lines until they’re frayed. My gouache layers look better if I can make my statement in a few brush strokes and then leave it alone. Once I start dabbing at it and pushing things around it becomes a muddy mess, quickly.
Gouache seems happiest, at least for me, in thin, succinct layers. It's better when I don't put too much paint down. Start wet and transparent on the base layers, and let the paint get thicker and more opaque on higher layers. Likewise, start with big shapes and tones, and finish with small touches, so that the ratio of transparent washes to opaque strokes is high, in favor of the former.
Finally, I’m learning there’s a fine balance to keep between getting it right and knowing when to stop. Lay down a layer, get it just right, let it dry completely, then start another layer. But if something is working, call it a day and leave it be. Often, my urge to add just one more little thing has pushed a promising drawing over a cliff.
So: next week I’ll start another one.