By the time I get home, I've thrown away several ideas and usually have a strong vision for the painting. Before I go to sleep, I sit in bed and sketch a few quick thumbnails while the idea is still fresh.

The next step is gathering reference—lots of it. I collect, photograph, or create whatever I need, then draw and redraw the characters and environment until the composition feels right.

At this stage everything is still loose. I often think about the lessons Andrew Loomis taught in Creative Illustration: solid design comes first.

I rarely produce highly finished drawings because I think like a painter more than a penciler. Once the composition is working, I begin exploring light and color. Most of this happens in Photoshop, where I keep experimenting until I reach what I call the "A-ha!" moment—the instant I know the painting is going to succeed.

From there, the rest is craftsmanship: refining, polishing, and bringing the image to life.

Finishing a painting has become more intuitive with experience. I've been creating professional illustration since my time in Ubisoft's marketing art department in the early 2000s, and over the years I've learned to adapt my process to whatever a project demands.

For high-realism franchises like Tom Clancy's Ghost Recon and Rainbow Six, I relied heavily on photography and 3D reference. For stylized toy advertising created for Mattel, I used digital airbrush techniques to achieve a bright, playful, cartoon-inspired look.

Every project requires a different approach, but the goal never changes.

At the end of every painting, I ask myself one final question: Did I capture what made me fall in love with this character in the first place?

If I can see it—and if fans can see it too—then I've done my job.

Then I reward myself.

Usually by starting to think about the next painting...

...on the drive back to Buc-ee's.

The important word, though, is reference. The danger is falling in love with the tool instead of the picture.

I learned this while illustrating a science-fiction book cover. The painting needed a futuristic cyber tank, and I decided to build the thing myself in ZBrush. I modeled it. Detailed it. Refined it. And refined it some more.

A month later, I had a very nice cyber tank.

I had also apparently changed careers and become a 3D modeler.

Once I rendered the tank and placed it into the painting, I realized something slightly embarrassing: all that detail wasn't nearly as important as the composition, lighting, action, color, and values surrounding it. The tank mattered, but the picture mattered more.

That lesson has stayed with me.

My job as an illustrator isn't to prove that I can model every spaceship, photograph every costume, or paint every raindrop individually. My job is to create an image that makes you stop and look. Technology is valuable when it helps me get there faster, more accurately, or in a way I couldn't otherwise achieve. But eventually I have to paint over it, change it, exaggerate it, simplify it, and make it serve the image I had in my head.

Which brings me back to Norman Rockwell.

Whatever tools he used, nobody remembers his paintings because he owned a camera. We remember them because they told stories. They made us laugh, recognize ourselves, feel nostalgic, or occasionally get a lump in our throats.

The tools helped him make the picture.

The picture was the point.

I try to remember that every time I open Photoshop.

And whenever I find myself spending a month modeling a cyber tank.