I wanted to write up each Landscape thumbnail that I painted, but I ended up needing to dedicate the time to the challenge itself in order to finish on time. I am very pleased with the results that I got from doing this, and I think I’ll be working thumbnails like this into my workflow for a while. It’s a really useful way to experiment quickly, and reduce a composition to just a few problems. Then, I can iterate on ways to solve those problems before making a finished version.
Just as an example – consider my first attempt, and the final version of each of the five compositions that I painted:
This highlights exactly why I am so often frustrated by painting with watercolor. I start out intending to paint a composition, and end up fiddling and scratching until the whole thing is just one big muddy mess. By putting all that trial and error into individual thumbnails, I find that I’m able to not just find out what the best composition will be, but what techniques will be appropriate for each element in the composition. And, the thing that excites me the most… these lessons really came to life when I painted one of the thumbnails on a quarter sheet. I’m very happy with this painting, and I am confident in saying that this painting was a success because I learned what not to do in the 34 thumbnails I painted before I painted this.
All of the noodling around that I would have done ordinarily all throughout the image happened in the area of focus. And that helped me keep a tight rein on the experimentation, which ultimately helped contribute to a feeling of freshness and the gestural qualities that I like so much about watercolor. In short, by dedicating time to the exercise of painting a whole slew of thumbnails before attempting a painting in full, I had much more confidence when it came time to paint the final version. And that confidence resulted in an image that is more lively, and more accurately expresses what I wanted to express.
For reference, here are all the thumbnails I painted while I developed this painting.