Dusk Eucalypti, 9 x 12 inches, oil on linen
One thing I like to do often - especially on small paintings - is to look at finished work and ask myself, "what if?". What if the mountains were larger? what if the trees were larger? what if the tops of the trees got cropped? what if there were more distance between the trees and the mountains? What if the sunset colors were more saturated? Less? What if the ground plane was inclined? You can come up with all kinds of compelling possible variations. Most of these go unanswered, and fewer still end up being paintings, but sometimes it starts a series of variations.
The second painting is one of these variations of the first one. I wanted to see a little more color on the mountains. When I'm doing these "variations on a theme", I make it point to change only one or two things from the original (in this case, saturation). That way, the effectiveness of the change is readily measured. If I make a lot of changes, I can't tell what the change does exactly, because the "control" element is gone. It's like comparing apples to oranges, you see, rather than comparing an orange to another, slightly different orange.
The new piece may now become the "original" and spawn variations of its own, or I can go back to the first piece and answer a different "what if?" question. Either way, the painting starts to evolve and I have all the pieces to compare and make decisions on what works and what doesn't.
It used to be that, if I had another idea based on a painting, I'd paint right on top of it and make the changes. While it sometimes resulted in a good painting, I no longer had the original to compare to. I don't think I learned nearly as much from this method as I do from my current method of creating variations on a theme.