Recently I started questioning whether it was important that the figures in my compositions continue to follow aspects of atmospheric perspective. In atmospheric perspective, as objects in space recede into the distance, their clarity, articulation, and light/dark contrast diminish as they move towards the horizon line. Objects in the foreground are clear and high in contrast, whereas objects in the distance are blurrier and low in contrast. I starting thinking about whether I could have the figures be any gradient of tone, regardless of where their location was in space. To figure this out, I did two digital sketches to allow for a side by side comparison.
The sketch on the left follows atmospheric perspective, whereas the one on the right does not. This is definitely an example of thinking in your head that something is a good idea, only to have it not work out on paper. Looking at the sketch on the right, the randomness of the gradients in the figures feels chaotic and fragments the space far too much in a way that is detrimental to the cohesiveness of the composition.
On the other hand, I did have one important discovery in pursuing this possibility. I realized that perhaps I was carrying this idea too far, and perhaps I could lend a subtle hint instead. In both of the above sketches, the figure in the center is the same gradient. However, that center figure does in some ways defy the guidelines of atmospheric perspective: the crouching figure to it’s left is darker, and yet further in the distance. From looking at this figure, it seems that defying the atmospheric perspective would work on figures that are separate and not layered on top of another figure. Without the atmospheric perspective, the layered figures become impossible to decipher. I think I’ve arrived at a better balance for my new “rule”: when figures are separate, they can potentially defy these guidelines depending on the composition, but when they’re layered, they need to follow the guidelines.