Finally understanding how to create relatively consistent characters across multiple poses/pictures is pretty satisfying