On the relation between ray theory and the banana-doughnut formulation