pre-computing shadow fields for multiple entities using a computing device, a shadow field representing shadowing effects of an entity at sample points in an empty space surrounding the entity, the multiple entities comprising at least one light source and at least one occluding object, a shadow field for a light source indicating radiance from the light source, a shadow field for an occluding object indicating occlusion of radiance on the occluding object as viewed from the sample points, the shadow fields being used to render shadows in dynamic scene configurations; and

wherein the pre-computing is performed independent of scene configuration.