We view these characteristics as resulting from a hierarchy of priorities. The edge lines and highlights are black and white, and provide a great deal of shape information themselves. Several studies in the field of perception have concluded that subjects can recognize 3D objects at least as well, if not better, when the edge lines (contours) are drawn versus shaded or textured images [1,3,5,22]. However, when shading is added in addition to edge lines, more information is provided only if the shading uses colors that are visually distinct from both black and white. This means the dynamic range available for shading is extremely limited. In most technical illustrations, shape information is valued above precise reflectance information, so hue changes are used to indicate surface orientation rather than reflectance. This theme will be investigated in detail in the next section.
A simple low dynamic-range shading model is consistent with several of the principles from Tufte's recent book . He has a case-study of improving a computer graphics animation by lowering the contrast of the shading and adding black lines to indicate direction. He states that this is an example of the strategy of the smallest effective difference :
Make all visual distinctions as subtle as possible, but still clear and effective.Tufte feels that this principle is so important that he devotes an entire chapter to it. The principle provides a possible explanation of why cross-hatching is common in black and white drawings and rare in colored drawings: colored shading provides a more subtle, but adequately effective, difference to communicate surface orientation.