----------------------------Original message---------------------------- Greetings all (and pardons for cross posting), I have done a couple maps with shaded relief. The relief image is generated by a fairly simple computer algorithm, but there is a fairly wide degree of control over things like sun angle, ambient light, specularity, etc. On the first map I placed the shading as a TIFF in FreeHand, assigning it an ink color of "30% black", effectivly compressing the dynamic range to 30% of the original. Printed as a halftone, it could only use a small portion of the possible range of the black ink. Nonetheless, this worked reasonably well. On my second map, I gave the shading its own spot color (the maps are printed in spot colors, no process) - a slightly blueish gray - with the idea that by using the entire range of halftone dot sizes, the image would be less muddy. It didn't work out that way. In fact, if anything it seemed more muddy, possibly because the blue in the shading did not work well with green of the vegetation overlay. Recently, thanks to a tip from a cartographic colleague of mine, I got a copy of "Cartographic Relief Presentation" by Eduard Imhof (of the Swiss Topographic Service). It was written in 1965, with translation to English in 1982 with some updating, and published by Walter de Gruyter (ISBN 3 11 006711 0). It's an absolutely wonderful book, and I recommend it highly to anyone interested in this subject, though at a cost of $165, your interest would probably need to be more than casual. Imhof, in an updated section, expresses strong skepticism about algorithmic shading. Weaknesses include: * A ridge that is in-line with the illumination will have both sides at the same shade of gray. Skilled hand-shading would bias the illumination to prevent this. * A less steep section on a shaded slope will appear too light (I can attest to this). I would add that algorithmic shading does a poor job of showing small landforms, for a variety of reasons (insufficient data density, inability to intelligently exagerate small forms). The book was written before process color had been widely adopted, and before the fidelity of halftone printing could be really relied upon. Imhof describes the need to use two blue ink colors plus black, for example, to achieve a full range of blue color, from the lightest tint to very deep blue. For shaded relief, he suggests using not one or two, but five ink colors! They are: gray, grey-violet, light blue, yellow and light red. Each one is printed from a slightly differently exposed version of the shading, resulting in the shading appearing tastefully "flavored" instead of a constant hue. The yellow is actually reversed - the light areas on the shading are printed with the strongest yellow. The subjects I want to discuss: * In this age of process color (4-color and occasionally hexachrome), is there still call for this kind of spot color handwork for shaded relief? * The color "flavoring" that Imhof accomplishes by using multiple spot colors on differently exposed negatives can be done in, for example, Photoshop. Does anybody do this? Care to share some tips? * What is the state of the art with regard to "intelligent" algorithmic shading? Is there anything out there which addresses these alleged weaknesses? -- Pat Dunlavey ////////////////////////////////////////////// // // // PATRICK DUNLAVEY // // Cartography and Photogrammetry // // // // 40 Oblong Rd // // Williamstown, MA 01267 USA // // phone 413-458-9273 fax 413-458-9812 // // // // // // // //////////////////////////////////////////////