Over at his blog, David Orr writes about a comparison of two reconstructions of Diabloceratops, asking how his perception of their difference could diverge so much from that of Jim Kirkland, who felt that Matt’s take was too close to Brad’s, the version which appeared with the release of the paper. I’m fairly certain this is a non-event, as both versions have their own merits and are obviously not infringing on the other. What I find worthwhile is how we can discuss the issues, and what vocabulary can be used to distinguish visual aspects of such illustrations – all with the goal of designing efficient artist-scientist collaborations. One person’s blue is another persons ultra-marine, so to speak. So, here goes. some terminology that might be of use…
Level-of-detail, or LOD, describes the amount of details per area. This can apply independently to both to volume and to texture. Pixar generally employs mid-level LOD in the volumes of its characters, combined with high-level surfacing details… as illustrated by Roz, below. In contrast, Pocoyo (right) has a low-level of detail both in volume and surfacing.
Squinting your eyes is often enough to assess a design’s surface LOD. Try it out… look at Brad’s and Matt’s Diabloceratops and squint your eyes. Brad’s has lost significantly more detail than Matt’s. All those finely articulated scales and wrinkles disappear. Matt’s hasn’t got that much to begin with, so it only looses a bit.
Another difference are tonal ranges and contrast. I took each image and isolated them from the white background, then called up the histogram in Photoshop. Brad’s is balanced whereas Matt’s is contrast-rich, with more pixels falling into the dark and lighter ranges than in the middle. You don’t need the histograms to see this, but its a solid way to describe the effect.
None of this is sufficient without recognizing that the stroke qualities and overall mood play a major role in influencing the viewer. Both artists have created great-looking images that inspire the imagination as to what those fossils represent.
But, what’s it mean?
All of these visual descriptors will never be as clearly defined as as anatomical descriptor such as dorsal or anterior. Nor will they alleviate different backgrounds and expectations. One person’s industrial-influenced thrash metal will be another person’s indistinguishable heavy metal. Most scientists I know also enjoy cool imagery that shows their focus of interest in a manner that the general public can get psyched about. But not all… some mistrust the attempt and bemoan a depreciation of the original bones. what I find interesting is that there are divergent volumes being presented in these two images.
Brad’s has a very different nose shape, for example. There are soft tissue divergences about the mouth that are interesting.