The upcoming Dinosaur Revolution offers some beautiful opportunities to compare visual results across numerous borders: 3D vs. traditional, artist vs. team. Namely, the artist Angie Rodrigues has a blog of her work including bits for the show. Lucky for us, she’s also a talented artist proficient in both traditional and digital techniques. So we can geek out on the differences in her work as it appears throughout all these different venues.
As a side note, I haven’t seen the show, not even the above-linked press materials due to <rant on> dumb-ass regional restrictions which keep me from doing so <rant off> so please read this in the intended spirit of analyzing what I’ve been able to view from the context of my own experiences in 3D productions. Even if I see problems with this and that, I’m overall looking forward to seeing the documentary and acknowledge the effort that’s been stuck into this… I think it’s safe to say from the graphics already released that the Revolution team raised the bar. I’m also going to making assumptions based on the materials and my experience… instead of leveraging each and every time, please allow me to boldly presume and call me out on it in the comments. This begins with the title of this post.
Shall we begin?
Compare these two images, both containing models done by Angie. Both consist basically of the same process, integrating a 3D model into a photographic backplate. The Shunosaurus on the left is from the series, the Hadrosaur on the right she did on her own. Compare the shadow areas on the creatures and background plates, both similar ranges from the camera. That sauropod is much too bright… he’s glowing. The histogram of the selection confirms this… the profile isn’t working. If the animal were a few centimeters large, this might be attributable to subsurface scattering, and the high saturation would encourage such an interpretation. But this is a large animal with thick, non-translucent bones and dense musculature. Alone, Angie got it right. The shadow profiles match quite nicely.
Here we see closer integration of the shadows, with a tendency towards the opposite effect: except for a dominant rim light, our subject of interest tends toward deep shadow while the eyes pop out of the picture as they reflect light. Well, skin also reflects light, so it’s a question of balance… here it tips towards iris.
Where’s this come from, and why do so many documentaries suffer poor lighting when single artists often get it right. Well, a common issue in 3D productions is how the countless, complicated issues intertwine with each other, while the artistic hand-off process often isolates each artist, preventing them from viewing their contribution within the final context in which it will appear.
Are you thinking: “maybe that’s the look they were after”? Perhaps. But the look is determined by the live-action plates. They determine how the viewer will interpret the color of each pixel – what mixture of luminescence, direct and indirect lighting does that rgb value represent? In and of itself, the saturated shadows are wonderful, and this comparison gets even better because Angie has done wonderful paintings of similar subjects. The Edmontosaurus above has very diffuse, saturated shadows as well as a darker shadow cast on the ground. Incompatible? Here it works wonderfully because the overall palette doesn’t dictate an interpretation – its there as a bi-chromatic contrast to bring out those blues. An unskilled adaptation of this effect into photoreal would likely place this fellow in the midst of a forest fire.
So – if you’re upset that the cg documentaries make your favorite dinosaurs drab or day-glo… you know that often it’s not the individual artists to blame, but the conglomerate pipeline.
Leave comments if you’d like to hear my thoughts about the marketing disconnect or animation.