Wednesday, March 4, 2009
A year and a half ago, I painted this.
Normally, I don't revisit or do remixes of anything this big or complicated, but there was always something unfinished and rushed about this piece in my mind, and I actually jumped at the chance to rework it.
Since I always save my file history, this is easy (at least at first). Oh, and if you're an artist working digitally, and you aren't doing this, you should. Saving often is great, but saving a new version every time a new change is made is even greater! For example: my current proportions don't favor a standard US Mail Postcard (which should be no bigger than 4.25" x 6.00"), but with layers, I can scale and arrange everything I want included until they do (with an eighth-inch bleed all the way 'round).
First, the difference between a restoration and a reconstruction (in the sense of William Elgin Swinton):
A reconstruction is an impression, model, or re-enactment of a past event formed from the available evidence (Oxford American Dictionary).
For restoration, the definition specifically mentions extinct animals ("a model or drawing representing the supposed form of the extinct animal, ruined building, etc."; ibid.).
I think it's a mistake to assume that because specializations are not present in a given skeleton, general actions like digging or (say) swimming are therefore ruled out. But it's actually fallacious to say that they are: absence of evidence is not evidence of absence.
Traits that would positively identify many many animals as diggers or swimmers are often absent (since behavior seldom fossilizes)--yet these are both things many animals can do (and some do rather well--without the need for evident specialization).
For the sake of a good artistic restoration, more than mere knowledge of anatomy is required here; the animal in a successful life reconstruction will be doing something or otherwise have some discernable motivation--this lends verisimilitude and detail and life to the restoration.
A restoration does not have to be a scientific hypothesis; it can posit behavior that is perfectly feasible, but might be specifically untestable. For example, maybe Psittacosaurus didn't dig at all, but how can anyone tell if it didn't? In the case of Pachyrhinosaurus, charging down a hill might not be something it did--but if it was not overtly incapable, then why not add a little dynamism to the image?
This is not, incidentally, a constraint of scientific conservativism; it's a constraint of having a limited understanding of animals in general--or not understanding the point of a life restoration, possibly. It might work for reconstructions that are technically valid but otherwise lifeless, but a good work of art will often go beyond what is overtly displayed in the skeleton of a given critter--and often must, if it is to be memorable at all.
Most of my sources come from skeletal material as well as close examination of other artists' restorations (as well as a liberal amount of reading from the technical literature).
But I also originally wanted to pay homage to the reason why I hold Pachyrhinosaurus in such high regard in the first place. This I did okay with originally, but here's where it all started for me: in 1996, I lived in Washington State, and was experiencing a resurge in interest in paleontology after a lengthy hiatus. When I was younger I had a series of Golden Guide books that I carried everywhere with me. (One was Rocks and Minerals).
And when I came across a 1990 edition of Dinosaurs in the same vein of design and style, I snagged it immediately.
I've lived in numerous places, thanks to my Dad being in the Navy. But Alaska and now Washington didn't yield much when it came to dinosaurs. So this page, in particular, made me do Pachyrhinosaurus:
The new remix finally satisfies me in that respect. In spite of this post's title, I don't think necessarily I've perfected the piece; far from it. But deadlines are deadlines, and eventually the artist absolutely must let it go.
Until, at least, the next time, should another second chance presents itself.
But now to it.
Step one was to erase the body and adjust the pose slightly. This was easy since I try to preserve everything in separate groups of layers in Photoshop.
Step two involved reworking the background extensively. Very little was retained, since I knew this would be printed out large-scale, I wanted the most texture and detail I could get away with.
Monkey-puzzle trees were a must this time around, and I redid every single horsetail, layering thousands of them in as efficiently as possible.
I used personal photo-reference for the trees, from photographs taken when I lived in Washington state; with the main painting on the Cintiq display and a preview file open on Harryhausen (my 5-year-old G4 laptop's display), I was able to paint freehand from the reference without having a printout of it. (Since I don't have a printer, this saves time.)
Quick tip for artists not necessarily new to tablets regarding the Cintiq pen-based display: look at the cursor, much as you might with a regular tablet, rather than where the point of your pen is, and painting and other operations will go much more smoothly. Some artists get frustrated by the change, and I think this must be the reason why. Once you adapt to that, you'll never want to go back.
The trees were also built in several layers: trunks, branches, and leaf-clusters (background clusters and highlighting clusters). For highlights, I made extensive use of the "lock transparency" toggle for the background leaf cluster layer. That way, my highlights would never appear anywhere other than where leaves were painted already.
The new trees now frame the left-hand side of the composition more, giving something for the silhouette of the animal to interact with.
Ground vegetation in the fore- and background were built up in a series of layers, with a shadow for the animal multiplied on top of the background layers (also on its own layer).
For the vegetation, transparency in brushes was used a lot as well. Ninety-percent works for most things that were green. Then I just layered in as many brush strokes as it took to convey greenery.
The new sky is a gradient (as a real sky is).
I don't use filters much, but the green wall in the distance needed some texture, so I faded in some faint layers of monochromatic uniform noise and rebrushed much of it.
The body and head of the animal took on a new color scheme. This time, I favored a model presented by an existing animal: the Okapi. I also redesigned the iris somewhat, making it look more natural. On the rough nose I acheived a more natural texturing in highlights to better emphasise the "pachy" in Pachyrhinosaurus.
The original purpose of the remix was to enable a traveling exhibit that would show in museums in Alaska. My hope is it will be met with much success. But I also finished it in time for a virtual exhibit, located here.