The character was posed in a variety of different poses to find the one that best represented the character. My test case, just like with the lightning mage, was to recreate my concept art in 3D. This meant reproducing the pose, lighting, etc., and taking it a step further than was done originally in the concept. A comparison can be seen in Fig.17.
Once confident, I experimented with more sophisticated poses. In the end I choose to do a contrapposto ("counterpose"), which is a pose used a lot in classical art where the figure puts his or her weight on one leg, called the standing leg. Since I had drawn these types of poses for a year now I was familiar with their characteristics. The cross was added to add some more implied story to the character, and also to give her something to balance on. The face is looking away on purpose; the idea was that she was looking at a church window in the distance. A floor and back wall for the church were modeled to give the entire scene better context (Fig.18).
Rendering was completely done in mental ray. My lighting setup was fairly traditional: I used a key, fill and two rim lights (all mental ray area lights). Final Gather was used to add additional bounce light into the scene. An HDR image of a church interior was used to cast the final gather, and be reflected in the metallic parts of the character (Fig.19).
The lighting was split up into multiple passes; diffuse, specular, reflection and bounce light were all separated to have greater control in post. Helper maps, such as material ID maps and a depth map, were also rendered out (Fig.20).
Post-work in Photoshop
In the end I composited the main layers (diffuse, specular, reflection and bounce light) in a very straightforward fashion: I just used Linear Dodge to add them on top of each other in Photoshop. Only the bounce light intensity was lowered to give a better sense of direct light coming through the window out of frame.
The depth map turned out to be invaluable in controlling all the atmospheric effects, and really making all the 2D effects integrate nicely with the 3D objects. Fig.21 shows the image going through some of the steps to get to the final image.
After that, it was just a matter of cropping your image nicely and color correcting it. In Fig.22 you can see the resulting final image. One of my biggest lessons learned from the last project that I took into this one is to always render big enough. If you ever want to make a nice print from your work you don't want to have to go back and re-render everything at higher resolution and redo all your post-effects in Photoshop.
< previous page