Tobia at work, part 2

Tobia's pal

This is the final rendering of the Tobia project, a collaboration between Anna Maria Bonanese and myself. The previous post dealt with the construction of the barn and a few props. This one focuses of the design and construction of Tobia's companion.

mouse forms by Anna

The first step in the mouse design was to research photos of mice and mouse anatomy. From those, Anna produced these sketches. Silhouettes are used, in order to resolve the form. The modeling environment is not the place to resolve rudimentary design issues, so we sketch until we're happy with its shape.

spec drawings by Anna

Then Anna designed these modeling guides.

Tobia at work, part 1

Tobia en garde (click to enlarge)

Tobia recently found employment as a sheepdog. Here, he’s seen preparing for work.

The concept was to use the same model of the plastic toy, and change the textures so it appears as an old discarded toy in an abandoned building. The new composition was conceived of and designed by myself and Tobia-designer, Anna.

 design stages, by Anna and me

early rendering - new scene being built

... and then with some textures applied ...

intermediate rendering, ready for the next stage
(click to enlarge)

next: a friendly visit

Tobia at home

Tobia at home, final composite

(click to enlarge)

I've been wanting to do some small product-visualization projects for a while, and have finally found a designer to work with.

production process, from sketch to modeling; drawings by Anna

(click to enlarge)

Anna Maria Bonanese designed the product and environment for our first project, Tobia the beagle. I constructed the scenes, using Maya, ZBrush, and Mental Ray. Anna then tweaked the materials and arranged the props to her liking.

Tobia, studio shot

Tobia at home, design by Anna

Anna loves to design many things, but children's products are her favorite. Tobia is inspired by a dog that she sees each day, during her morning walk.

Wonder Warthog, part 10 - head textures

(I'm redesigning, modeling, and rigging Gilbert Shelton's classic cartoon character, "Wonder Wart-Hog.")

Hello, handsome!

(click to enlarge)

Now the textures have been added, using UV, paint, and sculpt tools from ZBrush. Here’s a Mental Ray test render with roughed-out material settings.

head maps, rendered separately

(click to enlarge)

Currently, the head uses four different map types. The diffuse map holds the color. If the nose is pinker or the eyelids darker, it’s because of that map. The displacement map adds geometric detail at render time: wrinkles, bumps, pores, scars. This map saves me from having to create gigantic, geometrically-dense models, which most video cards can’t handle. The normal map adds further bumpy definition to the surface.

There’s also a specular map that distributes varying amounts of shininess. The eye, nose, and mouth areas are a little shinier than the rest of the face.

 eye & mouth test renders

(click to enlarge)

The head is always created first, to allow more time for scrutinization, refinement, and revision. Other parts of the character don’t need to be so perfect because audiences don’t examine them as closely.

During the rigging stage, I'll be creating another face map layer for animated wrinkles.

production video

This video documents the entire texturing process. The eye texture was created in Photoshop. The rest were done in ZBrush. UVs were mapped using ZBrush’s fantastic UV Master plug-in, and tweaked in Maya.

next: an inordinate amount of hair

Twitter Art

140Artist

Are you a visual artist with tons of images in your head but only time to develop a few? Maybe try them out on Twitter in the form of Unicode art. Because of the 140 character limit, you can’t waste too much time on each design, and – through replies and retweeting – people will tell you which subjects and compositions are successful.

hg47

MargaRlda

Testing the Waters, ep. 6 - rigid bodies

(experiments in fluid dynamics)

silly rabbit,

click to enlarge

As I continue to pick up skills, these fluid videos will gradually become more sophisticated; or that's the plan. This is the first experiment that resembles a practical application, i.e., something you might see in a movie or ad.

A rigid body is a non-deforming piece of geometry in a dynamic simulation. This means it can be thrown around by forces and collide with other objects, but can't be smashed, squished, or shmushed. (By contrast, a SOFT body is a deforming piece of dynamic geometry, like something made of fabric, paper, or rubber.)

In this test, the rigid bodies are breakfast cereal Os, a bowl, and a spoon. The bowl and spoon are static rigid bodies. They aren't affected by forces, but contribute to cereal collisions.

cereal sim : filling the bowl

If you have hundreds of objects that you need to arrange into a pile, the quickest way is through dynamic simulation. My first simulation fills the bowl with cereal, and isn't included in the final animation. The second simulation pours the milk.

my epic

The simulations are so time-consuming, that I eventually had to end the adjustments and settle for less-than-satisfactory results.

• The milk should permeate the cereal more quickly.
• The cereal shouldn't be thrown around by the milk so easily.
• The cereal should slowly rise and expand as the milk displaces it.
• The drops should not bounce off the bowl. They should stick and slide down, leaving a film, if possible.

Right now, I don't have the computing power and time to perfect this, but someday I'll revisit the project to fix the dynamics, and add some fruit slices.

Testing the Waters, ep. 5 - mixing paint

(experiments in fluid dynamics)

 

making a mess

The sister project of the immiscible test (ep 4) is this miscible test. Although I'm 80% happy with the results, it's not an ideal solution because the particles in this simulation don't inherit the hue of particles they come in contact with. For example, orange-looking fluid will revert to yellow and red if its particles become isolated. This results in splotchy - rather than smooth - mixing.

Maybe someday a better solution will be available. I'll keep looking.

DOF, before & after

Because of the distance between bowls, this was a good scene for introducing depth of field (DOF). Like motion blur and specular highlights, this is another blur-based post effect that raises the level of realism. DOF is usually the only post channel that I animate, as the focal distance often needs to bounce from object to object.

mixing video

Yes, the floor is moving in this animation. I wanted to see full room lighting in this video without the extra modeling and rendering time, so I used my IBL image for a background. Consequently, it doesn't line up with the geometry. It's just a test, so NBD.

Wonder Warthog, part 9 - refining the head geometry

(I'm redesigning, modeling, and rigging Gilbert Shelton's classic cartoon character, "Wonder Wart-Hog.")

head geometry, test render

I love doing this, but it sure takes a long time. Ideally, you want a team creating a character this complex.

Above is a test render of the "finished" head geometry. It will certainly require further geometric adjustments down the line, but this is good enough to move it into texturing. If we continue the architecture analogy introduced in the previous chapter, all the walls, floors, and staircases have just been installed.

construction of new topology

At this refinement stage, the first development was overall topology. This means that the edge flow needs to conform to the contours of the underlying muscles and bones, which aids in realism and good deformation during animation.

face areas: blocked out, then refined

The most extensive head modeling is usually done in the mouth, which includes lips, teeth, gums, and tongue. Following that, the neck and the eye area needed the most work. (The ears often need much work, but I did most of that in the blocking stage.)

mouth insides

turntable

This turntable video demonstrates the progress between the rough "blockhead" model and the finished geometry.

There's much more work to be done on the head.

next: texturing

Testing the Waters, ep. 4 - oil & water

(experiments in fluid dynamics)

composite

It's time to try interaction between multiple particle systems. This test simulates immiscible fluids, like oil and water.

combined MV data

Because the oil's motion vector and specular data are obscured by the water, each fluid needed to be rendered separately. The data was then combined in a compositer.

video

Testing the Waters, ep. 3 - wetmaps

(experiments in fluid dynamics)

composited rendering

Wetmeps are created when a particle leaves some record of its contact with a surface. Wetmaps can stain a surface, or wash grime from a surface when inverted.

simulation (top) and
map enhancement (bottom)

The wetmaps are generated in the simulator, then enhanced in an image processor. Finally, they're used as a mask to mix dry and wet ground shaders within the rendering software.

Here's my latest blockbuster.