Creating in Maya after using Blender for so many years is like performing an invasive medical procedure using only plastic silverware.

I reach for the Blender hotkeys and end up clumsily changing edit modes without realizing that I have also unintentionally deleted the vertexes I wanted to extrude.

But I realize that Maya is a very powerful tool, that learning it is helpful, and that these problems are the same as those I encountered when I began working with Blender. My Maya class assigned me to create a character, so I decided to make it in both Maya and Blender.

The initial sketch is not mine; it was drawn by my friend Etai Dvora. As can be seen from a highly recommend trip to his gallery, Etai works solely in 2D. He was kind enough to sketch a front and side views of an invented character for me to see the quality of the 2D to 3D translation.

The Original Sketch by Etai Dvora

The process is the same in both Blender and Maya. I begin with a cube, apply a mirror effect, and extrude the vertexes over the sketch.

Modeling in Maya

The front and side views are usually sufficient for creating the general outline and keeping consistent proportions. Since this is all the assignment required, I did not continue to add details with are only useful in 3D.

Modeling in Blender

Final Render

I hope this is the first of many 2D to 3D projects to become bilingual in the language of modeling.

An interview with Etai is scheduled for a future episode of The Blender Show.

Ray tracing is a computer graphics technique which helps create the illusion of a photorealistic image. It refers to the way the light simulated in a digital environment (or “ray”) interacts with digital objects. Under a normal, non-ray tracing technique, the individual particles of light simply end at the first surface they interact with and illuminate it. Under a ray tracing technique, the computer uses a model to simulate the illumination caused when light bounces off a surface.

In real life, not all the light particles react with the surfaces in the same way: some bounce off at each angle, some are absorbed by the surface, some shine right through. The model determines how each light particle will react based on the location of the light source and the properties of the object in question. The characteristic of the surface will determine whether the object will cast a shadow, refract the light all the way to the other side, or reflect the light back to the camera. It is useful to direct the computer’s resources to focus on the digital light since shadows, refractions, and reflections are all properties of light.

The image below highlights a normal, non ray traced technique. Notice the lack of shadows and the ghostly feeling that the colored balls are not three-dimensional objects. This is because the light is simply a flat projection and does not calculate how the light should react with the other objects.

In the ray traced image below, it possible to see the reflections, the shadows, and the refractions of light. The relationship of the individual objects to one another and each object’s position in digital space are more clear. The final ray traced image has greater definition and better subtle illumination than the non-ray traced image.

Ray tracing is computationally intensive. Because the system needs to calculate the path of millions of light particles, it takes time and computing resources to create the final image. As a result, the current application of ray tracing is limited to visualizations which can be created in advance, such as computer-animated films. Video games, for example, have a basic structure similar to a computer-animated film but cannot take advantage of ray tracing because the environment is being continuously updated by the user and cannot be pre-visualized. As computing power increases, however, it will be possible to utilize the technique in real-time.

Most people learn Blender by teaching themselves.

This was the case with me. While I was encouraged by online tutorials and inspired by other people’s work, the majority of my knowledge came from getting my hands dirty with the program. I pressed random keys and clicked vaguely-labeled buttons until I became competent enough to model and animate comfortably using Blender.

But this semester I decided to take a class in Maya. Even though I will still enthusiastically use Blender (and continue to update The Blender Show), I would like to learn other interfaces and methods. Because of my lack of formal training, there are plenty of 3D topics that I don’t know about or understand. I tend to stick to whichever method I am shown first, so I am sure I have inefficiencies when creating in the Blender interface.

The class has just begun. Maya seams clumsy and awkward so far; it feels like I am manipulating the objects from inside a space suit. Being able to identify operations based on function and translate that into various programs would inevitably make me a better animator.

I was recently asked to explain the differences between my amateur open source operation and a major animation studio like Pixar. The short answer is there are too many fundamental differences to even warrant comparison, but after thinking it over, I came up with the long answer.

Luxo in Blender

At the basic level, the open source programs I use are similar to the propriety software that Pixar uses. They both summon a relatively basic engine to matrix individual vertexes in 3D space and define their relationship to one another. From there, the rendering software takes all the vertex information, along with surface values and light sources, and creates an image.

Pixar and and other effects companies use orders of magnitude upon orders of magnitude more computing power than I do. It routinely takes my laptop an hour to render a single 30-frames-per-second frame. With more computer power, they can afford to do much more advanced simulations. For example, I use a raytrace approximation technique to illuminate my scenes, while Pixar’s computers will actually trace each of the millions of light particles as they bounce around the set. But things are improving. Former ILM employee Alex Lindsay podcasted in MacBreak Weekly about the machines used to render the finished shots of the first Star Wars prequel, noting that the iPhone in his pocket was more powerful.

Pixar is also highly specialized. The person who animates a character’s body movements is different from the person who animates the facial expressions; the person who sets up the lighting system is different from the person who creates the sets. My work is all my own, from the modeling and rigging to the textures and simulations. This holistic approach is one reason I became interested in animation.

Perhaps the most important difference, however, is the legacy that Pixar animators seem to understand. There was a great deal of mistrust between traditional animators and computer animators in the 80’s and 90’s: many hand-drawn animators thought that the computer animators would take their jobs and reduce their art form into a bunch of clean, shiny reflective surfaces while many computer animators dismissed 2D as useless and antiquated. The mistrust became so bad that Disney set up boot-camps for the animators, essentially forcing 2D animators to learn 3D and vice versa while elevating the animation itself above the means.

The legacy that I refer to comes from this lighter focus on the method and a heaver focus on the animation. It is so easy to get lost in the wonderful aircraft cockpit-style array of buttons in an animation software and forget about the reasons behind a character’s movement. In this way, computer animation is no easier than traditional methods. Sure, if you want to create a simple bouncing ball, it is no more complex than turning on a physics simulator. But when it comes to making movements that tell the audience something about the character, making different characters behave differently in the same situations, and (most difficult of all) knowing when stillness is a appropriate for a scene, the toolbox is irrelevant. A good animator is a good animator in any dimension.

From what I can see from afar, the industry has embraced these ideas and have become more open to different styles and methods, while not allowing the method to overshadow the story. As an aspiring animator, I can only except that the next few decades to yield fantastic changes in the possess of creating animation.

This is Blender on my new Asus EeePC 1005ha. It works well and the result is a kind of sketchpad for the Blender artist, better for roughing out ideas than for rendering.

I love the idea of taking Blender to a remote place and doing some inspired modeling and animation.