Archive notice: This article was originally published on August 18, 2014. Links and embedded videos are preserved as part of the historical record.
The third "Hobbit" instalment will be the first project in which Manuka is used as the primary renderer. The Weta team already used the renderer for several scenes from "The Desolation of Smaug" and "Dawn of the Planet of the Apes". For example, Manuka was able to handle the immense amount of CG hair for the scene in which the ape horde, around a thousand strong, rides on horseback to the humans’ camp. According to Weta Rendering Research Supervisor Luca Fascione, the tool is said to be particularly good at implementing hair and fur rendering, even for large crowds in one go. The scenes with this enormous CG complexity were a good test example for Weta, in which they noticed that Manuka could render all elements in one pass, which leads to a major memory problem with other rendering programs.
Gazebo
In addition to the renderer, the research project includes a hardware-based pre-lighting tool called Gazebo, which is intended to work almost in real time and likewise be physically based. This allows images to be moved consistently and with predictable results from one renderer to another – to achieve this, Weta worked to ensure that both tools use the same reflection models.
MIS
Manuka is both a uni-directional and a bi-directional renderer and includes Multiple Importance Sampling (MIS). The spectral MIS feature in the form of the new "Hero Spectral Sampling" technique is particularly interesting for the production of CG skin. Weta also placed a further emphasis on volumetrics: Manuka is able to realise complete Volumetric Scattering Simulations.
Deep Data
Because of Weta’s deep-data pipeline, Manuka was already written with these requirements in mind. A travelling matte for models and their components – such as a hand – can be created based on deep data. This saves lengthy roto work when the element is obscured or similar. The tool created for this, "Deep-CompC", was developed for Pixar’s RenderMan and is now also included in Manuka.
Camera
In terms of camera models, the renderer offers Orthographic and Projection Mapping, and Manuka also has sophisticated depth-of-field and motion-blur functions. Manuka currently does not yet support cameras that simulate a rolling shutter, but Weta wants to implement the Shutter Modeling feature soon so that these artefacts can be rendered accurately.
The system does not yet support Open Shader Language (OSL), as it was designed for Weta’s current shader pipeline.
Weta already won a Technical Oscar with its previous rendering solution for pre-processes, the PantaRay Renderer – which was used for the first part of "Avatar", among other things.
It is currently uncertain whether Manuka will ever be available as a public rendering solution. After the furious visual results that could already be admired in "Dawn of the Planet of the Apes", one can be more than excited about the quality of the third "Hobbit" instalment.
Where does the renderer’s name come from, by the way? Quite simple: Weta’s headquarters in New Zealand are located on Manuka Street.
All further information about Weta Digital and Manuka can be found here.
