This technical documentation relates to tips and workarounds for experienced users. The Hare can be used without resorting to this documentation and does not require a high degree of knowledge of the Blender interface in order to render still images or animated sequences. Simply, set up the camera and render as you would with any 3D scene. Lighting and camera are already setup within the .blend file.
The realtime model is lite enough to be used in games and apps. Whereas the high resolution model can be used for cinematic close-up to wide shots.
The scene is setup such that all animation is performed on the realtime model, this information is then propagated to the high resolution model. No further setup is required by the user.
Visualization of high resolution data within the 3D viewport is made possible through the Subdivision Surface and Shrinkwrap Modifiers. This technique does not compromise on system performance during animation.Systems matching Blender's minimal specs, will be able to utilize the realtime model but could struggle with the high resolution version. All other recommended spec (and higher) systems for Blender are fully compatible with the product.
The file consists of a particular setup with both realtime and high resolution meshes within the same scene. Regardless of the final output be that, realtime applications to high quality cinema, interaction, editing, animation and rendering is generally performed on the realtime model.
It's subsequently best to leave the high resolution model on a hidden layer during animation and use the setup detailed below to control LOD (Level Of Detail). This will also improve system performance.
The realtime model is setup to subdivide (through the Subdivision surface Modifier) at render-time. Increasing subdivisions within the Modifier stack will allow the realtime model to more closely match the look of the sculpted model.
Vertices generated through subdivision will match surface points of the high resolution, sculpted model through the Shrinkwrap Modifier.
A Normal Map has also been baked from the Sculpted data that can be used to replace this system. However, although this may be the conventional approach the detail obtained from utilizing Subdiv and Shrinkwrap will be far greater.
Deformation is then performed on the model through the Armature modifier.
Secondary animation simulates the effects of physical forces acting on the character while it is moving. For example, the softer parts of the ears swaying as it runs or the tail moving around while the character's center of gravity shifts from left to right.
This type of animation can be added with Blender’s Soft Body modifier. The modifier should be evaluated after the Armature modifier in order for it to be effected by the character’s animation. However, the Subdivision Surface modifier cannot be added after a Soft Body Modifier at present in Blender. As a result a destructive setup of the modifier stack is required, meaning that the Subdivision Surface modifier must either be applied or deleted before attempting to create dynamic secondary animation. Subsequently, the ability to subdivide the character at render-time will be lost. For reintroducing the sculpted detail a Normal map can then be utilized.
A dynamic simulation can then be performed once the Subdivision Surface modifier is resolved. Secondary animation can be baked into the model with the Soft Body modifier which should remain present after the Armature modifier within the character's modifier stack.