OpenGLContext.shadow.volume

Shadow-volume implementation

A volume is cast by a light from an edgeset, it's basically the volume of space which is shadowed by the given edgeset/object.

Classes

class Volume( object ):

A shadow-volume object

This object represents the shadow cast by a single light and a single occluder of that light. It is rendered (along with all other volumes for a given light) into the stencil buffer to determine what parts of the scene are lit by the light.

XXX doesn't yet handle single-edges for faces or (more cricitally) clock-wise windings

__init__( self , edgeSet , sourceVector )

Initialize the shadow volume

edgeSet: pointer to the edge set from which we retrieve most of our geometric data
sourceVector: the homogenous coordinates of the shadow-casting light.

_debug_render_silouhette( self )

Debug render of silouhette as lines with current colour

_render_boot( self )

Render the shadow-volume boot (backward-facing faces projected)

_render_cap( self )

Render the shadow-volume cap (forward-facing faces)

_render_edges( self )

Render the shadow-volume edges

calculate( self )

Calculate the shadow-volume's shape

Returns segments*2*3 array where each item in the array has the property that it defines a face of the shadow volume like so: A,B = segment (2 3-item arrays) for point lights (Quads): Ax,Ay,Az,1.0 Bx,By,Bz,1.0 Bx,By,Bz,0.0 Ax,Ay,Az,0.0 for directional lights: Ax,Ay,Az,1.0 Bx,By,Bz,1.0 0,0,0,0 Which is fed into the "equation 14" or "equation 15" of the article. Note that this is pre-calculated to not require switching calculations on the face when doing the later transformation to a shadow volume. (i.e. in the article there are two different cases for whether the "first" or "second" face is facing the lights, I've folded them together).

need certain things when we're done: set of light-facing-faces set of rearward-facing-faces (projected to infinity) set of edge-faces (silouhette projected to infinity) # the first two are currently missing # should do those as indices into the points array