Graphics Reference
In-Depth Information
we can use the
v_shadowCoord
as the texture coordinate to sample
into the shadow map by first transforming the coordinate from
homogeneous coordinate [-1, 1] space into [0, 1] space in the vertex
shader. Alternatively, we can avoid performing these calculations in
the vertex shader by pre-multiplying the MVP matrix from the light
position with the following bias matrix in the host code:
0.5, 0.0, 0.0, 0.0,
0.0, 0.5, 0.0, 0.0,
0.0, 0.0, 0.5, 0.0,
0.5, 0.5, 0.5, 1.0
In the fragment shader, we check the current fragment to determine
whether it is in the shadow by sampling the shadow map using the
v_shadowCoord
and
textureProj
call. We perform 3 × 3 kernel filtering
to further increase the effect of PCF (effectively 6 × 6 PCF when combined
with four hardware depth comparisons per tap). Then we average the
shadow map sample result to modulate the fragment color. When the
fragment is in shadow, the sample result will be zero and the fragment
will be rendered in black.
This chapter explored how many of the OpenGL ES 3.0 features presented
throughout this topic can be applied to achieve various rendering
techniques. This chapter covered rendering techniques that made use
of features including cubemaps, normal maps, point sprites, transform
feedback, image postprocessing, projective texturing, framebuffer objects,
vertex texture fetch, shadow maps, and many shading techniques. Next,
we will return to the API to discuss the functions your application can use
to query OpenGL ES 3.0 for information.
