Emulate Hardware Shadow Mapping (2)

Continue with my previous post. The key to generate correct result is to set the texture parameter GL_TEXTURE_MAG_FILTER to GL_NEAREST when trying to emulate hardware PCF by one’s own. The code used in my shader is showing here:

vec4 newTexPos = texPos / texPos.w;

vec2 texmapscale = vec2(1/512.0, 1/512.0);

vec2 shadowMapCoord = vec2(512, 512) * newTexPos.xy;
vec2 lerps = fract(shadowMapCoord);

float result[4];

float bias = -0.0;

result[0] = texture(depthTex, newTexPos.xy).r + bias > newTexPos.z + bias? 1.0 : 0.0;
result[1] = texture(depthTex, newTexPos.xy + vec2(1.0/512, 0)).r + bias > newTexPos.z ? 1.0 : 0.0;
result[2] = texture(depthTex, newTexPos.xy + vec2(0, 1.0/512)).r + bias > newTexPos.z ? 1.0 : 0.0;
result[3] = texture(depthTex, newTexPos.xy + vec2(1.0/512, 1.0/512)).r + bias > newTexPos.z ? 1.0 : 0.0;

float shadowCoeffs = mix(mix(result[0], result[1], lerps.x),
mix(result[2], result[3], lerps.x),

The above code does a bi-linear interpolation between the four samples, therefore, there is no need to turn on GL_LINEAR for GL_TEXTURE_MAG_FILTER. Also, the result here:

Figure 7

You may refer to the two threads, thread on OpenGL forum and thread on MSDN forum for further details.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s