OpenGL, diffuse shader

Question

I'm trying to implement very simple diffuse shader in GLSL/openGL. Here's what I got: Vertex shader:

#version 130

in vec3 vertPos3D;
in vec3 vertNormal3D;

uniform mat3 transpMatrix;
uniform mat4 projectionMatrix;
uniform mat4 viewMatrix;

uniform vec3 lightPosition;

varying vec3 vertNormal;
varying vec3 lightVector;

void main() 
{ 
    vec4 res_pos = projectionMatrix * viewMatrix * vec4(vertPos3D.xyz, 1.0);
    gl_Position = res_pos;
    mat4 pm = projectionMatrix * viewMatrix;
    vertNormal = (viewMatrix * vec4(vertNormal3D, 0)).xyz;
    lightVector = (viewMatrix * vec4(lightPosition, 1.0)).xyz - (viewMatrix * vec4(vertPos3D.xyz, 1.0)).xyz;
}

Fragment Shader:

#version 130
out vec4 color; 

varying vec3 lightVector;
varying vec3 vertNormal;
void main()
{
    float dot_product = max(normalize(dot(lightVector, vertNormal)), 0.0);
    color = dot_product * vec4( 1.0, 1.0, 1.0, 1.0 );
}

As soon as I multiply final color with dot_product, nothing displays. I remove dot_product, everything works (except diffuse lightning ofc.). I'm afraid it's something obvious I'm missing.

Dylan Holmes Dylan Holmes · Accepted Answer · 2013-09-19T19:15:13

A problem:

normalize(dot(lightVector, vertNormal))

dot in GLSL 1.3 returns a float.
normalize accepts a vector, not a float.
documentation for dot
documentation for normalize

A Solution, at least to this problem:

In Fragment shader, replace

float dot_product = max(normalize(dot(lightVector, vertNormal)), 0.0);

with

float dot_product = clamp(dot(lightVector, vertNormal), 0, 1);

It looks like you are using max and normalize to avoid negative numbers returned from dot. This is exactly what clamp is for. Here's the documentation for clamp

OpenGL, diffuse shader

2 Answers