所以我试图基于this教程实现一个guassian模糊着色器(用于练习)。我正在将他们的LWJGL代码翻译成JSFML代码,这有点棘手。
通过我的实现,只有背景精灵可见,鼠标周围不会出现模糊。
我的代码如下,为了便于阅读,提出了不相关的内容(如异常处理等)。
public class ShaderPractice {
private static final int RTSIZE = 1024;
private static float radius = 3;
private static float MAX_BLUR = 3;
public static void main(String[] args) {
RenderWindow rw = new RenderWindow();
rw.create(new VideoMode(800, 600), "Gaussian Blur!");
Texture t = new Texture();
t.loadFromFile(Paths.get("textures/bg.jpg"));
Sprite bgSprite = new Sprite(t);
RenderTexture blurTargetA = new RenderTexture();
blurTargetA.create(RTSIZE, RTSIZE);
RenderTexture blurTargetB = new RenderTexture();
blurTargetB.create(RTSIZE, RTSIZE);
Shader blurShader = new Shader();
blurShader.loadFromFile(Paths.get("src/shaderpractice/frag.glsl"), Shader.Type.FRAGMENT);
blurShader.setParameter("u_texture", t);
blurShader.setParameter("dir", new Vector2f(0, 0));
blurShader.setParameter("resolution", RTSIZE);
blurShader.setParameter("radius", radius);
while (rw.isOpen()) {
Vector2f mousePos = rw.mapPixelToCoords(Mouse.getPosition());
blurTargetA.clear();
blurTargetA.draw(bgSprite);
blurTargetA.display();
blurShader.setParameter("dir", new Vector2f(1, 0));
float mouseXAmt = mousePos.x / rw.getSize().x;
blurShader.setParameter("radius", mouseXAmt * MAX_BLUR);
blurTargetB.clear();
blurTargetB.draw(bgSprite, new RenderStates(blurShader));
blurTargetB.display();
blurShader.setParameter("dir", new Vector2f(0, 1));
float mouseYAmt = (rw.getSize().y-mousePos.y-1) / rw.getSize().y;
blurShader.setParameter("radius", mouseYAmt * MAX_BLUR);
rw.draw(bgSprite, new RenderStates(blurShader));
rw.display();
for (Event event : rw.pollEvents()) {
if (event.type == Event.Type.CLOSED) {
rw.close();
}
}
}
}
}
我的片段着色器如下(尽管我确定它是正确的,GLSL与教程中的相同)
uniform sampler2D u_texture;
uniform vec2 dir;
uniform float resolution;
uniform float radius;
void main() {
// RGBA sum
vec4 sum = vec4(0.0);
vec2 tc = gl_TexCoord[0].xy;
float blur = radius/resolution;
float hstep = dir.x;
float vstep = dir.y;
//apply blurring, using a 9-tap filter with predefined gaussian weights
sum += texture2D(u_texture, vec2(tc.x - 4.0*blur*hstep, tc.y - 4.0*blur*vstep)) * 0.0162162162;
sum += texture2D(u_texture, vec2(tc.x - 3.0*blur*hstep, tc.y - 3.0*blur*vstep)) * 0.0540540541;
sum += texture2D(u_texture, vec2(tc.x - 2.0*blur*hstep, tc.y - 2.0*blur*vstep)) * 0.1216216216;
sum += texture2D(u_texture, vec2(tc.x - 1.0*blur*hstep, tc.y - 1.0*blur*vstep)) * 0.1945945946;
sum += texture2D(u_texture, vec2(tc.x, tc.y)) * 0.2270270270;
sum += texture2D(u_texture, vec2(tc.x + 1.0*blur*hstep, tc.y + 1.0*blur*vstep)) * 0.1945945946;
sum += texture2D(u_texture, vec2(tc.x + 2.0*blur*hstep, tc.y + 2.0*blur*vstep)) * 0.1216216216;
sum += texture2D(u_texture, vec2(tc.x + 3.0*blur*hstep, tc.y + 3.0*blur*vstep)) * 0.0540540541;
sum += texture2D(u_texture, vec2(tc.x + 4.0*blur*hstep, tc.y + 4.0*blur*vstep)) * 0.0162162162;
gl_FragColor = vec4(sum.rgb, 1.0);
}
在有人询问之前,我只需要片段着色器,因为SFML在片段着色器中提供了正确的纹理坐标。
我在实施中做错了什么?我怀疑这是我处理我的FBO的方式(在这种情况下是RenderTextures)。
答案 0 :(得分:0)
我注意到你的程序中有两个令人困惑的地方
第一个是您在
中设置dir
(0.0, 0.0)
blurShader.setParameter("dir", new Vector2f(0, 0));
但是在你的着色器程序中,你将它用作倍增因子:
float hstep = dir.x;
float vstep = dir.y;
...
sum += texture2D(u_texture, vec2(tc.x - 4.0*blur*hstep, tc.y - 4.0*blur*vstep)) * 0.0162162162;
...
4.0*blur*hstep
的结果显然是0。所以你只是在采样同一点。
第二件事是你使用高斯的方式。不是这样的:
for(int i=0;i<Num;i++)
{
for(int j=0;j<Num;j++)
{
sum += weight[i][j]*texture2D(texture,vec2(texture_coor[i][j].u,texture_coor[i][j].v));
}
}
您只是在对角线上采样点