为什么GPU外的顶点变换与GPU内部的顶点变换不一样？

Question

我们知道如果我们想在世界空间中绘制某些东西,我们需要通过视图和投影矩阵来乘以我们的数据.现在我需要我的任务是在着色器外部进行顶点变换并直接发送转换后的顶点.像这样:

初始化矩阵

glm::mat4 view, projection;
view = glm::lookAt(this->cameraPos, this->cameraPivot, this->Up);
projection = glm::perspective(glm::radians(60.0f), (float)screenWidth / (float)screenHeight, 0.1f, 100.0f);

码

vector<glm::vec3> data;
glm::mat4 projection, view;
...
data.push_back(glm::vec3(projection * view * glm::vec4(0.0, 0.0, 0.5, 0.0)));
data.push_back(glm::vec3(projection * view * glm::vec4(0.0, 0.5, 0.5, 0.0)));
data.push_back(glm::vec3(projection * view * glm::vec4(0.5, 0.5, 0.5, 0.0)));
data.push_back(glm::vec3(projection * view * glm::vec4(0.5, 0.0, 0.5, 0.0)));
...
shader.set();
GLuint VAO,VBO;
glGenVertexArrays(1, &VAO);
glGenBuffers(1, &wVBO);
glBindVertexArray(wVAO);
glBindBuffer(GL_ARRAY_BUFFER, wVBO);
glBufferData(GL_ARRAY_BUFFER, data.size() * sizeof(vector<glm::vec3>), &data.data(), GL_STATIC_DRAW);
glEnableVertexAttribArray(0);
glVertexAttribPointer(0, 3, GL_FLOAT, GL_FALSE, 0, 0);
glBindBuffer(GL_ARRAY_BUFFER, 0);
glBindVertexArray(0);
...
glBindVertexArray(wVAO);
glDrawArrays(GL_LINE_LOOP, 0, data.size());
glBindVertexArray(0);

着色器

#version 330 core
layout (location = 0) in vec3 position;

void main()
{
    gl_Position = vec4(position, 1.0f);
}

但我得到错误的视图转换 - 看起来我的数据没有考虑投影矩阵,只考虑视图矩阵.但如果我像往常一样改变它:

码

 ...
 data.push_back(glm::vec3(glm::vec4(0.0, 0.0, 0.5, 0.0)));
 ...
 shader.set();
 glUniformMatrix4fv(glGetUniformLocation(shader.Program, "matrices"), 1,
 GL_FALSE, glm::value_ptr(projection * view));

着色器

#version 330 core
layout (location = 0) in vec3 position;

uniform mat4 matrices;

void main()
{
    gl_Position = matrices * vec4(position, 1.0f);
}

它按预期工作.我的错是什么？在我的任务中,需要知道在着色器中发送之前在世界空间中定位所有顶点的位置.我怎样才能做到这一点？

Answer 1

顶点的均匀坐标"w"应为1.0f,否则不会对顶点应用变换.