c++ - SDL_GL_SetAttribute 不设置颜色尺寸

Question

我正在尝试更改 *SDL_opengl* 中累积缓冲区颜色分量的大小，但 SetAttribute 命令似乎没有执行任何操作。这是我正在使用的代码。
(为了减少代码大小，我在这里只处理红色组件，但在实际代码中，我将所有 4 个组件都传递给颜色和累积缓冲区，效果是一样的)

#include <iostream>
#include <SDL/SDL.h>
#include <SDL/SDL_opengl.h>

int main(int argc, char *argv[])
{
    //Initialize all SDL subsystems
    if( SDL_Init( SDL_INIT_EVERYTHING ) < 0 )std::cout << "SDL ERROR!";

    // Try to Set the BitSize, while checking for errors
    int BitSize = 1; //This number never makes a difference!!
    int ErrorCode = 0;
    ErrorCode += SDL_GL_SetAttribute( SDL_GL_DOUBLEBUFFER, 1 ) // This one WORKS
        + SDL_GL_SetAttribute( SDL_GL_ACCUM_RED_SIZE, BitSize ) // These ones DON'T
        + SDL_GL_SetAttribute( SDL_GL_RED_SIZE, BitSize )
        ;

    if( ErrorCode < 0 )std::cout << "SDL ERROR!";

    // Create the Window
    int w = 1000, h = 700;  
    int bpp = 32;

    SDL_Surface* Screen = SDL_SetVideoMode( w, h, bpp, SDL_OPENGL | 
    SDL_NOFRAME |
    SDL_DOUBLEBUF );

    if( !Screen )std::cout << "SDL ERROR!";


    // Check if BitSize's are correct (they are not)
    // I'm using glGetInteger, but SDL_GL_GetAttribute yields the same output.
    glGetIntegerv( GL_ACCUM_RED_BITS, &BitSize );
    std::cout << "AccumBuffer color component size in bits is " << BitSize << "\n";

    glGetIntegerv( GL_RED_BITS, &BitSize );
    std::cout << "ColorBuffer color component size in bits is " << BitSize << "\n";

    ErrorCode = SDL_GL_GetAttribute( SDL_GL_BUFFER_SIZE, &BitSize );
    std::cout << "FrameBuffer BitSize is " << BitSize << "\n";

    if( ErrorCode < 0 )std::cout << "SDL ERROR!";
    if( glGetError() != GL_NO_ERROR )std::cout << "GL ERROR";

    return 0;
}

这个编译很好，并且总是打印以下输出:

AccumBuffer color component size in bits is 16
ColorBuffer color component size in bits is 8
FrameBuffer Bit Size is 32

无论我将 BitSize 变量设置为什么。这就像 SDL_GL_SetAttribute( SDL_GL_*_SIZE, int ) 没有任何效果。我可以理解 Color Buffer 组件可能被限制为 8 位，因为我用 32 bpp 初始化了窗口。但是我不应该能够编辑累积缓冲区颜色分辨率吗？

Answer 1

您设置的属性值只是对您期望的最小值的请求，但是给您更大的值是完全有效的。

顺便说一句:除非您拥有专业级 GPU(FireGL、Quadro)，否则累积缓冲区可能不是硬件加速的。请改用帧缓冲区对象。