OpenGL ES 3.0如何实现抗锯齿？求安卓端现代AA技术代码示例

以下两种方案均为移动端场景优化，兼容安卓全品类主流设备，可根据性能需求选择。

方案1：原生MSAA（硬件加速实现，优先推荐） #

OpenGL ES 3.0原生支持多采样抗锯齿，无需修改渲染管线逻辑，中高端设备上性能开销可控，画质表现最优。

• EGL配置阶段开启多采样

// 安卓GLSurfaceView EGL配置示例
int[] eglAttributes = new int[] {
    EGL14.EGL_RED_SIZE, 8,
    EGL14.EGL_GREEN_SIZE, 8,
    EGL14.EGL_BLUE_SIZE, 8,
    EGL14.EGL_ALPHA_SIZE, 8,
    EGL14.EGL_DEPTH_SIZE, 24,
    EGL14.EGL_STENCIL_SIZE, 8,
    // 开启4xMSAA配置，设备不支持可降级为2x
    EGL14.EGL_SAMPLE_BUFFERS, 1,
    EGL14.EGL_SAMPLES, 4,
    EGL14.EGL_RENDERABLE_TYPE, EGL14.EGL_OPENGL_ES3_BIT,
    EGL14.EGL_NONE
};
// 后续正常选择匹配的EGL配置即可

• 渲染初始化阶段开启多采样开关

GLES30.glEnable(GLES30.GL_MULTISAMPLE);

• 离屏渲染场景的MSAA FBO创建示例

int[] msaaFbo = new int[1];
GLES30.glGenFramebuffers(1, msaaFbo, 0);
GLES30.glBindFramebuffer(GLES30.GL_FRAMEBUFFER, msaaFbo[0]);

// 创建多采样颜色附件
int[] colorRenderBuffer = new int[1];
GLES30.glGenRenderbuffers(1, colorRenderBuffer, 0);
GLES30.glBindRenderbuffer(GLES30.GL_RENDERBUFFER, colorRenderBuffer[0]);
GLES30.glRenderbufferStorageMultisample(GLES30.GL_RENDERBUFFER, 4, GLES30.GL_RGBA8, width, height);
GLES30.glFramebufferRenderbuffer(GLES30.GL_FRAMEBUFFER, GLES30.GL_COLOR_ATTACHMENT0, GLES30.GL_RENDERBUFFER, colorRenderBuffer[0]);

// 多采样深度模板附件
int[] depthStencilRenderBuffer = new int[1];
GLES30.glGenRenderbuffers(1, depthStencilRenderBuffer, 0);
GLES30.glBindRenderbuffer(GLES30.GL_RENDERBUFFER, depthStencilRenderBuffer[0]);
GLES30.glRenderbufferStorageMultisample(GLES30.GL_RENDERBUFFER, 4, GLES30.GL_DEPTH24_STENCIL8, width, height);
GLES30.glFramebufferRenderbuffer(GLES30.GL_FRAMEBUFFER, GLES30.GL_DEPTH_STENCIL_ATTACHMENT, GLES30.GL_RENDERBUFFER, depthStencilRenderBuffer[0]);

// 检查FBO完整性，失败则降级采样倍数
if (GLES30.glCheckFramebufferStatus(GLES30.GL_FRAMEBUFFER) != GLES30.GL_FRAMEBUFFER_COMPLETE) {
    // 降级逻辑
}

• 多采样结果输出逻辑
  渲染完成后需要将多采样FBO的内容blit到目标帧缓冲：

GLES30.glBindFramebuffer(GLES30.GL_READ_FRAMEBUFFER, msaaFbo[0]);
GLES30.glBindFramebuffer(GLES30.GL_DRAW_FRAMEBUFFER, 0); // 0为默认窗口帧缓冲，可替换为普通FBO
GLES30.glBlitFramebuffer(0, 0, width, height, 0, 0, width, height, GLES30.GL_COLOR_BUFFER_BIT, GLES30.GL_NEAREST);

方案2：FXAA后处理抗锯齿（低配置设备兼容方案） #

无需修改EGL配置，仅需增加一个全屏后处理Pass，性能开销极低，适合低端设备或者性能敏感场景。

• FXAA片元着色器代码

#version 300 es
precision mediump float;
out vec4 FragColor;
in vec2 TexCoords;

uniform sampler2D u_SceneTexture;
uniform vec2 u_ScreenSize;

// 移动端优化参数，无需调整
#define FXAA_REDUCE_MIN   (1.0/128.0)
#define FXAA_REDUCE_MUL   (1.0/8.0)
#define FXAA_SPAN_MAX     8.0

void main() {
    vec2 rcpFrame = 1.0 / u_ScreenSize;
    vec3 rgbNW = texture(u_SceneTexture, TexCoords + vec2(-1.0, -1.0) * rcpFrame).xyz;
    vec3 rgbNE = texture(u_SceneTexture, TexCoords + vec2(1.0, -1.0) * rcpFrame).xyz;
    vec3 rgbSW = texture(u_SceneTexture, TexCoords + vec2(-1.0, 1.0) * rcpFrame).xyz;
    vec3 rgbSE = texture(u_SceneTexture, TexCoords + vec2(1.0, 1.0) * rcpFrame).xyz;
    vec3 rgbM  = texture(u_SceneTexture, TexCoords).xyz;

    vec3 luma = vec3(0.299, 0.587, 0.114);
    float lumaNW = dot(rgbNW, luma);
    float lumaNE = dot(rgbNE, luma);
    float lumaSW = dot(rgbSW, luma);
    float lumaSE = dot(rgbSE, luma);
    float lumaM  = dot(rgbM,  luma);
    float lumaMin = min(lumaM, min(min(lumaNW, lumaNE), min(lumaSW, lumaSE)));
    float lumaMax = max(lumaM, max(max(lumaNW, lumaNE), max(lumaSW, lumaSE)));
    
    vec2 dir;
    dir.x = -((lumaNW + lumaNE) - (lumaSW + lumaSE));
    dir.y =  ((lumaNW + lumaSW) - (lumaNE + lumaSE));
    
    float dirReduce = max((lumaNW + lumaNE + lumaSW + lumaSE) * (0.25 * FXAA_REDUCE_MUL), FXAA_REDUCE_MIN);
    float rcpDirMin = 1.0 / (min(abs(dir.x), abs(dir.y)) + dirReduce);
    dir = min(vec2(FXAA_SPAN_MAX, FXAA_SPAN_MAX), max(vec2(-FXAA_SPAN_MAX, -FXAA_SPAN_MAX), dir * rcpDirMin)) * rcpFrame;

    vec3 rgbA = 0.5 * (
        texture(u_SceneTexture, TexCoords + dir * (1.0/3.0 - 0.5)).xyz +
        texture(u_SceneTexture, TexCoords + dir * (2.0/3.0 - 0.5)).xyz);
    vec3 rgbB = rgbA * 0.5 + 0.25 * (
        texture(u_SceneTexture, TexCoords + dir * -0.5).xyz +
        texture(u_SceneTexture, TexCoords + dir * 0.5).xyz);

    float lumaB = dot(rgbB, luma);
    FragColor = (lumaB < lumaMin || lumaB > lumaMax) ? vec4(rgbA, 1.0) : vec4(rgbB, 1.0);
}

• 调用逻辑

1. 创建普通离屏FBO，将场景渲染到FBO绑定的颜色纹理上
2. 绑定FXAA着色器，传入场景纹理、屏幕分辨率参数
3. 绘制全屏四边形即可输出抗锯齿结果

选型建议 #

• 中高端设备优先选择4x MSAA，画质最优，性能损耗约10%~15%
• 低端设备或性能敏感场景选择FXAA，性能损耗约3%~5%，画质略逊于MSAA
• 禁止叠加使用多种AA方案，会造成无意义的性能浪费

内容的提问来源于stack exchange，提问作者user14587078