目录
Ray Traced Reflections
prepare
- createBottomLevelAccelerationStructure: 在本地创建模型底层加速结构信息
- loadFromFile：加载模型
- 创建两个VkDeviceOrHostAddressConstKHR,分别存放vbo和ibo的指针(deviceAddress)
- vertexBufferDeviceAddress : getBufferDeviceAddress
- indexBufferDeviceAddress: getBufferDeviceAddress
- 创建加速结构
- VkAccelerationStructureGeometryKHR
- 参数数模型的一些基本属性
- VkAccelerationStructureBuildGeometryInfoKHR: 几何信息
- VkAccelerationStructureBuildSizesInfoKHR:Get size info
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure: 创建加速结构buffer
- VkBufferCreateInfo:特殊的参数
- sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO
- usage = VK_BUFFER_USAGE_ACCELERATION_STRUCTURE_STORAGE_BIT_KHR
- vkCreateBuffer: 下面使用的参数结构体都使用加速结构相关参数
- vkGetBufferMemoryRequirements
- vkAllocateMemory
- vkBindBufferMemory
- VkAccelerationStructureCreateInfoKHR
- vkCreateAccelerationStructureKHR: 创建加速结构
- vkGetAccelerationStructureDeviceAddressKHR: AS device address
- VkBufferCreateInfo:特殊的参数
- VkAccelerationStructureGeometryKHR
- createScratchBuffer
- 在构建底层加速结构的过程中创建一个小的scratch缓冲区
- VkAccelerationStructureBuildGeometryInfoKHR:
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- createTopLevelAccelerationStructure
- 创建 top level 加速结构
- VkAccelerationStructureInstanceKHR
- createBuffer: for instance data
- VkDeviceOrHostAddressConstKHR
- getBufferDeviceAddress
- VkAccelerationStructureGeometryKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildSizesInfoKHR
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure()
- createScratchBuffer ()
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- instancesBuffer.destroy()
- createStorageImage(): 存放光追渲染结果:
- 释放image信息， 阴影教程里没有这部分
- vkDestroyImageView
- vkDestroyImage
- vkFreeMemory
- storageImage
- VkImageCreateInfo image = vks::initializers::imageCreateInfo();
- vkCreateImage
- vkGetImageMemoryRequirements
- vkAllocateMemory
- vkBindImageMemory
- VkImageViewCreateInfo
- vkCreateImageView
- createCommandBuffer
- setImageLayout
- flushCommandBuffer - createUniformBuffer
- createBuffer
- updateUniformBuffers
- createRayTracingPipeline
- VkDescriptorSetLayoutBinding:结构体
- VkDescriptorSetLayoutCreateInfo:结构体
- vkCreateDescriptorSetLayout
- VkPipelineLayoutCreateInfo
- vkCreatePipelineLayout
- Setup ray tracing shader groups
- VkPipelineShaderStageCreateInfo
- Ray generation group
- VkRayTracingShaderGroupCreateInfoKHR: 结构体,存放shader和其他配置信息
- shader: raygen.rgen.spv
- VkRayTracingShaderGroupCreateInfoKHR: 结构体,存放shader和其他配置信息
- Miss group: 光追未命中
- VkRayTracingShaderGroupCreateInfoKHR
- shader: miss.rmiss.spv
- Closest hit group: 光追命中
- VkRayTracingShaderGroupCreateInfoKHR
- shader: closesthit.rchit.spv
- Create the ray tracing pipeline
- VkRayTracingPipelineCreateInfoKHR:结构体
- vkCreateRayTracingPipelinesKHR
- createShaderBindingTable
- 创建用于加速结构的 shader 绑定表(Shader Binding Tables)-(SBT)
- vkGetRayTracingShaderGroupHandlesKHR
- createShaderBindingTable:创建三个SBT
- createBuffer: 创建SBT – raygen
- 使用字段VK_BUFFER_USAGE_SHADER_BINDING_TABLE_BIT_KHR
- createBuffer: miss
- createBuffer: hit
- memcpy
- memcpy
- memcpy
- createBuffer: 创建SBT – raygen
- createDescriptorSets
- vkCreateDescriptorPool
- vkAllocateDescriptorSets
- writeDescriptorSet: Ray tracing result image
- writeDescriptorSet: Uniform data
- writeDescriptorSet: Scene vertex buffer
- writeDescriptorSet: Scene index buffer
- vkUpdateDescriptorSets
- buildCommandBuffers
- handleResize
- createStorageImage
- storageImageDescriptor
- writeDescriptorSet
- vkBeginCommandBuffer
- 使用反射时，这里使用的pipe数增加，这里cmd的两次，应该是只反射一次，命令相同
- vkCmdBindPipeline
- vkCmdBindDescriptorSets
- vkCmdBindPipeline
- vkCmdBindDescriptorSets
- vkCmdTraceRaysKHR
- Copy ray tracing output to swap chain image
- setImageLayout
- setImageLayout
- vkCmdCopyImage
- setImageLayout
- setImageLayout
- vkEndCommandBuffer
- handleResize
render
- draw
- prepareFrame
- vkQueueSubmit
- submitFrame
- updateUniformBuffers
shader
- raygen.rgen：计算与顶层加速结构碰撞的点
#version 460
#extension GL_EXT_ray_tracing : require
layout(binding = 0, set = 0) uniform accelerationStructureEXT topLevelAS;
layout(binding = 1, set = 0, rgba8) uniform image2D image;
layout(binding = 2, set = 0) uniform CameraProperties
{
mat4 viewInverse;
mat4 projInverse;
vec4 lightPos;
} cam;
struct RayPayload {
vec3 color;
float distance;
vec3 normal;
float reflector;
};
//这里存储的结构是一个结构体
layout(location = 0) rayPayloadEXT RayPayload rayPayload;
//设置最大递归次数
// Max. number of recursion is passed via a specialization constant
layout (constant_id = 0) const int MAX_RECURSION = 0;
void main()
{
const vec2 pixelCenter = vec2(gl_LaunchIDEXT.xy) + vec2(0.5);
const vec2 inUV = pixelCenter/vec2(gl_LaunchSizeEXT.xy);
vec2 d = inUV * 2.0 - 1.0;
vec4 origin = cam.viewInverse * vec4(0,0,0,1);
vec4 target = cam.projInverse * vec4(d.x, d.y, 1, 1) ;
vec4 direction = cam.viewInverse*vec4(normalize(target.xyz / target.w), 0);
//判断是否透明的tag
uint rayFlags = gl_RayFlagsOpaqueEXT;
uint cullMask = 0xff;
float tmin = 0.001;
float tmax = 10000.0;
vec3 color = vec3(0.0);
//这里循环计算结果
for (int i = 0; i < MAX_RECURSION; i++) {
//计算反射值
traceRayEXT(topLevelAS, rayFlags, cullMask, 0, 0, 0, origin.xyz, tmin, direction.xyz, tmax, 0);
vec3 hitColor = rayPayload.color;
//miss中将distance设置为-1， 因为miss证明光线未与模型hit，不需要反射， 直接获取背景色，跳出循环
if (rayPayload.distance < 0.0f) {
color += hitColor;
break;
}
//rayPayload.reflector = 1证明光线与模型hit，并且白色需要反射，修改光线方向继续反射步骤
else if (rayPayload.reflector == 1.0f) {
const vec4 hitPos = origin + direction * rayPayload.distance;
origin.xyz = hitPos.xyz + rayPayload.normal * 0.001f;
direction.xyz = reflect(direction.xyz, rayPayload.normal);
}
//其他情况直接赋值颜色即可。
else {
color += hitColor;
break;
}
}
imageStore(image, ivec2(gl_LaunchIDEXT.xy), vec4(color, 0.0));
}
```
- miss.rmiss:
```cpp
#version 460
#extension GL_EXT_ray_tracing : require
struct RayPayload {
vec3 color;
float distance;
vec3 normal;
float reflector;
};
layout(location = 0) rayPayloadInEXT RayPayload rayPayload;
void main()
{
//没有hit赋值背景色即可， 并设置distance = -1.0f， reflector = 0.0f;不进行反射
// View-independent background gradient to simulate a basic sky background
const vec3 gradientStart = vec3(0.5, 0.6, 1.0);
const vec3 gradientEnd = vec3(1.0);
vec3 unitDir = normalize(gl_WorldRayDirectionEXT);
float t = 0.5 * (unitDir.y + 1.0);
rayPayload.color = (1.0-t) * gradientStart + t * gradientEnd;
rayPayload.distance = -1.0f;
rayPayload.normal = vec3(0.0f);
rayPayload.reflector = 0.0f;
}
```
- closesthit.rchit
```cpp
#version 460
#extension GL_EXT_ray_tracing : require
#extension GL_EXT_nonuniform_qualifier : enable
struct RayPayload {
vec3 color;
float distance;
vec3 normal;
float reflector;
};
layout(location = 0) rayPayloadInEXT RayPayload rayPayload;
hitAttributeEXT vec3 attribs;
layout(binding = 0, set = 0) uniform accelerationStructureEXT topLevelAS;
layout(binding = 2, set = 0) uniform UBO
{
mat4 viewInverse;
mat4 projInverse;
vec4 lightPos;
int vertexSize;
} ubo;
layout(binding = 3, set = 0) buffer Vertices { vec4 v[]; } vertices;
layout(binding = 4, set = 0) buffer Indices { uint i[]; } indices;
struct Vertex
{
vec3 pos;
vec3 normal;
vec2 uv;
vec4 color;
vec4 _pad0;
vec4 _pad1;
};
Vertex unpack(uint index)
{
// Unpack the vertices from the SSBO using the glTF vertex structure
// The multiplier is the size of the vertex divided by four float components (=16 bytes)
const int m = ubo.vertexSize / 16;
vec4 d0 = vertices.v[m * index + 0];
vec4 d1 = vertices.v[m * index + 1];
vec4 d2 = vertices.v[m * index + 2];
Vertex v;
v.pos = d0.xyz;
v.normal = vec3(d0.w, d1.x, d1.y);
v.color = vec4(d2.x, d2.y, d2.z, 1.0);
return v;
}
void main()
{
ivec3 index = ivec3(indices.i[3 * gl_PrimitiveID], indices.i[3 * gl_PrimitiveID + 1], indices.i[3 * gl_PrimitiveID + 2]);
Vertex v0 = unpack(index.x);
Vertex v1 = unpack(index.y);
Vertex v2 = unpack(index.z);
// Interpolate normal
const vec3 barycentricCoords = vec3(1.0f - attribs.x - attribs.y, attribs.x, attribs.y);
vec3 normal = normalize(v0.normal * barycentricCoords.x + v1.normal * barycentricCoords.y + v2.normal * barycentricCoords.z);
// Basic lighting
vec3 lightVector = normalize(ubo.lightPos.xyz);
float dot_product = max(dot(lightVector, normal), 0.6);
rayPayload.color = v0.color.rgb * vec3(dot_product);
//获取最大距离
rayPayload.distance = gl_RayTmaxEXT;
rayPayload.normal = normal;
//如果模型的颜色是白色，开启反射
// Objects with full white vertex color are treated as reflectors
rayPayload.reflector = ((v0.color.r == 1.0f) && (v0.color.g == 1.0f) && (v0.color.b == 1.0f)) ? 1.0f : 0.0f;
}
小结
反射是在光追Basic的基础上，通过循环增加渲染次数（反射）。 原理相同，丰富了miss和hit shader的用法。
Callable RayTracing
prepare
- createBottomLevelAccelerationStructure: 在本地创建模型底层加速结构信息
- 创建一个三角形的VBO和IBO
- createBuffer: Transform buffer
- createBuffer: Vertex buffer
- createBuffer: Index buffer
- 创建三个VkDeviceOrHostAddressConstKHR,分别存放tbo, vbo和ibo的指针(deviceAddress)
- vertexBufferDeviceAddress : getBufferDeviceAddress
- indexBufferDeviceAddress: getBufferDeviceAddress
- -transformBufferDeviceAddress: getBufferDeviceAddress
- 创建加速结构
- VkAccelerationStructureGeometryKHR
- 参数数模型的一些基本属性
- VkAccelerationStructureBuildGeometryInfoKHR: 几何信息
- VkAccelerationStructureBuildSizesInfoKHR:Get size info
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure: 创建加速结构buffer
- VkBufferCreateInfo:特殊的参数
- sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO
- usage = VK_BUFFER_USAGE_ACCELERATION_STRUCTURE_STORAGE_BIT_KHR
- vkCreateBuffer: 下面使用的参数结构体都使用加速结构相关参数
- vkGetBufferMemoryRequirements
- vkAllocateMemory
- vkBindBufferMemory
- VkAccelerationStructureCreateInfoKHR
- vkCreateAccelerationStructureKHR: 创建加速结构
- vkGetAccelerationStructureDeviceAddressKHR: AS device address
- VkBufferCreateInfo:特殊的参数
- VkAccelerationStructureGeometryKHR
- createScratchBuffer
- 在构建底层加速结构的过程中创建一个小的scratch缓冲区
- VkAccelerationStructureBuildGeometryInfoKHR:
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- 创建一个三角形的VBO和IBO
- createTopLevelAccelerationStructure
- 创建 top level 加速结构
- VkAccelerationStructureInstanceKHR
- createBuffer: for instance data
- VkDeviceOrHostAddressConstKHR
- getBufferDeviceAddress
- VkAccelerationStructureGeometryKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildSizesInfoKHR
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure()
- createScratchBuffer ()
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- instancesBuffer.destroy()
- createStorageImage(): 存放光追渲染结果:
- 释放image信息， 阴影教程里没有这部分
- vkDestroyImageView
- vkDestroyImage
- vkFreeMemory
- storageImage
- VkImageCreateInfo image = vks::initializers::imageCreateInfo();
- vkCreateImage
- vkGetImageMemoryRequirements
- vkAllocateMemory
- vkBindImageMemory
- VkImageViewCreateInfo
- vkCreateImageView
- createCommandBuffer
- setImageLayout
- flushCommandBuffer - createUniformBuffer
- createBuffer
- updateUniformBuffers
- createRayTracingPipeline
- VkDescriptorSetLayoutBinding:结构体
- VkDescriptorSetLayoutCreateInfo:结构体
- vkCreateDescriptorSetLayout
- VkPipelineLayoutCreateInfo
- vkCreatePipelineLayout
- Setup ray tracing shader groups
- VkPipelineShaderStageCreateInfo
- Ray generation group
- VkRayTracingShaderGroupCreateInfoKHR: 结构体,存放shader和其他配置信息
- shader: raygen.rgen.spv
- VkRayTracingShaderGroupCreateInfoKHR: 结构体,存放shader和其他配置信息
- Miss group: 光追未命中
- VkRayTracingShaderGroupCreateInfoKHR
- shader: miss.rmiss.spv
- Closest hit group: 光追命中
- VkRayTracingShaderGroupCreateInfoKHR
- shader: closesthit.rchit.spv
- Callable shader group: 这节重点
- shaderStages.push_back(loadShader(getShadersPath() + “raytracingcallable/callable” + std::to_string(i+1) + “.rcall.spv”, VK_SHADER_STAGE_CALLABLE_BIT_KHR));
- 这个shader没有在管线中
- Create the ray tracing pipeline
- VkRayTracingPipelineCreateInfoKHR:结构体
- vkCreateRayTracingPipelinesKHR
- createShaderBindingTable
- 创建用于加速结构的 shader 绑定表(Shader Binding Tables)-(SBT)
- vkGetRayTracingShaderGroupHandlesKHR
- createShaderBindingTable:创建四个SBT
- 最后一个是callable shader
- createBuffer: 创建SBT – raygen
- 使用字段VK_BUFFER_USAGE_SHADER_BINDING_TABLE_BIT_KHR
- createBuffer: miss
- createBuffer: hit
-
- createBuffer: callable
- memcpy
- memcpy
- memcpy
- memcpy
- createDescriptorSets
- vkCreateDescriptorPool
- vkAllocateDescriptorSets
- writeDescriptorSet: Ray tracing result image
- writeDescriptorSet: Uniform data
- writeDescriptorSet: Scene vertex buffer
- writeDescriptorSet: Scene index buffer
- vkUpdateDescriptorSets
- buildCommandBuffers
- handleResize
- createStorageImage
- storageImageDescriptor
- writeDescriptorSet
- vkBeginCommandBuffer
- 使用反射时，这里使用的pipe数增加，这里cmd的两次，应该是只反射一次，命令相同
- vkCmdBindPipeline
- vkCmdBindDescriptorSets
- vkCmdBindPipeline
- vkCmdBindDescriptorSets
- vkCmdTraceRaysKHR
- Copy ray tracing output to swap chain image
- setImageLayout
- setImageLayout
- vkCmdCopyImage
- setImageLayout
- setImageLayout
- vkEndCommandBuffer
- handleResize
render
- draw
- prepareFrame
- vkQueueSubmit
- submitFrame
- updateUniformBuffers
shader
- raygen.rgen：正常渲染
- miss.rmiss: 正常渲染
- closesthit.rchit
#version 460
#extension GL_EXT_ray_tracing : require
#extension GL_EXT_nonuniform_qualifier : enable
layout(location = 0) rayPayloadInEXT vec3 hitValue;
//注意这里返回一个callable数据
layout(location = 0) callableDataEXT vec3 outColor;
hitAttributeEXT vec3 attribs;
layout(binding = 0, set = 0) uniform accelerationStructureEXT topLevelAS;
void main()
{
// Execute the callable shader indexed by the current geometry being hit
// For our sample this means that the first callable shader in the SBT is invoked for the first triangle, the second callable shader for the second triangle, etc.
//待光线与面片相交，不直接返回该像素的颜色，而是调用callable函数，让callable shader 加工该命中像素的颜色，再输出。
executeCallableEXT(gl_GeometryIndexEXT, 0);
hitValue = outColor;
}
- callable1.rcall
#version 460 core
#extension GL_EXT_ray_tracing : enable
layout(location = 0) callableDataInEXT vec3 outColor;
void main()
{
//gl_LaunchIDEXT 当前像素坐标（不是uv）
// Generate a checker board pattern
vec2 pos = vec2(gl_LaunchIDEXT / 8);
//应该是奇数像素赋值白色，偶数像素赋值黑色
outColor = vec3(mod(pos.x + mod(pos.y, 2.0), 2.0));
}
- callable2.rcall
//这里单纯给了一个绿色作为颜色输出
#version 460 core
#extension GL_EXT_ray_tracing : enable
//
layout(location = 0) callableDataInEXT vec3 outColor;
void main()
{
outColor = vec3(0.0, 1.0, 0.0);
}
- callable3.rcall
#version 460 core
#extension GL_EXT_ray_tracing : enable
layout(location = 0) callableDataInEXT vec3 outColor;
void main()
{
// Generate a line pattern
vec2 pos = vec2(gl_LaunchIDEXT / 8);
//奇数行白色，偶数行黑色
outColor = vec3(mod(pos.y, 2.0));
}
小结
本节主要讲executeCallableEXT函数的使用方法。总结来说，感觉应该是为光追处理材质管线用的， 也可以做一些后处理。
Ray Query
prepare
- loadAssets: 加载模型， 这里使用阴影小节模型
- prepareUniformBuffers
- createBuffer
- updateLight
- updateUniformBuffers
- setupDescriptorSetLayout
- descriptorSetLayoutBinding
- vkCreateDescriptorSetLayout
- vkCreatePipelineLayout
- preparePipelines
- pipelineCreateInfo
- vkCreateGraphicsPipelines：这里绑定了正常管线的shader（VS\FS）
- createBottomLevelAccelerationStructure: 在本地创建模型底层加速结构信息
- vertexBufferDeviceAddress : getBufferDeviceAddress
- indexBufferDeviceAddress: getBufferDeviceAddress- 创建加速结构
- VkAccelerationStructureGeometryKHR
- 参数数模型的一些基本属性
- VkAccelerationStructureBuildGeometryInfoKHR: 几何信息
- VkAccelerationStructureBuildSizesInfoKHR:Get size info
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure: 创建加速结构buffer
- VkBufferCreateInfo:特殊的参数
- sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO
- usage = VK_BUFFER_USAGE_ACCELERATION_STRUCTURE_STORAGE_BIT_KHR
- vkCreateBuffer: 下面使用的参数结构体都使用加速结构相关参数
- vkGetBufferMemoryRequirements
- vkAllocateMemory
- vkBindBufferMemory
- VkAccelerationStructureCreateInfoKHR
- vkCreateAccelerationStructureKHR: 创建加速结构
- vkGetAccelerationStructureDeviceAddressKHR: AS device address
- VkBufferCreateInfo:特殊的参数
- VkAccelerationStructureGeometryKHR
- createScratchBuffer
- 在构建底层加速结构的过程中创建一个小的scratch缓冲区
- VkAccelerationStructureBuildGeometryInfoKHR:
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- 创建加速结构
- createTopLevelAccelerationStructure
- 创建 top level 加速结构
- VkAccelerationStructureInstanceKHR
- createBuffer: for instance data
- VkDeviceOrHostAddressConstKHR
- getBufferDeviceAddress
- VkAccelerationStructureGeometryKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildSizesInfoKHR
- vkGetAccelerationStructureBuildSizesKHR
- createAccelerationStructure()
- createScratchBuffer ()
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildGeometryInfoKHR
- VkAccelerationStructureBuildRangeInfoKHR
- accelerationStructureFeatures.accelerationStructureHostCommands
- 是
- vkBuildAccelerationStructuresKHR
- 否
- createCommandBuffer
- vkCmdBuildAccelerationStructuresKHR
- flushCommandBuffer
- 是
- deleteScratchBuffer
- instancesBuffer.destroy()
- setupDescriptorPool: 这里没有创建StorageImage
- vkCreateDescriptorPool
- setupDescriptorSets
- vkAllocateDescriptorSets
- VkWriteDescriptorSetAccelerationStructureKHR
- 注意这个表示，是正常管线与光追管线的桥梁
- writeDescriptorSets
- vkUpdateDescriptorSets
- buildCommandBuffers
- commandBufferBeginInfo
- loop
- vkBeginCommandBuffer
- vkCmdBeginRenderPass
- vkCmdSetViewport
- vkCmdSetScissor
- vkCmdBindDescriptorSets
- vkCmdBindPipeline
- drawUI
- vkCmdEndRenderPass
- vkEndCommandBuffer
render
- draw
- prepareFrame
- vkQueueSubmit
- submitFrame
- updateLight
- updateUniformBuffers
shader
- VS：正常渲染
- FS:
#version 460
#extension GL_EXT_ray_tracing : enable
#extension GL_EXT_ray_query : enable
//管线初始化的时候，设置了模型的加速结构，管线可以使用该标识引用加速结构
layout (binding = 2, set = 0) uniform accelerationStructureEXT topLevelAS;
layout (location = 0) in vec3 inNormal;
layout (location = 1) in vec3 inColor;
layout (location = 2) in vec3 inViewVec;
layout (location = 3) in vec3 inLightVec;
layout (location = 4) in vec3 inWorldPos;
layout (location = 0) out vec4 outFragColor;
#define ambient 0.1
void main()
{
vec3 N = normalize(inNormal);
vec3 L = normalize(inLightVec);
vec3 V = normalize(inViewVec);
vec3 R = normalize(-reflect(L, N));
vec3 diffuse = max(dot(N, L), ambient) * inColor;
//首先进行正常的光照计算
outFragColor = vec4(diffuse, 1.0);
//准备rayQuery实例
rayQueryEXT rayQuery;
//inWorldPos点向L方向发射光线
rayQueryInitializeEXT(rayQuery, topLevelAS, gl_RayFlagsTerminateOnFirstHitEXT, 0xFF, inWorldPos, 0.01, L, 1000.0);
//等该查询完成
// Start the ray traversal, rayQueryProceedEXT returns false if the traversal is complete
while (rayQueryProceedEXT(rayQuery)) { }
//如果管线与场景中模型有交点，即有阴影存在
// If the intersection has hit a triangle, the fragment is shadowed
if (rayQueryGetIntersectionTypeEXT(rayQuery, true) == gl_RayQueryCommittedIntersectionTriangleEXT ) {
outFragColor *= 0.1;
}
}
小结
本节主要讲，正常管线如何使用加速结构。使用场景会很多。与之前的raytracingshadows程序做了比较，该节程序在4K屏（3840*2160）中1200+FPS， 而raytracingshadows程序只有500+FPS。 初步估计（猜测）： closesthit虽然做了cullMask，剔除的顶点依然少于FS。所以碰撞检测的次数比FS多很多。后续复习时再深入分析一下原因。
974
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
