March 2010 3
How Does It Work?
Overview
In this sample we simulate a small number of hair strands (166 strands) and then
when rendering the final hair we interpolate between these strands to create many
more strands (over 18,000). Since the amount of final rendered hair is a lot larger
than the amount of simulated strands, this method saves the considerable cost of
having to simulate every final hair that is rendered while producing very acceptable
results. The creation of these additional strands can be done using the tessellation
engine, or can be done using other approaches as outlined in [Tariq 08] (code for
both paths is provided in the sample).
Rendering Hair Using
Tessellation
Generating Data on the GPU
Given the large amount of data that we are trying to render, it is more efficient to
generate this data directly on the GPU than to send it at each frame from the CPU
to the GPU. DirectX 11 introduce functionality to create large amounts of data
directly on the hardware; the tessellation engine. In this section we are going to be
talking about how to use the tessellation engine to easily and efficiently create hair
strands for rendering.
Note that functionality for creating data on the GPU existed even in D3D10, via the
Geometry Shader. However, the Geometry Shader is meant for only small amounts
of data expansion, for example for expanding a line into a quad. It is not meant for,
and is certainly not efficient at, creating the amounts of geometry that we would like
to create for all the hair strands.
There are a number of reasons why the tessellation engine is useful for creating hair
for rendering. The most important advantage is that it is faster to create data using
the tessellation engine than it is to create data on the CPU and then upload it to the
GPU, or in some cases even to render “dummy vertices” on the GPU and then
evaluate them in the Vertex Shader. It is also easier to create hair using the
Tessellation engine than using the dummy vertex method. Finally, using the
Tessellation engine we can have very fine grained and continuous control over the
level of detail.
剩余22页未读，继续阅读
amazon007008
- 粉丝: 0
- 资源: 11
我的内容管理
展开
最新资源
- 构建Cadence PSpice仿真模型库教程
- VMware 10.0安装指南：步骤详解与网络、文件共享解决方案
- 中国互联网20周年必读：影响行业的100本经典书籍
- SQL Server 2000 Analysis Services的经典MDX查询示例
- VC6.0 MFC操作Excel教程：亲测Win7下的应用与保存技巧
- 使用Python NetworkX处理网络图
- 科技驱动：计算机控制技术的革新与应用
- MF-1型机器人硬件与robobasic编程详解
- ADC性能指标解析：超越位数、SNR和谐波
- 通用示波器改造为逻辑分析仪：0-1字符显示与电路设计
- C++实现TCP控制台客户端
- SOA架构下ESB在卷烟厂的信息整合与决策支持
- 三维人脸识别：技术进展与应用解析
- 单张人脸图像的眼镜边框自动去除方法
- C语言绘制图形：余弦曲线与正弦函数示例
- Matlab 文件操作入门：fopen、fclose、fprintf、fscanf 等函数使用详解
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
