（转）游戏引擎中三大及时光照渲染方法介绍(以unity3d为例)

重要：在目前市面上常见的游戏引擎中，主要采用以下三种灯光实现方式：

顶点照明渲染路径细节 Vertex Lit Rendering Path Details

正向渲染路径细节 Forward Rendering Path Details

延迟光照渲染路径的细节 Deferred Lighting Rendering Path Details

以unity3d为例，以下将详细讲解三种灯光渲染方式的实现、原理及缺陷。

顶点照明渲染路径细节 Vertex Lit Rendering Path Details

Vertex Lit path generally renders each object in one pass, with lighting from all lights calculated at object vertices.

顶点照明渲染路径通常在一个通道中渲染物体，所有光源的照明都是在物体的顶点上进行计算的。

It‘s the fastest rendering path and has widest hardware support (however, keep in mind: it does not work on consoles).

顶点照明渲染路径是最快的渲染路径并且有最广泛的硬件支持（然而，请记住：它无法工作在游戏机上）。

Since all lighting is calculated at vertex level, this rendering path does not support most of per-pixel effects: shadows, normal mapping, light cookies, highly detailed specular highlights are not supported.

由于所有的光照都是在顶点层级上计算的，此渲染路径不支持大部分的逐像素渲染效果：如，阴影、法线贴图、灯光遮罩、高精度的高光。

正向渲染路径细节 Forward Rendering Path Details

Forward Rendering path renders each object in one or more passes, depending on lights that affect the object. Lights themselves are also treated differently by Forward Rendering, depending on their settings and intensity.

根据影响物体的光源的不同，正向渲染路径用单个或多个通道来渲染物体。在正向渲染中，光源本身也会根据他们的设置和强度受到不同的对待。

Implementation Details 实现细节

In Forward Rendering, some number of brightest lights that affect each object are rendered in fully per-pixel lit mode. Then, up to 4 point lights are calculated per-vertex. The other lights are computed as Spherical Harmonics (SH), which is much faster but is only an approximation. Whether a light will be per-pixel light or not is dependent on this:

在正向渲染中，影响物体的最亮的几个光源使用逐像素光照模式。接下来，最多有4个点光源会以逐顶点渲染的方式被计算。其他光源将以球面调和（Spherical Harmonics）的方式进行计算，球面调和技术计算很快但只能得到近似值。根据以下的规则判断一个光源是否为逐像素光源：

• Lights that have their Render Mode set to Not Important are always per-vertex or SH.  渲染模式被设置为不重要（Not Important）的光源以逐顶点或球面调和的方式进行计算
• Brightest     directional light is always per-pixel.  最亮的方向光源为像素光源
• Lights     that have their Render Mode set to Important are     always per-pixel.  渲染模式被设置重要(Important)的光源为像素光源
• If     the above results in less lights than current Pixel     Light Count Quality     Setting, then more lights are rendered     per-pixel, in order of decreasing brightness.  如根据以上规则得到的像素光源数量小于质量设置中的像素光源数量(Pixel Light Count)，为了减少亮度，会有更多的光源以逐像素的方式进行渲染

Rendering of each object happens as follows:

用以下的方法渲染每个物体：

• Base     Pass applies one per-pixel directional light and all per-vertex/SH lights.  基础通道渲染一个逐像素方向光和所有的逐顶点/球面调和光。
• Other     per-pixel lights are rendered in additional passes, one pass for each     light.  其他逐像素光在附加的通道中进行渲染，每个光源都需要一个通道

For example, if there is some object that‘s affected by a number of lights (a circle in a picture below, affected by lights A to H):

例如，如果有一个物体受到若干光源的影响（下图中的圆圈，受到光源A到H的影响）

Let‘s assume lights A to H have the same color & intensity, all all of them have Auto rendering mode, so they would be sorted in exactly this order for this object. The brightest lights will be rendered in per-pixel lit mode (A to D), then up to 4 lights in per-vertex lit mode (D to G), and finally the rest of lights in SH (G to H):

假设光源A到H都有相同的颜色和强度，且它们的渲染模式都为自动的(Auto)，那么它们严格的按照其名字排序。最亮的光源以逐像素光照模式的方式进行渲染（A到D），然后最多有4个光源以逐顶点光照模式进行渲染（D到G），其他光源以球面调和的方式进行渲染（G到H）。

Note that light groups overlap; for example last per-pixel light blends into per-vertex lit mode so there are less "light popping" as objects and lights move around.

注意不同的光照组间有重叠，如，最后一个逐像素光源也以逐顶点光照模式的方式渲染，这样能减少当物体和灯光移动时可能出现的"光照跳跃"现象。

Base Pass 基本通道

Base pass renders object with one per-pixel directional light and all SH lights. This pass also adds any lightmaps, ambient and emissive lighting from the shader. Directional light rendered in this pass can have Shadows. Note that Lightmapped objects do not get illumination from SH lights.

基础通道用一个逐像素方向光和所有球面调和光渲染物体。此通道还负责渲染着色器中的光照贴图，环境光和自发光。在此通道中渲染的方向光可以产生阴影。需要注意的是，使用了光照贴图的物体不会得到球面调和光的光照。

Additional Passes 附加通道

Additional passes are rendered for each additional per-pixel light that affect this object. Lights in these passes can‘t have shadows (so in result, Forward Rendering supports one directional light with shadows).

附加通道用于渲染影响物体的其他逐像素光源。这些通道中渲染的光源无法产生阴影（因此，前向渲染支持一个能产生阴影的方向光）。

Performance Considerations 性能注意事项

Spherical Harmonics lights are very fast to render. They have a tiny cost on the CPU, and are actually free for the GPU to apply (that is, base pass always computes SH lighting; but due to the way SH lights work, the cost is exactly the same no matter how many SH lights are there).

渲染球面调和光很快。它们只花费很少的CPU计算时间，并且实际上无需花费任何GPU计算时间（换言之，基础通道会计算球面调和光照，但由于球面调和光的计算方式，无论有多少球面调和光源，计算它们所花费的时间都是相同的）。

The downsides of SH lights are:

球面调和光源的缺点有：

• They     are computed at object‘s vertices, not pixels. This means they do not     support light Cookies or normal maps.  它们计算的是物体的顶点而不是像素。这意味着它们不支持投影遮罩和发现贴图。
• SH     lighting is very low frequency. You can‘t have sharp lighting transitions with     SH lights. They are also only affecting the diffuse lighting (too low     frequency for specular highlights).  球面调和光只有很低的频率。球面调和光不能产生锋利的照明过渡。它们也只会影响散射光照（对高光来说，球面调和光的频率太低了）。
• SH     lighting is is not local; point or spot SH lights close to some surface     will "look wrong".  球面调和不是局部的，靠近曲面的球面调和点光和聚光可能会"看起来不正确"。

In summary, SH lights are often good enough for small dynamic objects.

总的来说，球面调和光的效果对小的动态物体来说已经足够好了。

延迟光照渲染路径的细节 Deferred Lighting Rendering Path Details

Deferred Lighting is rendering path with the most lighting and shadow fidelity:

延迟光照是一种当前最高级的能实现光线和阴影保真的渲染路径

• There‘s     no limit how many lights can affect any object.  对于能影响任何物体的光线数量没有上限
• All     lights are evaluated per-pixel. Which means that they all interact     properly with normal maps etc.  完全采用以每像素的方式评估光线，这等于意味着全部将以正常贴图的方式正确的和物体交互
• All     lights can have Cookies.  所有光线都能拥有信息缓存
• All     lights can have Shadows.  所有的光线都能产生阴影

Deferred Lighting‘s advantages 延迟光照的优点：

• Lighting     cost is proportional to light size on screen. Does not matter how many     objects it shines on. Small lights = cheap!  光照的开销与屏幕的光线尺寸成正比，不用担心光线所照射的物品的数量，少量光线     等价于     廉价的花费
• Consistency.     All lighting for all lights is computed per-pixel; there are no lighting     computations that break down on large triangles etc.  一致性，所有的光线的光照采用按像素为计算分割单位来计算。比如，不会有在大规模三角形情况下光照计算使计算性能发生崩溃的情况发生。

Disadvantages 缺点：

• No     real anti-aliasing support.  没有实时抗锯齿支持
• Deferred     Lighting can‘t handle semi-transparent objects. Those are rendered using     Forward Rendering.  延迟光照不能处理半透明物体，也不能用在哪些使用前向渲染的物体之上
• Limited     lighting model support (Blinn-Phong). All lighting is computed the same     way; you can‘t have drastically different lighting models on different     objects.  有限的光照模式支持（Blinn-Phong）。所有光照以同样的方式计算，你不能够在不同的物体上采用完全不同的光照模式
• No     support for "receive shadows" flag and limited support light     Culling Masks.  没有对接收阴影特征的支持和对光线遮罩剔除有限的支持

Requirements for Deferred Lighting 延时光照的需求

• Requires Unity Pro.  需要Unity专业版
• Graphics     card with Shader Model 3.0 (or later), support for Depth render textures     and two-sided stencil buffer. Most graphics cards made after 2004 support     it: GeForce FX and later, Radeon X1300 and later, Intel 965 / GMA X3100     and later.  显示卡支持Shader Model 3.0(或更高)，深度纹理渲染和双面模板缓冲特性。许多2004年后的显卡都支持：如Geforce     Fx或更高，Radeon X1300或更高 Intel 965/ GMA     X3100 或更高
• Currently     does not work on mobile platforms.  目前在移动平台不支持。

Performance Considerations 性能注意事项

Cost of realtime lights in Deferred Lighting is proportional to number of pixels the light shines on; and not dependent on scene complexity. So small point or spot lights are very cheap to render. Point or spot lights that are fully or partially occluded by some scene objects get their pixels skipped on the GPU, so they are even cheaper.

延迟光照中实时光线的开销和光线照亮的像素值的数量成正比。而不取决于场景的复杂性。微小的点光源和聚光灯光源非常容易渲染。点光源或者完全或者部分被场景物体遮挡的聚光灯光源所照射的像素则被GPU所跳过，因此更加廉价。

Of course, lights with shadows are much more expensive than lights without shadows. In Deferred Lighting, shadow casters still need to be rendered once or more for each shadow-casting light. And the lighting shader that applies shadows is also more expensive than one without shadows.

当然，拥有阴影的光源比没有阴影的光源要昂贵许多。使用延迟光照，光影投射器仍然需要为每个阴影投射渲染一次或者多次。而且产生阴影的光线着色器也比不产生阴影的光线着色器要昂贵许多。

Implementation Details 实现细节

When Deferred Lighting is used, rendering process in Unity happens like this:

当延迟光照生效时，在Unity中发生的渲染过程如下：

1. Base Pass: objects are rendered, producing screen-space buffers with depth, normals, and specular power.  基本渲染：被渲染的对象产生带有深度，法线，和反射量的屏幕空间缓冲
2. Lighting pass: lighting is computed using the previous buffers. Lighting is computed into another screen-space buffer.  光照渲染：使用上一步的缓冲计算出光照。结果放入另一个屏幕空间缓存
3. Final pass: objects are rendered again. They fetch computed lighting, combine it with color textures and add any ambient/emissive lighting.  最后渲染：物体再次渲染。取来已经计算好的光线和颜色纹理混合在一起，然后再加上环境光以及散射光照。

Objects with shaders that can‘t handle Deferred Lighting are rendered after this process is done, using RenderTech-ForwardRendering path.

不能采用延迟光照技术的带阴影的物体在延迟光照渲染完后使用前向渲染路径处理。

Base Pass 基本渲染阶段

Base pass renders each object once. View space normals and specular power are rendered into single ARGB32 Render Texture (normals in RGB channels, specular power in A). If platform& hardware supports reading Z buffer as a texture, then depth is not explicitly rendered. If Z buffer can‘t be accessed as a texture, then depth is rendered in additional rendering pass, using shader replacement.

基本渲染将每个物体都渲染一次。视图空间法线和高光强度被渲染进单一的ARGB32渲染纹理（法线在RGB通道，高光强度在A通道）中。如果平台和硬件支持将Z缓冲按纹理读取，那么深度不会被明确的渲染。如果Z缓冲不能被以纹理的方式访问，那么深度将在额外的渲染处理中被使用着色器替代技术渲染。

Result of the base pass is Z buffer filled with scene contents and Render Texture with normals & specular power.

基本渲染的结果是被屏幕内容填满的Z缓冲和带有法线和高光强度的渲染纹理。

Lighting Pass 光照渲染阶段

Lighting pass computes lighting based on depth, normals and specular power. Lighting is computed in screen space, so it‘s independent of scene complexity. Lighting buffer is single ARGB32 Render Texture, with diffuse lighting in RGB channels and monochrome specular lighting in A channel. Lighting values are encoded using logarithmic encoding to provide extended dynamic range than usually possible with ARGB32 texture.

光照渲染基于深度，法线和高光强度计算光照。光照是被屏幕空间被计算的，因此和屏幕复杂性无关。光照缓冲是一个单一的ARGGB32渲染纹理，纹理的RGB通道带有漫反射的光照信息，在A通道带有单一特定颜色的光照。光照值采用对数值编码以产生比通常ARGB32纹理所能达到的动态扩展范围。

Lighting model is fixed to Blinn-Phong.

光照模式固定为Blinn-Phong。

Point and Spot lights that do not cross camera‘s near plane are rendered as 3D shapes, with Z buffer test against scene enabled. This makes partially or fully occluded Point and Spot lights very cheap to render. Directional lights and Point/Spot lights that cross the near plane are rendered as fullscreen quads.

不能跨越临近平面的点光源和聚光灯光源被作为带有开启测试场景的Z缓冲3D形状渲染，这部分和完全屏蔽的点光源和聚光灯光源可以非常廉价的渲染。 跨越临近区域的平行光或者点光源能作为全屏四边形。

If a light has shadows enabled, they are rendered and applies in this pass as well. Note that shadows are not "free"; shadow casters need to be rendered and a more complex light shader needs to be applied.

如果一个带有阴影的光源生效，在这个处理过程中会被很好的渲染。注意阴影并不免费，阴影投射器需要开销来渲染，同时一个更加复杂的光线着色器需要应用。

Final Pass 最后渲染阶段

Final pass produces final rendered image. Here all objects are rendered again; with shaders that fetch the lighting, combine it with textures and add any emissive lighting.

Lightmaps are also applied in the final pass. Close to the camera, realtime lighting is used, and only baked indirect lighting is added. This crossfades into fully baked lighting further away from the camera.

最终渲染阶段产生最后渲染后的图像，到这一步，所有的对象都将被再次渲染，其中着色器将混合前一步生成的光源和纹理以及所有自发光照明。

在最后渲染阶段光照贴图也被应用。靠近相机，使用实时光照，并仅烘焙间接光照。

其他：lightingmap 烘焙贴图不在及时光的技术范围内。烘焙灯光是通过贴图记录光照信息来模拟固定的光照效果，场景中本身不包含及时灯光。

Rendering Paths Comparison 渲染路径比较