高级着色器语言

高級着色器语言（即 HLSL，缩写自 High Level Shader Language^[1] 或 High-Level Shading Language^[2]），是由微软拥有及开发的一种着色器语言，最初的开发是为了辅助 Direct3D 9 的着色器汇编语言，后成为 Direct3D 10 以来统一着色器模型（英语：Unified Shader Model）所必须的语言。

HLSL只能供微软的Direct3D以及XNA使用。HLSL是GLSL的先辈，不能与OpenGL标准兼容。它跟Nvidia的Cg非常相似，是因为两个开发者曾经紧密合作。^[3]

HLSL的主要作用为将一些复杂的图像处理，快速而又有效率地在显卡上完成，与组合式或低阶Shader Language相比，能降低在编写复杂特殊效果时所发生编程错误的机会。

Shader model 比较

Pixel shader 比较

Pixel shader 版本	1.0 to 1.3[4]	1.4[4]	2.0[4][5]	2.0a[4][5]	2.0b[4][5]	3.0[4][6]	4.0[7]	4.1[8]	5.0[9]
Dependent texture limit	4	6	8	Unlimited	8	Unlimited	Unlimited	Unlimited	Unlimited
Texture instruction limit	4	6*2	32	Unlimited	Unlimited	Unlimited	Unlimited	Unlimited	Unlimited
Position register	No	No	No	No	No	Yes	Yes	Yes	Yes
Instruction slots	8+4	8+4	32 + 64	512	512	≥ 512	≥ 65536	≥ 65536	≥ 65536
Executed instructions	8+4	6*2+8*2	32 + 64	512	512	65536	Unlimited	Unlimited	Unlimited
Texture indirections	4	4	4	Unlimited	4	Unlimited	Unlimited	Unlimited	Unlimited
Interpolated registers	2 + 8	2 + 8	2 + 8	2 + 8	2 + 8	10	32	32	32
Instruction predication	No	No	No	Yes	No	Yes	No	No	No
Index input registers	No	No	No	No	No	Yes	Yes	Yes	Yes
Temp registers	2	6	12 to 32	22	32	32	4096	4096	4096
Constant registers	8	8	32	32	32	224	16x4096	16x4096	16x4096
Arbitrary swizzling	No	No	No	Yes	No	Yes	Yes	Yes	Yes
Gradient instructions	No	No	No	Yes	No	Yes	Yes	Yes	Yes
Loop count register	No	No	No	No	No	Yes	Yes	Yes	Yes
Face register (2-sided lighting)	No	No	No	No	No	Yes	Yes	Yes	Yes
Dynamic flow control	No	No	No	No	No	24	Yes	Yes	Yes
Bitwise Operators	No	No	No	No	No	No	Yes	Yes	Yes
Native Integers	No	No	No	No	No	No	Yes	Yes	Yes

• PS 2.0 = DirectX 9.0 original Shader Model 2 specification.
• PS 2.0a = NVIDIA GeForce FX-optimized model.
• PS 2.0b = ATI Radeon X700, X800, X850 shader model, DirectX 9.0b.
• PS 3.0 = Shader Model 3.0.
• PS 4.0 = Shader Model 4.0.
• PS 4.1 = Shader Model 4.1.
• PS 5.0 = Shader Model 5.0.

"32 + 64" for Executed Instructions means "32 texture instructions and 64 arithmetic instructions."

Vertex shader 比较

Vertex shader 版本	VS 1.1[10]	VS 2.0[5][10]	VS 2.0a[5][10]	VS 3.0[6][10]	VS 4.0[7]	VS 4.1[11]	VS 5.0[9]
# of instruction slots	128	256	256	≥ 512	4096	4096	4096
Max # of instructions executed	Unknown	65536	65536	65536	65536	65536	65536
Instruction predication	No	No	Yes	Yes	Yes	Yes	Yes
Temp registers	12	12	13	32	4096	4096	4096
# constant registers	≥ 96	≥ 256	≥ 256	≥ 256	16x4096	16x4096	16x4096
Static flow control	???	Yes	Yes	Yes	Yes	Yes	Yes
Dynamic flow control	No	No	Yes	Yes	Yes	Yes	Yes
Dynamic flow control depth	No	No	24	24	Yes	Yes	Yes
Vertex texture fetch	No	No	No	Yes	Yes	Yes	Yes
# of texture samplers	N/A	N/A	N/A	4	128	128	128
Geometry instancing support	No	No	No	Yes	Yes	Yes	Yes
Bitwise operators	No	No	No	No	Yes	Yes	Yes
Native integers	No	No	No	No	Yes	Yes	Yes

• VS 2.0 = DirectX 9.0 original Shader Model 2 specification.
• VS 2.0a = NVIDIA GeForce FX-optimized model.
• VS 3.0 = Shader Model 3.0.
  • Note that ATI X1000 series cards (e.g. X1900) does not support Vertex Texture Fetch, hence it does not fully comply with the VS 3.0 model. Instead, they offer a feature called "Render to Vertex Buffer (R2VB)" that provides functionality that is a superset of Vertex Texture Fetch.^[12]
• VS 4.0 = Shader Model 4.0.
• VS 4.1 = Shader Model 4.1.
• VS 5.0 = Shader Model 5.0.

范例

灰阶贴图用

• 灰阶值 = 0.3 x 红 + 0.59 x 绿 + 0.11 x 蓝

sampler2D Texture0;

float4 ps_main( float2 texCoord  : TEXCOORD0 ) : COLOR
{
   float4 _inColor = tex2D( Texture0, texCoord );
   float gray = 0.3*_inColor.x + 0.59*_inColor.y + 0.11*_inColor.z;
   float4 _outColor = float4(gray, gray, gray, 1.0);
   
   return _outColor;
}