一些光照模型

一些光照模型

1.  Lambert

Diffuse Light

与视角无关

float   NdotL = max( 0.0 , dot(s.Normal, lightDir));
float4 finalColor;
finalColor.rgb = s.Albedo * _LightColor0 * NdotL;
finalColor.a = s.Alpha;
return   finalColor;

 折叠源码

fixed diff = max( 0 , dot(s.Normal, lightDir));

fixed4 c;
c.rgb = s.Albedo * _LightColor0 * diff * atten;
c.a = s.Alpha;
return   c;

2. Half-Lambert

半条命起源

可以看清物体暗的部分的形状，因此物体显得更加扁平

float   NdotL = max( 0.0 , dot(s.Normal, lightDir));
float   HalfLambertDiffuse = pow(NdotL *  0.5   +  0.5 ,  2.0 );
float4 finalColor;
finalColor.rgb = s.Albedo * _LightColor0 * HalfLambertDiffuse * atten;
finalColor.a = s.Alpha;
return   finalColor;

3. Phong

表现高光效果

diffuse light and specular light

float   NdotL = max( 0 , dot(s.Normal, lightDir));

float3 lightReflectDirection = reflect(-lightDir, s.Normal);
float   RdotV = max( 0 , dot(lightReflectDirection, viewDir));
float   spec = pow(RdotV, _SpecPower/ 4 ) * _SpecularColor;

float4 color;
color.rgb = (s.Albedo * _LightColor0.rgb * NdotL) + (_LightColor0.rgb * _SpecularColor.rgb * spec) * atten;
color.a = s.Alpha;
return   color;

4. Blinn-Phong

建立在Phong的基础上。Phong计算light reflection vector ，而Blinn计算Half Direction替之，即灯光方向和视角方向的半角向量。
好处是高光更加柔和。

float   NdotL = max( 0 , dot(s.Normal, lightDir));

float3 halfVector = normalize(lightDir + viewDir);
float   NdotH = max( 0 , dot(s.Normal, halfVector));
float   spec = pow(NdotH, _SpecPower) * _SpecularColor;

float4 color;
color.rgb = (s.Albedo * _LightColor0.rgb * NdotL) + (_LightColor0.rgb * _SpecularColor.rgb * spec) * atten;
color.a = s.Alpha;
return   color;

5. PBR

完全基于物理的

6. Banded-lighting

将连续的灯光变为离散的一段一段的

float   NdotL = max( 0.0 , dot(s.Normal, lightDir));

float   lightBandsMultiplier = _LightSteps /  256 ;
float   lightBandsAdditive = _LightSteps /  2 ;
fixed bandedNdotL = (floor((NdotL *  256   + lightBandsAdditive) / _LightSteps))
    * lightBandsMultiplier;

float3 lightingModel = bandedNdotL * s.Albedo;
float3 attenColor = atten * _LightColor0.rgb;
float4 finalDiffuse = float4(lightingModel * attenColor,  1 );
return   finalDiffuse;

7. Minnaert

特别适合天鹅绒或月亮，这类多孔可渗透性或纤维状的表面，这类表面会导致大量光反向散射

float3 viewDirection = viewDir;
float   NdotL = max( 0 , dot(s.Normal, lightDir));
float   NdotV = max( 0 , dot(s.Normal, viewDirection));

float3 minnaert = saturate(NdotL * pow(NdotL*NdotV, _Roughness));

float3 lightingModel = minnaert * s.Albedo;
float3 attenColor = atten * _LightColor0.rgb;
float4 finalDiffuse = float4(lightingModel * attenColor,  1 );
return   finalDiffuse;

8. Oren–Nayar

Lambert 光照模型，是一个让光线向各个角度都均匀辐射的模型。这个均匀实在太不可思议了，真实物体表面理应不是这样的。

在灯光方向不变的情况下，观察物体的角度不同，物体表面光强、颜色也会发生变化（一般粗糙物体正光光强比背光光强要强）。就是光线向各个角度并非均匀辐射。

Oren-Nayar model 这个主要来自于Michael Oren和Shree K. Nayar在SIGGRAPH94上发表的论文Generalization of Lambert’s Reflectance Model。里面通过统计的等手段总结出比较接近真实粗糙表面的数学公式

http://www1.cs.columbia.edu/CAVE/publications/pdfs/Oren_SIGGRAPH94.pdf

float   roughness = _Roughness;
float   roughnessSqr = roughness * roughness;
float3 o_n_fraction = roughnessSqr / (roughnessSqr + float3( 0.33 ,  0.13 ,  0.09 ));
float3 oren_nayar = float3( 1 ,  0 ,  0 ) + float3(- 0.5 ,  0.17 ,  0.45 ) * o_n_fraction;
float3 viewDirection = viewDir;
float   cos_ndotl = saturate(dot(s.Normal, lightDir));
float   cos_ndotv = saturate(dot(s.Normal, viewDirection));
float   oren_nayar_s = saturate(dot(lightDir, viewDirection)) - cos_ndotl * cos_ndotv;
oren_nayar_s /= lerp(max(cos_ndotl, cos_ndotv),  1 , step(oren_nayar_s,  0 ));


//lighting and final diffuse
float   attenuation = atten;
float3 lightingModel = s.Albedo * cos_ndotl * (oren_nayar.x + s.Albedo * oren_nayar.y + oren_nayar.z * oren_nayar_s);
float3 attenColor = attenuation * _LightColor0.rgb;
float4 finalDiffuse = float4(lightingModel * attenColor,  1 );
return   finalDiffuse;

转载于:https://www.cnblogs.com/CloudLiu/p/10855354.html