LeoYeAI/agent-master-skills / skills/animation-shader
animation-shader
VerifiedREAD this skill when implementing or configuring animation-style shaders (Toon/Cel Shaders) — including outlines, rim lighting, toon shading, MatCap, emission, dissolve, hatching, or any stylized rendering effect. Contains preset styles and feature-to-reference mappings for lilToon, Poiyomi, UTS2, RToon, SToon, and ToonShadingCollection. Works as a domain knowledge plugin alongside workflow skills (OpenSpec, SpecKit) or plan mode of an agent.
references/RToon/Details/01_Core_Shading.md
View on GitHub# Core Shading (核心光照)
RToon 的核心光照模型主要由 "Self Shadow" (自阴影/二值化光照) 和 "Smooth Object Normal" (平滑法线) 组成。
## 1. Self Shadow (自阴影)
这是 RToon 实现赛璐璐风格(Cel-Shading)的基础。它不依赖贴图 Ramp,而是通过计算 NdotL 并进行阈值切割来生成硬边阴影。
### Properties (参数)
* `_SelfShadowThreshold` ("Threshold", Range(0, 1)): 决定亮部和暗部的分界线。值越大,阴影区域越大。
* `_SelfShadowHardness` ("Hardness", Range(0, 1)): 控制阴影边缘的软硬度。1.0 为绝对硬边,0.0 为非常柔和。
* `_SelfShadowRealtimeShadowIntensity`: 控制实时投影(ShadowMap)对自阴影的影响程度。
* `_SelfShadowAffectedByLightShadowStrength`: 自阴影是否受光源的 Shadow Strength 属性影响。
* `_VertexColorGreenControlSelfShadowThreshold`: **[重要]** 使用顶点色的**绿色通道**来局部控制阴影阈值。
* *原理*: 绿色通道值会偏移全局的 Threshold。
* *应用*: 让某些部位(如脖子、腋下)更容易产生阴影,或让脸部更不容易产生阴影。
### Implementation (实现原理)
在 `RT_URP_Core.hlsl` 中的 `RT_SS` 函数实现:
```glsl
// RT_SS Function Analysis
half RT_SS(float4 vertexColor, float3 RTD_NDOTL, half attenuation, float dim_val)
{
// 1. 硬度重映射: 将 0-1 的硬度参数映射到 smoothstep 的下限范围
half RTD_SS_SSH_Sil = lerp(0.3, 1.0, _SelfShadowHardness);
// 2. 阈值重映射: 将 0-1 的阈值参数映射到 -1 到 1 的范围 (用于偏移 NdotL)
half RTD_SS_SSTH_Sli = lerp(-1.0, 1.0, _SelfShadowThreshold);
// 3. 顶点色控制: 如果开启,顶点色绿色通道会改变阈值
half RTD_SS_VCGCSSS_OO;
if (!_VertexColorGreenControlSelfShadowThreshold) {
RTD_SS_VCGCSSS_OO = RTD_SS_SSTH_Sli;
} else {
// 绿色通道值越小,阈值越低?这里逻辑是乘法混合
RTD_SS_VCGCSSS_OO = RTD_SS_SSTH_Sli * (1.0 - vertexColor.g);
}
// 4. 核心计算: Smoothstep(min, 1.0, NdotL * ThresholdFactor)
// 这里并没有直接比较 NdotL > Threshold,而是通过乘法缩放 NdotL 然后用 smoothstep 切割
half RTD_SS_SST = smoothstep(RTD_SS_SSH_Sil, 1.0, ((float)RTD_NDOTL * lerp(7.0, RTD_SS_VCGCSSS_OO, RTD_SS_SSTH_Sli)));
// 5. 结合光照衰减 (Attenuation / ShadowMap)
// dim_val 通常来自 _MainLightShadowData.x
half RTD_SS_SSABLSS_OO = lerp(RTD_SS_SST, lerp(RTD_SS_SST, 1.0, (1.0 - dim_val)), _SelfShadowAffectedByLightShadowStrength);
// 6. 最终输出
half RTD_SS_ON = lerp(1.0, (RTD_SS_SSABLSS_OO * attenuation), _SelfShadowRealtimeShadowIntensity);
return RTD_SS_ON;
}
```
### Analysis (分析)
* **非标准 Ramp**: RToon 没有使用传统的 `tex2D(_RampTex, NdotL)` 采样,而是使用数学公式计算。
* **优点**: 省去了一张贴图采样,性能稍好,且参数调整更直观。
* **缺点**: 无法做复杂的色彩渐变(如明暗交界处的红色散射线),只能做纯粹的明暗过渡。
---
## 2. Smooth Object Normal (平滑法线)
用于解决低模(Low Poly)角色面部或身体上的丑陋阴影块。
### Properties (参数)
* `_SmoothObjectNormal` ("Smooth Object Normal", Range(0, 1)): 混合权重。0 为使用原始法线,1 为使用平滑法线。
* `_VertexColorRedControlSmoothObjectNormal`: **[重要]** 使用顶点色的**红色通道**控制平滑程度。
* `_XYZPosition`: 可以微调计算平滑法线时的中心点偏移。
### Implementation (实现原理)
RToon 在 Vertex Shader 中通过算法实时计算一个近似的平滑法线,或者通过混合法线来实现。
在 `RT_URP_Core.hlsl` 中的 `calcNorm` 函数(用于计算平滑法线)和 `RT_SON` 函数:
```glsl
// 1. 计算近似平滑法线 (在 Vertex Shader 中可能被调用)
// 这是一个非常简化的算法,实际上是基于位置的叉积,模拟球形法线
float3 calcNorm(float3 pos)
{
float3 vecTan = normalize(cross(pos, float3(1.01, 1.0, 1.0)));
float3 vecBitan = normalize(cross(vecTan, pos));
return normalize(cross(vecTan, vecBitan));
}
// 2. 混合逻辑 (RT_SON)
float3 RT_SON(float4 vertexColor, float3 calNorm, float3 normalDirection, out float3 RTD_SON_CHE_1)
{
// 计算混合权重
float RTD_SON_VCBCSON_OO;
if (!_VertexColorRedControlSmoothObjectNormal) {
RTD_SON_VCBCSON_OO = _SmoothObjectNormal;
} else {
RTD_SON_VCBCSON_OO = _SmoothObjectNormal * (1.0 - vertexColor.r);
}
// 混合原始法线 (normalDirection) 和 计算出的/传入的平滑法线 (calNorm)
// TransformObjectToWorldNormal(-calNorm) 说明 calNorm 可能是模型空间的
float3 RTD_SON_ON_OTHERS = lerp(normalDirection, TransformObjectToWorldNormal(-calNorm), RTD_SON_VCBCSON_OO);
// 返回最终用于光照计算的法线
return RTD_SON_ON_OTHERS;
}
```
### Analysis (分析)
* **用途**: 在不修改模型几何体的情况下,让角色的脸部阴影看起来像是一个圆润的球体,消除棱角分明的三角面阴影。
* **限制**: `calcNorm` 算法是一种近似(基于位置向量),假设物体中心在原点或通过 `_XYZPosition` 修正。对于形状复杂的非球体物体,效果可能不理想。建议仅在脸部等接近球体的区域使用顶点色开启此功能。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |