Gamedev Hub

Shadow Mapping

Shadow mapping is the standard technique for rendering shadows in 3D games. It works by treating the light as a camera and determining which surfaces are “visible” to the light and which are blocked.

1. The Two-Pass Process

Pass 1: The Shadow Pass

Render the scene from the Light’s point of view.

• Instead of colors, we store only the Depth (distance from the light) in a special texture called the Shadow Map.
• This tells us exactly how far light can “travel” before hitting an object.

Pass 2: The Lighting Pass

Render the scene from the Camera’s perspective. For every pixel on screen:

1. Convert its world position into Light Space.
2. Calculate its current distance from the light (pixelDepth).
3. Sample the Shadow Map to see the closest thing the light hit (mapDepth).
4. The Comparison:
   • If pixelDepth > mapDepth, something is blocking the light. The pixel is In Shadow.
   • Otherwise, it is Lit.

2. Common Challenges

Shadow Acne

Tiny precision errors cause surfaces to shadow themselves, creating “Z-fighting” stripes.

• Fix: Add a small Bias (an offset) to the depth comparison.

Peter Panning

If the bias is too high, the shadow “detaches” from the object and appears to float.

• Fix: Use “Normal Bias” which pushes the shadow inward based on the surface normal.

Resolution & Cascades

Standard shadow maps look blocky in the distance.

• The Solution: Cascaded Shadow Maps (CSM). We split the camera’s view into different “frustum slices.” Each slice gets its own shadow map. Close-up slices get high resolution, while distant ones get lower resolution.

3. Implementation (HLSL Snippet)

float4 lightSpacePos = mul(_WorldToShadowMatrix, float4(i.worldPos, 1.0));
float depthFromMap = tex2D(_ShadowMap, lightSpacePos.xy).r;

// The Comparison
float shadow = lightSpacePos.z > (depthFromMap + _Bias) ? 0.5 : 1.0;

return col * shadow;

4. Summary

• Pass 1: Save depth from light perspective.
• Pass 2: Compare pixel depth against that map.
• CSM: Scale resolution based on distance.