Unity uses a technique called shadow mapping to render real-time shadows.
Shadow mapping uses textures called shadow maps. Shadow maps are similar to depth textures. A Light generates a shadow map in a similar way to how a Camera generates a depth texture. If you imagine a Camera at the same location as the Light, the areas of the Scene that the Camera cannot see are the same areas of the Scene that rays from the Light cannot reach; therefore, they are in shadow.
Unity populates the shadow map with information about how far rays from the Light travel before they hit a surface, and then samples the shadow map to calculate real-time shadows for GameObjects that the Light hits.
For more information on shadow mapping, see the Wikipedia Page on shadow mapping.
To calculate the resolution of a shadow map, Unity:
Determines the area of the screen view that the Light can illuminate. For directional lights, the whole screen can be illuminated. For Spot Lights and Point lights, the area is the onscreen projection of the shape of the light’s extent: a sphere for point lights, or a cone for Spot Lights. The projected shape has a width and height in pixels on the screen; the larger of those two values is then taken. This value is called the Light’s pixel size.
Performs the following calculation, and then clamps the result to the maximum size:
| Light type: | Formula: | Maxium resolution, in pixels: |
|---|---|---|
| Directional | NextPowerOfTwo (pixel size * shadow quality multiplier * 3.0) | 4096 x 4096 when Shadow Resolution is Very High Quality and/or if the GPU has 512MB or more of RAM, 2048 x 2048 otherwise. |
| Spot Lights | NextPowerOfTwo (pixel size * shadow quality multiplier * 2.0) | 2048 x 2048 if the GPU has 512MB or more of RAM, 1024 x 1024 otherwise. |
| Point Lights | NextPowerOfTwo (pixel size * shadow quality multiplier * 1.0) | 1024 x 1024 if the GPU has 512MB or more of RAM, 512 x 512 otherwise. |
Point lights have a lower limit on size than the other types because they use cubemaps for shadows. This means that six cubemap faces at this resolution must be kept in video memory at once. They are also quite expensive to render, as potential shadow casters might need to be rendered into all six cubemap faces.
In the Built-in Render Pipeline, you can set the resolution of a Light’s shadow map by setting the Light.shadowCustomResolution property to a value greater than 0. When this value is greater than 0, Unity performs the following calculation for all Light types:
NextPowerOfTwo (shadowCustomResolution)
It then clamps the maximum resolution based on Light type and hardware, as shown in the table above.
Note that the shadowCustomResolution property is supported only in the Built-in Render Pipeline. It is not supported in the High Definition Render Pipeline (HDRP) or in the Universal Render Pipeline (URP).