Hi folks,

Imagine we have 8 different light sources in our scene and want dynamic shadow map for each of them. The question is how do we generate shadow maps? Do we render the scene for each to get the depth data? If so, how about performance? Do we deal with the performance issues just by applying general methods (e.g. frustum culling)?

Thanks,

Additionally to frustum culling you can use the light cut off to cull anything too far away.

If the light is static you can precompute depth for static scene and render only dynamic objects.

You can update shadow maps only each Nth frame, risking some self shadowing on moving objects.

You can do shadows for dynamic objects only near the camera, and far away use only static shadows.

So, nothing magic - it is like you expected

Edit:

Actually there is a lot attention spent to the idea of a voxelized scene representation converted to signed distance field. You need to voxelize only once to support any number of lights, but the visibility test becomes ray tracing instead of a quick texture fetch.