momodot, scaling the shadow cams' view can be much more efficient in controlling the resolution and edge feathering of the shadows than is increasing the shadow map size. Also saves memory and render time. Keep in mind that to get rid of jaggy shadows, you just want more shadowmap pixels defining the shadow edges. So zooming the shadowcams in (scaling) can often do this much more effectively, and in much higher multiples, than using a larger, memory hogging shadowmap.