We estimate this by taking the difference in size between the current position and the worst-case reachable position after T.
Again, we assume textures are still required, and incremental transmission is assumed not very effective because of infrequent updates, so JPEG coding.
For the polygons we assume perfect incremental coding, so we use
PolygonSimpN[r,d2,DI,θ,k] - PolygonSimpN[r,d,DI,θ,k]
![[Graphics:../Images/bytesize_gr_47.gif]](../Images/bytesize_gr_47.gif)
![[Graphics:../Images/bytesize_gr_48.gif]](../Images/bytesize_gr_48.gif)
![[Graphics:../Images/bytesize_gr_49.gif]](../Images/bytesize_gr_49.gif)