Why can't we have combined automatic vertical+metadata mode? So if metadata produces slanted model i could fall back to automatic vertical heights? Now it seems that vertical position is locked with horizontal so if one is less accurate it also moves other dimension.
↧