We are currently trying out some features of deep zoom and are stumped by a seemingly simple problem: vertically align a fully zoomed in subimage into the whole frame (so the deepzoomobject's viewportwidth = multiscalesubimage's viewportwidth, with space
leftover on top and bottom of the subimage).
If we set the viewportwidths equal and the viewportorigins equal we kind of have what we want, except for the vertical align, so we need to detract an additional deltaY from the viewportorigin.Y. I think we have misunderstood something really basic about
the coordinate system.