Good call. Looking into this further, looks like you're right.[1] For the Mercator projection, translate, scale, and rotate are sufficient for the georectific

Good call. Looking into this further, looks like you're right.[1] For the Mercator projection, translate, scale, and rotate are sufficient for the georectification of small images (read: floorplans). You'd need more of the affine toolbox if you used something like Mollweide (mostly sheer, right?), and then you might need non-affine stuff to deal with crazier projections. I guess, then, that begs the question of why the Google floorplan submission UI uses that weird double-pane thing? [1] http://blogs.esri.com/esri/arcgis/2011/03/24/tissot-s-indicatrix-helps-illustrate-map-projection-distortion/ http://blogs.esri.com/esri/arcgis/2011/03/24/tissot-s-indica...