Mr. Fred Wilson <br />October 5, 2016 <br />Page 3 <br />Applicant's Allegation 4: Vector-Based Diagrams Preferable to Raster Images <br />The issue of raster image vs. vector-based diagram was initially introduced in September, <br />2015, during the first open record period, by the applicant's insistence on using an <br />extremely poor-quality scan of the Metro-Plan Diagram. Their first submission during the <br />2015 zone change application had been presented on vector-based Metro-Plan Diagram <br />images obtained from LCOG's GIS Metro Plan Layer. Only after LHVC presented <br />alternative maps based on the same LCOG GIS layer, did the applicant submit maps using <br />the poor-quality scan of the "adopted" Metro Plan diagram insisting that these were the <br />only acceptable version. Since that time, it has been the applicant who has insisted that <br />ONLY raster-based scanned images of the original Metro-Plan Diagram are <br />admissible! And now they want to complain about the use of raster-based images by <br />LHVC. It was actually LUBA who indicated in their opinion that LCOG's GIS layer <br />would likely result in an acceptable Metro-Plan diagram base map onto which to the place <br />the applicant's property. So the limitation of using raster-based scans of paper copies of <br />the Metro-Plan Diagram remains entirely at the applicant's insistence and not due to any <br />limitations from LHVC, City Staff, or LUBA. <br />Nonetheless, LHVC is at least in partial agreement with the applicant's argument, which <br />postulates that "vector-based" diagrams (which locate points in space by direction and <br />distance from an origin) can be more accurate than pixel-based images. The latter are <br />limited in their accuracy to the size of the pixels. If Oregon were represented by a raster <br />image with pixels 300 miles wide by 300 miles tall, it would not be possible to locate any <br />points within the state, as one pixel would essentially encompass the entire state. The <br />entire United States would be approximately 9 pixels wide, from Washington State's <br />Olympic Peninsula to the eastern tip of Maine, and 5 pixels high, from the northern border <br />of Minnesota to the south tip of Texas. Such an image would clearly not be useful to the <br />question before us. <br />However, the accuracy of raster (pixelated) images is entirely dependent on the size of the <br />pixels used, also called the "resolution". For generating LHVC Sheet 9/2/15-04, LHVC <br />worked on an image of the Metro Plan Diagram (1" = 7,000') at a resolution of 1,200 dpi <br />(dots - or pixels per inch, re-calculated from the original scan). Therefore, on LHVC's <br />Metro Plan Diagram, each pixel represents 7,000 feet/1,200 dots = 5.83 feet per pixel, or <br />each pixel represents an area of 5.83' x 5.83' = 34 ft'/pixel. Once the Metro Plan Diagram <br />had been scaled up to the 1" = 200' scale used by the applicant's maps, the resolution was <br />recalculated to 300 dpi (pixels per inch or 2007300 dpi = 0.667feet per pixel), which does <br />not change the coloration of the scanned Metro Plan Diagram (5.83 feet/dot) but improves <br />the workability of the map and the image quality of the applicant's Sheet ZC-4. <br />As outlined in our submission for the I" Open Record Period (on September 28, 2016), the <br />applicant used a Metro Plan Diagram with a pixel size of 17.5' x 17.5' = 306 ft2/pixel. <br />This means that nine pixels on the Metro Plan Diagram underlying LHVC Sheet 9/2/15-04 <br />fit into one pixel of the applicant's Metro Plan Diagram scan. Along those same lines, it <br />should be pointed out that on their maps (1" = 200'), the applicant represents their property <br />Laurel HillValleyCitizens@gmail.com . 2585Moon Mountain Drive . Eugene, OR 97403 <br />www.lhvc.org <br />