User:Northwesterner1/photomaps

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Route Maps: A How-To[edit]

I've been experimenting with making photo maps (examples below), and I'm posting this to keep track of my methods, as well as to assist anyone who wants to edit my maps or make similar maps using a consistent style. The maps are displayed in articles using Template:Wide image, which is designed for panorama photos, but I've found that it works well for a certain type of linear photo map. The key innovation here is splitting the photo map into panels and orienting each panel separately, so that your subject can be represented as a horizontal line.

The particular method I've devised is suitable for urban creeks in the United States with a length of 10 to 30 miles:

  • I'm especially interested in urban creeks because they are often not shown on widely available sources (city road maps, Google Maps). They are usually mapped extensively in urban and regional planning documents, but those documents can be hard to find and often have too much detail to be useful for general readers of the encyclopedia. I'm trying to make maps that would enable regular folks to find access points to urban creeks for walking, nature study, restoration, etc.
  • Anything longer than about 20 or 30 miles has too many panels at this scale, and the image size becomes a problem. You could, of course, change the scale if you want to experiment with other types of creeks. But then you lose some street detail.
  • To follow my methods, the creek needs to be within the United States urban areas mapped by the color USGS Digital Ortho satellite survey at a .25 meter resolution. For rural creeks, you would need to use the b&w USGS satellite photos, which are at 1 meter resolution.

This is just what I've worked out for urban creeks. You could adapt these methods to anything that you want to map along a line (highways, rivers, etc.). However, I'm not sure if photo maps would be useful for highways and rivers, which are easy to find on existing maps, and at a large scale you lose most of the relevant detail. I can imagine many other uses, though. (Lewis and Clark expedition, anyone?)

Methods[edit]

Scale: I find it useful to work in Photoshop with an image that is 600 pixels high. At the end of the process, I downsize the image to 300 pixels, which looks good in an article. If your subject allows a narrower map, try working in a document of 500 pixels and downsizing to 250 (or 400/200). Keeping the file size small is important to help with page load times.

  1. I use NASA World Wind. It's kind of like Google Earth, but it's free and open-source and it contains a lot of public domain imagery that can be used on Wikipedia. Some of the data layers are copyrighted and cannot be used commercially. But the most broadly useful layers -- NASA Landsat 7 imagery, USGS 1 meter photos, USGS .25 meter urban area photos, and USGS topo maps -- are in the public domain.
  2. Unless you are attempting a 3D model, turn off your vertical exaggeration (View > vertical exaggeration > 0.0x). You probably also want to turn off "motion momentum," "planet inertia," and "point go-to" under the View menu.
  3. Keep a constant altitude. I find that for urban creeks of this length, 2500-3500 m is a good level. (Lower altitude works for shorter and curvier creeks; straighter and longer creeks require a higher altitude to reduce the number of panels in the composite image). I copy the coordinates (Edit > Copy Coordinates) to a notepad file so that I can paste them back into World Wind if I lose my bearings and need to return to the same altitude easily (Edit > Paste Coordinates). You also need to make sure your tilt is constant. I've been using a 0-degree tilt (directly overhead angle) but I can imagine other uses.
  4. Turn on the USGS Digital Ortho layer. Find your first point of interest, arranging the subject horizontally across the screen. Wait until all the tiles have downloaded.
  5. Take a screenshot. (I use Windows Vista's "snipping tool", but there are many freely available programs that do the same thing, such as Gadwin Printscreen.)
  6. Open a new photoshop document and size it to 1000 x 600 pixels.
  7. Paste your screenshot in photoshop and align it how you want it.
  8. Download this compass image. Copy and paste it into your Photoshop document as a new layer. Rotate it in the proper direction. If you turn on the "3D Compass" option in World Wind, you can get a numerical reading. Then it's easy to rotate the compass in Photoshop (Edit > transform > rotate > type value in toolbar at top). I put the compass in the top left corner of the panel.
  9. In World Wind, switch to the USGS Topo Map of the same area. By switching back and forth between the satellite image and the topo map, you can get an idea of where you need to draw your creek.
  10. Open a new layer in photoshop (Layer > New). Add the main stem of the creek and major tributaries. I draw them freehand with the paintbrush tool. Diameter 10 pixels. Opacity 100%. Flow 25%. The main stem is #ff0000 (red) and the tributaries are #ffffcc (pale yellow). As a general rule, I indicate major tributaries, but not the intermittent streams that are shown on the USGS topo map.
  11. After I've drawn the creek, I zoom out to 50%, which makes the document easy to manage as I add points of interest.
  12. Text on my maps is #ffffcc (pale yellow), Calibri font. Creek tributaries are bold italic 30 pt. Cities are indicated in parentheses, bold, 30 pt. All other text is bold, 25 pt. I've been adding major street intersections, parks, natural areas, neighborhoods, bike trails, railroads, and occasionally other landmarks. I look for points of interest on Google Maps, city maps, and also by turning on the "place names" layer in World Wind. Planning documents will often contain the names of smaller tributaries and parks.
  13. Tributaries and streets are rotated as appropriate (Edit > transform > rotate). Generally, parks, neighborhoods, etc., are not.
  14. Merge all your text layers into one layer. (Layer > Merge Layers)
  15. You should now have four layers: your background satellite photo, the creek route, the compass, and the text. For the top three layers, I like to add a drop shadow so they stand out a little from the background. (Layer > Layer Style > Drop Shadow.) I just use the default drop shadow settings. It's a subtle thing that helps the maps look better and makes them easier to read. In my first maps, I didn't add a drop shadow to the creek route, but I'm going to start doing that in the future.
  16. Now you've finished Panel 1. Find your next point of interest in World Wind (rotating the perspective so that you can fit as much of your subject as possible in a horizontal frame), take another screenshot, and make Panel 2 in a new photoshop document.
  17. At the end, you take all your panels and paste them into one document with a little white space in between.

Northwesterner1 (talk) 07:05, 4 May 2008 (UTC)

Image Size[edit]

I'm still experimenting with the best file size as a compromise between image clarity and download time. After discussion at Talk: Johnson Creek (Willamette River), I recommend uploading three versions to Wikimedia commons.

  1. High-resolution version for printing. Cut the image size to 75%, so that it is 450 pixels high. (Image > Image size). Then flatten the image (Layer > Flatten Image). Then save as .png. Ideally, this file should be 5 to 10 MB.
  2. Medium-resolution version for online reference. Cut the image size to 50%, so that it is 300 pixels high. "Save for web" as PNG-8 or PNG-24, depending on how big your image is (File > Save for Web). Ideally, this file should be 1 to 2 MB.
  3. Low-resolution version for display in articles. Cut the image size to 25%, so that it is 150 pixels high. "Save for web" as JPG at 60 quality. Ideally, this file should be 50-120 KB.

All three files should be linked to one another in the image summaries. Also, a preview of the medium-resolution version should be included on the low-resolution and medium-resolution image pages. The wide image template code is {{Wide image|Image (medium).png|DISPLAYWIDTHpx|Caption}}. Readers who click on the low-res version in the article will be taken to the low-res image page, where they will see a preview of the medium-res image. Readers who click on "zoom in" in the caption will be taken to the medium-res page, where they will see the same thing. From either of those places, they can click to the high-res version for printing.

For an example of how to format your image summaries, see

See below for examples of how the "zoom" works.

Northwesterner1 (talk) 09:55, 8 May 2008 (UTC)

Examples[edit]

Low Res (displayed in article)[edit]

Map of Johnson Creek. The route has been artificially straightened by orienting each of the panels differently. The compass rose marks north for each. (Zoom in)

Medium Res (displayed on image page when the reader clicks to see image)[edit]

Map of Johnson Creek. The route has been artificially straightened by orienting each of the panels differently. (The compass rose marks north for each.)

Articles using these photo maps[edit]