Eye-candy for the map nerd

[Adam White]

Well, I guess this belongs in “On the Web”, because there’s no sub-forum titled “The Intersection of Math Geekery and the Sierra Nevada”.

I’ve had an un-explicable urge to render USGS elevation data with rendering software for a while now. I don’t know where the urge comes from–Google earth already kind of does this. So do other things. But they’re not pretty.

The basic equation is:

I already had the USGS elevation data. The challenge was getting that in Blender, my rendering software of choice (because free. Also, because Python).

That step took a little doing, but eventually I did it. The full-resolution mesh for a single USGS quad is 10,812/8 x 10,812/8 faces, or around 1.8 million mesh elements. Yikes! A million is a big number!

Fortunately, computers are also big these days; although sluggish to manipulate, I was able to import the mesh in full resolution with no optimizations. Hooray for brute force!

I started with a simple one: the Mono Mills topo. Simple, because it’s at the upper right corner of the (-120, 37) (-119, 38) elevation dataset.

I threw on the texture from the 90s era quad, added some lighting, rendered, and… voila!

(if you click on any of the following, they should open in a new window in full-and-obscene resolution)

I wandered over to a place with a real mountains: a Mammoth Mountain!

I can put two together…

…or many (at the cost of mesh resolution–these and the above are reduced by a factor of around 10x in mesh size):

I can take the USGS quad that I’m using as a texture, and export it to a shader map, then add that to the model such that the water has a specular reflection:

I can visit an old friend, like Mt. Ritter:

Then, because I never know How Far Is Too Far, I can create an animation, with the camera flying around Mt. Ritter and the minarets with moving sun-like and moon-like lighting (now with more alpenglow!):

(Realists may note that my sun rises and falls in the wrong directions. I noted this too–after 40 hours of rendering. You get what you pay for.)

In summary: neat! Not particularly useful for route-planning. But useful to stare at, useful to ogle. Perhaps useful to make a pretty awesome wall-poster out of.

Please enjoy! And if you think of a way that this can make me instantly wealthy, please notify me. My endeavors seem to always lack this trait.

[Ken Thompson]

Very cool!

[Alex H]

Beautiful!

[Jennifer Mitol]

AWESOME!!!!!

[Simon Kenton]

I wish you owned a 3D printer.

[Garrett McLarty]

Adam,

Love your maps.   I have been trying to figure out a way to have something like this on my phone to use offline when on a hike.  Here’s a couple screen shots, definitely not as nice as what you have.  These  were done with oruxmaps, and caltopo data.   Along with oruxmaps download of the dem files for elevation.

[Garrett McLarty]

A few more, first two using USGS topo data and then a one more with caltopo data.  Again, not nearly as nice as the large renders you have but these take 5 min in oruxmaps and can be used offline in the field.   One my phone, i can pan, tilt, and zoom these as well as place and view tracks and waypoints.    The last one has a larger dem area than my map overlay, oops:)

Also, BPL limits uplaod size to 1mb, so details arent so good.

[Alex Wallace]

Very cool, indeed. Thanks for sharing.

@Paul, “I wish you owned a 3D printer.” Have you seen this?

Nice Trails hiking memories 3D Printed mountains.

[Steofan M]

Adam, the fuzzy google maps have been useful but the clarity of yours are just amazing!  Thanks for lighting the fuse on this math geekery!

Garrett, I have just begun to dig into oruxmaps and the offline potential is huge!

And to think that I was once held spellbound by a fancy feast stove on the old BLP site… you two have just kicked the door wide open.

[Greg Mihalik]

“Realists may note that my sun rises and falls in the wrong directions. ”

But … the first cartographers rendering three dimensions by hand were the Swiss. They accomplished their magic with various forms of hash marks – as if the sun were shining from the upper left of the view.  Culturally it continued that way, so we are “conditioned” to understand map renderings with a northwest sun. This is not the case for modern architectural renderings, and our brains realize the difference.

If you take one of your renderings and turn it upside down, locating the sun in the southeast,  you will “see” valleys instead of ridges.  You can change your sun angle to be “correct” but people will be confused by what they see.  Bizarre, but very interesting.

[Author]

Posts