Physics + fashion from @scanlabprojects

Die Traumdeutung

ScanLAB Projects and Brigitte Stepputtis, Head of Couture at Vivienne Westwood, produced this video piece following a recent fashion shoot with one rule - no cameras. The shoot instead used terrestrial laser scanning to capture the surface of the models, textiles and architecture with forensic precision. This collection of over a hundred million perfectly measured points form a frozen moment in time. The figures, garments and crumbling building remain, monumental but revealing their own fragility by sometimes fading away. The figures are frames by their own data shadows, voids of emptiness in a perfectly recorded world.”

(Source: vimeo.com)

Peter Richardson's Going Into Detail | edgeca.se

"Mapmakers have long attempted to use various geographical features to define Europe’s eastern border for their own purposes, citing rivers, mountains, and so on. These days, the Urals are most often given as the boundary; this is partly because they happen to be the unusually well-preserved result of two sub-continents running into each other, and this carries a lot of weight with people who base arguments on precedent. But it’s also because it allows Moscow to be European without defining all of Russia as part of Europe, the prospect of which irritates cartographers."

Peter Richardson's essays on relief maps, local context, and information


alt

Wired: A Cloudless Atlas — How MapBox Aims to Make the World’s ‘Most Beautiful Map’, by Tim Maly

I really love Tim Maly’s account of Mapbox’s process. It shows us the power of combining computation power with human creativity in the production of perception. Even maps, are socially and mathematically constructed. Here’s a large chunk from his article, which should be read in its entirety! 

"Zoom in close and you’ll start to find more strange things. Seams sometimes appear. In some places the colour of the landscape changes dramatically. In other places the land is blurry like it was shot with a webcam.

Blue Marble NG July v. Cloudless Atlas

This is the image that got Charlie Loyd his job at MapBox. On the left is the northern tip of Greenland based on NASA’s Blue Marble project as it appears in Google Maps. On the right is a prototype output from his algorithms.

These oddities are reminders that the maps we see are images that have been stitched together, often from a variety of sources. This is a big data problem, and it often results in errors and other imperfections. For some people, finding these glitches is a hobby. For MapBox, eliminating them is a mission. This is a behind the scenes look at how that’s being done.

MapBox grew out of a personal need for better custom mapping tools, says CEO Eric Gundersen. The company began as a series of open source projects to work with OpenStreetMaps’ data. It shifted into a business when it became clear that other people would need similar services. The opportunity to make a big business out of it came when Google started charging for access to the Maps API. Today, MapBox powers the maps of services like Foursquare and Evernote, and their past clients include NPR, The Guardian, Greenpeace and the FCC. “Where the map is a central component to the presence, that’s where we’re fitting in,” he says.

Until recently, MapBox’s maps were drawings made from OpenStreetMap vector data, says Gundersen. Clients could modify these maps and add data using the open source design studio TileMill and a CSS-like language called CartoCSS to customize the look and feel and use them as data visualization tools, or simply nicely branded maps. In December 2012, led by data analyst Chris Herwig, MapBox released its first version of a satellite imagery layer. Loyd joined the team to help perfect their output.

Raw MODIS Terra imagery stitched together

This is the raw material of a beautiful map. This composite from MODIS Terra show the world captured on March 28th, 2013. It’s covered in clouds, there are strips where the satellite didn’t scan, and light areas around the equator where the sun glints. Images: NASA LANCE-MODIS.

How do you go from the chaotic data that a satellite captures to the beautiful idealized images that MapBox is producing now? Pixel by pixel.

MapBox begins with public domain data provided by NASA’s LANCE-MODIS data system. The images come from a pair of satellites called Terra and Aqua which have been orbiting the planet since 1999 and 2002, respectively. They capture data at a wide variety of wavelengths including the visual field. This is what MapBox is using.

“For the new release we’re processing two years of imagery, captured from January 1, 2011 through December 31, 2012,” says Loyd, “this amounts to over 339,000 16-megapixel+ satellite images, totaling more than 5,687,476,224,000 pixels. We boil these down to a mere 5 billion or so.”

The first problem is even getting the data. It’s all available in the public domain, but just transferring it over to MapBox’s servers was a major task because of the volume. To do this render, they needed to download two thirds of a terabyte of compressed data. “We’ve got 30 to 40 servers pulling down data from NASA,” says Herwig. “We called them up and said, ‘hey we’re going to hit you hard, what’s the best way we can do it for you?’”

NASA worked with the team to ensure that there was a way for them to grab the data from their servers without overwhelming the network. “Hats off to NASA for putting this out there,” says Gundersen. “When it comes to open government, there’s all this talk of APIs. What we really need is government infrastructure for bulk download.”

Once the image data is in MapBox’s hands, the problem is sifting through those images to filter out the clouds, sun glints and atmospheric haze to get a clear image of the ground.

Normally, the approach here would be to find each region’s clearest days and quilt them together. “Unfortunately, this leaves seams,” writes Loyd. “Adjacent images may clash (for example, if they’re from different seasons) and draw attention to the base layer in a way that a mapper rarely wants.”

To solve the problem, MapBox takes a much finer-grained approach. It takes all the images it has of an area and stacks them on top of each other. Then, it reorders each column of pixels in the stack based on how cloudy it thinks it is. “We do that for every pixel in the world,” says Loyd.

Making a Cloudless Atlas, Step 1. These are the input images for 2012 for a small region of the world. If you look closely you might be able to tell where. “As you can see, there are only a couple days that are mostly clear for the whole region over the entire year, and if you zoomed in you would see local clouds even in them,” says Loyd.

Once MapBox has reordered the pixels, it takes the average of the least cloudy ones, and that average becomes the canonical pixel for that particular spot on the map. The scale is dizzying. Loyd says that when he and his team were about 40 percent of the way through the job, he calculated that if they printed out their work to that point, it’d cover 2 acres of land at 300dpi.

Making a Cloudless Atlas, Step 2. Once it has the images, MapBox’s algorithms go through and sort them pixel-by-pixel from darkest to lightest. The terrain begin to separate itself from the clouds, and you should now be able to recognize what you are seeing.

MapBox has to pull some other tricks too. The color of the landscape changes throughout the year as summer green leaves turn to oranges in the fall, then snow falls in the winter, then new growth returns in the spring. Average all that together and you’d get a muddy brown. So the team uses some techniques to ensure that they’re capturing peak growth, which is May/June in the northern hemisphere and December/January in the southern. In addition, because the process favors darker pixels, the first output can seem very dim and underexposed, says Loyd.

“It’s a completely natural product,” says Loyd. “Every pixel is a real pixel captured by an camera in the sky. But it’s also completely synthetic.” The goal for the map is to capture roughly what the naked eye can see from space, but for an idealized cloudless planet trapped in eternal summer. “Our goal is to make the most beautiful map,” says Gundersen.

“It’s a balancing act between wanting to be accurate and do right by data and making it look like everyone thinks the earth looks,” says Loyd.”

(via This Is What Informal Transit Looks Like When You Actually Map It - Emily Badger - The Atlantic Cities)

(via This Is What Informal Transit Looks Like When You Actually Map It - Emily Badger - The Atlantic Cities)

How Uber's Taxi App is Changing Cities - Derek Thompson - The Atlantic

In a wide-ranging interview with James Fallows of The Atlantic, Kalanick introduced his company as part of a new wave of tangible technology that is changing urban policy and city protectionism, starting with the taxi lobby. In Washington, he said, a late-night attempt to pass a law to effectively ban Uber prompted a voracious social media response, including 37,000 tweets, which eventually defeated the so-called Uber Amendment. The experience created a “playbook” that Kalanick is taking across the country, and overseas, as he fights to popularize his app, which connects wannabe passengers with on-demand drivers.

‘Geography of Hate’ Maps out Racism on Twitter
(via ‘Geography of Hate’ Maps out Racism on Twitter - COLORLINES)

‘Geography of Hate’ Maps out Racism on Twitter

(via ‘Geography of Hate’ Maps out Racism on Twitter - COLORLINES)

roomthily:

Zeitgeist Google Borders - mapping the borders of google’s autocomplete (note: the site is disabled now)

roomthily:

Zeitgeist Google Borders - mapping the borders of google’s autocomplete (note: the site is disabled now)

Reblogged from roomthily with 4 notes

@metaActivism - How “Slacktivism” Revealed a New Political Map of America

The data scientists at Facebook – particularly Eytan Bakshy – have produced an excellent set of public analytics on Human Rights Campaign‘s equality avatar initiative, which some have called slacktivism.  Facebook’s full report is here.  Of the many graphics Facebook produced, the one below particularly caught my eye:

It reveals the likelihood that a person in a given US county would change their profile pic to HRC’s red and pink equality symbol.  The darker the shade of red, the more Facebook users in the county were likely to adopt the equality symbol.  Of course, seeing that map made me think of this map:

The map above, created by The Washington Post, shows 2012 voting behavior by county.  Here, the strength of  the red or blue tone of the county indicates the strength of the Republican or Democratic win in that county.

First of all, I love that the Facebook data people, perhaps unconsciously, used the image of the red county to represent strong support for a socially liberal cause instead of strong support for a socially conservative party.

But what I really love is how the map shows that Americans are a lot more tolerant and liberal than electoral maps indicate.  Based on the electoral maps we have all seen so often, we think of the US as having liberal coasts and cities and a conservative “heartland.”  The Facebook map of avatar changes doesn’t show such clear geographic distinctions.  Though the South is a notably paler than the rest of the country, the Southwest, West, Northwest, and Great Lakes region are all pretty rosy.

Let’s take Wyoming as an example.  Looking at the electoral map, we see a deep red state with a blob of dark blue on the western edge, along the border with Idaho.  Looking at the Facebook map, we again see this area is darker red than the other Wyoming counties, but the darkest red county is actually in the southeast corner of the state, and there are a number of other counties that are distinctively pink.  None of that support for progressivism is captured in the electoral map.

So what does this mean?  Clearly this is a map of one particular instance of digital activism, and it is a model, not a count of actual profile changes.  Still it reveals that Americans all over the country support marriage equality and that the red and blue of electoral differences fails to capture the subtle distinctions of political opinion.  Though the man in Cheyenne may vote Republican, he may also have gay co-worker whom he goes out to lunch with every day.

Thanks to Facebook for mining their data on digital activism, and for making the results public.

Astonishing Map Of Tweets By Language In New York

“Twitter NYC” gives you a bird’s–eye view of the multilingual conversation — Spanish, Portugese, Japanese, Russian, Korean, Turkish, Arabic and Italian — happening across the five boroughs.”

NYC

Map by James Cheshire, Ed Manley, John Barratt, and Oliver O’Brien.