Archive

Posts Tagged ‘Animation’

Tours and 3D animation

February 19th, 2009 9 comments

Looking around the various tech reviews of the new Google Earth 5, one new feature seems to have slipped under the radar – the KML tags describing the new touring function include a seriously powerful method for defining smooth motion 3d animation.

I’ve been messing around with this a little, and hacked together a few basic (not much KML code), slightly weird 😉 , but dramatic examples. These are tours – view them by loading them into GE and double clicking on play me! in the side panel.

Fly wingman with a 3d model plane (also see the vid above)

Polygon morph France into Spain!

Dot-to-dot tour of European cities

How do these work ?

With the new <gx:AnimatedUpdate>. A container for the <Update> element. Which allows you to temporarily modify KML, and specify a length of time for that action to take place over.

Switching features on and off during a tour is the obvious example. Not so apparent, because it’s not directly available through tour recording inside Google Earth, is that for many other changes GE will smoothly interpolate (tween) the difference.

The example Google provide gradually increases the size of a placemark. But this effect can be applied to any KML attribute taking a range of values – orientation, colour, scale and coordinates. Significantly, these transitions will work in parallel, sequentially, across multiple features and alongside scripted camera movements.

This is powerful stuff! A much more convenient and concise method of animation than time slider hacks, where every intermediate step needs coding in KML. Only real downside in all this – and I guess it could be a biggie depending on your needs – you can’t fly around at will during tour playback.

About the examples

  • The first uses a single <gx:AnimatedUpdate> to gradually change location of a 3d model over a 60 second duration. Code for this below the fold if anyone wants to check it out.
  • Second does the same to transform all 38 coordinates in a polygon.
  • Last example combines a series of linestring coordinate transformations with color and scale changes for placemark icons.

Read more…

When two satellites collide, in Google Earth

February 12th, 2009 No comments

Satellites Colliding in Google Earth

Satellites Colliding in Google Earth

This is a quick Google Earth timeline animation of the recent satellite collision between the Russian Cosmos 2251 and the US iridium 33.

Update 2009-02-14: Two ways to view this now.
– A Google Earth browser plug-in version
– Or the original KML file for regular Google Earth below

Cosmos 2251 and Iridium 33 Collide

Use the time slider at the top of the screen to play the file back. The animation tracks the objects during the 6 minutes leading up to impact. Also included are the orbital paths (ground tracks) of the two objects between 16:00 and 17:00 GMT on February 10th.

Location data is based on the orbital elements for the two satellites, from Celestrak.

Visualizing Twitter Activity Inside the Google Earth Plugin

February 9th, 2009 No comments

Twitter, Google Earth PluginHere’s a browser based version of the (Feb 2009) UK Snowstorm Twitter activity, animated inside the Google Earth Plugin.

This mashes up 3800 Tweets over the course of a week. Users posting their location and the amount of snowfall they were seeing on a scale of 0 to 10. You can see how Twittersphere activity peaks on Monday February 2nd, the day the country ground to a halt. With sporadic outbursts every day over the following week.

Embedded, below the fold.
Read more…

#uksnow Twitter Animation – Google Earth 5

February 3rd, 2009 8 comments

A much more significant event than the release of Google Earth 5.0 yesterday was the fact the much of Britain (particularly S. England) was covered by a sprinkling of snow. 😉

As a knock-on of this there’s been a large amount of media buzz about the use of Twitter as a mob-sourced geo-mashup generator to show snowfall.

Twitter users have been posting the first half of their postcode, plus the amount of snow in their location. 0 = no snow, 10 = bucket loads. ie: “#uksnow W12 9/10”

Developers have been mashing-up that data. Check out Ben Marsh’s neat, live Google Map mashup of this: #uksnow Tweets

Google Earth 5.0, Twitter uksnow

#uksnow Twitter Animation, Google Earth 5

Now that the flurry (sorry) of activity has died down, I’ve trawled the Twitter API for the past few days of #uksnow hash tags (2200 so far) and put them into a KML time animation tailored to run in Google Earth 5:

#uksnow Tweet Timeline (40kB)

Download the kml file, and use the time slider at the top left of the screen to play it back.

Couple of new techy things to notice if you have the new Google Earth version.

  • I’ve used the KML gx:TimeSpan element in my Document level lookAt so that by default the time slider displays one hours worth of Tweets. Notice how GE flys back in time after the KML loads.
  • The icons expanding and contracting during playback is a built-in feature of the new version of GE.

Growth of Wal-Mart, Google Earth Animation

January 27th, 2009 No comments

Here’s a Google Earth time animation showing 3176 Walmart store openings in the USA, from 1962 to 2006. See how the business started out with a single retail location in Arkansas, then spread at an accelerating rate across different regions.

Video showing the animation running in Google Earth:

Grab the KML file for Google Earth here:

Walmart 1962-2006 (300kB)

  • Once it’s loaded, use the time control slider top of the main Google Earth window to play.
  • Because this is Google Earth, you can fly around and zoom in and out during the animation.

For anyone thinking this visualization looks vaguely familiar – hands-up, it’s not my idea, the animation is inspired by FlowingData‘s very slick, slippy-map animation of the same data, released last year, here.

To replicate that I’ve used the same KML carpet bombing technique as the flight path animations I knocked together a while back – each time a placemark appears I rapidly cycle through a sequence of Icons of decreasing size, to provide some visual punch.

Note: The location data isn’t all that precise in this – I think it’s just based on zip code area locations, not street addresses. But the stores generally aren’t too far away from the placemarks – look for the massive grey buildings with a zillion cars parked out front.