Random Etc. Notes to self. Work, play, and the rest.

Posts Tagged ‘Google-Earth’

The one with crazy 3D bar charts coming out of the Earth

Over the last few weeks I've been collecting circular visualisations. This week, I seem to keep running into 3D globes and heatmaps. I'm all for 3D done right, but even with my fondness for circular visualisations I'm still wary of these things. (Of course I'm not denying the eye candy appeal of any of it!).

First up, Twingly, which I think is a Swedish blog aggregator/search site of sorts, but I came across it as a screensaver video on YouTube:

Twingly blogoshpere visualisation

The person who posted it to YouTube helpfully describes the blogosphere for those of us not in the inner circle:

"The blogosphere is the total sum of all blogs connected into a social network. The term was cool a year ago but is too widespread now for the general blog crowd to use it. But since it's actually a useful term it is still referred to by the inner circle. From there it will work it's way back into the common language, acheiving a renaissance around febtuary next year."

So that's cleared that up then. Next up, an academic one, G-Econ from Yale (via Things Magazine), which plots geographically based economic data:

G-Econ from Yale

I admire their thoroughness in doing the whole world (check the site for country by country breakdowns), and their multi-megabyte eye candy movies. It's a shame it's all based on a GDP-like measure, which isn't the most intuitive or easy to visualise thing itself. I'm reading their papers now to see what the story is.

Lastly, I'm really pleased Dan Catt over at Flickr/Geobloggers can't resist plotting his interestingness heatmaps in 3D inside Google Earth. When the sky goes pink you know it's because Yahoo's Dubai office decided to build it for real.

Early Flickr Google Earth heatmap

Update: Eric found this one from ESRI:

Results 2004

Google Earth London Update

Google Earth just updated its imagery, bringing London up to date with 10cm resolution photography from Blue Sky2lmc did the shadow-forensics to date it around February of this year, so it's a crisp winter morning.

One artefact of Google's variable coverage is the patchwork effect their updates have on the world.  The UK has strips of photography from this provider or that provider, but some of the coverage remains plain old 15m landsat.  Ironically some of the low-resolution stuff actually looks the best when viewed at a distance.  The high-resolution stuff has been over-processed and resampled beyond recognition, and that's a real shame.

Greater London is particularly bad at the moment, coming out a muddy brown that looks like the result of aggressive high-pass filtering (useful for making tiling textures in games, and apparently fpr getting aerial photos to stitch together).

Coming in closer you realise the Thames is completely obliterated.

The only reprieve is that close up the detail is fantastic, and these shadows on the Thames are gorgeous, like architectural elevation drawings.  See also the amazing crowds in Trafalgar Square.

Perhaps the resampling system could be tweaked to fix the brown patch though.  I mean, in some ways it reflects reality (London is a horrible confusing mish-mash until you concentrate on one particular area for a while), but in other ways it rather spoils the illusion of a seamless view from space.  Of course there are no seams if everything is one colour!

Software that speaks for itself, or: are you sure you need Powerpoint?

Euro Foo began with introductions to the group: name, affiliation and three words (tags) that describe you. So tough! I picked simulation, architecture and design (I think), but that was way too narrow. Conference introductions are tricky beasts... Aaron Swartz asks "what have you been thinking about?" and Simon Willison asks "what are you excited about?". That's better!

I've been thinking about software that - for the content it generates - is a better presentation tool than Keynote or PowerPoint. Stop, wait, come back! This isn't a Powerpoint bashing post. Of course it will depend what you are presenting and who is doing the presenting. Powerpoint and Keynote are great tools in the right hands, but I'm coming to believe that they often involve doing work for presentation's sake, when you already have good presentation tools in your workflow or - more importantly - when your work could speak for itself.

A colleague once talked me through a detailed 3D model that was built in Sketchup, using Sketchup's views and geometry tools to show things from different angles and manipulate the data as she spoke. Stills in Powerpoint would never have done this justice. Animations would have to be scripted to perfection. She had constructed a solid narrative inside the software, and the ability to manipulate things and show/hide different components came for free inside Sketchup itself. The same narrative could have been retold in a slideshow format, but why create extra work for a less effective result?

In the case of Sketchup, it's easy to see why the CAD software says more about the design than presentation software does - in many cases the Sketchup model is the design. The same thing happens with software like Photoshop, where stepping through the layers and changing filters or moving things around can often tell a better story than a selection of stills.

Google Earth is a great example of software that speaks for itself. With a bit of practice and a little planning you can make a presentation within the software that is far richer and more persuasive than out of context screen shots. Likewise with Excel, almost the entire point of the software is that everything is out there for everyone to see. And a live spreadsheet has currency too - you can pass it around - just like KML in Google Earth. Sure you can pass around a presentation or report with charts and tables to explain your analysis, but the live spreadsheet data is the analysis, and there are a few simple tools (like Juice Analytics' slider for Charts) that make the analysis even more accessible.

My own work in architectural simulation arose out of a dislike of "black box" models, where assumptions and specifications would be prepared and then a simulation would be run by an external agency who would deliver their findings in a report. Our reaction to this was a granular, individual-based model presented in a dynamic interactive way - not sure about a particular result? Interrogate the model to find out what went wrong. The result of the model was an experiential way to understand a particular scenario, a story telling tool. Just like most of the software you have on your desktop.

Perhaps people do want just the answer, and not the journey, but it's still worth asking: are you sure you need Powerpoint?