Smile – you’re live on Flitter

Last year I started a project called Datama – using vvvv as a playback device for remixing media on the web. There’s been no time to push on with it, but this morning I rolled up the sleeves and started to remember the vvvvay forward.

Last week I did a few petty hacks on Yahoo!Pipes, creating mashups with twitter, wordie, 43things with youtube, then using Wikidot to render the results. You can see them here: TwitterTube WordieTube 43Tubes. (You might need to refresh the pages a few times as the wikidot page renders faster than Pipes can serve up the feed.)

So now we have Flitter (beta, of course). A mashup between Twitter and Flickr: Using Yahoo!Pipes to aggregate the lastest Twitterings of the public timeline, then it matches up these texts with Flickr image descriptions. This creates an RSS feed which I pull into VVVV running on my local machine, which basically creates a slide show of the resulting matches.

Not the most fascinating animation, but the role of Twitter to edit/select/sequence the images from Flickr is what I’m interested in here. Unlike the inspirational Ad Generator, which is a random matching of image and text, or We Feel Fine, which creates a soup of matches, Flitter aims at creating media that reflects what is happening as close to realtime as possible. Not as an ontology of the present, but towards the significance of relationships within the audiences ‘loose couplings’ . Less the uncanny,more like a live semantic broadcasting.

If you want to use the feed from Pipes – it’s here. If you want the vvvv patch it’s here. (vvvv is Windows only but free to use.) Because there isn’t always perfect matches between Twitter and Flickr (mainly because of Japanese character sets) you get nothing. Bug or feature? Heck – it’s a beta..

Here’s a couple of quick video renders to see the results.