There was a flurry of interest in Microsoft’s Photosynth this week. I’m not sure why, since it’s been around for a while, and was one of the WPF/e showcases at Designertopia last year. The engine for Photosynth is Seadragon (acquired by Microsoft last year I think), explained here in more detail.
Photosynth (or at least it’s primary concept) comes alive when it’s pointed at Flickr. So I was at first mystified as to why the public demos of Photosynth all used photos taken by one person, but the video explains that they were not able to use a Flickr feed for legal reasons.
However, whether or not the photos used are heterogeneous, there is a problem I think. Spatially relating the images is of course very clever, but if we ignore this and look at what it’s like to use the interface, there is clearly a “keyhole” feeling to it. You are, at any one time, simply flicking though similar photos. Despite the occasional panorama that jumps out at you, it is far too easy to become disorientated (even with the homogeneous photos, so I assume even more with the heterogeneous ones). I thought at first that this may have been due to my unfamiliarity with the UI, but I’ve been playing with it quite a bit today, and I still feel as if I’m looking though the wrong end of a telescope while walking on a high-wire. Overall it mainly delivers the same experience as sifting though a stack of photos grouped by place.
There is, however, something of the Bladerunner here. The promise of discovering something hitherto unknown about a place (cf the example in the video using the poster of Notre Dame). It’s all quite intriguing, but I have my doubts about its actual utility.