Dreadful timing for me, because I really want to be honing the presentation I’m doing later today – but I was up late looking at this and it still feels exiting.
So here’s a problem that’s always bugged photographers: You can only get so much in a photo. The camera is at one point in 3D space, aimed and focused on another point in 3D space, and has a given angle of view. Now … you can keep the camera at the same place and take pictures aimed at different points then stitch the resulting pictures together to form a Panorama (I’ve been pleasantly surprised what a good Digital Image Suite does stitching pictures from my phone camera) and for some time we have had tools which let you move a computer viewpoint around a large panorama. But really that only let you over come the limitation of looking at a computer screen instead of standing in front of a print. You can move the camera around a single object and stich the results together to get what I call a “peel” – which is trickier than a Panorama, but are trying to produce a continous photograph which gives you a single impression of a bigger scene.
The painter David Hockney experimented with making collages of photographs which he called “Joiners”. A joiner isn’t smooth,its lot of individual impressions which you also give you a single impression of a whole – even if that couldn’t be seen from one view point,or different parts existed at a different times.
Although you can make Joiners and Panoramas on the computer you’re not doing anything which exploits the computer as a medium – or in less pompus terms, the computer overcomes the limitation of the frame size of your camera, and it replaces pasting up prints to form a collage, but it’s not a “multimedia experience” – letting you pan and zoom round an image which is too big for your screen doesn’t count, sorry. Cue Photosynth. It’s such a visual experience that I’m not sure I can verbalize it properly; but let me try. It takes a lot of photos of the same place (one of the Demos is St Mark’s in Venice) and relates them together. The pictures can be taken from different points around the square, with different angles of view; so we couldn’t make a panorama out of them; but photosynth works out where the camera was in 3D space and where the parts of the images are, in 3D space, and which pictures overlap with each other, so that you can look around the scene and get all those different impressions linked to each other so that they have a kind of context. It’s not virtual reality (which is a relief) but like VR it doesn’t have any kind of experience away from the interactive one on the computer… You have to try it, but be warned: it’s a beta and currently it only supports recent Microsoft Browsers on Recent OS’s – it needs a 5MB download to install the active X control. But it’s worth it. Go there now !
Update: I had a chance to watch one of the Videos where they show this stuff in action and cover it’s history a bit. Rick Szeliski from Microsoft Research talks about this as a new medium in much the way I was thinking about it above. Also look out for the book of Hockney pictures about 3 minutes in.