Noel, if you read the entire 'Behind All the Buzz: Deblur Sneak Peek' blog post, it rather deflates a lot of the awe and wonder the Adobe Max demonstration initially inspired. In fact the full size shots of the before and after Plaza image are even more incredible, and the detail gained is barely short of magic!
However the article makes it clear that the Plaza shot was synthetically applied and a ‘simple’ motion blur. Real life blurry images apparently tend to have more complex blur going on, and that is currently beyond the deblur algorithms employed in the Adobe Max demonstration. It also made it clear that those algorithms need strongly contrasted edges to lock onto. The last bugbear stated was the heavy computing employed in making it work.
The blog article, it strongly stresses that the team behind deblur are from Adobe’s Advance Technology Labs, and makes it clear that there is a way to go before we see it in Photoshop. But in actual fact, I bet that most of us would be more than pleased to have a flawed version of deblur in CS6, and I think it would work on enough images to be worthwhile. I use a tripod for important non dynamic shots, but still see evidence of motion blur. But because of the tripod, that would be the sort of simple, one directional, blur that deblur would work very nicely on. Incidentally, before anyone comments, my grip gear probably cost more than some of the cameras people here use. I mainly shoot with a Canon 1DsMK3, (I also own a 1DMK4) and with very sharp lenses, you get to be hyper-critical when it comes to acceptable sharpness.
