the future of lenses, part 1

Consider the implications of higher resolution. With increased megapixels, there is typically an increase in resolution, up to the point where the lens limits the resolution.

Putting aside the lens problem momentarily, what would be the implications of a 1 Gigapixel sensor? (In case you think that's unrealistic, there was a time when I thought my 4.77MHz IBM XT was fast enough, just like I think my 6.3Mpixel Canon Digital Rebel is sharp enough now. Well, I'm writing this on a 1 GHz Pentium, and it's four years old.) This corresponds to roughly 25,000x40,000 pixels at the 35mm aspect ratio. We could crop almost 90% of the photo in each dimension (i.e., 99% of the area) and still have an image comparable to the Canon 1Ds's 11-megapixels.

For one, this means you could throw away your massive super telephoto. Since 10x digital zoom would still produce excellent results, most people could get by with a wide angle zoom, say 17-40mm on a 22mm sensor (27-640mm equivalent w/cropping). If you really needed shots of the moon or small birds, you could still make a compact but high-quality lens up to, say 100mm (1600mm w/cropping). Note that, as far as my logic goes, the cropped center of a wide-angle shot is virtually the same image as a zoomed-in telephoto shot at lower resolution. Please correct me if I'm wrong.

Problem: Your viewfinder would need to zoom as well, or else you couldn't see your 1/10th scale subject.
Solution: Use a digital viewfinder, like digicams do. Of course, LCDs/DLPs are going to have to get a lot better to be useful for focusing, but they will.

Next post, we return to the problem of getting lenses to resolve more.