Thx Kirbic - I think I get what you're saying - just hard for me to wrap my mind around this technology sometimes. Theoretically, a gp camera using the same technique would be commensurately higher resolution, if you had a processor large enough to handle the job. But, a single comparable picture taken with the same gp camera could not possibly include the detail captured in the lower mp camera's individual shots.
You really only need those extra pixels if you want to create larger, high resolution prints. However, you can't pack that many pixels on todays sensors, and even if you could the pixel density would be so high as to render airy disks visible at even the smallest apertures. For example, with a 36x24mm sensor and 1G in pixels (roughly), your 300dpi print size of 140x93 inches will render airy disks visible below f/1.8 (you can just get f/1.8 @ 200ppi though). Even large format only gets you f/4 @ 300ppi (f/7.1 @ 200dpi); your print size changes but that's generally the point.
For very large prints (measured in feet) and very close viewing distances (approx 10"), you definitely want to use stitching. Otherwise, they better start making really, really big sensors.
...For very large prints (measured in feet) and very close viewing distances (approx 10"), you definitely want to use stitching. Otherwise, they better start making really, really big sensors.
A silly thought exercise: if you have a 10µm pixel pitch (similar to today's DSLRs) and an 8"x10" sensor, that's still only 0.5Gpx!! And a single 8"x10" sensor won't quite fit on a 300mm silicon wafer... so much for manufacturing!
There's an 82 gigapixel image of Southern California; it does need a plugin to view it though...
Nice find, i was able to find my house, my school and some of my friens houses. The picture is fairly old tho, i found places that since then have been built up
