This is a neat technology. My understanding of how it works is that each final image pixel actually contains many sub-pixels of focus and/or depth information, made possible by micro-lenses in front of the CCD. The software can change focus by recombining the sub-pixel data differently. The limitation is inherently low resolution, since the sub-pixels are normal CCD pixels. This can be overcome with larger optics and detector arrays, but that is expensive. In the short run, if high resolution is desired, one is probably better off stacking images with Helicon or the equivalent. But this light field technology is likely to scale up nicely.
-Gary
