On Tue, Jan 24, 2012 at 11:18:39AM -0500, Paul Stenquist wrote: > > On Jan 24, 2012, at 11:10 AM, Matthew Hunt wrote: > > > On Tue, Jan 24, 2012 at 10:45 AM, Paul Stenquist > > <pnstenqu...@comcast.net> wrote: > > > >> Sensor arrays are used on large format backs. I watched a shooter work > >> with a 4 x 5 camera and a digital back tethered to a computer. The > >> complete image, without gaps, resolved almost immediately on the monitor > >> screen. I don't know if the back was gap free or if software corrections > >> were made. But the results were outstanding. > > > > You're sure it was a mosaic of sensors, and not a scanning back? > > Yes, the shooter told me it was a four-sensor array. This was about a dozen > years ago. We were shooting a Dodge Nascar vehicle in a Michigan studio.
I've heard of these - that's mostly a mechanical (or cost) limitation. It's all but impossible to make a reasonanbly defect-free sensor above a certain size (although the limiting size changes with the technology). But it is possible to design a sensor so that all the connections, &c. are made along one or two edges, with only a very thin gap between the edge of the package and the actual sensor cells on the other two edges. That means you can combine four sensors into a larger composite unit. This is almost as good as a single sensor of the same size, but has the advantage that it's a whole lot cheaper. For example, I think the sensor used in the 645D was initially priced at around $1000. That means you could probably make a four-sensor array of comparable chips for somewhere around $5000. I'd expect a single sensor design to be at least an order of magnitude more expensive, simply due to the mathematics of yield rate in chip manufacture. -- PDML Pentax-Discuss Mail List PDML@pdml.net http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
