On Wed, Jul 29, 2009 at 09:48:14PM +0200, Ralf R. Radermacher scripsit: > Graydon <[email protected]> wrote: > > I strongly suspect that long exposures are junk without it. > > My suspicion as well. Could it be that this Samsung sensor is so rotten > that they just can't afford to let us see the terrible unmasked truth?
Everything is trade-offs. Can't speak to the K-7 version, but the K20D version strikes me as an excellent sensor for the more usual -- under 10s, say -- range of exposures, and it has a number of virtues . > I mean, it would fit in nicely with the general noise problems they seem > to be having ever since they turned away from using Sony CCD sensors. Are there, though? Noise seems very well controlled in short exposures; it's that the CMOS architecture is sensitive to heat, which is a problem for long exposures. (Thermal management in IC design is something of a black art, even now; I suspect there's a very nasty tradeoff between short exposure responsiveness and long exposure stability built into the thermal design.) I'd generally expect that the CMOS sensor architectures are superior to the CCD ones in terms of potential sensitivity. It may be that this potential won't be fully realized until someone figures out how to integrate a Peltier cooler with the in-body anti-shake, but as a long term architecture choice I think it's entirely defensible. It just happens to suck for astro. -- Graydon -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
