On 17 Dec 2003 at 16:30, Ryan Lee wrote: > Thanks kindly for the link. It's a good overview, though I think it helps > illustrates the question I posed (under 'system noise', the article > identifies CCDs to have 'low' while giving CMOSes 'moderate to high'). Ergo, I > find the dpreview *ist D/10D comparison with regard to noise (*ist D had > slightly lower noise than the 10D up to ISO800, where the *ist D's noise levels > got higher) curious. Just to rehash for everyone else, I'm wondering why this is > the case, if it's a sensor or processing issue, or perhaps could the dpreview > test have been flawed and inconclusive?
Ryan, We see the noise in an image after the output from the sensor has passed through amplifiers, AD convertors then and interpolation and filtering algorithms. On top of that the sensors that you are comparing likely have difference in photosite area which in its self leads to a difference in noise performance and dynamic range. So again it's like comparing apples to oranges. Sensor types have family traits but there are still plenty of deviations. For instance the following article contradicts the data in the previous link that I posted: http://www.smalcamera.com/pressreleases/smal_auto_release.html > And I was also wondering, the article seems to favour CCDs when it comes to > image quality- how far off would one be in suggesting that a 10D with a CCD > instead of a CMOS (forget the extra cost or battery life or what it'd look like > or who'd buy it) might be capable of producing higher quality captures? If you have to make comparisons then compare the performance of the complete cameras (entire system), try to ignore the sensor type and you'll probably suffer fewer brain aches :-) Cheers, Rob Studdert HURSTVILLE AUSTRALIA Tel +61-2-9554-4110 UTC(GMT) +10 Hours [EMAIL PROTECTED] http://members.ozemail.com.au/~distudio/publications/ Pentax user since 1986, PDMLer since 1998
