On 23 Oct 2004 at 15:07, Shel Belinkoff wrote: > Hasselblad says the 22 Mpixels sensor is 37x49mm, which, if my math is > correct, is slightly larger than two 35mm frames. In addition they say, > "Thanks to its unique Double Duration Circuit technology (patent pending), > the H1D runs more than twice as cool as similar digital camera systems, for > cleaner images, true 16bit colour depth and double the battery life." Of > course, I haven't a clue as to how accurate this all is, but I've no reason to > doubt the claims at this point. > > http://www.hasselblad.se/products/level3.asp?secId=1135&itemId=3362
Having a slightly increased sensor area per pixel and cooling on the sensor would I expect lead to a better signal to noise ratio so I assume they have utilized a greater bit depth to take advantage of any extra information. WRT the 12bit resolution of the *ist D etc, if the information isn't there to encode (ie the information is noise) then there is no point encoding it. The difference in cost to implement 12 or 16 bit digitization would have been negligible so I can only assume that the designers of the camera selected an optimal match between sensor (which has an analogue voltage output) and analogue to digital convertor. Scanner sensors are linear (ie a single row of pixels (or three, RGB in a tri- sensor array) not a matrix like camera sensors and so interconnection circuitry within the package can be optimised to reduce noise (and increase latitude/stop range). And lastly the size of the pixels isn't so constrained as in the case of camera sensors so they can be made larger and more sensitive plus there may be more opportunity to apply heat sinking and control to reduce thermal noise. 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
