> > It is this last point that is the bone of contention - manufacturers are > > saying "ours is 14 bits so our density range is 4.2 wow isn't that a > good > > figure", and that is probably crap in the case of the consumer level > > scanners we are talking about. It MAY be 4.2 but is most likely is much > > less than that - that is all I am saying ... I think. > > It's certainly crap if they don't use a special technique like split > exposure > multi-scanning. > All the available CCDs on the market today are limited to a dynamic range > of > 5000:1 (~12 bits) at normal temperatures. > I'm not sure if the upcoming CMOS sensors will do any better. > > It's called "How to lie with statistics". When the method used to claim a dMax (or dRange or ...) is not precisely defined, you can make up anything that makes sense. One way the manufactures can "lie" is to measure two patches, one with an OD of 4.2 and one with an OD much larger (opaque). The opaque patch might have an average value of 100 (out of 2^14 = 16,384) while the 4.2 patch might read 101. Never mind than the pixels in each patch have a range of 50 to 150, the average values are different. The two patches _can_ be distinguished, but only because a large number of pixels are in each patch. For example, if each patch contains 100 pixels and the noise is random, then the noise would be reduced by a factor of 10 (this is similar to multi-sampling 10 times) and the manufacturer might claim that the dynamic range is not just 5,000 but rather 50,000 (dRange = 4.7)! Just be glad that the manufacturers don't carry this to its absurd limit - over the entire 4,000 by 6,000 pixel slide this would reduce the noise by a factor of 4,900 for a dRange of 24,500,000:1 = 7.4! Hey, I think I will start my own company. ;-)
