> In theory I think you could gain 4 stops. One stop is x2 > more light, so x2^4 should translate to 4 stops.
Yep, 1 stop == 1 bit, conveniently. > However, that requires a linear intensity response from > the sensor and the same size of the steps. If you just put > 16bit within the same dynamic range that is covered by > the 12 bit you may get better nuances but not dynamic range.... Well, technically you *do* get more dynamic range. Dynamic range is basically the ratio between the smallest change measurable, and the largest value measurable. So for 12 bit you basically have 4095:1 dynamic range; for 16 bit 65535:1 dynamic range. What you choose to be maximum determines if you have over-exposure lattitude or not, which is a different thing altogether. So in 16 bit, you could shoot 4 stops underexposed, and get the same nuances as 12 bit, but with 4 stops of over-exposure latitude. Nice in theory anyway.... Of course, in the real world the linearity comes in to play as you say, but, as discussed previously, theres a point where sensor noise makes more bits basically meaningless. Love, Light and Peace, - Peter Loveday -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net
