Garbz
No longer a newbie, moving up!
- Joined
- Oct 26, 2003
- Messages
- 9,713
- Reaction score
- 203
- Location
- Brisbane, Australia
- Can others edit my Photos
- Photos NOT OK to edit


In Analogue to Digital conversion the top value is defined by a reference, and the dynamic range is then defined by how many bits are used for quantisation. So more is better the more bits you have the more detail and information you can pull out of your shadows. ... till a point. This assumes an analogue source of arbitrarily large dynamic range. Digital sensors are not that. The lowest value in most sensors is defined by the noise floor of the everything leading up to quantisation, not the quantisation process itself (which can in itself generate noise).
So you could have a 16bit ADC, heck throw in a 24bit ADC what have you got to lose. The problem is those extra 2 or 10 bits will be measuring nothing but noise, and be utterly useless. Worse still they will present an extra load to the processing circuitry wasting power, and more importantly speed. It's non-trivial reading out and then processing and storing 10million 24bit values several times a second.