14 bit. Any use?

Started Jan 1, 2008 | Discussions thread
bobn2 Forum Pro • Posts: 53,189
Re: In theory...

Blade93012 wrote:

Some have argued that the difference between 14-bit and 12 bit is
just finer quantization of the information, and that the range
remains the same.

That may be true if you kept the full scale reference to the A/D the
same. But what if you can change (increase) the FS reference for the
A/D converter when you are in 14-bit mode? In theory, this can give
you anywhere from 0 to 4X the dynamic range as compared to 12-bit,
depending on how much you increase the FS.


Probably best not to rely on 'theory' if you don't know it. The FS reference of the A/D has to correspond to the FS output of the photosensor, if you increase the A to D reference, all that happens is the samples captured are in the bottom 12 bits of the 14 bit sample. 'In theory' the extra DR given by an A to D bit depth above the DR of the input signal results from the reduction in the quantisation noise, and is vanishingly small the more unnecessary bits you have. For todays sensors with 9-10 stops (bits) DR, the theory (depending on the assumptions you make) predicts about 1/3 stop advantage for 14 bit over 12 bit, which seems to check out pretty well with real measurements. The 'finer gradations' argument is baloney. The finest gradation you can detect (mechanically, as had been pointed out, our eyes can only detect about 7 bits) is determined by the noise.

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow