E-M5 has DR 13!

Started Apr 13, 2012 | Discussions thread
Detail Man
Detail Man Forum Pro • Posts: 16,674
Re: hard to get 13 bits of DR from 12-bit RAW files [n/t]

Dr_Jon wrote:

Okay, I confess I haven't read all this thread, but I want to reply to Detailman by saying:
(1) No, I don't believe the OM-D has thet much dynamic range.

(2) Modern sensors tend to use variable slope ADCs, so you can exceed the number of bits in DR.

It would be interesting if you could elaborate on your item (2) above. It is indeed true that variable-slope ADC's exist (and the "active-pixel" designs are potentially amenable to such further, albeit complex, machinations). Any such varying of the slope of, say a delta-modulation ramp waveform, any "adaptive" measures taken, will result in the translation to the encoding representing powers of another base other than 2. Yes, that can indeed compress an input range.

Each ascending (more "significant") bit in a common ADC corresponds to a power of 2 - which is equal to 6.02 dB (or 1.0 EV, or 1.0 "stop") change in whatever input signal is converted. There do exist "logarithmic ADC" designs weighting the bits to the power of a different base other than 2.

Around 10 years ago, I designed, developed, built, and tested an ADC that does just that. The resultant output codes, however, while meaningful, have to be interpreted (decoded) into a form down the line that is amenable to binary arithmetic operations. Not unlike the compression-curve applied to the Sony NEX-7 image-data, where a separate decoding stage is required to be applied:


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