How are dynamic range and sensor size related?

[NormanBates]

Because of the plateau at low ISO, I think the problem with the DR on the 5D3 comes because of the ADC

http://www.clarkvision.com/articles/digital.sensor.performance.summary/
"The dynamic range is often limited by the A/D converter and other electronics in the system, illustrated by the measured data falling below the model at lower ISOs."

 

[jrista]

Because of the plateau at low ISO, I think the problem with the DR on the 5D3 comes because of the ADC

http://www.clarkvision.com/articles/digital.sensor.performance.summary/
"The dynamic range is often limited by the A/D converter and other electronics in the system, illustrated by the measured data falling below the model at lower ISOs."

He explicitly called out "other electronics" as well. The A/D converter is not always an integrated part of the sensor, however the sensor contains a significant amount of complex electronic elements that can create noise on a variety of levels. An off-die ADC must usually process more pixels than say an on-die CP-ADC, so it requires a higher clock, which has a greater potential to introduce its own electronic noise into output units. Once the ADC has actually converted an electron charge into an ADU, you lose the ability to eliminate those forms of nose...they are now essentially "baked into" the output units. Sony Exmor sensors eliminate several forms of ELECTRONIC noise in the sensor before they send the charge that represents any given pixel to the ADC (which, in the case of Exmor, is also on the sensor die). Sony has two advantages...they eliminate as much electronic noise as possible before ADC, and they have one ACD per pixel column, which gives then the added advantage of eliminating vertical FPN (which is actually apparent...the D800 only seems to exhibit horizontal FPN unless it is pushed by 10 stops or more, at which point it might finally start exhibiting full cross-hatch FPN).