View Single Post
Old 03-09-2014, 01:28 PM
clive milne
Registered User

clive milne is offline
Join Date: May 2011
Location: Freo WA
Posts: 1,443
This thread is not entirely unrelated to a conversation I had with Ray over the last day or so, the gist of which relates to optimising colour channel exposure times.

Consider this; conventional wisdom suggests that if a camera's QE in the blue is half that of the red, then the blue channel requires double the total accumulated exposure time of the red. However, even though this will ensure both blue and red channels have equivalent read-out and shot noise, there will be additional noise contributed by effectively doubling the dark current.
It then stands to reason that in situations where dark current is significant, conventional exposure ratios derived from estimates of relative spectral sensitivity are not an optimum use of exposure time. This is perhaps more relevant to Kodak chips and narrow band imaging. What it potentially translates to is that the results obtained with linear colour/QE scaling could be achieved x% more quickly accounting for dark current shot noise in areas of the spectrum where any given CCD is inherently less sensitive.

Anyway... just thought I would put it out there for public discussion and contemplation.

Reply With Quote