Originally posted by heckflosse First of all, thanks for sharing!
Second, what is the meaning of the "nRead" values that differ per camera model and why do they not form monotone sequences? For instance, for the K-3, the "nRead" value for ISO 400 is smaller than both ISO 320 and ISO 500 values. I only had a very quick look at the code and just stumbled across this. I wonder whether it allows any conclusions regarding how the cameras deal with various ISO settings.
Why do all Pentax models have different "ePerIso" values?
Interesting.
Not sure why you have a need for flood filling. When is this needed?
I may figure some of this out by looking at the code a bit longer, but I have to go to bed now and thought why not ask you?
Third, it is of course not exactly the code Phase One would be needing. Not sure Phase One is using C++ but in any event, they'd have to adapt the code to their own data representation. If you are keen for Phase One to adopt the code (you might be completely indifferent, I'm not making assumptions) then I think a few more comments would go a long way to make the code easier to adopt by others.
When I wrote "
the code has to be written" I meant the code that Phase One has to add to their C1 code. I didn't realise that your code has such a generous licence associated to it but even if Phase One were to look at it, they couldn't use it as is.