Jack Hogan

Yummy, please do show.

Good summary Juan. I would just add that these days the better raw converters (e.g. RawTherapee) use floating point internally before displaying/saving the rendered image as an integer. ...

No, the Metadata is fine.  Matlab is having problems extracting the raw cfa data from camera-native DNG files. Jack

Yes, but perhaps not so far off. As far as DR is concerned, what channel is the most objectionable to you? It's then easy to calculate the DR with a few simple assumptions. Jack

Hi Mark, I think the easiest way to look at it is starting from white balanced raw data, assuming a linear process until gamma is applied - and further assuming that gamma is perfectly undone by ...

Could be, it depends on the camera and I am not familiar with those RAW files. I am on the other hand familiar with DNG Specification version 1.4.0.0 (sometimes used natively by consumer cameras) ...

RAW files have their own proprietary formats. If the format is unpacked then it's no longer a RAW file :-)

It depends on the camera. Many current cameras have the option of capturing data at 14 or 12 bits and that's often what's written in the raw file (perhaps after additional compression). Matlab ...

Ah, good to know theory is reflected in practice, thanks Erik. Jack

That's BS and you know it.  Plus this is the busiest time of year for me, I am not on holidays, things get better in a month or so.  In the meantime I do not have time to explain in detail the ...

Joofa, take a step back and look at your (lack of) contribution to this thread: 8 posts so far and not one, not a single one, on topic. Every single one off topic - imprecise, vague and ...

No time to answer properly Joofa but I don't think the correct answer is either of those - and I think you are being pedantic without being forthcoming with a usable solution.  There is more than ...

I appreciate your comments and knowledge Joofa but I believe you are barking up the wrong tree. Cheers, Jack

Ok, thanks Brandon, bed time reading for the fall. Ciao, Jack

Nothing, I've already explained that the W terms are to be considered a rough plug to a simple model :-)   I could explain why this sort of makes sense in the context of this thread but I do not ...

I hear you but don't forget the title of this subthread.  And I am still interested in your suggestion for a better x-axis label in the OP. Jack

I'd be willing to 'make' such a plot if I were familiar with Zernikes, which currently I am not. I will have time to play with them after the summer. In any case I am not clear what you are showing ...

Ok, good. So given the mentioned setup, what relationship between W020 and W040 would you recommend? Jack PS The model is not broken, it is just 'simple' with very few terms. So don't fault it for ...

As already mentioned I think you are expecting too much out of the simple model in its current form. The best solution to your comments imho would be to replace the Wxxx wording in the legend with ...

Good point Jonas.