[argyllcms] Re: profiling camera
- From: edmund ronald <edmundronald@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 5 Aug 2011 09:57:32 +0200
On Fri, Aug 5, 2011 at 9:09 AM, Nikolay Pokhilchenko <nikolay_po@xxxxxxx>wrote:
> 05.08.2011, 04:18 nino loss wrote:
>
> I still get a peak error of 18.381637 and an average of 7.896481 !
>
>
> IMHO, It's OK for camera profiling. This number depends of the uniformity
> of illumination, of lens and sensor uniformity and metamerism phenomenon.
> You can check uniformity of target illumination and lenses by shooting
> clean white paper sheet in place of target.
>
Yes.
Anyway, there is little point in fully profiling a modern SLR, in my
experience, they are superbly linearized. What you really need is just to
establish the primaries, this will perfectly solve even nasty lighting for
you. The hard part is just getting the Raw data, and writing the profile
out. The core code is 3 lines, it can be done in anything including
Javascript and Python.
I know about this because it attracted my attention some years ago and I did
it all in Matlab, worked perfectly.
Does Argyll provide a way to write out an input profile with given
primaries?
Edmund
Other related posts:
