[argyllcms] Re: Camera calibration: LUT only as good as matrix?
- From: Ben Goren <ben@xxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 3 Jul 2013 12:43:04 -0700
On Jul 3, 2013, at 12:16 PM, Torsten Bronger <bronger@xxxxxxxxxxxxxxxxxxxxx>
wrote:
> And correcting exposure and WB are mere linear operations, which are
> hardly error-prone.
They're linear operations, yes, but how do you determine what the linear
factors are?
Using the profiling mechanism is the only guess-free method I'm aware of for
precisely determining those factors.
Worse, if you *don't* perfectly nail those factors, you'll bake your errors
into your profile.
Traditional methods of determining white balance amount to a crude
oversimplification of profiling, using a target with a single patch. And
they're as error-prone as you'd expect. Using the profiling mechanism will give
you results as good as your profile itself.
Cheers,
b&
Other related posts:
