[argyllcms] Re: dcraw linear workflow?

  • From: Ben Goren <ben@xxxxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sat, 12 Sep 2009 14:24:40 -0700

On 2009 Sep 12, at 2:07 PM, edmund ronald wrote:

On Sat, Sep 12, 2009 at 10:52 PM, Ben Goren <ben@xxxxxxxxxxxxxxxx> wrote:
On 2009 Sep 12, at 1:11 PM, edmund ronald wrote:

what camera is this ?

Canon 5D Mark II.

usually really ugly to profile. There is some strong signal processing
applied to the "raw" files

Well, as I mentioned, I was able to get a profile with an average error of ~0.8 using an XYZ LUT. And that was with a handheld moderately-high ISO snapshot with ambient indirect window lighting; I'm hoping my usual studio flash setup for art reproduction can do even better.

It's quite entertaining to compare the results of the workflow I described with ACR with all its sliders, etc., set to their neutral positions. Quick glance, they're basically the same. Look for longer than a couple seconds, and you spot an orange blown to hot pink here, and a yellow that's way too orange there, and....

Of course, I now have a dilemma. I love ACR for artistic manipulations of an image -- toning, dodging and burning, etc....it's easy to work with and produces excellent results. But I *have* had to deal with some problems, such as a little girl's pink shirt that becomes just too unrealistic while the rest of the scene is merely ``punchy.'' I'm afraid I might have to come up with some sort of hybrid workflow for most of my stuff. Colorimetric reproduction of artwork is easy (if tedious) once you get your formula down pat, but I don't do a whole lot of that sort of thing.



Other related posts: