[argyllcms] Re: Making a scanner profile
- From: Sameli Rantanen <sran@xxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sun, 30 Oct 2016 13:50:58 +0200
I’d like to characterize the scanner. Which Argyll commands/options should I
use to make a scanner profile (not a specific film profile)?
Of course, start with the procedure outlined in the documentation:
Yes, I have read it.
Since you're worried about the spectral characteristics of your targets and
the fact that you don't have a target with the same spectral characteristics
as your subjects, you might consider merging multiple charts in order to
sample a broader range of spectra.
So it’s not possible to read just the real colors on a film and compare them
with the measured values? Is it because the dyes are always involved?
For making a merged measurements file should I use the average command (with
I'd urge against the natural tendency to go for as much precision as
possible. Not only does that way madness lie, but there are hard-and-fast
mathematical limits to what you can do with a scanner.
Sound advice, I agree totally :-)
You might also take a step back and ask yourself what it is you're intending
to do. Do you want to reproduce the _photograph_ or the _original_scene_? You
should be able to do the former readily enough with the basic workflow and a
decent profile. The latter...will quickly drive you crazy.
Trying different profiles to a scan of an older slide produced very different
looking results and none of them looked good (to be fair, the slide looked bad:
it was underexposed, the weather had been gray and maybe even the film roll had
been mistreated). Then I thought I’d need a neutral uncorrected starting point.
With the scanner profile I thought I’d like to be able to scan the film’s
interpretation, not the original colours found in the scene. Like the warm
Agfa, cool Kodak or the exotic Tasma. Probably I’d use mostly the actual film
profiles but it’d be good to have a ”neutral” profile in my toolbox.
If I take the ”choose one” route, which film profile would be good to start
Or is it better to just make a gamma correction to the scan and then apply the
working profile, without too much worrying about the scanner or film profiles?
Other related posts: