[argyllcms] Re: Different approach to profiling?
- From: Ben Goren <ben@xxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Tue, 4 Sep 2007 19:12:19 -0700
On 2007 Sep 4, at 1:38 PM, Lars Tore Gustavsen wrote:
> On 9/3/07, Ben Goren wrote:
>
>> 1) tiffgamut -v -c md ProPhoto.icm Flattened.tif
>
> icclink: Warning - Image gamut is wrong colorspace for link (Lab
> != Jab)
Sorry 'bout that -- I realized I had left out ``-i j'' from that
line when I tried to follow my own instructions later that
evening. As Graeme has already pointed out, you need to specify
the Jab color space.
One other gotcha: If you tell Argyll that the image is in some
color space than the one it's actually in, or if you specify the
worng output profile, you might get a print that's mostly right
but with some weird artifacts. That seems to be an especially
likely occurrence if you're working first thing in the morning
after the cat woke you up too early after you went to bed too
late....
And, yes, you'd ideally want to create a gamut map for every image
you want an especially high-quality print of. Otherwise, you can
get still-darned-good prints for all images using the same source
and destination space by not specifying the gamut. The result will
be a linked profile that skips the multiple extra conversion
steps, but it won't be ``extra optimized'' for the range of colors
in that particular print.
If you're a photographer who's already gone to the trouble of
using Agryll in the first place, chances are excellent you'll want
to do the gamut map bit, at least for anything you consider a
finished print. If you're some other kind of graphic artist, I'd
suspect that a simple linked profile would be exactly what you're
looking for for all but your most critical jobs. Of course, I'm
sure many people on both sides of that divide would disagree with
me, with most suggesting that you've gotta be nuts to type
anything on a command line in the first place....
Ah, well. Their loss.
Cheers,
b&
Other related posts:
