Re: ImagePrint Answer
Re: ImagePrint Answer
- Subject: Re: ImagePrint Answer
- From: bruce fraser <email@hidden>
- Date: Tue, 28 Jan 2003 10:45:42 -0800
At 8:16 PM -0400 1/24/03, Cris Daniels wrote:
This is a case where a newbie reading our posts would wind up very
confused. I know that the profile knows nothing. But using your logic, if you
preview an image with a not-so-hot profile and don't like the result, that
hammering the image into shape in Photoshop is going to allow me to get it
printing properly which is simply not true. If the profile determines that
print X can only print color X, you can't make the case that using Photoshop I
can manipulate the image to print better reds for example. If the profile
sucks trash it. If the profile is excellent, an image that is properly
prepared shouldn't need any (or very minimal) work.
<snip>
Cris,
Any profile, no matter how good, still does a one-size-fits-all
conversion for all images. Take an extreme case: you have an image of
a black cat in a coal cellar, and another of a polar bear in the snow.
I can guarantee you that, given an accurate soft proof, I can edit
the source images to produce better prints than the blind conversion
will give me. The blind conversion has to be a compromise that will
produce equally good, or equally bad, results on both images.
We aren't talking about editing images to address deficiencies in the
profile -- that is indeed a mug's game. We're talking about going the
extra distance to make the difference between good and great. Even
when we eventually get adaptive CMMs that address the gamut of the
specific image, there's still the question of the tonal mapping. I
haven't seen any developments that lead me to believe that
intelligent human intervention won't still be required to make the
difference between good and great.
Bruce
--
email@hidden
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.