Re: Lost in ICC Land
Re: Lost in ICC Land
- Subject: Re: Lost in ICC Land
- From: Darrin Southern <email@hidden>
- Date: Sun, 22 Apr 2001 13:16:04 +1000
>
"Howard Newby" <email@hidden> wrote:
>
>
> 2. Should the densities of the inks of the inkjet proofer match or be close
>
> to the densities of the inks of the target device, (like the press or analog
>
> proof)? Or does that matter at all?
>
> It seems like the black on the inkjet is blacker than the black from
>
> the press or the analog proof?
>
>
I posted an 'Inkjet profiling ABC' in January, maybe I should dig it
>
out and repost it with some extras, if you can't find it in the
>
archives.
That might be a good idea to update your excellent post (I got a lot out of
it), although I would like to add my 2 cents.
>
The short and sweet answer is that inkjet densities and press
>
densities have NOTHING to do with one another.
Sorry, I would have to disagree with you on this one Henrik.
>
Also the press inks and your inkjets inks have different colors. Even
>
the blacks are a different color.
I agree here, in principle though.
You need to give your profile some room to work in, to get a better mix of
the actual 'visual' color and the 'density' values you are trying to match.
>
Don't tweak any densities, don't even think of tweaking any
>
densities. Set up each device to its maximum gamut volume and best
>
visual reproduction capability. Characterize / fingerprint / profile
>
that. And then leave the rest to the ICC framework.
Back to disagreeing again.
The idea of the 'contract proof' is to simulate the output of any given
printing method, whether the output is an analog proof (agfa, cromalin or
fuji) or the actual printing press itself.
With the latest range of 6 color (CYMKcm) inkjets, their densities are much,
much higher than the above mentioned printing or analog proofing methods.
Just because the inkjet has the ability to provide such high densities of
color, it does not mean that the application (rip in this case) has the
ability to do the color management necessary. It also does not mean that the
ink and paper combination being used can accept the densities and amount of
ink that the printer can provide.
This fact is still true with the manufacture supplied inks and stock, and it
can even be said that some third party stocks work better . . .
'Personally', I have found that setting the ink densities, and then the
total ink limits BEFORE building the profile, is actually 80% of the work
needed, and then how you build your profile (UCR/GCR - Black generation) is
the other 20% of the work.
We then get back to the 'idealized proof' and 'contract proof' theories you
pointed me to, with the 'ECI guidelines' which I actually found at
http://www.color.org.
I have only began to read the document, although I do have an issue with
some of the 'ideals' and 'ideas'.
'Ideals' unfortunately get corrupted by politics and financial gain.
It would be great for all concerned to agree on the Apple CMM as the only
CMM, but even in your posts you mention the Heidelberg CMM, which is build
into the BEST rip, and it makes sense that if I am going to build profiles
in PrintOpen for the BEST rip, I would build the rest of my workflow on the
Heidelberg CMM.
BEST themselves moved from the Logo CMM to the the Heidelberg CMM in a
recent version upgrade, and this caused a lot of grief for those who were
building profiles with software based on the Logo CMM.
On the 'ideas' issue, I have been in contact with the separate raster and
vector profile workflow when setting up a PosterShop 5 rip which we applied
separate raster and vector profiles.
The problems started when a customer had built his Quark document with an
image that had a background of certain CYMK values. This image was then
placed into a picture box that also included these same CYMK values, to
extend the 'background' panel across the top of the document, without
creating a bigger raster image.
Of course, you could see where the raster background finished, and the
vector background continued.
The response may be to get artists to change the way they build their files,
but this could be just one more issue for the preflight check . . .
>
> I am in search of reference material for self training or more formal
>
> training on ICC setup.
>
>
Setting up ColorSync and building an ICC proofer profile :
>
http://www.iccabc.com. A little aging on the update side, but it'll
>
do fine for starters.
These were the first 'TWO' documents I read on profiling, and I am still
waiting on the 'THIRD' to be finished . . . ;)
Thanks again Henrik,
Darrin.