On 05/12/02 11:43, Andrew Woods wrote:

> the iris print is to Euroscale from known CMYK target file.

You realise of course that there is no such thing as Euroscale, except as a
general concept?
The Photoshop Euroscale v2 profile supplied with Photoshop is based oin the
ISO 12647-2 Fogra 1 standard and then manually tweaked.
No press I have even seen use the exact same ink colours or densities as
this standard. Some (the good ones) adhere to this ISO standard and in this
case the profile is a good choice, the others well...

> My print is 
> from the  same file in  ps7 with Euroscale as the working space and
> working CMYK as the proof setup. Printed by 2100 with source space as
> proof and print space as thomas' profile with a relative colorimetric
> intent. Is this what you meant?

First ask yourself this, and I mean this question dead seriously, I'm NOT
making fun in any way:
How do you KNOW when you digital file (which consists of several millions of
One and Zero numbers in a cleverly defined pattern) is printed correct?

Do you assume it looks correct on print because it is pleasing?
Or because it matches the monitor?
Or because it was what you imagined it should look like?
None of these solutions are proof that the print you have it a correct
representation of the contents of your file.

The only way you can possibly know when something is correct in print is by
verifying your workflow by adhering to certain standards, and calibration
procedures, and verify EVERYTHING in your workflow over and over again.
Unfortunately this is a rare practice, although a lot of people seem to
think that have done it, but are in error.
Just a bit of food for thought.

If you want to be able to (correctly) simulate one device on another you
basically have three options, and one prerequisite:

The device you are simulating MUST be stable from day to day/hour to hour.
If not what exactly is it you are simulating?
Printing presses are notoriously unstable, but then a skilled press operator
can tweak the ink on press to compensate somewhat.

The old fashioned proofing: Adjust the proofing device to match the output
device via calibration and closely matching substrate and ink colours on
both devices. This is the way film based contract proofs like Cromalin,
Matchpront, Agfaproof etc. was made. The separation films are/were used to
make the proof, put it in a lightbox expose and "develop" after previously
set guidelines and you have a proof simulating the press at a given
condition. If the press is stable the proof will yield a reasonable match.
If the press for some reason changes ink type/brand, paper or has a higher
or lower dot-gain the proof is off.

Digital proofing:
If you want to simulate a press on something like an inkjet printer (IRIS,
Epson, Digital Cromalin or whatever), you calibrate that printer to make it
simulate a certain kind of press condition. This will generally take care of
gradation and grey balance, but certain colours may be off due to
inconsistencies between ink used on press and in the printer. The more
similar substrate and ink are, the better the results. Certain systems also
employ selective corrections of course, but if 100% magenta is a different
hue, anything from 90-95% magenta WILL be different.
Just because your IRIS did actually perform the way it should in the past,
don't assume it still does unless it is calibrated regularly (bi monthly or
even more often). And even if it is calibrated, print a test image and
compare with the previous version, and the version before it. Also measure
the solid colours to make sure the density of the colours are the same.

ICC proofing:
To be able to simulate something by using ICC profiles you simply need a
profile describing the press/proofer you want to simulate, and a profile
describing the proof printer you use. So if you want to simulate your IRIS,
you should first make sure that your IRIS is actually simulating the device
it is set up to simulate CORRECTLY. Don't just assume that it is, I have
seen Cromalins, IRIS, Approval, Matchprint, Digital Cromalins, Rainbow and
what have we from reputable shops claiming to simulate "Euroscale" be WAY,
WAY off from what they should be (think 25-30% too much cyan in a midtone
grey on the proof, not to mention the other colours). And many of they
supplied a profile not even close to their proofing standards.
That said the same limitations as above holds true; if the inks are
different the solid colours can never match.

Assuming that your IRIS is actually simulating your press ok (check it,
don't assume it), you need to know what CMYK profile was used to convert the
material that appears correct on press and on the IRIS. This profile will be
the source profile, you need to convert from.

Many people skip the fence and just declare that "My IRIS IS correct!"
Assuming this, you need to have an ICC profile describing the IRIS to be
able to cross render it on your Epson. Just using a standard ICC profile
isn't good enough if you choose this "the easy way out". Effectively you
need to limit the gamut of the Epson, to that of the IRIS, and the only way
of reliably doing that is to have a profile for both devices.

Then there is the matter of substrate. If the Epson profile is on a darker
substrate than the IRIS the two will never match perfectly. If the Epson
uses a lighter substrate or one with a different colour, you need to use
Absolute colorimetric when converting from it's profile to the Epson
profile, to simulate the paper colour and grey balance of the IRIS on the
Epson...

And then only evaluate the results in a D50 daylight box, as this is the
place where metamerism will be virtually eliminated (at least with the
profiles I do). Not so coincidentally this viewing environment is also the
place where the proof are judged by the press man at the printing stage.

Happy proofing.


Best Regards

Thomas Holm / Pixl ApS

- Photographer & Colour Management specialist
- Adobe Certified Training Provider in Photoshop�
- Imacon Authorized Scanner Training Facility
- Remote Profiling Service (Output ICC profiles)
- Seminars speaker and tutor on CM and Digital Imaging etc.

- Home Page: www.pixl.dk � E-mail: [EMAIL PROTECTED]
-- 



===============================================================
GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE

Reply via email to