Re: Following standards ('Generic CMYK Profile' analysis)
Re: Following standards ('Generic CMYK Profile' analysis)
- Subject: Re: Following standards ('Generic CMYK Profile' analysis)
- From: "Andre Schützenhofer" <email@hidden>
- Date: Sat, 29 Dec 2001 06:12:56 +0100
Thinking of the 'Generic CMYK Profile' as source for LEGACY data makes it
an unimportant profile, a profile which is of little consequence.
Back when I started to manage color we made our printer to print some test
charts. Contrary to the normal recommendation to create the profile based on
an average production condition they cleaned the whole press, which means
press sheets as well as the conditon of the press in general.
This was not the case when we made further approaches in creating profiles
with other printers, because we used the method "creating an average
production condition". It turned out that by following this method it was
really difficult to establish a condition which I'd call worth profiling
for.
Despite normal/standardised density the dot gain in the primaries
substantial differed from each other, which resulted without fail in an
inharmonious graybalance, not to mention in an inreproducable condition.
I think the reason for this is up to the inconstant condition during the
phase of the test chart printing procedure by itself, and therefore only
reproducable if exactly this special inconstancy like the one in the phase
of the test can be restored.
Interestingly the profile based on the condition first mentioned has a
graybalance almost ideal. It turned out that this profile can be easily
assigned to data if there is no source profile or if it is likely that
assignment has not happened with knowledge to that file (there are
unbelievable plenty of files in germany going around tagged with
SWOPcoated).
So the condition described by the profile with the good graybalance can be
satisfactory reproduced and very good as well, even by printers who never
had heared of profiles - dependent on specific situations like for example
inks used and relative humidity, of course.
I think this is possible because the profile is based on a condition which
was linear, coherent an harmonious by itself. So the colorspace was not
unnecessarily distorted and skewed, as much parameters as possible were "set
to zero". The coherence of the primaries among each other is very important
and only possible to achieve by the approach having every color deck bahave
as similar as possible to the other.
Every deviation interferes the graybalance and thus the integrity of the
whole process to record. Whereby I don't think that this comes out to a sort
of idealism or some kind of far from production everyday life - in fact this
is the opportunity to create a reference, based on reality, almost perfect
comprehensible in the event of all parameters are coherent to each other.
At least we are normally satisfied with less than that, due to compromises
involving impositional reasons or the reverse showing through.
Another issue considered to be an advantage is that profiles with a natural
graybalance are well accepted by lithographers (especially in RGB retouching
matters), because the readout in the info palette matches with decades of
experiences gathered by professionals.
In the case of if for example 80c/60m/51y/50k or 83c/79m/83y/8k results in a
neutral dark gray then this might be correct in a colorimetrical sense, but
it indicates too that during printing the test chart it is likely that there
was something wrong.
Conclusion: I think the aim should be to create references based on reality.
Not completely generic profiles and no fingerprints of any condition either,
but rather generating conditions enough coherent worth making fingerprints
of, and that's what reference profiles should be based on.
Expressions of agreement and/or denial appreciated.
- Andre Schuetzenhofer
_________________________________________________________________
Werden Sie Mitglied bei MSN Hotmail, dem grv_ten E-Mail-Service der Welt.
http://www.hotmail.com/de