RE: Images for print
RE: Images for print
- Subject: RE: Images for print
- From: Roger Breton <email@hidden>
- Date: Sat, 11 Oct 2008 12:20:46 -0400
Hi Marco,
> ... but it seems logical that in
> order to increase printed gamut one has to increase both ink densities
> and purity/concentration.
Ink densities is just a measure of log reflectance. As with inkjet inks, the
more is added to the substrate, the more colorant is being laid down to
derive color from but there is an upper limit to the saturation of the hue
that ensues. Akin to the type of calibration practiced in the past by the
ColorBurst RIP, it is clear that the law of diminishing returns applies; at
some point, any further increases in density adds no further chroma to the
color. The same thing applies to a printing press. With the additional
restraint that, beyond a certain density, the ink/water balance becomes
unstable. The pressmen collectively calls this "running in mud". No one
wants that. Not you, not me, no customer would accept paying for such a poor
quality of printing. Instead, what they want is to print with a minimum
level of water. And the optimal ink film thickness in any printing press
approaches 1 micron. I'll submit that no press operator would allow setting
a press to 2.03C, 1.85M, 1.56Y, 2.24K in *StatusT*. That is just not
physically feasible. So density alone, here, does not tell the whole story.
It's like me telling you that I have this wonderful RGB image on my
computer, looks brilliant on my screen, and when you receive it, you call to
tell me that it just looks like shit. Until we agree on how to interpret the
RGB data we'll be arguing all night long on how good or bad the image really
is. Same with density information. Reporting density information *without*
its corresponding Status function is like me telling you that you have 100
dollars in your pocket; is it 100 dollars "Canadian", "US", "Australian"?
The absolute amount is meaningless without some knowledge of its currency.
> ink can only be as high as the substrate will allow.
Don't forget the printing press: they, too, have an upper limit.
> The results in Bill's book are undeniably more brilliant and saturated
> than the best standard sheetfed output.
Have you purchased a copy, btw?
Not to sound contrarian but I would have like to see what the output of a
good Fogra39L press would have been, side by side. I tend to think that the
comparison would not have suffered all that much. But I can't really talk
because I have not purchased a copy of the book, in all honesty.
> I have little doubt that the colorimetric information would confirm that.
Allow me some reservation.
> I agree that I too would have liked to see the colorimetric data
> relative to
> the primaries, secondaries, etc., as well as examine a copy of the
> spectral
> measurements or the ICC profile of VANFU's press output. That would
> provide
> a picture of the achievement that speaks more directly to me than the
> density values (as impressive as those are in their own right).
Bingo.
> I have not yet been able to see that kind of information anywhere
> regarding Bill's book.
Me neither, why? Clearly, the current PDF does not paint a complete picture.
> On page 113 of the PDF, Bill compares the VANFU densities to the
> standard
> SWOP densities. It would not make sense to compare the two sets if they
> weren't both expressed in the same scale (I assume Status T).
That's exactly my point and again, no offense to the author, but, as you
responded, it would *NOT* make sense to compare them directly if they
weren't both expressed on the same scale.
> Skepticism that they were not highly concentrated? Or that the
> densities were not as high as claimed? Or both?
Both. I have recently seen a swatch of magenta ink, from a local ink
supplier, just regular sheetfed magenta ink. The density was 1.85 StatusT. I
told the supplier how silly I thought it was from them to even bother
showing an ink draw with such a ridiculously high amount of ink film
thickness given that no experienced pressman in their saint mind would ever
allow running their press at such ridiculously high value. If it happened,
it would ever happened, it would be an unfortunate consequence of a bad CMYK
separation or of a proof heavily distorted along the red dimension (typical
effect of OB), and the only way to improve the match for the client, on
press, short of going back to Photoshop or reoutputting proofs would be to
bring the magenta density up like crazy. Clearly, to my knowledge, no one in
any company in the world would allow calibrating a printing presses for
everyday, commercial work at these supposedly StatusT levels of ink
densities.
But it was interesting in this case to see what the corresponding CIELab
value of that Magenta swatch was. Curious? L= 44 a*= 84 b*= -1. Imagine, a*
of 84! A dream come true for Edmund Ronalds, when the ISO-12647-2 and
GRACoL2006_C1 specs (the same, in fact) are both at a* = 74. So, imagine
what a magenta of a* = 84 do on a press? Huge gamut boost in reds and blue.
And we're not talking "high pigmented" inks, here, regular, every-day
StatusT process magenta. That is not say that a press would be able to
handle such super high ink film thickness for any amount of run length,
though.
That's why I am a little skeptical at a combination of 2.03C, 1.85M, 1.56Y,
2.24K in StatusT (as compared to "SWOP"). And, apparently, that's how this
Japanese printer has been operating eversince. How silly? StatusG, perhaps,
but not StatusT. Can't compare apple with oranges.
> I don't recall reading information anywhere on which RGB working space
> was used for the master files, but it would make sense if it was ProPhoto
> RGB.
I call that an omission on the part of the author. Unless, the PDF was
destined to wow uninformed readers.
But I have to recognize that at the time the book came out, ICC color
management was not as widespread as it is today in NorthAmerica. In
fairness.
> Good question. I don't know how the proofing was done. Perhaps the
> inkjet
> proofs were seen as just an approximation, with the knowledge that the
> final
> outcome would be brighter still.
Right. That's how think the book was ran. In my opinion, all they did was to
bring back the press up to the same (StatusG?) density levels that was used
to characterize -- no more. Whathever colors came out of the press then was
the colors that went into the final book. Granted, the author was standing
by the press when the sheets came out and probably saw color he liked but
there is nothing in the PDF that says that some kind of colorimetric match
was achieved.
> Also, the subject matter is abstract enough that there are no "memory
> colors" involved (skin tones, skies, grass, etc.). So there is really
> no way to be "wrong" in the results, and one can let the chips fall where
they
> may.
Depends. Suppose I'm the next client at this Japanese printer and I agree to
use the author's custom profile for my separations. Would I accept turning
my job over to the printer without some kind of assurance prior to running
the job what to expect the color to look like? I don't think we're there
yet. That's like handing anyone a blank check : you write the amount while I
sign blindly at the bottom. Sounds like the present worldwide financial
crisis...
> > Perhaps the author could post the measurements somewhere for people
> > to examine?
>
> That would be a treat.
I know I would. But that's me.
Roger Breton
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden