Spot channels are just selections/ channels like any other. Obviously the trick in your case would be to retain the fine structures in the CMYK. Seems to be mostly on the black side in your example. If you already have a separated CMYK version, it would probably be pretty simple to extract the info using a selecion from the M and C chanells and then do some voodoo with image calculations....
The reason I am doing this is because the artist’s works were represented on many publications over the years from many different countries. And we found out that CMYK color alone cannot duplicate the blue color very well (I mean the blue only, no problem with other colors). I know, it is hard to believe, but if you see all the publications and compare it to the original art, you would see the huge difference in color saturation and luminosity.
That is not hard to believe at all for people with experience in prepress and/or printing.
Getting a halfway decent Spot Channel (possibly with the help of a Lab duplicate of the image) may be possible, but determining how much K or also C and M to keep under that could prove difficult.
And you should realize that in such a case neither the on-screen preview nor digital proof are likely to be fully trust-worthy, because while they may be able to finely represent full tone spot colors, percentages of that mixed with other colors are a complex issue that would be difficult to simulate perfectly.
So your best bet would be to demand a press proof (on the machine where the run will be printed) and on the actual paper – and that will probably run into time and money.
Just adding a spot channel is hardly a solution for arbitrary images.
In your application you may print a selection by Pantone Reflexblue,
but that would be more an artificial graphic instead of a correct repro-
duction (you had expressed the need for additional CMYK, I know).
One needs a multicolor printing process, using a fixed set of inks, and
the appropriate ICC profile for a printing process, probably on coated
paper. More inks can be used to increase the gamut.
One of these systems is Pantone Hexachrome. Orange and green inks
are added to CMYK and surprisingly this extends the gamut for blues
considerably as well.
According to doc (1) below, a direct support by 'Pantone Hexware'
is actually not available, but it seems there a workarounds.
I don't have experiences with Hexachrome or similar systems, but I
wanted to point out, that printing photos with large gamuts is hardly
possible by adding an arbitrary spot ink. A multicolor workflow has
to be established, including the appropriate ICC profile.
Best regards --Gernot Hoffmann
(1) Features and history of Hexachrome
(2) How to apply Hexachrome in Photoshop
"if you just need to separate RGB to Hexachrome, you can download
a generic Hexachrome ICC profile from Pantone's website
Install profile, and you can use Photoshop CS5 to convert color, choose
'Convert to Profile', click the Advanced button, and choose your profile
from the Multichannel profiles."
(4) Hexachrome gamut
(5) Hexachrome primaries (p.15+16)
Thanks for the replies. Most people suggested me that getting a decent result from adding a spot color will require too much trials and tests. I think you are probably right.
Hoffman's links to Hexachrome printing was quite interesting. But I wonder how easy it is to find a print shop that does Hexachrome printing. We are thinking about printing in Korea as their printing quality seem to be quite good and the price is not over the head. I don't know if there are Hexachrome printing shops in Korea.
Also, I was thinking how adding orange and green will make a better Blue color in printing. Theoretically Orange is opposite of blue and when mixed, they will cancel each other. And there isn't any green mixture in the blue color that we are trying to reproduce. I am sure hexachrome will make better colors in other areas such as green shades and skin tones etc...
Honestly saying, I didn't really understand the color patches in this document. I don't know if it is about LAB colors or Hexachrome colors. What does CIE stand for in this document?
Understanding color space diagrams for CIE(1931) and CIELab isn't easy,
but gradually one gets used to interpret them by intuition.
(1) helps understanding CIE(1931) chromaticity.
The horseshoe contour contains all possible colors with luminance left
out. Indicated colors are just for orientation. Yellow is the 'locus' for
bright yellow and brown (dark yellow) as well. White is the locus for
white and black - colors without saturation at a certain center. For two
colors on the same ray to the center, the color with larger distance is
more brilliant or vibrant.
You're right - adding green and orange ink in Hexachrome doesn't
extend the gamut at the blue side. But Cyan and Magenta (and Yellow
and Black) are different to common CMYK inks as well, and this
delivers more brilliant blues.
(2) shows the effect, now in CIELab in a horizontal slice for constant
lightness. Besides theoretical aspects, the diagram can be (again) inter-
preted by intuition.
The complete diagram is threedimensional. A shown color is not just
a placeholder (as in the CIE chromaticity diagram) but a more or less
correct reproduction, as good as possible, depending on the medium.
It seems indeed that there are few print houses using Hexachrome,
but (3) is one of them.
What's to do for the actual catalog? Printing by inkjet would be a
solution, because inkjets can use additional inks like Green, Orange,
Blue, mostly by replacing standard inks like LightCyan, LightMagenta,
Gray. The inkjet can be calibrated by GretagMacbeth ProfileMaker's
Multicolor Module. That would be very expensive. Even without
additional inks the blues can be reproduced fairly good.
As already said - just adding a Spot Blue would cure the problem
for some paintings, but there is no systematical workflow.
If Hexachrome is not an option, then one may try to modify the
blue in the image by shifting it towards cyan with less lightness,
which is better printable.
Of course wrong, but what finally counts is the impression.
A friend of mine is a famous German photographer for calendars
and tourist guide books which contain plenty images with blue skies.
So far he got almost always pleasant print results - by applying
appropriate image processing with Soft Proofing in Photoshop.
Examples are in (4).
It would be nice, if somebody who is practically working with
Hexachrome or other Multicolor processes could contribute.
About this question:
I don't know if it is about LAB colors or Hexachrome colors.
The CIE (1931) color space appears mainly by two representations:
1. CIE xyY Chromaticity diagram (horseshoe)
2. CIELab = Lab
These are color spaces which contain all possible colors. Spot inks
and primary inks (Pantone, CMYK, Hexachrome CMYKOG) can
be shown in all diagrams.
Best regards --Gernot Hoffmann
http://www.google.de/url?sa=t&rct=j&q=&source=web&cd=1&ved=0CDUQFjAA&url=http%3A%2F%2Fwww. pantone.com%2Fdownloads%2Farticles%2Fpdfs%2Fart_hex_primer.pdf&ei=CTJQUKnOEaWr0QWfg4DoDg&u sg=AFQjCNFx7-5XXealXZGPTQ5ek-A7FWb8gQ&cad=rja
Edited by author
This is the original:
The next image shows the best approximation by CMYK, using the process as
defined by profile ISOCoated v2 eci
Magenta areas indicate colors which are still out of gamut, just to show that
the settings are marginal.
The original blue has Lab values Lab = 25 42 -82. That's near to Pantone
Blue 072C with Lab = 19 40 -79. It seems this and your other images were
painted with such an extreme paint (painting color), a kind of pure pigment?
That's not only very far from the gamut of offset inks. Furtheron an auto-
matic conversion would lead inevitably to a magenta cast.
The next image shows a more favourable situation, using a calibrated inkjet
Hope this helps a little.
Best regards --Gernot Hoffmann
Sorry for the mysterious line breaks...
Thank you for guiding me to links with good information. I tried to read them but some of them seem to be a way over my head. I have never worked with LAB colors before, so I don't really understand them. But most of these links is suggesting me that Lab colors have some advantages over RGB. If there are some advantages can you tell me what they are. Most photographers work in RGB and most printers take RGB files, as well as CMYK.
You suggested me to use Inkjet Printer for the catalog. However, I think it is not an option for us because the book is going to have 300 pages and it will be printed 1000 copies. Plus, I think, offset printing ads a value to the book, by looking more factory made vs. hand made. But it is just my opinion. In other words, the book has to be printed in offset but the matter is how to print the blue color right. Again we need to do it in Korea.
The painter didm't really use any type of extreme paint or color. It is just a regular tube of Golden acrylics with French Ultramarine color. The problem is that mixing CMYK ink doesn't make the blue as much saturated as a tube of paint. I think CMYK had its limitation from the time they invented it, but I am not accepting the limitation. That is probably my problem.
It has been quite long. But I wanted to ask few more questions. I am still working on the book project of the artist and the color issue is still a problem. I have paid for a hexachrome test run in China with less than satisfactory result. I should have predicted it in advance since the additional two inks were Orange and Green which has nothing to do with the blue color. I am thinking to use Pantone spot color such as Blue 072C with the mixture of CMYK. I tried to make a spot channel in Photoshop by selecting the Blue 072C color range and making it as the fifth channel. The problem is that when I convert the image to CMYK, the cmyk colors are still under the spot channel. That means if I print it just like that, the Blue colored area of the image will come out very dark. I don't know how to adjust the CMYK channels behind the spot channel. You mentioned something about IMAGE CALCULATIONS... Can you explain me a little more about that and how it works?
c.pfaffenbichler said something important to consider. I didn't notice it previously.
Although it will cost us more money, we still want to make it look as best as possible. Because with the normal CMYK printing, it doesn't look half as good as the original work.
It has been very long since I started this thread. But finally I was able to make a test print using the Spot Blue colors. The quality of the blue color has improved a lot. But the tonality and the integration of the spot color with the rest of the standard colors is a mess. I have paid a lot of money to run this test. And I want it to work. If I can only solve the problem of integrating the spot color with the rest of the 4 colors, my job will be successful. Is any one willing to discuss about this further? I can provide the sample images and talk more in details. I hope there is a solution.
Just found this, it's probably way too late.
But what you're looking for is Spot Process (or Simulated Process) separations, where spot colors are blended seamlessly (with or without CMYK) being involved to create photo-realistic gradients.
This is the default separation type for photo-realistic screen prints on textiles. It's also used frequently in flexographic printing for packaging.
These aren't easy to do - professional artists train on these extensively and most seps done this way are done more "by hand" by experienced artist than "mathematically correctly by software."
But there is software that does these seps, what you're looking for is Special Color ICISS. It's not cheap, so if you only need one sep, what you're really looking for is an offset-press separator who already owns that software. They should be able to generate seps for you that nail the blue because it will be printed primarily with a spot ink, but where it also blends beautifully with the rest of the image and has proper contrast and detail.