Tuesday, April 10, 2018

Expanded gamut - when an idea's time has come

I enjoy researching the history of innovation.

Last week's post was the history of the creation of the board game Monopoly. It was a tale of innovation and deception; about good old capitalism at its best and at its worst. Today I will track the development of another innovation, but there is no deception and no villain in today's story. There is no moralistic message, just a practical message for would-be innovators.

So, mix up a Printer's Delight cocktail, and enjoy today's history lesson. For those not familiar with it, this drink is a combination of blue Curacao, Malbec, Yuengling, and Kahlua. The proportions are adjusted to match the appropriate color.

Try all 1,617 variations with the IT8 sampler!

The past few years, there has been a lot of hoopla about expanded gamut printing, which is to say, printing with inks beyond the standard four colors (cyan, magenta, yellow, and black). To the casual observer, one might conclude that this was a relatively recent idea. This is true... well... if one is willing to define recent in a more geologic sense.

Hallmark Cards

The earliest instance that I have found of expanded gamut printing dates all the way back to 1960 with a company called Hallmark Cards. If you are are not familiar with Hallmark, they invented the idea that you had to feel guilty about not sending a card to your mother on Mother's Day. I haven't found the original patent for the guilt thing, but it must have been filed somewhere around 1920.

Hallmark's unique printing problem is illustrated below. Greeting cards have a lot of pastel colors. These colors are often outside of the normal printing gamut. It is not always appreciated, but CMYK just can't get you highlights that are both very bright and highly saturated.


Karl Guyler (of Hallmark) said that Hallmark was in production with a six color process by 1962. The inks included fluorescent pink, yellow, and magenta. The system was so cool, that they had to give it a cool name that they trademarked: BigBox ColorTM. Although (as I mentioned before) Hallmark patented the guilt associated with their gilt, for some reason they did not patent this particular method of expanded gamut printing.

Why did this innovation happen? Hallmark had special niche needs that were not met by the existing technology.

Why didn't this innovation spread like wildfire? At the time, these were niche needs, and others didn't see a benefit.

Shoichi Shimada

The earliest patent I found on expanded gamut printing was from Shoichi Shimada of the Japanese company Dianippon Screen, filed in 1968. This patent describes a method where printing plates "are produced for reproducing color images with inks other than the standard inks …" They use three primary inks (cyan, magenta, and yellow), three secondary inks (for example, orange, green, and violet), along with black ink.

The diagram below shows in simplistic terms how the color separation is to work. Color space is divided into six pie slices. Actually, since color is three-dimensional, think of these as three apple slices. Each slice of the apple is assigned one of the process colors (CMY), one of the extra colors (in this case OGV), and black. The system determines which slice of apple the given color belongs in, and uses the appropriate inks for that apple slice to create the given color. 

Shimada-sam's color separation strategy

Shimada-san's patent mentions that printing of calicoes on fabric often uses a variety of non-standard inks to make images, but the color separation for this is "obtainable only by very complicated hand works [sic]". In patent parlance, printing of calicoes with special ink sets was admitted as prior art. To get a patent over prior art, one needs to demonstrate a novelty over the prior art which is non-obvious. One of the novelties in this patent is that they laid out an automated mechanism for replacing this very complicated hand works [sic].

Simple, easy to implement color separation technique

I was unsuccessful in finding out whether this invention was ever put to use. Google doesn't seem to know much about it. I have found precious little mention of the patent in my extensive search of technical papers on expanded gamut. I found the patent only through some pretty deep patent searching. The idea may have been actually turned into a product, but it certainly didn't become wildly successful.

Why did this innovation happen? According to the patent: "Such the special color inks [sic] are demanded frequently for the printing of color images which are difficult to produce by the combination of the standard inks."

Why didn't this innovation spread like wildfire? I can only guess, but it might be that the invention lacked a strong market driver. Yeah, it's nice to make more colorful images, but you can't build a better mousetrap until the world beats a path to your door.

Hang on to this thought, though. This patent describes something very similar to what we today call expanded gamut. The color separation is a bit different than what is done today. Currently, colors that are within the CMYK gamut are often printed with just CMYK, but otherwise, this looks a lot like what is done today.

Harald Küppers

Harald Küppers (often spelled either Kueppers or Keuppers, and rarely as kippers) developed a strategy for printing with more than CMYK, and filed for a patent in 1985. His method called for printing "… whereby the elemental surfaces which form the chromatic component are printed with a maximum of two of six chromatic printing inks, yellow, magenta-red, violet-blue, cyan blue, green and black…"

This deserves a bit of explanation. Kueppers' technique was not the traditional halftone technique as we all know and love. In his system, there is no overprinting of the halftone patterns of multiple inks. Instead, Kueppers divided the printed page into square cells, with each cell being divided into rectangular areas of up to four non-overlapping inks. Each rectangular cell is printed with certain rectangular areas of white and black -- the white area may be actually printed, or it may be the color of the substrate showing through. This gives the lightness. Each rectangular cell also is printed with up to two chromatic colors including yellow, magenta-red, violet-blue, cyan blue, and green.

The image below was taken from Kueppers' patent (4,812,899). It demonstrates the conversion of a color value (in this case described as a combination of violet, green, and orange) into a mosaic of printing inks. The colored box in the lower right corner is my own clarification of the box in his patent labelled Fig. 5c. Yeah, the colored box that looks like a Mondrian painting.

Keuppers' printing is a tiling of Mondrian blocks

Note that the little box labelled "S 25" is black (K 25). The German word for black is schwarz. In case you hadn't guessed, Kueppers lived in Germany.

If you squint real hard, this Mondrian of Kueppers tiles will look dark yellowish green

Anyone who has run a printing press will readily recognize that this technique is not practical for litho or flexo printing. Any small amount of misregistration will cause a color shift. For example, if the green ink were to be shifted a bit to the left, it would overlap with black. The color of the overlap would also be black, so misregister in that direction would cause a loss of green.

It is clear that Shimada-san's patent is more closely related to today's expanded gamut printing than Kueppers' patent. And yet, I have seen several references that claim that Kueppers' work is the forerunner of modern expanded gamut printing. Clearly that's wrong, since Shimada-san's invention not only predates Kueppers' work by almost 20 years, but Shimada-san's invention works a lot more like modern expanded gamut printing.

Why has history incorrectly attributed Kueppers with the invention of expanded gamut printing? I have several possible explanations:

1) Stigler’s Law of Eponymy: “No scientific discovery is named after its original discoverer.”

2) Kueppers' idea saw some market success. He produced a color matching book using his process, and apparently a bunch of these were sold. Hallmark's idea was forgotten since it was a niche solution to a different problem. Shimada-san's technique apparently did not create much hubbub. 

3) Kueppers' patent is a bit hard to understand. Much like Bob Dylan's lyrics, I think if Kueppers' expanded gamut printing were better understood, it would have gotten less credit. I readily admit to not understanding the patent the first 413 times I read it. I had been reading it with the assumption that it used normal halftone printing, so the diagram above (Fig. 5) was confusing. I assume that other print historians made the same mistake that I did. 

I didn't catch the idea of Kueppers' patent until I was reading through Kiran Deshpande's doctoral thesis, and saw his Mondrian diagram. Thank you, Kiran!

So, now my two innovation questions for the Kueppers approach to expanded gamut printing.

Why did this innovation happen? According to Kueppers' patent, he sought to solve the problem of moire patterns that are seen with conventional halftone printing, and he wanted solve the problem with conventional printing that "pure and luminous colors cannot be produced well." That is, he wanted to expand the gamut.

Why didn't this innovation spread like wildfire? The references below state that Kueppers' method languished because the color separations were manually intensive. I don't think that this is the case. The patent shows a scanner and a circuit for color separation. It is not necessary to have built an invention in order to get a patent, but it seems that the mechanics were fairly well developed when the patent was filed.

I think the more likely reason that the technique did not flourish is that, as stated before, it just was not practical for typical printing presses because of color shift due register.

Everything you thought you knew about the history of expanded gamut printing is wrong

Clearly, expanded gamut's time had not yet come. I'm getting tired now, so I will continue this in another blog post, which will be called the Heyday of Expanded Gamut Printing Patents

References

Bernasconi, Mathew, Color Printing Process and Product, US Patent 5,751,326, filed April 5, 1995

Boll, Harald, Color-to-colorant transformation for a seven ink process, Proc. SPIE 2170, Device-Independent Color Imaging, (15 April 1994)

Boll, Harold, and Scott Gregory, Color-to-ink transformation for extra-quarternary printing processes, US Patent 5,563,724, filed Oct 21, 1994

Cooper, Ted, Process for creating five to seven color separations used on a multicolor press, US Patent 5,687,300, filed March 27, 1995

Deshpande, Kiran, N-colour separation methods for accurate reproduction of spot colours, PhD thesis, University of the Arts London, May 2015

Guyler, Karl, Visualization of Expanded Printing Gamuts Using 3-Dimensional Convex Hulls, TAGA 2000

Herbert, Richard and Al DiBernando, Six-color process system, US Patent 5,734,800, filed Nov 29, 1994

Hutcheson, Don, Hi-Fi Color Growing Slowly, GATF 1999
http://www.hutchcolor.com/PDF/HiFiupdate98_2000_04.pdf

Küppers, Harald, Printing process where each incremental area is divided into a chromatic area and an achromatic area and wherein the achromatic areas are printed in black and white and the chromatic areas are printed in color subsections, US Patent 4,812,899, filed Jan 29, 1989

Küppers, Harald, Process for manufacturing systematic color tables or color charts for seven-color printing, and tables or charts produced by this process, US Patent 4,878,977, filed Nov. 7, 1985

Hamilton, Jim, High Fidelity Seven Ink Printing, Technical document from Linotype-Hell, 1994
http://www.greenharbor.com/LHTIfolder/lhti9405.pdf

Shimada, Shoichi, Apparatus for production of color separation records, US Patent 3,555,262, May 7, 1968

Viggiano, J A Stephen  and William J Hoagland, Colorant Selection for Six-Color Lithographic Printing, Proceedings of the IST/SID 1998 Color Imaging Conference, p 112 - 115

5 comments:

  1. Hi John, I remember at the Graphic Arts Technical
    College where I work, about 15 years ago we used Opaltone
    software to produce seven colour separations on the Crosfield Celsis drum scanner. CMYK plus Red, Green and Blue/violet to print in the saturated areas to extend the colour range. Results were definately an improvement.

    ReplyDelete
  2. I began working in Graphic Arts Enginering at Hallmark Cards right out of college in the mid-1960's. Hallmark was indeed printing many of their cards and other products with an expanded color gamut, using 6-colors. (long scale: Lt. cyan, fluorescent yellow, and fluorescent pink, and short scale: cyan, magenta, and black) This system was developed to allow our printed product to match the artwork from the 600+ artists in the Creative Division. (Oil, watercolor, assemblages, fabric, etc. Transparencies were separated for 4-c or 5-c, with a short scale red) Every quarter, random samples were pulled from our Distribution Center and placed in a display room along side the original artwork. Senior management would then grade how well the Graphic Arts Division matched and produced the artwork for that group of samples. In the 1960's, 6-c separations were created with custom, internally made, process cameras using our own electronic exposure controls created by our Graphic Arts Research Department. For several years the 6-c printing was done on 2-c presses, because 4-c presses of that era printed with too much dot gain. Working with our press suppliers, it was later possible to develop 3-c, then 6-c, later 8-c perfecting and finally 10-c perfecting presses, that could print the heavy ink film required for our very clean inks, without significant dot gain, usually on uncoated board stock. One of the challenges of Graphic Arts Engineering was making sure the 13 printers printing our products at that time, all printed to Hallmark's halftone and solid ink density standards, which allowed us to produce the needed 8-12 million cards per day.

    ReplyDelete
    Replies
    1. Thanks Jim! I greatly appreciate hearing this first hand account!!!

      Delete
  3. nice information. keep posting this blog to help the beginner. Color Separation

    ReplyDelete