Thursday, May 17, 2018

Do you remember a logo?

I stumbled across a quote the other day that I found interesting. This was on the Coca-Cola website:

"There is no Pantone color for Coca-Cola red, but when you see it, you know it."

Ah! To be tanning under the Coke Red Sun!

This sounds like one of those factoids that everyone knows is true, so nobody would be crazy enough to actually test it. Well, guess what? I know a few crazy people. In fact, one of my best friends, Eddy Hagen, has recently tested this very thing with an online test: how well can you pick out Coke red?

(As an aside, here is the process for joining the John the Math Guy's Best Friend Club: Connect with me on social media. Contact me somehow or other with a message that does not contain the phrase "John the Math Guy is a doofus." Then you're in. If you just want to get on my email list, then send an email to john@johnthemathguy.com to subscribe.)

There are two tests in Eddy's blog. In the first test, Eddy tests your short-term color memory . You are shown a color, and then asked to pick it out of a line-up later. That one is kind of a warm-up to the real test. In the second test, he shows you a bunch of colors and asks you to pick out Coke red.

Do people know Coke red when they see it, as the Coca-Cole website suggests? He shares the results in another blog post. I don't wanna give anything away, but the title of this post is You can’t correctly remember an iconic color, not even Coca-Cola red.

Which one makes you thirsty?

Who is right??!?!? Let's get to the bottom of this!

Brand color is important

Brand colors are important, especially if you have a brand to sell. Here is what Axel Kling (Print Quality Assurance Manager for Coke) has to say about the importance of brand colors:

In today’s marketplace of unlimited beverage choices, a brand’s first point of contact is most likely to be at the point of purchase. And how well your product stands out on shelf could determine whether it’s put in the shopping cart or left behind.

I know most of my readers have private chefs who do their grocery shopping, but imagine if you will, being in the snack aisle of a grocery. You are trying to find your favorite bran cereal with raisins. Just reading that line, I'm gonna guess that you're thinking "purple". Am I right??! Of course, the image below wasn't any clue.

When I am old, I shall eat cereal out of purple boxes

The bran owners of the various raisin brands have trained their cereal boxes to be distinctive colors so that they can jump off the shelf into your shopping cart. And lets, face it. Nothing says "raisin bran demographic" quite like the color purple.

This is an aside, but how can Kellog's and Post and Trader Joe's and Total and John the Math Guy Breakfast Foods all use the name Raisin Bran? Interesting trademark factoid: The Skinner Manufacturing Company was the first to sell raisin bran, back in 1926. It trademarked the name, but in 1944, the Supreme Court rescinded the trademark, saying that you can't trademark a simple description of a product.

Speaking of trademarks, the color purple, and brand colors ... In 2004 Cadbury applied for a British trademark for the color purple (Pantone 2685C)  "applied to the whole visible surface, or being the predominant colour applied to the whole visible surface, of the packaging of the goods." Nestle objected, and their application was denied. It seems you kinda have to have a mark, if you want to have it trademarked. But, this trademark application in 2004 was a revision of an earlier trademark from 1995, which is still in force, at least until Nestle contests that trademark. 

Imagine my surprise when I found out that the chocolate wasn't purple!

This is in the UK. I apologize in advance to my British friends and enemies, but I'm not all that excited about British law. I mean, back in 1492, we fought the Spanish-American war to get away from having to follow your laws about tacks in our tea. What about US trademark law and colors?

I read up a bit on Wikipedia about color trademarks. In the US, you can trademark a color so long as it serves no other purpose other than to distinguish your product. So, Johnson & Johnson can trademark the name Band Aid, but not the color, since that serves as camouflage on certain people's skin.

There are a number of colors that are trademarked in the US, as shown in the image below. I compiled these from the Wikipedia article and the Business Insider article Can You Identify These 12 Brands By Their Trademarked Colors Alone?


I am gonna conclude that at the very least, brand owners think that brand colors are important.

It's not just about being able to find your favorite cereal

Color is about brand recognition, It helps you find a specific product within a dazzling array of colors. But the prevailing wisdom is that it also communicates something about a product. Red universally means romance or hookers, except when it's used on a fire truck or a stop sign. And of course, it doesn't mean romance if you are in China, where red signifies joy and luck. Or on one of my earlier blog posts where I decided it just signifies excitement, which explains why double-decker buses are red. But trust me. The meaning of a color is universal and unambiguous.

John spent the better part of an afternoon looking for his cereal

I have heard several presentations at conferences where the speaker says something like "color accounts for 86.3% of our buying decisions". As a math guy, I know that 95.4% of all statistics are made up, so, is there any definitive research behind the importance of brand color? Or is this just one of those statistics that gets quoted enough so that it becomes established fact?

Here is a quote from Daivata Patil that sounds authoritative:

Color is ubiquitous and is a source of information. People make up their minds within 90 seconds of their initial interactions with either people or products. About 62‐90 percent of the assessment is based on colors alone. 

Authoritative, with numbers and everything. But the article does not describe how these numbers were determined or even give a reference to where they came from. Hmmm.... urban legend?

Here is similar quote from Axel's presentation. Remember Axel? The color guy with Coke? He attributes this quote to Jill Morton's Color Matters website. Both attribute it to Loyola University.

Color increases brand recognition by up to 80%. 

I googled this quote to try to find a link to the actual study. Note that I put quotes around the words so that Google knew that I was looking for those exact words in that order. Goggle told me there were "About 2,170,000 results"! I admit to not reading through them all. I looked at the first ten hits, trying to find the title or author of the study, or maybe a link. All of them mention Loyola, and several of the web pages reference Jill Morton. None of them give any more information about the study.

Time for an infamous John the Math Guy tirade. This is not Science. I'm not saying that I have reason to doubt the statement, or that the various places that provide this quote are required to track down and report the original source. It's just that, for me, I would like to assess the strength of the argument. Was this an undergrad student who made up the numbers the night before the term paper was due? A professor who assembled twenty students for a little test? Or was this a master's thesis with hundreds of volunteers following a rigid experimental protocol?

Gregory Ciotti expresses my concern a bit more emphatically than wishy-washy me:

Most of today's conversations on colors and persuasion consist of hunches, anecdotal evidence and advertisers blowing smoke about "colors and the mind."

Getting back to the topic

Let me take a minute to try to remember where I was going with all of this. Oh yeah. Eddy Hagen's experiment about Coke red recognition.

Eddy's online experiment carefully explains the methodology and the results. It's Science, but I'm not gonna claim that Eddy's online experiment is good solid Science, and I don't think Eddy would either. He acknowledges that not all monitors are calibrated, and surveys where the participants are self-selected are a bit less rigorous that random selection. It could be that zealous PepsiCo employees deliberately failed the test to discredit their competitor. Or it could be that some of the individuals clicked at random just cuz it was late at night and they were waiting for the pizza guy to arrive.

Signs.com did another test of people's ability to recall brand logos. They brought in 156 people, and had them draw the logos of ten well-known companies from memory. This involved recalling not only color, but the shape and text of the logo.

Can you draw these from memory?

They have some stats on various aspects of the logos, but I did my own counting. I looking only at whether they got all the right colors, without adding extraneous ones. The results below are not all that fabulous, especially for multi-colored logos.

Company
Correct
Colors
7-Eleven
14%
Green, Orange, Red
Burger King
20%
Blue, Orange, Red
Foot Locker
38%
Black, Red
Walmart
49%
Blue, Yellow
Dominos
53%
Blue, Red
Starbucks
53%
Green
Ikea
56%
Blue, Yellow
Apple
72%
Black
Target
84%
Red
Adidas
92%
Black

I will point out that I was rather lenient about allowing different shades of the correct color. I allowed an orange flavored yellow to count as a yellow, or for Ikea blue to be too light or too dark. The image below shows the variation in color for Satyrbucks, which uses only green in the logo.

156 guesses at Starbucks green

At the far left, you see all 156 logos as drawn by the participants in the survey. (You can see a full sized version of this on Signs.com website.) In the middle drawing, I pulled all of the green pixels from each drawing, and averaged them together to show the green that the participant chose. At the far right, I show the 21 contestants that came within 10 DE2000 of the true Starbucks green. For reference, a common tolerance for commercial printing a color is 3.0 DE2000. Only two people out of the 156 participants were able to create a color from memory that would have been deemed acceptable printing of that logo.

Caveats...

I have made the assumption that there was an unbroken chain of proper color management throughout this process. If I had to put money on that, I would say that I would prefer to not put money on that. I don't say that to disparage Signs.com at all. I just know that the bar for rigor in Science is pretty high. But, looking at the middle image above... I rather doubt that any deficiencies in the rigor of this test could have caused that much variation in color.

Another caveat is in the interpretation of the results. This is a test of the participant's ability to recall the proper color from memory (as in Eddy's Coke red test), but also a test of the participant's ability to reproduce that color using the software provided. So, the logo drawing test is harder than the task of trying to find your favorite raisin bran. 

Pondering

Eddy provided me with an interesting anecdote: "To put that unique Coke red in perspective: in the LinkedIn ‘Printing Production Professionals’ one of the printers that works for Coca-Cola shared that in the X-mas edition, the Coke red is slightly darker… (which I checked in my collection of Coke cans and it is correct) So if color is soooooo important, how does this different Coke red impact sales?"

I'm still kinda pondering why Eddy has a Coke can collection... but these two experiments beg the question about how precisely a brand color needs to be defined. Both experiments are well above the level of urban legend expressed by the statement "Color increases brand recognition by up to 80%". But neither experiment quite fulfills the high bar of rigor required to be accepted as peer-reviewed Science with a capital S. I don't expect to see either in the next edition of Color Research and Application.

But, the two experiments are suggestive, and that suggestion is a contradiction between the brand owner's expectations of what is needed and the psycho-physics of the color that we see.

In the next installment in this series, I will take a closer look at the Science that has been done, especially the Science having to do with our memory of colors. If you want a bit of a foretaste, look through the references below. I am going to pretend to have digested them in the next blog post.

References

Bae, G. L., M. Olkkonen, S. Allred, and J. Flombaum, Why some colors appear more memorable than others: A model combining categories and particulars in color working memory, J Exp Psychol Gen. 2015 Aug;144(4):744-63

Belcher, Teri, and Kevin Harvey, The Influence of Color, ANTEC 2007

Bartleson, C. J., Memory Colors of Familiar Objects, Journal of the Optical Society of America, Vol 50, No 1, Jan 1960

Burnham, Robert W., and Joyce Clark, A Color Memory Test, Journal of the Optical Society of America, Vol 44, No 8, Aug 1954

Ciotti, Gregory, The Psychology of Color in Marketing and Branding
https://www.entrepreneur.com/article/233843

Cunningham, Meagan, The Value of Color Research in Brand Strategy, Open Journal of Social Sciences, 2017, 5, 186-196

Elliot, Andrew J., Color and psychological functioning: a review of theoretical and empirical work, Frontiers in Psychology, April 2015, Vol 6, Article 368

Goguen, Kate, The Influence of color on purchasing decisions related to product design, Master's Thesis, Rochester Institute of Technology, Feb 20, 2012

Javed, Saad Ahmed and Sara Javed, The impact of product’s packaging color on customers’ buying preferences under time pressure, Marketing and Branding Research 2(2015) 4-14

Kling, Axel, The Importance of Color Management for a Consumer Product Company, Printing Industries of America Color Management Conference, 2011

Patil, Daivata, Coloring consumer`s psychology using different shades the role of perception of colors by consumers in consumer decision making process: a micro study of select departmental stores in Mumbai city, India, Journal of Business and Retail Management Research (JBRMR) Vol 7 Issue 1 October 2012

Mohebbi, Behzad, The art of packaging: An investigation into the role of color in packaging, marketing, and branding, International Journal of Organizational Leadership 3(2014) 92-102

Morton, Jill, Color & Branding, Color Matters
https://www.colormatters.com/component/content/article/62-color-a-marketing/240-color-a-branding

Satyendra Singh, Impact of color on marketing, Management Decision, 2006, Vol. 44 Issue: 6, pp.783-789

Tuesday, May 8, 2018

The heyday of expanded gamut printing patents

In the previous installment of this series on the history of expanded gamut printing, I chronicled three times where augmenting CMYK with a few extra colors was independently invented. There was such an uproar to my post that I had to write an addendum to add all the examples that I had missed.

At Hallmark Cards, the technique served a niche need for that time and place. The work of Harald Kueppers seems to have found a different niche, and gained some attention, but it saw limited use. And the developments at Dainippon, while they were very similar to what we see today, have left little trace in the history books. All the examples cited by friends also wound up being niches.

Sad fact: expanded gamut did not go mainstream during this time period.

One of the commentators on my post commented a comment about the futility of doing patent searches to dig up history. So today, I look exclusively at the patent record. In this blog post, I look at a period of twelve months in 1994 and 1995. These 372 days rocked the world of CMYK printing to its the very foundations. You think I'm being overly melodramatic? Consider this: These 372 days saw not one, not two, not three, four, or five, not six, but seven filings for patents on expanded gamut printing. CMYK printing. World of CMYK printing, consider yourself rocked.

Expand my expanded gamut, baby!

Hutcheson, Du Pont, March 29, 1994

If one cyan print unit is good, then two must be better, right? And if you print with two cyan print units, why not two magenta, and two yellow, and two black? The idea is to give a double bump anywhere that you need more ink than a single print unit can provide. This was invented by the very modest Don Hutcheson, and marketed by Dupont under the name HyperColor. I guess someone vetoed the name HyperDon.

Don'cha just love the cute drawings in patents?

The technique can be considered an expanded gamut process, since it does expand the gamut. It just uses CMYK as the additional colors, instead of OGV or some other collection.

Some of my readers may have met Don Hutcheson. He is still in the business, and is actually still working on this project. Idealliance's XCMYK project is a way to expand the gamut by pushing the standard CMYK to higher densities. After 24 years, one would hope that he will be making some progress soon.

A note on patents: Patents generally have a section at the beginning that describes the background of the invention or alternately, the prior art, that is, the existing stuff related to the new stuff being invented. The title prior art is actually shorthand for prior art bashing, since this section usually highlights the deficiencies of what is already out there. The prior art section is followed by an obligatory section  where the inventor explicitly states the purpose of the invention. This section is obligatory because patents have to be for something useful. I found that out the hard way when I tried to patent a wind tunnel with left-handed ear flaps!

What was the purpose for this (I mean Don's) invention? A quote from the patent: "... it is believed to be advantageous to provide a method for extending the color printing density range of a printing device without introducing special or non-process printing inks or unconventional pre-press proofing systems..."

Plettinck and Van de Capelle, Barco, April 29, 1994

Technically, this is not a printing method. It is a way to convert one color separation (based on CMYK) into another separation based on non-standard inks. What do they mean by "non-standard"? Here is an example from the patent:

For example, a chocolate manufacturer will prefer an ink set wherein brown ink plays a more dominant part.... So for example PANTONE (registered trademark) red, process yellow, and PANTONE brown form a set of non-standard inks that are used for printing packaging material for chocolate.

Reading the patent makes my mouth water!

Well... ok... maybe this isn't really expanding the gamut, although it could. I just couldn't pass up a patent that talked about chocolate. Those of you who are chocolate fanatics will understand.

What is the purpose of this invention? "The object of the invention is to provide a method for generating printing data wherein the second colour separation is determined in a more efficient and non-empirical way and a result is that the printing quality remains unchanged or is even increased."

Eder and Maerz, Eder Repros Offset Repro GMBH, May 19, 1994

This patent is in German, so I admit to not having read it in full. Well, actually, I didn't read any of it. But, I can tell you that Eder has been described by Anastasios Politis as: "One  of  the  most  significant  pioneers  in  processing  CMYK  +  x  colors...". I also know that Linotype-Hell marketed the Eder software under the name Eder MCS (Multi-Color Separation). More on that in a bit...

The company Eder still exists, and is doing software under the byline "product communication in the digital age".

Printing of the King Eider duck may benefit from ederMCS color separation

What was the purpose of the invention? I did some OCR on images from the pdf of the patent, and translated the German text into English: "It is therefore an object of the invention to provide a method for creating a color print image, with the help of which create high-brilliance color images, the required printing effort is reduced compared to the seven-color printing."

What is the purpose of me asking that question all the time? Please be patient. I am actually going somewhere with this. Suffice it to say that, so far, making prettier pictures has been the main goal so far.

Boll and Gregory, Eastman Kodak, October 21, 1994

One of the many things I like about writing patents is that the patent writer is allowed to be his or her own lexicographer. That means they can make up words! The title of this patent contains the word extra-quarternary, which I take to mean "beyond four".

Some comments on this uber-cool word. First, Harold Boll told me in an email: "I longingly love that word too, mainly because it should have been in the title of my first patent!" In the body of his patent, he used the word extra-quaternary. Due to a clerical error, an r was added to the word: extra-quarternary

Yoko was an extra-quaternary

Second comment on the word: several writers have used the term extra-trinary to connote expanded gamut printing. This is just plain wrong!

Kodak first got their feet wet working on a profile for Pantone's expanded gamut product, Hexachrome. More on Hexachrome later... 

Kodak had at least one major customer for their software product, Hallmark Cards. If you remember all the way back to the first blog post in this series, you will recall that Hallmark was big into expanding their gamut in the 60's.

I'm getting ahead of myself a bit here, but Kodak introduced an expanded gamut product called Spotless in 2011, 17 years after the Boll and Gregory patent. It is likely that there is not direct connection between the work of Harold and Spotless. Why would Kodak jump back into the expanded gamut ballpark? Hang onto that thought. I will come back to it.

What was the purpose of this invention? "It is another object of the present invention that it is uniquely capable of exploiting all of the attainable color gamut afforded by an n-ink (n>4) printing process and thereby achieves maximum colorfulness for rendered colors."

Maximum colorfulness... yum.

Herbert and DiBernardo, Pantone, November 29, 1994

Everyone in the print industry knows of Pantone. Lawrence Herbert is the guy who started Pantone. His son, Richard Herbert, took over the reigns. Lawrence and Al DiBernardo are the guys who invented Hexchrome, which was perhaps the most well-known of the mid 1990s commercial offerings for expanded gamut printing.

This system uses orange and green as the additional colors (there is no additional blue or violet ink). They wanted to keep the number of inks down to six, so as to make it usable on more presses. The ink set also includes richer CMYK inks, and some of the inks are fluorescent so as to make them more vibrant.

One of the things that distinguished Hexachrome is that they had a special fandeck for the Hexachrome colors. These guides had all the colors in their regular book, but with one difference. The regular Pantone guides have a recipe for how to mix each color in a bucket of ink. The Hexchrome guides have a recipe for how to mix halftones on press to make the color.

 Still available on ebay

If the fan deck of expanded gamut colors came from any other company, I would say that this was a brilliant marketing move. It certainly raised the awareness of the product to have a physical sample of the system. But since Pantone was kinda in the business of making fan decks, it wasn't so much brilliant as it was obvious.

What is the purpose of this invention? The first few words of the summary are: "A printing system for high fidelity printing of an image is provided..." 

According to the patent, Hexachrome is all abut making high fidelity colors. But (important point here for my narrative) the Hexachrome book really can't be used to make pictures.  

Seinfeld's 100th episode, February 2, 1995

The 100th episode of Seinfeld aired during the 372 days that rocked the world of CMYK printing. Coincidence?

Jerry Seinfeld has yet to comment
on his alleged links to expanded gamut printing 

Cooper, Linotype-Hell, March 27, 1995

This patent is a two-step process. First the CMYK separation is created, and then a correction is determined. This is all pretty obvious when looking at the diagram below from the patent.

The patent office desperately needs a service for
colorizing gorgeous drawing from old patents

Linotype-Hell released this as HiFi Color 3000 in 1994. In 1995, they announced that they would be selling the ederMCS package. It would have been interesting to have been a fly on the wall for the discussions they had about switching over to someone else's product.

What is the purpose for this invention?  I will skip the patent, and go to a press release for High Fidelity Color Printing:

Why would anyone want to print seven inks?
    • Seven inks can print a larger color gamut which includes colors that the four process color inks cannot achieve.
    • Seven inks can achieve a brighter color appearance and improved modulation of color.
    • Seven inks allow closer color matches to the original.
    • Printing with seven inks produces cleaner reds, greens, and blues.

Here is an interesting quote from the patent: "Spot colors are not considered in this application." Hang onto that thought. I will get back to it. Really. I am getting to something.

Bernasconi, Opaltone, April 5, 1995

Mathew Bernasconi developed a system of expanded gamut which uses CMYK+RGB. This is one of the few systems patented in the heyday of expanded gamut printing patents that has survived. This patent covers a device which scans a photograph and determines a set of color separations. Conceptually, there are two scans. The first scan is that of a traditional scanner, where a CMYK separation is done. The second scan creates a separation for the expanded inks to make up for the limitations of the first separation.

One difference between this patent and the others is that the extra-quaternary colors are preferably red, green, and blue instead of orange, green, and violet. Bernasconi explains the use of red over orange,

Orange is not a primary color, it’s a secondary (i.e mixed from red & green light). Therefore using orange ink in an expanded system actually restricts the color gamut. A red primary mixed with a yellow primary creates pure orange (see overprint image below) thus expanding the gamut whereby the red primary is also mixed with magenta to create “scarlet” reds. The hue angle difference between 100% overprint (R+Y) & (R+M) should be >30º. This hue angle difference cannot be achieved with 100%  (O+Y) & (O+M) because the orange is too yellow from the outset.

This is not the NBC peacock

Just in case you are getting a bit confused about which set of colors are being used as primaries, I provide the comparison in the image below. Which one is correct? The concept of primaries is based on RGB color theory, which is a simplification of color science. So, primaries are not really defined in color science. If, however, you seek your primaries based on color engineering, then the correct primaries are whatever set of pigments give you the biggest gamut. Finally, if you are a color practitioner, then the correct primaries are the ones that get you the colors that you want.

Comparison of the chosen ink sets

Another note on patents in general: the body of any patent describes specific embodiments of the invention. In the case of Bernasconi's patent, the addition of red, green, and blue inks to CMY is an embodiment. But the teeth of a patent is in the claims. The claims are generally much broader, covering many different embodiments. In this case, the first claim refers to "a plurality of data channels", instead of listing a specific set of inks. This means that the Opaltone patent could cover CMY+RGB (the preferred embodiment), or it could cover CMYK+OGV or RGB+CMYK.

In a much later patent (2011), Bernasconi described a CMY+RGB variation on this invention. Instead of using black ink, this system mixes red, green, and blue inks to make black. In this way, expanded gamut printing can be done on a six color press.

What was the purpose for this invention? "... saturated colours such as deep reds, greens and blues cannot be reproduced satisfactorily due to the limited print range of four colour process."

One more expanded gamut effort

Mark Mazur acquainted me with another expanded gamut effort in this time frame. He says that it was the first product in the packaging industry that allowed the user to select his own set of pigments.

The company is called Specialcolor. According to their website, they started selling expanded gamut color separation software (under the name ICISS) in November of 1995. This is just after the 373 days that shook the very foundations of the CMYK world, but I would argue that, had Glynn Hartley decided to file a patent, it would have been in the critical time period.

I did search for patents from Glynn. Couldn't find any in the US or the European database. His website doesn't list any patents, so I am guessing he never filed. That's not to say that he didn't invent anything that was patentable. I think it's a pretty good bet that there was something is this effort that would be inventive enough to get a few claims in a patent.

So what happened?

In 1991, Don Carli made a bold prediction "High Fidelity methodologies ... Represent a revenue opportunity potentially accounting for as much as 15 - 20% of the $150 billion dollar world-wide color printing market by the end of the decade." Speaking of cool made-up words, I should mention that Don Carli coined the phrase HiFi printing. I also remember hearing him refer to stochastic screening as sarcastic screening. Love the guy.

These predictions were enabled by technology in the mid 1990s. Back in the old days, the thing that made plates was a combination scanner (to scan the films), computer (digital or analog, to do the math for color conversion and screening, and platesetter (to make the physical plates). This is a pretty closed system. Only a few engineers really got a chance to play with the cool stuff inside. This changed in the early 1990s.

Apple provided affordable workstations that could play with images before they went to plate -- desktop publishing. In 1994, Creo introduced the first Computer To Plate (CTP) system. With these two pieces, a larger group of engineers could play with the way color is separated, and then make plates.

Gary Field points to another necessary technology that enabled the Heyday of Expanded Gamut Printing Patents: stochastic screening, AKA FM screening. When additional inks are added to CMYK, moire patterns show up. Icky, objectionable moire patterns. FM screening is a way to avoid these icky, objectionable moire patterns. Now, FM screening goes way back to 1976, but Gary argues that "it wasn't until the introduction of Agfa's CristalRaster in 1993, that this technology became suitable for high quality work."

Thus, desktop publishing, CTP, and FM screening were the final enablers that made it possible for engineers to scan in image files, play with them on a computer with enough horsepower to do interesting stuff, send out the files to have plates made, and use those plates for high quality printing. The playground for innovation was opened.

By the late 1990s, we had all the technology in place from multiple vendors for expanded gamut printing. Separation software was available from Kodak, eder, Opaltone, ICISS, and Pantone. Inks were available through Pantone or Opaltone, or from your local ink vendor. Even Adobe jumped on the bandwagon. Postscript 3, which became available in 1997, included support for HiFi color.

Look out!! The expanded gamut ink train is coming through!

But...

Don Hutcheson wrote a "state-of-the-market" article for GATF World in 1999. His first sentence: "Despite a splashy introduction in the early 1990's, HiFi color printing has grown very little in the last five years."

Hexachrome was well known, but was it a commercial success? It was estimated in 1999 (Hutcheson's article) that Hexachrome was in use by only a few hundred printers. Bear in mind that at this time, there were tens of thousands of printers. Another article (from Glynn Hartley) said in 1999 that "there is a perceived low take up of Hexachrome".

In the same article about Glynn, he reported that there were "over 100 ICISS users currently operating in the UK." Maybe the software sold for the equivalent of $1000 a copy? I would call this a moderately successful small business. I don't want to appear to disparage him, but this is still a small business.

Hexachrome was discontinued in 2008, but Opaltone is also still around. They have their niche in the digital printing market. But they are not a huge company. ICISS is also still around today, but I don't see 100 employees on LinkedIn.

So what happened?

Expanded gamut was showing so much promise. Why didn't it fulfill the hype and become the default printing technology?

Here is an adage which is important to developers of new products: People are generally not willing to pay more for higher quality. There may be niches where the extra cost is justified, but if you want a product to hit prime time, look for ways to make it cheaper. Better yet, look for ways that it can save your customer money.

Adding a few more inks may make prettier pictures, but it will cost more. Prettier, but more expensive pictures are definitely in the niche bucket.

Kevin Bourquin has pointed to another issue that held expanded gamut back in the 90s: "I think the problem in 1994 was that while there were patents about how to do separations and some software to help, it was not well integrated into the production workflows. This made it cumbersome for companies to keep streamlined workflows." Having software to do the color separation is super cool, but you also have to be able to design, create a proof, do the RIP (with FM screening), set up profiles and plate curves, and do process control at the press side. Finding a collection of software together from multiple vendors to do something new can be a challenge.

So what finally happened?

In 2013 Mark Mazur conducted a survey that estimated that 10% to 20% of printers in the flexo world were using extended gamut. Don Carli's prediction came true, but about 15 years later than he predicted, and only within one segment of the print market.

More recently, the percentage has been soaring. Dawn Connell (Brand Marketing of Snyder’s Lance, who own Snyder’s pretzels, Jays, Kettle, Pop Secret, and Archway, to name a few) spoke at the Flexographic Technical Association forum in spring of 2016. In her presentation she said that 85% of their work is expanded gamut. 


In 2016, Kevin Bourquin of Cyber Graphics told me that they have 5500 SKUs separated for expanded gamut. I just checked back with him. As of April 30, 2018, the number is 8615. I should also mention that Kevin spoke on expanded gamut at the FTA Forum conference in Indianapolis on May 7, 2018. (Rumor has it that he mentioned my blog.)

Kevin's presentation isn't the only presentation on expanded gamut at a high profile conference. I just got news that Mike Strickler will be speaking on the same topic at another big print conference at the end of September / beginning of October. This won't be just a quick twenty minute thingie. He has a whole seminar. Smart guy, this Mike fellow. We taught each other everything we know.

Having guys with these credentials... speaking at such prestigious conferences... How can you say that expanded gamut is not a big thing now?!??!

It's not about pretty pictures

Why this huge recurrence?

Kevin points to another enabling technology: "But the first real tipping point was about 2004. Digital flexo plate had gotten a lot better and could print somewhat consistent if you tightly controlled to process. At the same time Esko and Kodak at the Drupa show, committed development resources to ingrate these tools into the workflows that people used to push files." As you can tell, Kevin is big on this whole workflow thing.

Mark and Kevin both pointed to one major snack food company that was an early adopter. Frito Lay was aggressively pushing to drive cost down and quality up. It's tough to meet both of those goals without doing some retooling.


But enabling technology doesn't necessarily translate into market success. Companies need a reason to want to invest in change. According to Mark Samworth of Esko, "The number one way to reduce costs in packaging printing is to reduce the use of custom spot colors." He has no idea whether this is true or not, but he did say I could quote him on this.

It probably seems like ages ago that I mentioned that Kodak jumped back into the expanded gamut in 2011 market with Spotless. (Scroll back if you don't remember.) The name is pretty clever, really. The word means clean, but literally, it means without spots. The pun refers to the fact that expanded gamut printing can be used to replace the icky-dirty practice of spot color printing. Roughly 90% of the Pantone book of spot colors can be printed as a halftone of CMYKOGV.

This saves money. In an old-school print shop, the printer would print the first job of the day with CMYK plus a couple of spot colors. To switch over for the second job, the print units with the spot colors need to be cleaned out to put in a few other spot colors. Cleaning out the print units takes time. Furthermore, the left-over ink can't be just poured down the drain. It has to be stored in buckets for future jobs. I have seen shops that have invested a lot of money just in shelving units to store leftovers. 

Cleaning up after a spot color ink party takes time

With expanded gamut printing, the mixing of inks to make spot colors occurs not in the ink kitchen in buckets, but rather, on the press with halftone dots. Hence, there is no need to clean out the CMYKOGV print units between jobs.

I spoke with Steve Balschi (who is a prepress guy at PrintPak, huge packaging printer), who said that they have plants where all they print is expanded gamut. Steve went on to explain that they had three type of expanded gamut jobs: 1) jobs where only spot colors are printed expanded gamut, and images are left CMYK, 2) jobs where spot colors and images are converted to expanded gamut, and 3) jobs that are a mixture. Whenever possible, they do not convert the images. But why would they want to? They're trying to match an image that was printed with CMYK. The best way to do that is to print CMYK. This underscores my point that it ain't about prettier images.

A further savings comes from the ability to gang jobs, as illustrated in the image below. Multiple related products are printed on the same press as one run, rather than as multiple smaller runs. The same amount of printing, but with only one make ready.

Choco Lotta is one of my biggest sources of snack foods

Spot color replacement is big not only in and of itself, but it enables this gang printing which is like, way big. John Elleman commented on LinkedIn: "[Spot color replacement] is most commonly used for creating flavor/form coding across multiple packages allowing gang printing all on one form versus sequential printing with spot colors, which increases cost for extra printing plates and change over time on press." Kevin Bourquin had a similar comment: "The true benefit is the economics involved in running multiple jobs in combo after replacing all the spot colors."

When has an idea's time come?

Thanks for sticking it out through this long and boring dissertation about the history of expanded gamut printing. We finally get to the moral of this series of blog posts.

In the previous installment (and the addendum) we see that just having a clever idea doesn't make you a millionaire. Unless of course that clever idea is to marry into a hugely wealthy family. In the first part of this blog post we see that a clever idea with a slick implementation is also not necessarily a ticket to the Filthy Rich Club. 

Here is the moral: An idea's time comes when the idea meets up with both the enabling technology and the need. I put that in italics to remind people to quote me on this. The idea of printing with inks in addition to CMYK is a clever idea. Desktop publishing, FM screening, high quality plates, and a full workflow solution are all enabling technologies. Replacement of spot colors was the need that made this idea worthwhile.

When an idea's time has come

Acknowledgements

Normally, I just make stuff up for my blogs. In this case, I thought I might try something a little different. I would like to thank the following folks for making sure my facts were as factual as possible: Don Hutcheson, Mark Mazur, Steve Balschi, Kevin Bourquin, Gary Field, Robin Myers, Mike Strickler, and Mathew Bernasconi.

References

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

Bernasconi, Color separation and reproduction method to control a printing process, US Patent 8,064,112, filed November 22, 2011

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

Carli, Don, and L. Mills Davis, High Fidelity Color Rendering and Reproduction, TAGA 1991

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

Eder and Maerz, Producing colour printed image from scanner, German Patent #4,417,449, filed November 23, 1995

Hartley, Glynn, PrintWeek, Bespoke HiFi provides value added market for print films, December 10, 1999

Herbert and DiBernardo, Six-color process system, US Patent 5,734,800, filed November 29, 1994

Hutcheson, Extended density color printing, US Patent 5,528,377, filed March 29, 1994

Hutcheson, Dom, HiFi Color Growing Slowly, GATF World magazine, 1999

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

Plettinck and Van de Capelle, Method and a device for generating printing data in a color space defined for non-standard inks, US Patent 5,689,349, filed April 29, 1994

Politis, Anastasios, et al., Extended Gamut Printing: A review on developments and trends, 1st International Printing Technologies Symposium (PrintInstanbul 2015)

Wolf, Kurt, PS imagesetter: a reasonably priced entry with the Linotronic Mark series, 1995
http://xmedia.biz/home/page.aspx?page_id=550&archive_type_id=92&person=0&categories=&archive_id=2981&from=611&keyword=

Thursday, April 26, 2018

Expanded gamut - when an idea's time has come, addendum

Welcome John the Math Guy blog fans. Today, you can color me embarrassed. 

Here is a quote from my recent post on the history of expanded gamut printing: "The earliest instance that I have found..." Did I get called out on my lack of scholarly research on that topic! Not just from one person, but from three people! With multiple examples that significantly predated my lousy excuse for research!

The Math Guy, suitably humiliated

Gary Field

The first person who took me to task was Gary Field, professor emeritus from Cal Poly. Being corrected is embarrassing, of course, but being corrected by Gary Field is almost an honor. I am not saying he has been in print for a long time, but before God carved the Ten Commandments into stone tablets, he hired Gary as a lithography consultant. Gary is known as the author of the printostorical book The Color Printing Revolution: Productivity! Creativity! Quality!, and is a co-author of Pioneers of Modern Offset Lithography.

In his first response, Gary traced expanded gamut 70 years further back than I did. Here is a response from Gary on one of my LinkedIn posts.

Expanded gamut printing goes way back to the early days of process color printing - the 1890s. If you can locate early editions of the Penrose Annual, you will find many beautiful examples. 

There were three reasons for employing extra colorants: poor purity process pigments, additivity failure, and proportionality failure. In the early days (pre stochastic screens) moire avoidance was a key constraint. I used to make 6-color separations in the early 60s with proportionality failure correction the objective. Light cyan and light magenta were the extra colors (today they use the same colors in photo-quality inkjet printers). Light and regular magenta (for example) were placed on the same screen angle and the tone scales adjusted such that each colorant provided the highest purity for its respective part of the tone scale.

I looked for information on the Penrose Annual. Wikipedia agrees with Gary. The Wikipedia article says: "Penrose Annuals remain the quintessential record for the development of mass media, advertising, photography, design and typography throughout the 20th century..." 

The cover from the second issue of Penrose Annual

So I read though a few issues. The second issue of the Penrose Annual (1896) had an article on the history of three-color printing which is prophetic in view of this blog post:

The three-color process had to undergo the same experiences and difficulties as every new invention. At first one man claimed to have invented quote a new thing, then several others arose and claimed the first right of invention for this; and at last everyone got to know that the new invention was nothing but the old thing known years and years ago, thus confirming the old proverb "there is nothing new under the sun."

(This may be true, but I believe that I was the one who first said this old proverb, I mean, ecclesiastically speaking.)

I did a rudimentary search of the early issues of the Penrose Annual. The first few had articles on printing with three colors. The earliest article I found that referred to more colors was an article comparing the three-color and four-color processes in the 1898 issue. I quote: "the litho printer is taught to produce his ten or twelve color print". Oh! I guess expanded gamut lithographic printing was commonplace in 1898!

Gary goes on to chastise me for my poor research techniques, and also for my prodigious collection of patents:

Most patents, frankly, are worthless. It is not that difficult to locate the "prior art" of some enterprising photolithographer (often in the Penrose Annual).

Of course, over the course of several emails back and forth, Gary felt the need to demonstrate some further prior art. He found a patent for expanded gamut photography that was filed in 1950 in Great Britain: Improvements relating to multicolour photographic reproduction, by Joseph Arthur Ball. I dunno how I managed to miss this one. Pretty lazy of me, really.

Apparently he felt the need to outdo himself -- I mean this is a competition, after all -- and played the trump card: Process of photomechanical reproduction of colors and the resultant article (Charles Zander) which was filed at the US Patent Office in 1905. This is a four-color photographic process, which may not sound all that impressive since printing uses four inks. But his process was with four chromatic pigments: magenta red, lemon yellow, emerald green, and ultramarine blue.

So. I concede. Gary won.

Robin Myers

As if this weren't enough, another good friend, Robin Myers, had his own commentary on my research:

Your latest post on wide gamut printing is very interesting, but it exposes a flaw in performing historical searches using the Internet alone. The Internet is a wide pool, but for historical information, mostly shallow.

Boy, have I been called out on the carpet!

Robin is an archeo-bibliophile and collector of old books. He thanked me for not calling him a pack rat. Robin is also the proprietor of Chromaxion, a repository of color information. If that were not enough, he is the author of SpectraShop, a color acquisition program that I have actually used.

After reading my recent blog, Robin pulled out his copy of Dictionary of Color, published in 1930 by Maerz and Paul. This book was one of many efforts that attempted to provide official definitions of color names by way of color patches. (Earlier color-naming books were from Albert Munsell (1915), Robert Ridgway (1912), Milton Bradley (1895), Johann Ferdinand Ritter von Schönfeld (1794), and A. Boogert (1692). Boy, that sounds like grist for a future blog post!)

Available to pack rats through Amazon

I should clarify the images above. The image on the left is a page of delightfully pretty color patches from the book that were printed with two different inks. The image at the right is from the facing page, with a grid, and names of some of the corresponding colors.

Here is what Robin had to say about Maerz and Paul:

After reading your article, I decided to check a copy of “Dictionary of Color” by Maerz and Paul, published in 1930. This book was printed using many more than the standard 4 colors. ... So the techniques of printing with expanded gamuts on press were known well before the 1960’s. I suspect that they were not widely employed for economic reasons.

The authors claimed, and indeed used, 8 chromatic inks and 8 achromatic inks. This I confirmed by spectral measurement with an i1Pro 2 and visual observation with a Beta Color Proofing Viewer II (modified to use white LED illumination). The charts were printed using 150 lpi screens (determined by a Screen Pattern Analyzer and Rescreening Key from RIT).

I was provided with spectra of the inks that were used, that clearly show eight different inks. Robin is nothing if not thorough.

We had an interesting discussion (not a surprise, our conversations are often interesting) about what constitutes expanded gamut printing. The world's oldest book with multi-color printing was the Manual of Calligraphy and Printing, which was first printed in China in 1633. The images were printed with up to ten different inks. Should we consider this gorgeous collection of prints to be expanded gamut printing? We decided "no". This book was block printed, and we decided that to qualify as expanded gamut, it must have halftones.

I could not find a copy of this from Amazon

Robin also dug up some real gems -- early books that showcased a lot of early printing beyond CMYK: A Half Century of Color by Louis Walton Sipley (1951), Practical Color Simplified by William J. Miskella1 (1928).

Mike Strickler

I have relied on Mike for years to help me keep one foot outside the ivory tower. He told me about another expanded gamut effort:

Since you’re on about the history of ECG, I should tell you about a very rational system that was developed about 1978 by a guy I came to know in 2014, when I was hired to replace the system at Shorewood Packaging in North Carolina. The man’s name is Ken Reddick, and apparently around 1978, without any outside inspiration he decided that 4C lithography could be much improved if secondary inks could be added. 

He had 8C presses at his disposal at the Queens Litho plant in Indianapolis (later acquired by Shorewood). I believe they printed album covers at the start. Ken reasoned that if he could find strong colors in between the hues of C, M, and Y he could do something useful. He though about it for a moment and then realized that the Pantone base colors Warm Red, Pantone Green, and Reflex Blue fit the bill pretty well.

(He was not interested in abstractions like the greatest overall gamut—he had no way of measuring this in any event—but specific needs of certain customers—and he may have had only 6 colors to start.) 

Then he separated a bunch of Pantone-like patches varying the percentages of the (max) 3 likely constituent colors, building a sort of ring-around chart, printed these, holding density and dot gain steady, chose the best matches visually, and built his lookup tables. I still have one of those color books. Given the limitations of the era—no spectro to measure color difference, no color conversion software--he succeeded wildly. Designs had to be rebuilt object by object, and there was no solution for images other than (if anyone asked for it) making touch plates conventionally. But hey, this was great stuff!


I guess this trounces the myth that Hallmark Cards invented the whole expanded gamut thing. Who started that silly myth, anyway!?!?

Don Hutcheson

You have not lived until you have been corrected by Don Hutcheson. Here's his contribution:

The one outstanding ECG pioneer to whom you give no credit is the diminutive, be-monicled Moulin Rouge poster child, Henri de Toulouse-Lautrec.

For the Moulin Rouge (and probably other clients) Lautrec hand-drew posters with a grease pencil and chalk on litho stones that were then printed with an early approximation of CMY inks. After pulling a proof, he would often add extra stones printed in custom-mixed pastel inks, until he’d achieved his desired effect.

It’s not that Lautrec was an inventor, it’s just that, as an artist, he was using the then most common approach to color printing.

IS this expanded gamut printing or not? I dunno.

Larry Goldberg

The mention of the Beta Color Proofing Viewer leads us to the next person to expose my inadequacy in the field of the history of science, Larry Goldberg. Larry runs Beta Industries, which was mentioned by Robin. Note that this email was the second email that I received that mentioned a rat.

I thought I smelled a rat, or at least a fish.

Mr. Kippers (his pseudonym didn't fool me for a minute) color system reminded me of a system that I actually saw, and met the inventor thereof.

When there used to be a very good printing trade show in Long Beach, CA called the Gutenberg Festival a fellow came around with a sample of an additive color printing method.

It used fluorescent inks and a black keyline.  Newspapers, always famous for their print quality, could run cartoons in full additive color, and just keep running the black-only image if they ran out of dayglo ink.

The inks are quite opaque, as the substrate offers no benefit.  White was the result of tri-color adjacent bits.

The guy's name was WEAKLY or WEEKLY, but my  USPTO searches never produced a hit.  He also had a patent that he said was the ONLY one that used the term "Dick Tracy" in describing a wrist-worn electronic signalling device.  Little did he know the Apple Watch was just 30 years around the bend.

Larry eventually found the patent. Here is a link: Mosaic additive reflectance color display screen.

The Mosaic Screen Plate

I am going to explain Weekley patent, but lest one get confused, first I need to explain Larry's comment about the Kueppers patent. Kueppers' patent combined two interesting concepts. The first concept was about using more inks than just CMYK for printing. The second concept was about creating color my placing little color patches adjacent to one another, but not touching. This is known as additive printing. Larry was commenting that he remembered hearing about a patent for another process that used non-overlapping patches. He wasn't claiming that this invention had much to do with going beyond CMYK. (Although, Weekley did use fluorescent versions of CMY. Odd that he didn't mention fluorescent black ink.)

Weekley's patent was an improvement to a process that was invented in 1868, called the mosaic screen plate. This is an interesting bit of history and technology.

Ducos du Hauron was a pioneer in the field of color photography. In 1862, he invented the idea of the mosaic screen plate as a way to take color photographs using black and white film. This glass plate is a set of really tiny filters of red, green, and blue. The back side of this plate was covered with a normal photographic emulsion, containing silver halide, which would normally produce a black and white photograph.

Mosaic filter plate
(color scientist's conception)

When taking a picture, the emulsion was exposed from the front side, that is, through the filters. Thus, on the back side, there were areas where the emulsion had been exposed with only red light; others where the emulsion had been exposed by only green light, and still other areas where the emulsion had been exposed with only blue light. The individual areas were thus indicative of the amounts of red, green, and blue light in the image.

The emulsion on the plate was then developed with a positive process -- areas where light hit would be white, areas with no light would be black. When the developed plate was viewed (with the filters still intact) the intensity of the light reflected through the red, green, and blue filters would be in accordance with the amount of red, green, and blue light that was in the exposing image.

I was careful in my wording before. I said that du Hauron invented the idea of the mosaic screen plate, for which he was awarded a French patent in 1868. As you might imagine, those mosaic filters are kinda hard to build. Here is an interesting factoid about patents: you are not actually required to build the invention to get a patent. All you need to do is describe it in enough detail so that "one skilled in the art" could build it.

The first commercially successful implementation of du Hauron's invention didn't occur until 1903, with the Autochrome Lumière. The Lumière brothers figgered out how to make this mosaic screens. They sifted potato starch grains to a uniform, microscopic size. For the logophiles in the crowd, this process is known as elutriation. For the people who aren't really all that fanatical about words, the process is known as sifting.

One of the Lumiere brothers, proudly posing with the US version of their patent

They then dyed batches of the grains in each of the three primary colors. These grains were then mixed, and deposited, one layer deep, on a glass plate with black pitch between the particles. This mixture was pressed between glass to flatten out the grains to cover more area. Thus, they had a mosaic. It did not have a regular pattern as in my drawing above, but that didn't matter. Each grain was too tiny to see anyway. All that mattered is that the silver halide on the back side remained in register to the colored grains on the front side.

One problem with the mosaic screen plate was that the final film was rather dark. Imagine that the original exposure was with white light. The red pixels would all be bright red; the green pixels would be all green; the blue pixels would be bright blue.  But when viewed under white light, two-thirds of the light that hits the plate gets wasted. Any green or blue light what hits the red filter would get absorbed by the red filter. 

This is the problem that Weekley addressed through various methods. Fluorescent inks was part of that. I won't get into the rest.

Addendum to the Addendum

I want to give my sincere thanks to each of today's contestants of John Doesn't Know His Mosaic Filter Plate From a Hole in the Ground! All seriousness aside, I think this history stuff is pretty cool, and appreciate the additional information.

In order to at least partially redeem myself, I feel the need to challenge Weekley's comment about being the only patent with the term "Dick Tracy". I did a bit of searching and found a few that predated Weekley's patent. 


The dates listed are when the patents were granted, that is, published for all the world to see. Weekley's patent was filed in 1979, so he gosh darn shoulda known about these! Was that too harsh? Maybe I'm just jealous that none of my patents are cool enough to mention Dick Tracy.