Discussion

By Gordon Pritchard on Apr 24, 2019

It’s not the consumer nor is it the printer who is concerned about color consistency - it is the (unmentioned) brand owner. It is the brand owners that define the tolerances for color variation and it is they that design the labels and packaging. It would be interesting to know if the brands have done any research into the impact of color variation on sales or brand recognition.

By Raymond Prince on Apr 24, 2019

Who is the customer for packaging? Is it not the manufacturer that is purchasing the packaging material. I have listening to manufacturers of food, cosmetics, hard goods and they do put great emphasis on color consistency. It is not realistic to tell the buyer of millions of folding cartons that his customer does not care about the color variation of his packaged product thus neither should they - good luck on that one. A very poorly done "survey".

By Heidi Tolliver-Walker on Apr 24, 2019

I don’t think it so much that customers don’t care. I think the whole point of the survey is they don’t notice.

By Gordon Pritchard on Apr 24, 2019

If they cared they’d notice.

By Raymond Prince on Apr 24, 2019

With all due respect, I live in the packaging world
and the buyers notice, complain, ask for credits,
reject work, ask printer to sort, etc. No printer is
going to state how many jobs they have had
rejected and no customer will share that
information.
I keep very busy consulting regarding printing
complaints including color variation . I have done
quite a bit of legal work in court on these type of
issues. My best advice for a printer is to purchase
closed loop color control and defect detection
online inspection equipment for defects.

By Eddy Hagen on Apr 24, 2019

Thanks for reporting on my survey Heidi.

Print buyers (the customers of printing companies) are very demanding, that's out of question. But are their motives the right ones? I always had the impression that it is mostly a power game between buyer and supplier (e.g. to get discounts), that it has nothing to do with what how consumers (the buyers of packaged products) look at it. Over the past three decades I have seen many, many examples confirming that: e.g. print buyers without any technical knowledge, who sometimes even get the explicit instruction to reject the first (or even more) press proofs, 'to put the pressure on'. And many people play along with this game. That's what I want to challenge. Let's get real about color.

And I'm not the only one: if you look at color perception studies from outside the printing industry, it's clear that our color vision, color perception is flawed. I recently came across a study showing that even in an 'undelayed comparison' (so two samples next to eachother), the color perception was not correct, but biased towards the center of color categories (like 'red'). You can read it here: https://www.insights4print.ceo/2019/02/confirmed-our-color-memory-is-seriously-flawed-and-what-this-means-for-brand-colors/

I want to stress again that I'm certainly not advocating bad print quality, every printer should be proud of his job and adhere to ISO-standards, G7, PSO. These are very good, they are achievable for everybody with recent, decent equipment and skilled operators.

But if you want deviations that are half of the tolerances specified by these standards, you're in trouble: when added up, your color quality tools can already have higher deviations. E.g. two spectrophotometers measuring the same color sample can be almost 2 dE00 apart... (https://www.insights4print.ceo/2018/11/brand-color-quality-in-print-the-chain-of-tool-tolerances/) And two Pantone color guides can be 4 dE00 apart, for 90% of the colors, in theory. In reality, the deviations are even higher: https://www.insights4print.ceo/2018/11/your-color-guide-and-you-first-results/

@Raymond Prince: did you read the complete survey (questions, sampling)? What about it is very poorly done?
The advice on closed-loop color control, defect detection: I can only agree with that. They are excellent tools to stay within the tolerances stated by ISO standards, G7, PSO. They have a high ROI, they are (or should be) a no-brainer.

By Raymond Prince on Apr 24, 2019

With all due respect. The issue of color memory has nothing to do with the subject at hand namely color variation. One cannot remember color especially the accuracy that printing industry demands. When comparing one to one even my eyes can see in many colors a delta E of 1.0. I know of no major folding carton printer that would use a PMS book
on press for anything. Not knocking the PMS System
but most packaging printers want to see ink on the substrate they are going to print on and it must match the customers supplied Lab numbers. The printer would supply the proof (ink on substrate) to the customer to sign off on. Now you have a valid standard. In regard to inter instrument agreement, maintenance, age, model to model differences, brand, etc. all play a part. In the grand scope of things, ink on the substrate sample with a sample at press and match it basically works. All one would use the
spectrophotometer for is to insure that the current sheet matches the customer ok'd proof of ink on substrate.Is that not what we want - match what the customer ok'd? I am off to Ecuador and will let Gordon pick up the conversation.

By Gordon Pritchard on Apr 25, 2019

Cue the second fiddle LOL

Eddy Hagen wrote:

"I always had the impression that it is mostly a power game between buyer and supplier (e.g. to get discounts), that it has nothing to do with what how consumers (the buyers of packaged products) look at it."

GP: I have not seen it as a power game between buyer and supplier to get discounts. In my experience, professionals don't play such games.

In a way it does have nothing to do with how consumers (the buyers of packaged products) look at it. It has to do with brand consistency and integrity for the brand owner.

EH: "print buyers without any technical knowledge, who sometimes even get the explicit instruction to reject the first (or even more) press proofs, 'to put the pressure on'"

GP: I have seen art directors/graphic designers in general commercial work reject the first pull of a press run - however, not for the nefarious reason you state. Sometimes they want to "beat" the proof and get creative on press. But most times there are inline color issues or OBA issues that affect the print to proof alignment. It is then helpful for the buyer to assist the press operator in understanding which colours are a priority and which can be compromised. And so the press is adjusted to bias the colors as needed.

EH: "if you look at color perception studies from outside the printing industry, it's clear that our color vision, color perception is flawed."

GP: It is not flawed. Color perception is all in the mind - which just means there is tremendous variation in color perception between people. That's why we use instruments and color models to help objectify and bring greater consistency to perception.

EH: "every printer should be proud of his job and adhere to ISO-standards, G7, PSO."

GP: I would disagree - but that's a different topic.

EH: "when added up, your color quality tools can already have higher deviations. E.g. two spectrophotometers measuring the same color sample can be almost 2 dE00 apart"

GP: True, which is why some of the major brand owners provide certified print suppliers with their own equipment for quality assurance.

EH: "And two Pantone color guides can be 4 dE00 apart, for 90% of the colors, in theory."

GP: That's a supplier/buyer communication issue. Most brands either have their own color palette in addition to, or instead of, Pantone's. And they will specify the desired color by Lab values which are unambiguous and not dependent on the integrity of a swatch book. Ink draw downs on the actual stock with coatings applied are pretty standard rather than a Pantone swatchbook.

EH: "stay within the tolerances stated by ISO standards, G7, PSO."

GP: Most brands specify their own tolerance values and couldn't care less about ISO standards, G7, or PSO to maintain brand consistency.

By Eddy Hagen on Apr 25, 2019

Isn’t the whole idea behind brand owners demanding very tight tolerances their fear that consumers will switch brands if colors vary, that consumers will not trust it when a package looks ‘different’ from what they consider the brand color? Then color memory does matter, doesn’t it?

That was also the idea I wanted to challenge. And in the survey, most people would still buy the product, even if it ‘looked different’. Only two claimed they would buy a competing brand, one of them is active in printing, the other doesn’t go shopping often. I also included a question on which one of the six variations of Coca-Cola red they wouldn’t trust. And only upwards from 8 dE00 the disapproval rates become significant. Interesting: the right color (0 dE00) had a higher ‘distrust rate’ than the sample with a 3 dE00 deviation…

Re color discrimination: with a P1 illumination (2000 lux, ISO 3664) and when placing samples on top of each other, with the colors touching, everybody will probably be able to see a 1 dE00 difference. With a P2 illumination (500 lux), it already is a bit more difficult. When placing the samples a bit apart (e.g. 1 cm), it becomes (much) more difficult. X-Rite had an interesting example of that: when they launched the online version of the Farnswell Munsell 100 Hue test (https://www.xrite.com/hue-test) the different colors were touching. No problem for me to get a perfect score! Now it has a black border around the color patches, just a few pixels wide, and it has become a lot harder. And these are all conditions where the samples are flat and in the same plane. If they are not, like folded boxes in the supermarket, it becomes really hard to discriminate (small) color differences. That’s what I showed in another research. Look at these two graphs, showing the results of two sides of the same boxes, the one side with a bleeding color, the other side non-bleeding. https://www.insights4print.ceo/wp-content/uploads/2018/05/2018_05_Jens_NonBleeding_difference.png
https://www.insights4print.ceo/wp-content/uploads/2018/05/2018_05_Jens_Bleeding_difference.png The results look very different.

Re the use of Pantone guides / color references: in a small survey I did a few months, 63% said they are using Pantone guides as a physical reference for print jobs. Maybe not the printers/print buyers that Raymond and Gordon are working with, but others do use them as physical references. With ink on paper, you also have to be sure that the ink draw down is good. I have seen samples in the past that had huge density / color variations in that ink sample (double digit dEab between the areas with the lowest and highest density). So what color is then the right color? The most extreme example was BTW from one of the largest FMCG companies in the world. The ink sample was printed on a white, glossy paper, while the job was on brown corrugated board… That doesn’t seem professional to me.

Another example of the professionalism of certain print buyers: the print buyer signed of the print job at the press, but rejected the job when it was delivered a few days later: the ink gloss was different, due to the drying of the ink, and she liked the wet gloss better than the dry one, making her reject the job… So if you have only met professional print buyers, you are really lucky. I have heard a lot of horror stories.

Re the execution of the survey: a friend of mine is a university professor and does a lot of surveys. The only remark she had on my survey, was that I used ‘convenience sampling’, not a completely random sample of participants. Which isn’t a huge issue, as long as you are aware of that fact. But for the rest, she had no remarks.

By Timothy Baechle on Apr 25, 2019

"Beware of false knowledge; it is more dangerous than ignorance."

By Heidi Tolliver-Walker on Apr 25, 2019

As I read through these comments, there are a couple of things that strike me.

First, I think it's healthy to challenge conventional wisdom. It needs to be done. We shouldn't be afraid of doing it or be threatened by the results. (I'm not saying that's what's happening here. Only making a general point.)

Second, there are always imperfection in surveys, even the best ones, and I think it's valid to examine variables that might be impacting results. That's just responsible analysis. But unless a study is completely and utterly flawed to the point if it being irrelevant, which it doesn't strike me that this one is, the data have something to say to us.

What does Eddy's research have to say? I think it opens a lot of legitimate questions that are worth discussion (respectfully). Regardless of whether you take the research as gospel or not, his ongoing results confirm what most of us probably already know (but may not want to admit): that consumers don't have the color memory we'd like to think they have, and they don't necessarily make decisions based on slight variations in color to the extent that perhaps we've come to expect.

That doesn't mean that it's not important to adhere to strict color standards or that great color isn't important. It's just acknowledging the reality that consumers don't care about the details as much as the people in the industry do.

I'm glad that the study has generated discussion, because it's discussion that I think is valuable to be had. If others in the industry have different experiences that lead them to different conclusions, great! Let's talk about it. It's only by getting comments and observations from a variety of different perspectives that we begin putting together the pieces of what is clearly a much larger puzzle.

By Gordon Pritchard on Apr 25, 2019

One has to be careful to avoid one's own biases color one's interprepations of what one sees.

Eddy Hagen wrote:

"Isn’t the whole idea behind brand owners demanding very tight tolerances their fear that consumers will switch brands if colors vary, that consumers will not trust it when a package looks ‘different’ from what they consider the brand color?"

GP: I don't think brand owners fear that consumers will switch brands if colors vary. The tight tolerances are there to maintain brand integrity and consistency. It's one of the reasons that makers of counterfeit products spend the money to get the brand colours correct on their bogus products. They need to align their color as closely as possible to the real product so the buyer does not start asking questions. It's also why, when you see a certain robin egg blue on a jewellery product you immediately recognize it as Tiffany (Tiffany Blue is a trademarked color). It may not be the exact Tiffany blue as far as your eyes are concerned but by having strict color tolerances Tiffany has made the blue instantly recognizable as their brand.
What happens on the shelf, for example with grocery items, is that an item where colors deviate from the same surrounding products may remain on the shelf and end up in the clearance bin along with perfectly good products in dented cans.

Re color discrimination:

It is well established that the perception of color is affected by the observer, by the illuminant, by the spectral composition of pigments and substrates, and by context. That is one of the reasons why standards and specifications have been defined for evaluating and communicating color in a production environment. It's also why some brands (e.g. perfumes) try to control the environment in which their products will be viewed by consumers. It's also why some printshop have viewing areas with several different lighting types so that the brand owner can view the presswork under lighting that is similar to the lighting under which their products will be viewed by consumers.

Re the use of Pantone guides

The Pantone guides are more typically used in general commercial print production rather than in labels and packaging. Proctor and Gamble, for one example, provides their print suppliers with a numbered custom swatch book as their brand color reference. The swatch books also typically include Hi/Lo ink density guides. ( Here for example is Kodak's https://1.bp.blogspot.com/_lNzlqVMcIj8/So29g2AW5nI/AAAAAAAABIg/h04YL2NXuTQ/s1600/Swatch+Card.jpg)


RE EH wrote: "The most extreme example was BTW from one of the largest FMCG companies in the world. The ink sample was printed on a white, glossy paper, while the job was on brown corrugated board… That doesn’t seem professional to me." and "the print buyer signed of the print job at the press, but rejected the job when it was delivered a few days later: the ink gloss was different, due to the drying of the ink, and she liked the wet gloss better than the dry one, making her reject the job…"

There are always going to be idiots. So, good printers recognize this and take pains to effectively communicate color issues with their clients in order to correctly set expectations. They don't assume that their customers know what they're talking about. So, the ink sample being printed on a white, glossy paper, while the job was on brown corrugated board to me is simply a failure to communicate and properly set expectations. Printers can be idiots just like their customers can be. Either the printer learns from the experience and mends their ways or ends up going out of business.

By Eddy Hagen on Apr 26, 2019

Color differences as a reason for switching brands is an argument that is often used, by a lot of people, and it's used to justify tight tolerances.

Color perception is indeed affected by these factors. That's what brand owners should know and realize: they can't control the appearance and perception of their brand color all the time. Sunlight e.g. varies all day long, it depends on the time of the year, it depends on your location on the globe. And if the illuminant is different, color perception will be different. The lighting in my living room is e.g. different from the dining room, the fridge has a very different light. So the same package will look different on each occasion. Brands can't control the spectral power distribution of sunlight, nor the light in my dining room (and very often also not in supermarkets).

Regarding the 'not tech-savvy' (I don't like calling some people idiots in public): the one with the ink sample for brown corrugated on white, glossy paper; in that case the printer did try to communicate effectively, stating what the customer should bring as a sample, including the offer to do ink draw downs inhouse. But the customer - again: one of the largest FMCG companies in the world - refused that. It was either accepting that bad sample, or no job at all (and we're talking about a huge job...). The sales manager of another large packaging company told me several years ago that he saw a frightening evolution with FMCG companies: they were replacing tech savvy print buyers by 'not tech savvy' product managers doing the press checks. The reason according to him: preventing discussion about technical (im)possibilities and forcing printers to comply with the quirks of the product managers, leaving only 'price' as a common theme for discussion. Trying to educate the people doing press checks was not appreciated.

It seems that we have very different experiences, that we have been in contact with rather different types of companies.

Nice (and useful) concept of the Kodak swatch book. But what was it used for? The name on it says 'visual targets for trade dress', so is this for textile? Then there is the difficulty of comparing two very different substrates (paper vs textile).

By Gordon Pritchard on Apr 26, 2019

@Eddy

You're right. They may not be idiots. They're incompetent.

RE: "It seems that we have very different experiences, that we have been in contact with rather different types of companies."

I have visited hundreds of print shops in Europe, N America, and south east Asia. I've also worked in print shops as technical director. That is the experience I draw from.

RE: "Nice (and useful) concept of the Kodak swatch book. But what was it used for? The name on it says 'visual targets for trade dress', so is this for textile? Then there is the difficulty of comparing two very different substrates (paper vs textile)."

That visual target I linked to is used for offset printing of Kodak packaging and is fairly typical method used in the packaging industry for communicating color. More info about the trade dress colors is included on the inside of the swatch target.

The term "Trade dress" as used in labels and packaging production, is a legal term of art that refers to the visual characteristics/appearance of a product or its packaging that signifies the source/brand of the product to consumers. Trade dress is a form of intellectual property. As any packaging/labels printer would know, it has absolutely nothing to do with textiles.

Have you had any actual experience in labels and/or packaging production?

By Eddy Hagen on Apr 27, 2019

Since you asked: I’ve worked a long time for the printing industry association in Belgium and a very long time for the printing industry innovation center. I have visited and worked with hundreds of companies, both commercial printing and packaging, including a few label printers; next to well-known suppliers and research institutes. I had the fortune that people were always eager to tell me their stories, about technical topics and special customers. I guess they considered me as a trusted person who might also be able to help them, next to the fact that I’m a good listener. Some contacts were very knowledgeable and seasoned. I’ve also had the opportunity to attend many conferences, read and study a lot, to be involved in research projects. I didn’t operate a printing press, but I think it’s fair to say that my knowledge of color and printing is probably above average.

Next to giving courses and presentations, doing research, I’ve been advocating standardization and best practices for a very long time. Even those things that aren’t popular, usually because ‘too expensive’, like climate control in the press room and water treatment.

My presentations, e.g. on color and printing technology, were packed with interesting ‘real life’ anecdotes that were drawn from my contacts with printers, from practical examples, which invited the attendants to share their stories with me. An example: how a shift in registration of the plates can induce color shifts. How many people are aware of that? The most extreme example I was involved with, was a package printer where the gray background of a box started shifting towards blue after 5000 to 10000 sheets. It was on a hot summer day, the press room had no climate control, the press had no cooling system; combined, this made the press running hot and shifting or expanding the cyan plate a tiny bit, enough to be visible as a color shift. A microscopic image of a good and a bad copy clearly showed that.

Some of the stories that people shared with me led to research projects, tools or educational material. E.g. the first version of the GWG ‘9 reasons why you need preflight’ was designed by me, when I was trying to convince a (large) label printer to use preflighting software (they did that manually). Also, the Live Preflight profiles, which are now under the umbrella of the GWG, were initiated by me, many years ago, when the GWG wasn’t interested yet in this kind of profiles. As a result of conversations with printers, I also designed and shared a PDF file that tells users if the ‘overprint preview’ of their PDF viewer is on or not. Next to a panel for Adobe InDesign that streamlines the document and PDF creation process, eliminating expensive errors.

Among the research projects I initiated or executed, was the first large scale study of spectrophotometers that are used in the field. If you search for ‘spectrophotometer nightmare’, you can still find references to it. Another research project, about 8 years ago, was the result of a conversation with a food packaging printer, who had troubles with ink drying (UV-inks were blacklisted at that moment). I wanted to find out how low you can go in TAC before you have a visual degradation of the image (his profiles were 320% or higher). And this was much lower than common wisdom said. One of my colleagues created ICC-profiles with a TAC between 320% and 180%, with a 20% increment. I selected several ‘difficult’ images and got them printed. I asked a group of seasoned printers if they noticed something. When I told them afterwards what they were looking at, they wouldn’t believe me. There was no visual difference between the 320% and the 260% images. (please note that I placed a 1 cm white border around the images, when cut out and placed on top of each other you could notice that the darkest black was slightly different; see my earlier remark on the Farnsworth-Munsell 100 Hue test). The same test was done on a print exhibition, under standardized conditions, with dozens of participants: nobody saw the difference. The package printer immediately started using the 260% ICC profile, all his drying problems were gone. In a few occasions he used the 220% profile, customers never noticed. You can still find the 260% profile in the ICC profile registry.

I’ve also been involved with the certification of print standardization in the Benelux. An interesting anecdote from that era: color shifts due to drying. A company had sent a few samples for certification via airplane, they were within specified tolerances. When the rest of the job arrived via truck a few days later, a color shift of a few dEab was noticed… And the job was out of tolerance.

I'm will never claim that I know everything, I’ve certainly made mistakes in the past, especially when I was much younger. But I learned from that. I started questioning and investigating things more thoroughly and checking everything from multiple sides, asking others – especially some seasoned and knowledgeable people that were open for discussion – how they looked at things. I’m always open to looking at reports and research that contradict my view, that’s the only way we as a society can move forward.

Over the years, I also developed the annoying habit to notice things that others don’t see, don’t want to see. E.g. during a tour in the new building of a packaging prepress house I noticed that the D50 light booth was very close to a huge, South oriented window, probably influencing color perception. During an after-hours tour in another packaging prepress house I once noticed that their monitors weren’t calibrated correctly: I was standing in the back of the – at that time desolate – room, with the same image on a few dozen monitors: they all looked different.

And one of those annoying things I noticed, is the fact that consumers really aren’t looking at color the way we do, they only register color in very general terms, certainly not small differences. Take the pictures of the two packages in this article: https://www.insights4print.ceo/2017/08/uncertainty-principle-visual-color-evaluations/ And look at the conversation I had with my girlfriend, especially how long it took before she noticed the – for us rather obvious – color difference. Which lead me to ‘the uncertainty principle of visual color evaluation’: you can’t be objective about color (and color differences) if you know that you are judging color. Think about that (preferably with a glass of wine or beer, it’s weekend after all ?? )

By Chris Lynn on Apr 30, 2019

Thanks to Heidi and the commentators for a stimulating and educational thread.

Gordon's point that "makers of counterfeit products spend the money to get the brand colours correct on their bogus products" reminds me of a recent experience. A very large consumer goods manufacturer suffered from Chinese knock-off products that had the brand's (monochrome) logo printed on the fake product with HIGHER print quality than the genuine branded article. The difference related to resolution and density, not color. When senior management realized this, they authorized the six-figure purchase of a custom inkjet printer to fix the disparity.

By Eddy Hagen on Jan 15, 2023

It's an old thread, but here is new information, from Quad's Package Insight division.
That division helps companies understand how people shop, how they react to packaging. They have a setup which resembles a real shop and use eye-tracking to register how long, how frequently a person looks at a package. Which results in statistics like 'Total Fixation Duration' and heatmaps.

Guess how long a shopper looks at a package: on average less than a second... The top package(s) in their tests got around 2 seconds, half however less than a second.

One of their tests is with soda cans. The total fixation duration for Coca-Cola? One-tenth of a second... And even with that short duration, it was the second most sold product in the test, after dr. Pepper.
(https://packageinsight.com/canned-soda/ - the full graph that includes Coke is in the download)

That very short timeframe makes that it's a different part of our brain that processes that information compared to a press check. This relates to 'Dual Process Theory' from behavioral economics, also known as System 1 (the fast but inaccurate part) and System 2 (the slow but accurate part). This concept is explained in the highly recommended book 'Thinking, Fast and Slow' by Nobel price winner Daniel Kahneman.

More: https://www.insights4print.ceo/2023/01/how-our-brain-perceives-evaluates-color-system-1-vs-system-2-dual-process-theory/

By Heidi Tolliver-Walker on Jan 16, 2023

Thanks, Eddy! This sounds like a follow-up post to me. Thank you for sharing!