Discussion

By William Ray on Aug 27, 2018

Printers think in reflective color and, typically in terms of a polychromatic metaphor. LED light that arises from Stokes shifting material is a pseudo white and lacks significant amounts of many frequencies.

The most common Stokes material is YAG phosphor. This yields 450nM blue from the LED die and yellow from the shifted blue.

Some lighting manufacturers add a red shift to make emitted light warmer. This still is a long way from polychromatic output.

This is not a trivial problem.

By David Avery on Aug 27, 2018

Penneys opened a store in Binghamton with Pulsed Xenon lighting many years ago.
Had to close and redo the lighting because of the high rate of returns on clothing.

By Dave Dickey on Aug 27, 2018

Another consideration when talking about lighting in both textile and printing environments is the UV component of the light source in use. While an LED source may produce D50 or D65 light in the visible spectrum, that does not mean that it will necessarily produce the amount of UV required for UV brighteners in paper stocks, or fabric treatments that rely on them to produce the whiter whites customers have come to expect.

By Peter Goldberg on Aug 28, 2018

Serious issue when LED industry proclaims CRI numbers that exclude RED, GREEN and other hard colours. CRI is measured using only 8 pastel colours causing issues in the retail environment. A far better measure would be the Color Quality using 15 colours. This provides a real description of CRI

By Danny Rich on Aug 28, 2018

We have been raising this issue for the past five years. I was at the AATCC conferences that you mentioned. The lighting and standards industry were completely clueless. One report stated that his purple flowers turned blue under his LED lights but ... no one cares about purple. I raised the issue of Kraft purchasing Cadbury for their brand - based on purple. We spend a lot of time making inks and printing brand colors only to have them change on the shelf.

The textile retail business is very sensitive to this as they lost millions of dollars in garment returns when the lighting industry introduced the tri-phosphor fluorescent lamp light. This is definitely something on which the print industry should be focused.

By William Ray on Aug 28, 2018

Another little thing to contemplate. If you buy several of the same brand of LED light, say an A19 bulb form factor as an example, and then fire them up side by side you will see that none of them will appear to be the same color.

This is due to the fact that the epitaxial film that is deposited on the wafer from which LEDs are made is not uniform due to crystal lattice strain. So the die that are cut from the edge if the wafer have different characteristics than die cut from the center.

Better vendors try to mitigate this by "binning" die of similar characteristics. For a lot of reasons this ($$) this is imperfect.

By Cary Sherburne on Aug 28, 2018

Danny, thanks for your comment. You are so right. Hoping this article -- and a longer, more detailed one that ran in the AATCC journal -- will help with the education process

By Gordon Pritchard on Aug 28, 2018

Lighting standards for the printing industry aren't there for color "accuracy" - they're to provide lighting constancy to enable better color communications in the production chain.
In-store lighting is usually determined by the building lease holder and running costs - not color "accuracy".
It's not feasible to control nor predict the lighting conditions in the home where the shopper examines their purchase. So, led lighting standards intended to provide brand color consistency and reduce customer returns simply will not work - except, of course, to provide work for standards groups.

By Peter Goldberg on Aug 28, 2018

Gordon I totally disagree. It is possible to achieve very high CQS if the right LED's are selected to achieve the cool, natural or warm white. This aspect is being discussed amongst all the leading USA and European standards authorities and hopefully theu will conclude that CRI is not the way to select lighting when colour reflection is of importance

By Cary Sherburne on Aug 28, 2018

Plus the work AATCC is doing to bring together all the stakeholders, brands, engineering, manufacturing etc. Will be a big help

By Gordon Pritchard on Aug 28, 2018

@ Peter

What exactly do you totally disagree with in what I wrote?

• That "Lighting standards for the printing industry aren't there for color "accuracy""?
• That they're to provide "lighting constancy to enable better color communications in the production chain"?
• That "It's not feasible to control nor predict the lighting conditions in the home where the shopper examines their purchase"?
• That "led lighting standards intended to provide brand color consistency and reduce customer returns simply will not work"?

Or that they're "to provide work for standards groups."? (Cary S has given the example that one is already in operation)

It's not about the leds output being "cool, natural or warm white" as you put it (BTW as I'm sure you know, those are meaningless terms).

It's quite possible for standards groups to come up with specifications for led output. Unlikely, however, they may even allow for led lights to be a metameric match with current proofing media. If not then the printer might have to support two media/lighting types. More confusion and opportunitis for error.

By William Ray on Aug 29, 2018

Keep in mind that "color" is a result of the Stokes shifting phosphor, the thickness of same, the accuracy of the LED emission centroid and the state of the epitaxial layer. It is a sum of errors problem.

Then add into the mix that there is both built in process variation based upon cost issues in the LED die selection process and wafer (as opposed to die) variation.

Each correction point in this chain of events will raise costs for a certified end product. Also, don't expect the result to be as stable as some would have you believe. Thermal degradation of the die and phosphors for, particularly, high efficacy devices has been well demonstrated. So called blue death -- the shift of emission color to blue -- will occur over effective real life of about 10,000 hours.

By Chris Lynn on Aug 29, 2018

Interesting discussion. Most of the women I know take clothes to a store window to check color before buying because they know that the artificial light does not render accurately, so I disagree with Gordon's assertion of the futility of the effort. And since retailers have an economic incentive to minimize returns, one would expect them to pressure the LED lighting vendors to address both both the spectral shortcomings of LEDs, and the manufacturing consistency & stability issues that William raises.

By Gordon Pritchard on Aug 29, 2018

@Chris
An interesting anecdotal rationale which doesn’t actually address the issues I raised.
Paint manufacturers faced a similar issue with folks deciding on paint colors based on the paint chips they provided customers.
Their solution was quite simple - they put a disclaimer about the impact of lighting on color in their swatch sample books.
Some folks in print industry apply RHEM indicators on their proofs for similar reasons - the assumption being that their customers already have color knowledge and simply need lighting verification.
Perhaps similar solutions to the led issues is not complicated enough?
If you board the wrong train it does not matter how many train drivers are on board or how skilled they are - it’s unlikely you’ll arrive at your desired destination.

By Cary Sherburne on Aug 29, 2018

Chris, if you are in a large store, I don't think most people head for the windows. Plus some of them don't even have outside windows. You would think brands and retailers would pressure LED lighting vendors, but the issue has been the fact they are operating in siloes and many are not even aware of these issues. Ideally, they should put D50 or D65 lights in, but even then, as Ann points out in the quote from her, there are so many other factors affecting color perception that while that would help, a complete solution seems a bit out of reach.

By Ann Laidlaw on Aug 29, 2018

There are two different issues with regard to new lighting technologies and product color in retail.
1) We need consistent, well-defined viewing environments throughout the supply chain for evaluating and approving color. The lighting cabinets are not "mini-stores", and are not intended to mimic the retail environment (which is a mix of multiple sources, signage, products, windows, etc).
2) The sources in the consistent, well-defined viewing environment need to be selected with consideration of the retail environment. I typically recommend 3 sources: A (incandescent), D65, and a fluorescent lamp that is used in the retail environment. These 3 sources are very different from each other, and therefore will provide a robust assessment of metamerism and color inconstancy.

When the CIE publishes LED sources, then users may use a selected LED illuminant that is aligned with the LED source in their retail store instead of the fluorescent source. Visual (lightboxes) and instrumental (computer color systems) processes must be aligned with regard to sources and illuminants.

We will encounter this same challenge with the next advancement in lighting technology. LEDs are giving us a lesson in aligning fast-paced technological improvements with the slow process of international standards, coupled with the need for consistency throughout the supply chain.

By Chris Lynn on Aug 29, 2018

Perhaps my sample of clothes shoppers skews to the high street; but implicit in my comment is the thought that, when the consumer gets home, she will view her purchase under both daylight and indoor lighting and will decide whether or not to take it back.

I think we can agree that the retailer (who - pace Gordon - cannot typically make a paint-shop-style disclaimer) would prefer not to give a refund, and so has an incentive to ensure the in-store viewing environment is as accurate as possible.

As Cary and Ann make clear, this has technical challenges, but it seems a worthy goal, and not merely a job creation scheme for standards bodies.

By Gordon Pritchard on Aug 29, 2018

The problem with your clothes shopper example is that it is a logical fallacy. To rephrase it so that's clear, if I wrote: "Most of the women I know say they've seen Sasquatchs", does it follow that you should disagree with my assertion of the futility to taxanomically classify them?

Yes, we can agree that the retailer would prefer not to give a refund. However, ensuring that the in-store viewing environment is as accurate as possible, is not possible if the target environment is the shopper's home viewing conditions since the retailer has no control over the lighting in that environment. So, despite print industry led standards and specifications the volume of goods returned due to color issues will not change.

Also, setting standards and specifications for led lighting is redundant since there are already standards and specifications for lighting within the print production environment. And setting standards and specifications for led lighting cannot solve the problem of color evaluation at the consumer level in their home viewing environment no more so than setting standards and specifications for lighting in printer viewing booths does.

BTW, I didn't say that color disclaimers should be affixed to consumer goods. What I hope I said was that was one very simple solution to a similar problem. The print industry seems averse to developing simple solutions though.

By Danny Rich on Aug 30, 2018

Perhaps, I can shed some light on the issue of why pirnt shops do not want to put disclaimers on their products.

In clothing retail or paint retail, the consumer has an idea of what color they wish to acquire. So in the store they see a rack of purple sweaters from vendor A and perhaps a second rack from vendor B. She has a pair of slacks at home and this will be to complete an outfit. At home, the sweater from vendor A is not a good complement to the slacks and so she may go back and exchange for the one from vendor B.

In print, the printed sheet goes to a packaging plant and a consumer product goes inside. In the store, the logo colors and images are now considered a contract between the product vendor and the consumer. If the images on the carton are not rendered as expected, the product is left on the shelf. If the product is on a shelf among a large number of competing products, the consumer will look for the brand that they like or recognize and if they do not see that brand's color they will look for a another product. reproduction color quality is a hugely important financial property to the Consumer Product Company. This is so critical that they often send a brand manager into the printing factory to watch and sign-off on the production, even if the printing plant is on the other side of the world.

By Cary Sherburne on Aug 30, 2018

Danny makes a good point. In this instance I was not talking about packaging as much as I was talking about clothing ... so textiles, not print/packaging production. Most experts recommend examining color under three lighting conditions that have different spectra to avoid metamerism and color inconstancy under conditions most likely to be experienced "in the wild." It's not clear to me that is a standard in textiles. Adding LED to the mix -- without clearly defined "standards" (it's actually going to be recommended spectra) and the rapidity with which manufacturers change specifications -- and as has been pointed out here, both the differences in materials used and their propensity to deteriorate (changing color perception) over time, just complicates things even more. The point of the article -- and the longer version that was published in the last AATCC journal -- was to help get all of the stakeholders together so at least they could jointly come to some conclusions about how to handle this while addressing the needs and goals of each as best as possible.

By Barry Brown on Aug 31, 2018

Unless LED lighting can be developed to provide a full/fuller spectrum, perhaps the solution would be
a)to educate the consumer on the limitations of LED lighting (perhaps "real" colour rendition indices compared to full-spectrum natural daylight would be a good start),
and b) for retailers to provide colour evaluation light booths with which customers could evaluate and compare colour under a choice of lighting conditions.

Until then I suspect canny shoppers will do what my wife does, take the product to a window or door - or even outside - to evaluate colour in natural light.

By Robert Arena on Jul 10, 2019

You might be interested in what this company is up to. They have many good articles regarding spectral accuracy.
https://www.waveformlighting.com/absolute-series